Loading...
The URL can be used to link to this page
Your browser does not support the video tag.
Home
My WebLink
About
Staff Report 8984
City of Palo Alto (ID # 8984) Policy and Services Committee Staff Report Report Type: Action Items Meeting Date: 10/23/2018 City of Palo Alto Page 1 Summary Title: Fire Department Emergency Medical Services Future Needs Assessment Title: Review and Acceptance of Fire Department Emergency Medical Services Future Needs Assessment From: City Manager Lead Department: Fire Recommendation Staff recommends that the Policy and Services Committee recommend that Council review and accept the Fire Department Emergency Medical Services Future Needs Assessment, and provide feedback and direction to move forward in exploring viable alternative service delivery models. Background In Fiscal Year 2016, City Council authorized one-time funding for the Fire Department to contract with an expert consulting firm to conduct a study to understand the health risks of the community members based on age, demographics, and utilization of emergency medical services. The Department contracted with Actionable Insights, a respected consulting firm that conducts studies for hospitals and health organizations on the Peninsula, to complete a Community Health Needs Assessment (CHNA). A CHNA is a data-driven examination of the health status indicators that is used to identify key problems and assets in a community, and draws upon research on evidence-based interventions to make recommendations specific to the emerging needs of the community. Typically utilized by hospitals and health providers, the Fire Department worked with Actionable Insights to tailor the research focus and data projection methodologies for the Department with a specific focus on emergency medical needs. This included qualitative and quantitative data collection and analysis; predictive data modeling on the anticipated health needs of the community; literature review of emergency health service provider innovations and best practices; and recommended next steps for the Department and City to consider. City of Palo Alto Page 2 Discussion The Community Health Needs Assessment report is the first of a two phase initiative to identify the current and future community health needs that have an impact on the Department and develop innovative alternative services and programs to meet those needs more effectively and in a less costly manner. The CHNA report includes three primary elements: Information gathering: This included analysis of two years of call volume and patient statistics, and community input from focus groups and key informant interviews. Predictive analysis: Utilizing the two population models from the City Comprehensive Plan and a period life table approach, a prediction of future demographics and call volume was estimated. Distinct predictions were made for daytime and nighttime. The limited data available for Stanford’s daytime population restricted the predictive modeling for Stanford campus to nighttime residential needs only. Literature review of evidence-based interventions and programs: This focused on scientific program evaluations conducted on alternative models for emergency medical service delivery. The programs found to have an impact on reducing reliance on 911 and emergency response are highlighted for consideration of replicating locally. The data shows that the overwhelming need for emergency response is for medical incidents, and with the older population expected to age in place there will be a significant impact on the draw of resources by 2030. The study found a statistically significant correlation with age and engagement of emergency resources for medical needs. Adults age 65 and older represent seventeen percent (17%) of the resident population of the City of Palo Alto, yet make up forty-seven percent (47%) of all Fire Department EMS patients. The predictive analysis estimates that all call types will increase by 2030, with EMS experiencing the largest proportional increase. Emergency Medical Calls are expected to increase between twenty-five to thirty-four percent (25 – 34%), while all other call types are projected to see an increase between ten and sixteen percent (10 – 16%). Proportionally, the largest increase is expected in EMS calls during nighttime hours (8PM to 8AM). However, this increase is not enough to impact the historical trend of most incidents occurring during the daytime (8AM to 8PM). The future of the community needs from the Fire Department will be overwhelmingly for emergency medical services. Some of these calls for service are for incidents that could be treated or prevented without a trip to the hospital. Other jurisdictions already grappling with swells in emergency medical incidents have found great success by City of Palo Alto Page 3 implementing alternative service delivery models. The literature review provides a number of examples that have been proven successful at diverting and reducing the need for emergency response or hospital transport. This includes models such as mobile integrated healthcare; use of nurses or physician assistants in triage; in-home preventive care; or targeted outreach efforts. After the consideration of the data and projected future increases in demand for emergency medical services, the Department will begin the second phase of this project. The Community Impact Planning phase will bring key leaders of health and emergency service agencies together to assess the feasibility of bringing these alternative service delivery models to Palo Alto and explore multi-agency collaboration. At this time, the Department is seeking to inform City Councilthe Policy and Services Committee about the planned future direction and requests feedback as it begins to explore which alternative services would be the best fit for the community. Resource Impact Funding for the second phase of the project, the Community Impact Planning phase, has been identified, and no resource impact is identified at this time. However, the final resource assessment to implement recommended EMS strategies, as described in the study and/or determined during the Community Impact Planning phase, will most likely require a realignment of existing resources and/or new resources to fund these new services. It is anticipated that upon completion of the second phase, additional services will need to be evaluated in the context of the City’s financial outlook, which is currently focused on addressing the projected gap of $2.6 million in the General Fund outlined in the FY 2019-2028 Long Range Financial Report. Policy Implications There are no immediate policy implications at this time, however, as the Department moves forward with the second phase of the project it will be seeking to collaborate with local health service providers to explore alternative models of service delivery. If any innovations are identified as viable, the Department will bring forward proposals to the Policy and Services Committee and the City Council for consideration. Environmental Review Not applicable. Attachments: PAFD_CHNA Final STUDY OF THE HEALTH NEEDS OF THE PALO ALTO/STANFORD AREA AND PAFD EMERGENCY RESPONSE February 1, 2018 1346 The Alameda, Ste. 7-507 San Jose, CA 95126 www.actionableLLC.com PAFD 2017 Report Final Draft © Actionable Insights, LLC ii February 1, 2018 Acknowledgments The Palo Alto Fire Department acknowledges the contributions of the following people and organizations toward this study of the health needs of the Palo Alto/Stanford area and PAFD emergency response: Palo Alto Fire Department Fire Chief Eric Nickel Strategic Operations and Initiatives Manager Amber Cameron EMS Chief Kim Roderick Deputy Fire Chief Geo Blackshire Deputy Fire Chief Catherine Capriles Battalion Chief Kevin McNally Training Battalion Chief Shane Yarbrough Fire Marshall James Henrikson Community partners American Muslim Voice Foundation Stanford Health Care Stanford University Public Safety Stanford University Student Affairs Stevenson House (affordable housing for seniors) Vaden Health Center at Stanford University Veterans Affairs Palo Alto Health Care System Actionable Insights, LLC Co-Founder and Principal Melanie Espino Co-Founder and Principal Jennifer van Stelle, Ph.D. PAFD 2017 Report Final Draft © Actionable Insights, LLC iii February 1, 2018 Table of Contents Acknowledgments ................................................................................................................. ii Table of Figures .................................................................................................................... vi Table of Tables .................................................................................................................... viii Executive Summary ............................................................................................................... 1 Methods .................................................................................................................................................... 1 Pressing Health Needs .............................................................................................................................. 2 Predictive Analysis..................................................................................................................................... 2 Recommendations .................................................................................................................................... 4 Introduction .......................................................................................................................... 4 Study Team ............................................................................................................................................... 5 Situation Overview .................................................................................................................................... 5 CHNA Methods Summary ...................................................................................................... 7 Secondary Statistical Data Collection ....................................................................................................... 7 Sources of Secondary Statistical Data Used in the Assessment ............................................................................ 7 Methodology for Collection, Interpretation, and Analysis of Secondary Data ..................................................... 7 Community Input ...................................................................................................................................... 8 Community Input Gathered Via Key Informant Interviews and Focus Groups ..................................................... 8 Methodology for Collection and Interpretation of Community Input .................................................................. 9 Data Limitations and Information Gaps ................................................................................................. 10 Predictive Analysis Methods Summary ................................................................................ 11 Data Used for Predictive Analysis ........................................................................................................... 11 City of Palo Alto................................................................................................................................................... 11 Stanford .............................................................................................................................................................. 13 Analytical Methods Used for Predictive Analysis .................................................................................... 14 Palo Alto/Stanford Area Geography and Demographics ....................................................... 16 City of Palo Alto ....................................................................................................................................... 17 Ethnicity, Origin, and Language .......................................................................................................................... 18 Age ...................................................................................................................................................................... 18 Stanford................................................................................................................................................... 19 Ethnicity, Origin, and Language .......................................................................................................................... 20 Age ...................................................................................................................................................................... 21 Community Health Needs Assessment (CHNA) ..................................................................... 22 Research Questions ................................................................................................................................. 22 Pressing Health Needs ............................................................................................................................ 22 PAFD 2017 Report Final Draft © Actionable Insights, LLC iv February 1, 2018 Behavioral Health ................................................................................................................................................ 22 Health Care Access and Delivery ......................................................................................................................... 24 Housing ............................................................................................................................................................... 25 Older Adult Health .............................................................................................................................................. 25 Transportation and Traffic .................................................................................................................................. 27 Youth and Young Adult Health ............................................................................................................................ 28 Equity and Cultural Competency ............................................................................................................. 28 Additional Community Concerns ............................................................................................................. 30 PAFD Calls and Medical Emergency Patients: City of Palo Alto, 2015 and 2016 ..................... 31 PAFD CAD Calls Originating in the City of Palo Alto ................................................................................ 31 Current Call Volume and Type ............................................................................................................................ 31 Current Medical Emergency Calls by Time of Day .............................................................................................. 33 PAFD Medical Emergency Patient Volume Originating in the City of Palo Alto ...................................... 33 Current Medical Emergency Patient Volume by Age .......................................................................................... 34 Current Medical Emergency Patient Volume by Gender .................................................................................... 35 Current Medical Emergency Patient Volume by Incident Location (ZIP Code) ................................................... 35 Current Medical Emergency Patient Volume by Type of Illness/Issue ............................................................... 37 Predictive Analysis Findings: City of Palo Alto, Year 2030 ..................................................... 39 Projected Population Overall and by Time of Day .................................................................................. 39 Predicted Medical Emergency Call Volume and Other Call Types by Time of Day ................................. 42 Predicted Medical Emergency Patient Volume Originating in the City of Palo Alto ............................... 44 Predicted Medical Emergency Patient Volume Overall and by Age ................................................................... 44 Predicted Medical Emergency Patient Volume by Time of Day ......................................................................... 45 Predicted Medical Emergency Patient Volume by Type of Illness/Issue ............................................................ 46 PAFD Calls and Medical Emergency Patients: Stanford, 2015 and 2016 ................................. 49 PAFD CAD Calls Originating in the Stanford Area ................................................................................... 49 Current Call Volume and Type ............................................................................................................................ 49 Current Medical Emergency Calls by Time of Day .............................................................................................. 50 PAFD Medical Emergency Patient Volume Originating in the Stanford Area ......................................... 51 Current Medical Emergency Patient Volume by Age .......................................................................................... 52 Current Medical Emergency Patient Volume by Gender .................................................................................... 53 Current Medical Emergency Patient Volume by Incident Location (ZIP Code) ................................................... 53 Current Medical Emergency Patient Volume by Type of Illness/Issue ............................................................... 54 Predictive Analysis Findings: Stanford, Year 2030 ................................................................. 56 Projected Population ............................................................................................................................... 56 Predicted Medical Emergency Call Volume and Other Call Types .......................................................... 57 Predicted Medical Emergency Patient Volume Originating in the Stanford Area .................................. 61 Predicted Medical Emergency Patient Volume Overall and by Age ................................................................... 61 Predicted Medical Emergency Patient Volume by Type of Illness/Issue ............................................................ 62 PAFD 2017 Report Final Draft © Actionable Insights, LLC v February 1, 2018 Recommendations & Next Steps .......................................................................................... 65 Addressing the Expected Increase in EMS Calls ...................................................................................... 65 Successful Strategies to Reduce EMS Calls and/or Reliance on Emergency Transport ...................................... 65 Implementation Challenges and Best Practices in CP/MIHP .............................................................................. 70 Addressing Community Health Needs ..................................................................................................... 72 Next Steps in Addressing Future Needs and Capacity ............................................................................. 74 Appendix 1 .......................................................................................................................... 75 Sources .................................................................................................................................................... 75 Appendix 2 .......................................................................................................................... 84 Methods .................................................................................................................................................. 84 Appendix 3 .......................................................................................................................... 86 Mapping of Primary Impressions to Categories ...................................................................................... 86 Appendix 4 .......................................................................................................................... 92 Call Volume and Call Type Predictions .................................................................................................... 92 Appendix 5 .......................................................................................................................... 94 Population Figures by Age – Current and Predicted ............................................................................... 94 Current Figures, Including Populations by Age ................................................................................................... 94 Year 2030 Predictions, Including Predictions by Age .......................................................................................... 95 Appendix 6 ........................................................................................................................ 101 2030 Predicted ME Call/Patient Volume, Based on CAD Calls and ESO ME Patient Volume................ 101 City of Palo Alto................................................................................................................................................. 101 Stanford ............................................................................................................................................................ 104 Appendix 7 ........................................................................................................................ 107 Total Medical Emergency Patient Volume by Incident ZIP Code by Year: Percentage and Number .... 107 Appendix 8 ........................................................................................................................ 109 Current 2015/2016 Average Patient Volume by Primary Impression and Age ..................................... 109 Appendix 9 ........................................................................................................................ 111 Current 2015/2016 Average and Predicted 2030 Patient Volume by Primary Impression, by City and Time of Day ........................................................................................................................................... 111 Appendix 10 ...................................................................................................................... 113 Palo Alto/Stanford CHNA Data Dashboard ........................................................................................... 113 Appendix 11 ...................................................................................................................... 127 Palo Alto/Stanford Neighborhood Data Dashboard ............................................................................. 127 PAFD 2017 Report Final Draft © Actionable Insights, LLC vi February 1, 2018 Table of Figures Figure 1, Map of the Cities of Palo Alto and Stanford .................................................................. 16 Figure 2, City of Palo Alto Daytime Population by Resident Status ............................................. 17 Figure 3, Ethnicity by Geography, Santa Clara County versus City of Palo Alto Neighborhoods . 18 Figure 4, Population by Geography and Age Range, Santa Clara County versus City of Palo Alto ....................................................................................................................................................... 19 Figure 5, City of Stanford Daytime Population by Resident Status .............................................. 20 Figure 6, Ethnicity by Geography .................................................................................................. 20 Figure 7, Population by Geography and Age Range ..................................................................... 21 Figure 8, Proportion of Population Spending 30% or More of Income on Housing, 2011-2015 . 25 Figure 9, Alzheimer’s Disease Mortality Rate ............................................................................... 26 Figure 10, PAFD Calls for Service Originating in the City of Palo Alto, 2015/2016 Average ........ 32 Figure 11, PAFD Calls by Type and Time of Day, 2015/2016 Average, City of Palo Alto .............. 33 Figure 12, Current Medical Emergency Patient Volume by Age Category, 2015/2016 Average Counts and Percentages, City of Palo Alto ................................................................................... 34 Figure 13, Current Medical Emergency Patient Volume by Age Category, 2015/2016 Average Percentages, Compared to Residential Population by Age Category, City of Palo Alto ............... 35 Figure 14, Map of Current Medical Emergency Patient Volume by Incident ZIP Code, 2015/2016 Average ......................................................................................................................................... 36 Figure 15, Current and Year 2030 Projected Populations by Age Category, Daytime, City of Palo Alto ................................................................................................................................................ 40 Figure 16, Current and Year 2030 Projected Populations by Age Category, Nighttime, City of Palo Alto ........................................................................................................................................ 41 Figure 17, Predicted Change in Average Annual Call Volume from 2015/2016 to 2030, Major Call Types, City of Palo Alto ................................................................................................................. 42 Figure 18, Predicted Change in Average Annual Call Volume from 2015/2016 to 2030, Major Call Types, Daytime, City of Palo Alto .................................................................................................. 43 Figure 19, Predicted Change in Average Annual Call Volume from 2015/2016 to 2030, Major Call Types, Nighttime, City of Palo Alto ............................................................................................... 44 Figure 20, Current and Year 2030 Predicted Age-Adjusted Medical Emergency Patient Volume by Age Category, City of Palo Alto ................................................................................................ 45 Figure 21, Comparison of Actual Average Annual Patient Volume versus Predicted Change in Average Annual Patient Volume from 2015/2016 to 2030, by Primary Impression, City of Palo Alto, Daytime ................................................................................................................................ 47 Figure 22, Comparison of Actual Average Annual Patient Volume versus Predicted Change in Average Annual Patient Volume from 2015/2016 to 2030, by Primary Impression, City of Palo Alto, Nighttime .............................................................................................................................. 48 Figure 23, PAFD Calls for Service Originating from Stanford, 2015/2016 Average ...................... 50 Figure 24, PAFD Calls by Type and Time of Day, 2015/2016 Average, Stanford .......................... 51 PAFD 2017 Report Final Draft © Actionable Insights, LLC vii February 1, 2018 Figure 25, Current Medical Emergency Patient Volume by Age Category, 2015/2016 Average Counts and Percentages, Stanford ............................................................................................... 52 Figure 26, Current Medical Emergency Patient Volume by Age Category, 2015/2016 Average Percentages, Compared to Residential Population by Age Category, Stanford ........................... 53 Figure 27, Current and Year 2030 Projected Residential Populations by Age Category, Stanford ....................................................................................................................................................... 57 Figure 28, Predicted Change in Average Annual Call Volume from 2015/2016 to 2030, Major Call Types, Stanford ............................................................................................................................. 58 Figure 29, Predicted Change in Average Annual Call Volume from 2015/2016 to 2030, Major Call Types, Daytime, Stanford .............................................................................................................. 59 Figure 30, Predicted Change in Average Annual Call Volume from 2015/2016 to 2030, Major Call Types, Nighttime, Stanford ........................................................................................................... 60 Figure 31, Current and Year 2030 Predicted Age-Adjusted Residential Medical Emergency Patient Volume by Age Category, Stanford .................................................................................. 61 Figure 32, Comparison of Actual versus Predicted Absolute Change in Average Annual Patient Volume from 2015/2016 to 2030, by Primary Impression, Stanford, Daytime ........................... 63 Figure 33, Comparison of Actual versus Predicted Absolute Change in Average Annual Patient Volume from 2015/2016 to 2030, by Primary Impression, Stanford, Nighttime ......................... 64 PAFD 2017 Report Final Draft © Actionable Insights, LLC viii February 1, 2018 Table of Tables Table 1, Details of Key Informants Interviewed ............................................................................. 9 Table 2, Details of Focus Groups Conducted .................................................................................. 9 Table 3, Summary of EIR Scenarios: Population and Employment Parameters(1) ....................... 12 Table 4, All Medical Emergency (ME) Patients, Categories of Primary Impressions, 2015/2016 Average, in Order of Frequency for the City of Palo Alto ............................................................. 37 Table 5, Predicted Year 2030 Medical Emergency Patients, Categories of Primary Impressions, by City, in Order of Frequency, City of Palo Alto .......................................................................... 46 Table 6, All Medical Emergency (ME) Patients, Categories of Primary Impressions, 2015/2016 Average, in Order of Frequency for Stanford ............................................................................... 54 Table 7, Predicted Year 2030 Medical Emergency Patients, Categories of Primary Impressions, in Order of Frequency ....................................................................................................................... 62 Table 8, Total Medical Emergency Patient Volume by Primary Impression by Year: Percentage and Number .................................................................................................................................. 86 Table 9, Stanford Projected Growth in Academic Year Residential Population Through 2035 (from Stanford GUP Tab 5.5) ........................................................................................................ 99 Table 10, Stanford Projected Growth in Worker Population Through 2035 (from Stanford GUP Tab 8) .......................................................................................................................................... 100 Table 11, Current 2015/2016 Call Volume, City of Palo Alto ..................................................... 101 Table 12, Current 2015/2016 and Predicted 2030 Call Volume, City of Palo Alto ..................... 102 Table 13, Current 2015/2016 and Predicted 2030 Daytime Call Volume, City of Palo Alto ...... 102 Table 14, Current 2015/2016 and Predicted 2030 Nighttime Call Volume, City of Palo Alto .... 103 Table 15, Current 2015/2016 and Predicted 2030 Medical Emergency (ME) Patient Volume, City of Palo Alto .................................................................................................................................. 103 Table 16, Current 2015/2016 Call Volume, Stanford ................................................................. 104 Table 17, Current 2015/2016 and Predicted 2030 Call Volume, Stanford ................................. 105 Table 18, Current 2015/2016 and Predicted 2030 Daytime Call Volume, Stanford .................. 105 Table 19, Current 2015/2016 and Predicted 2030 Nighttime Call Volume, Stanford ............... 106 Table 20, Current 2015/2016 and Predicted 2030 Patient Volume, Day versus Night, Stanford ..................................................................................................................................................... 106 Table 21, Total Medical Emergency Patient Volume by Incident ZIP Code by Year .................. 107 Table 22, Current 2015/2016 Average Patient Volume by Primary Impression and Age, City of Palo Alto ...................................................................................................................................... 109 Table 23, Current 2015/2016 Average Patient Volume by Primary Impression and Age, Stanford ..................................................................................................................................................... 110 Table 24, Current and Predicted Year 2030 Medical Emergency Patients, Categories of Primary Impressions, by Time of Day, City of Palo Alto ........................................................................... 111 Table 25, Current and Predicted Year 2030 Medical Emergency Patients, Categories of Primary Impressions, by Time of Day, Stanford ....................................................................................... 112 PAFD 2017 Report Final Draft © Actionable Insights, LLC ix February 1, 2018 Table 26, Data Employed for Community Health Needs Assessment: Combined Palo Alto/Stanford Area ...................................................................................................................... 113 Table 27, Data Employed for Community Health Needs Assessment: Cities of Palo Alto and Stanford (Separately) .................................................................................................................. 127 PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 1 February 1, 2018 Executive Summary In the spring and summer of 2017, the City of Palo Alto Fire Department (PAFD) conducted a study of the Palo Alto/Stanford area and PAFD emergency response to determine current health needs of the area and to predict future needs. Health needs were defined as physical health conditions and outcomes, mental and behavioral health, and social determinants of health. The study involved three elements: 1. A community health needs assessment (CHNA), which included collection of statistical data and community input; 2. A predictive analysis, which included a review of current data from PAFD and a modified or “period” life table approach to predict future PAFD volume using demographic projections; and 3. A brief literature review to understand alternative models for emergency medical services (EMS) delivery, including prevention efforts. This data analysis and report is the first step towards implementing changes in the community. Following presentation of this report, including some suggestions for consideration, PAFD hopes to convene community partners that play a part in preventative health to develop recommendations and an action plan. Methods Actionable Insights, a professional health research company contracted to complete this study, gathered both primary and secondary data for this study. For the CHNA, our firm blended primary and secondary data to better understand the community’s health needs. The health needs that were identified include physical health, behavioral health (wellbeing), and social determinants of health (including economic factors and education). The secondary data came from a variety of sources (see Appendices 10 and 11 for statistical data and sources). We also conducted primary research for the CHNA by soliciting input, via interviews and focus groups, from community members and professionals who serve the Palo Alto/Stanford area and who have insight into the community’s needs. For the predictive analysis, looking separately at the City of Palo Alto and at Stanford, we analyzed current emergency medical needs using secondary data from the PAFD, and predicted how the need for EMS will change by 2030 based on current information and expected demographic trends. The methods for the predictive analysis are further detailed in the Predictive Analysis Methods Summary section of this report. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 2 February 1, 2018 Finally, we conducted a brief literature review on new models for EMS service delivery, using both scholarly and journalistic literature, in order to make informed recommendations to the PAFD. Based on these research activities, we created a list of health needs, predictions of future call volume by city and call type, and a set of recommendations for action. Pressing Health Needs The CHNA determined the pressing health needs in the Palo Alto/Stanford area to be (in alphabetical order): Behavioral health (mental health and substance abuse) Health care access and delivery (including equity and cultural competency) Housing Older adult health Transportation and traffic Youth/young adult health Predictive Analysis Actionable Insights conducted a predictive analysis of PAFD’s future (year 2030) call volume by call type and EMS patient volume, using population projections and current data from PAFD. The current data, averaged across the years 2015 and 2016, indicate: On average, across the two years 2015/2016, 67% of calls to PAFD originating in the Palo Alto/Stanford area were for EMS. Breaking that down: o Fully 70% of calls originating in the City of Palo Alto were for EMS. o An average of 50% of calls from Stanford were for EMS.1 The most common reason for EMS in the Palo Alto/Stanford area across the two years was altered level of consciousness (ALOC), followed by injury/hemorrhage. More specifically: o The top three reasons for EMS in the City of Palo Alto were ALOC (29% on average), injury/hemorrhage (23% on average), and cardiac-related issues (average of 11%). o The top three reasons for EMS in Stanford were ALOC (29% on average), injury/hemorrhage (28% on average), and behavioral health-related issues (average of 9%). 1 The average of both is higher than one might expect because there are far more calls from the City of Palo Alto than there are from Stanford. Note that these figures are based on PAFD computer-aided dispatch (CAD) unduplicated data for 2015 and 2016. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 3 February 1, 2018 Generally, the patient volume for EMS increases with age. However, Stanford has a larger population of younger residents due to its university, explaining the spike in patient volume for Stanford among those aged 15-24. We calculated year 2030 predictions based on six growth scenarios provided by the City of Palo Alto and on the latest Stanford University General Use Permit Application. The City of Palo Alto’s growth scenarios for the years 2014 to 2030 “test the impacts of various policy changes which would presumably yield different amounts of job and population growth” (Hillary Gitelman, Director, Planning & Community Environment Department, City of Palo Alto, personal communication, December 21, 2016). Findings indicate that the total number of calls will increase between now and 2030 for all call types, and the volume of patients will increase between now and 2030 for EMS. More specifically: The EMS call increase is expected to be proportionally higher than the increase in other call types (for the City of Palo Alto, depending on the growth scenario utilized, between 25% and 34% compared to all other call types’ combined increase of 10% to 16%; for Stanford, about 36% compared to all other call types’ combined increase of between 30% and 42%). For the City of Palo Alto, Scenario 6 (which the Planning & Community Environment Department calls “Housing Tested II”) generally yields the highest increase in EMS volume while Scenario 2 (called “Slowing Growth”) generally yields the lowest increase. The proportion of daytime calls from the City of Palo Alto is predicted to remain the same over time (67%). The predictive analysis suggests that proportion of nighttime calls from Stanford may increase over time, such that daytime call from Stanford is reduced to 58%.2 Finally, the prediction is that by 2030, EMS patient volume will increase between 24% and 33% from the 2015/2016 annual average for the City of Palo Alto, depending on the growth scenario utilized, and about 30% for Stanford. PAFD’s data show that older individuals are more likely than younger individuals to have a medical emergency in the combined Palo Alto/Stanford jurisdiction. Comparatively, a much greater proportion of older adults are being served by PAFD for medical emergencies compared to the proportion of older adults in the resident population. Both cities will have more older medical emergency patients in 2030 than they have had in 2015/2016. However, because Stanford has – and will continue to have – a smaller and younger set of medical emergency patients, despite the increasing median age of area populations, the City of Palo Alto will see both a much greater absolute number and a greater skew towards older adult patients in 2030 than Stanford will. For example, in the City of Palo Alto, in the highest-growth scenario, the volume of annual EMS patients is expected to rise from 4,330 to as many as 5,766, with 68% of those being age 2 We caution that Stanford’s daytime call predictions were made without access to age projections for that population. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 4 February 1, 2018 55+, while in Stanford, the volume of annual EMS patients is expected to rise from 933 to 1,213, with only 32% of those being age 55+. Recommendations The projected increase in future population (especially the senior population) demands that cities across the Bay Area begin planning now to address these changes. The local community’s concerns about today’s health needs are a bellwether of future needs and, as such, deserve attention. Almost all of the community’s top concerns are issues that PAFD currently addresses through EMS. The predicted corresponding increase in the number of EMS patients by the year 2030 raises the question of how the City of Palo Alto can improve its preparedness to address the community’s needs and concerns. The community has expressed a need for preventative health services, health education, and outreach to address the health needs. Many communities have invested in community integrated health care to improve health outcomes for residents and prevent the need for EMS. Actionable Insights recommends implementing interventions that increase access to preventative health services, including a range of options from low-cost efforts to the creation of large-scale programs. These interventions could include services delivered in collaboration with community partners, other health and service agencies, or by the Fire Department itself. Some best practice models to consider are community education through health workers (such as those utilized in the promotores model, see Centers for Disease Control and Prevention, 2016), remote health services (such as firehouse clinics), and/or mobile integrated health units that provide non- emergency treatment and proactive health services. Introduction The City of Palo Alto Fire Department (PAFD) conducted a study of the health needs of the Palo Alto/Stanford area and PAFD emergency response in 2017. The results of this study will inform PAFD’s planning and strategy for serving the Palo Alto/Stanford community. The study builds upon a countywide community health needs assessment (CHNA) conducted in Santa Clara County by Stanford Health Care and other nonprofit hospitals, but with a focus on the Palo Alto/Stanford area specifically. The Palo Alto/Stanford study focused on three questions: 1. What are the community’s current health needs, including emergency medical needs? 2. How are the community’s emergency medical needs predicted to change by the year 2030? PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 5 February 1, 2018 3. What does the research say about successful models for meeting community health needs, including emergency medical needs? Study Team Fire Chief Eric Nickel of the PAFD, who has served as Fire Chief since 2012 and has a background not only as a firefighter but also as a paramedic, initiated and directed the study. Amber Cameron, Strategic Operations Manager for the Public Safety Department, assisted Chief Nickel. PAFD contracted Actionable Insights, LLC, a Bay Area consulting firm, to conduct the study based on its experience in conducting CHNAs in Santa Clara County and other California counties, and to make recommendations for action such as those found in this report. Jennifer van Stelle, Ph.D. and Melanie Espino, Actionable Insights’ principals and co-leads on the Palo Alto/Stanford study, have experience conducting a variety of research and evaluation projects on topics including health and wellness, wellbeing, social justice, equity and diversity, and higher education. Situation Overview According to National Fire Protection Association (NFPA) data, the proportion of total fire department calls nationwide that were solely for medical aid increased from 54% in 1986 to 64% in 2015 (NFPA, 2016). Fire departments nationwide have reported this trend of increasing emergency medical services (EMS) calls (see, for example, Cooper, 2016; Keisling, 2015; Los Angeles Fire Department, 2015; Luthern, 2014). In the two years 2015 and 2016, emergency medical calls to the PAFD from the Palo Alto/Stanford area were 67% of all PAFD calls (based on PAFD CAD unduplicated data). Breaking that down, fully 70% of calls originating in the City of Palo Alto were for EMS, while 50% of calls from Stanford were for EMS.3 It has been projected that the City of Palo Alto’s resident population will continue an ongoing trend of increasing median age (age 32 in 1970, age 42 in 2010), which is similar to the national trend (City of Palo Alto, 2014). However, it has been noted that the City of Palo Alto’s older adult resident population is “larger than typical” (City of Palo Alto, 2014). Indeed, the median age for most of the counties in California is lower (i.e., younger) than the City of Palo Alto’s median of 41.9 years of age (Governing.com, 2017; City of Palo Alto, 2014). The City of Palo Alto’s median age is also higher (i.e., older) than that of all Bay Area counties except Marin (Governing.com, 2017). Older adults tend to require more care, and research reflects the common-sense notion that “ED [Emergency Department] resource use intensity increases with age” (Latham & Ackroyd- 3 The average of both is higher than one might expect because there are far more calls from the City of Palo Alto than there are from Stanford. Note that these figures are based on PAFD CAD unduplicated data for 2015 and 2016. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 6 February 1, 2018 Stolarz, 2014). It is logical to assume (and we do in fact find in PAFD’s current data) that older adults call 911 for medical emergencies more often than younger adults. The Public Safety Department, which includes police, fire service, and EMS for the City of Palo Alto, is in a unique position to take a broad look at the needs of the community and craft a strategy for addressing those needs now and in the future. By bringing together statistical analyses and the wisdom of the community to understand the community health needs, and leveraging existing research on the changing demographics of the Palo Alto/Stanford area, the PAFD can determine how its current structure and staffing aligns with forecasted changes. Specifically, the PAFD wants to understand how well-prepared it is to respond to the growing proportions of emergency medical calls, as well as the expected rise in absolute numbers of all types of calls over time and how the PAFD can increase its preparedness. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 7 February 1, 2018 CHNA Methods Summary In 2017, the PAFD identified community health needs in a process that mirrors that of the IRS requirements for nonprofit hospitals per the Affordable Care Act. The PAFD CHNA identified health conditions and, secondarily, the drivers of those conditions (including health care access and delivery), via both statistical data and community input. Note that experts and community representatives were asked to comment about the entire Palo Alto/Stanford area, and thus we present the CHNA data for the entire area, blended rather than separated out by city. The CHNA data collection process took place over four months and culminated in this written report in January of 2018. Secondary Statistical Data Collection Sources of Secondary Statistical Data Used in the Assessment Actionable Insights collected secondary statistical data on the combined cities of Palo Alto and Stanford from the Community Commons (2017) data platform, a publicly available source that includes a compilation of over 150 health need indicators. The Community Commons data served as the foundation for PAFD’s statistical data-gathering on community health. In addition, Actionable Insights collected secondary statistical data from the Santa Clara County Public Health Department, including their city and neighborhood profiles. Data from the California Department of Public Health (2013) and other online sources were also collected. Demographic data on the daytime population and resident population were collected from the U.S. Census Bureau. Details about specific sources and dates of the data used may be found in Appendix 1. Methodology for Collection, Interpretation, and Analysis of Secondary Data Actionable Insights used a Microsoft Excel spreadsheet to list indicator data. As described, data were collected primarily through the Community Commons data platform and other statistical sources. Actionable Insights generally retained the health need categories used in the Community Commons data platform spreadsheet (rubric) and integrated data indicators from other sources into the spreadsheet/rubric. Refer to Appendices 10 and 11 for the health needs assessment data collected through the two primary sources, Community Commons and Santa Clara County Public Health Department City and Small Area/Neighborhood Profiles. Specifically related to secondary statistical data, PAFD requested that Actionable Insights examine the following: How do the data perform compared to Healthy People 2020 (HP 2020) aspirational goals4 or 4 Healthy People is an initiative of the U.S. Department of Health and Human Services. For nearly 40 years, it has proffered 10-year national aspirational goals for improving the health of Americans based on existing scientific data. The most recent set of 2020 objectives was updated in December 2014 to reflect the most accurate population data available (www.healthypeople.gov). PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 8 February 1, 2018 against Santa Clara County and California rates or percentages? Are there disparities in outcomes and/or conditions for people in the Palo Alto/Stanford community? Actionable Insights compiled all the combined Palo Alto/Stanford area data indicators and compared them with existing benchmarks (either HP 2020 aspirational goals or statewide averages, whichever was more stringent) in order to evaluate how the indicators performed against these benchmarks. Indicator data were reviewed by gender, age groups, race/ethnicity, and/or geographies when available in order to discover any health disparities. Actionable Insights presented these data to PAFD and its analysis of which indicators failed the benchmarks in an April 2017 telephone meeting with PAFD. Community Input PAFD contracted with Actionable Insights to conduct primary research. The Actionable Insights team collected community input via key informant interviews and focus groups with health experts, community representatives, and community members. Community Input Gathered Via Key Informant Interviews and Focus Groups Following the guidance provided by the IRS for community health needs assessments conducted by nonprofit hospitals, Actionable Insights consulted with individuals with expertise in health and individuals who represent the Palo Alto/Stanford community, and with Palo Alto/Stanford residents. See the lists in Tables 1 and 2. Table 1 includes key informants’ titles and expertise as well as the date each was interviewed. Table 2 briefly describes each focus group and the date and location each took place. Actionable Insights interviewed these experts and community leaders/representatives by telephone for approximately one hour each, May through September 2017. Actionable Insights also conducted focus groups that ran for approximately one hour each in June and July 2017. In these interviews and focus groups, Actionable Insights asked experts and community members the following: What are the health needs that are most important and pressing in the community? Are there any groups or geographies that may have greater or special needs? What solutions do you suggest for addressing those health needs? Are all members of the community benefiting equally from available services, or are there concerns about equity and cultural competency related to physical and mental health? How do you think health care will change in the future? What do you think about alternative service delivery models that could prevent emergency incidents, such as mobile and remote health models focused on prevention and early intervention? PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 9 February 1, 2018 (Professionals only) What are your ideas about how to address the issue of 911 “super- users?” In all, Actionable Insights consulted with six key informants via individual interviews and 22 community members via focus groups. Table 1, Details of Key Informants Interviewed Title and Organization Date Interviewed Public Safety Chief, Stanford University 5/27/2017 Chief Medical Officer, Stanford Health Care 6/2/2017 Vice Provost for Student Affairs, Stanford University 6/14/2017 Associate Vice Provost for Student Affairs and Director of Vaden Health Center, Stanford University 6/14/2017 Manager of Aging Adult Services, Stanford Health Care 7/14/2017 Director, American Muslim Voice Foundation 7/24/2017 Deputy Chief of Staff, VA Palo Alto Health Care System 9/1/2017 Source: Actionable Insights, LLC, 2017, unpublished data. Table 2, Details of Focus Groups Conducted Group Focus Location Date Nonprofit Professionals Mitchell Park Community Center 6/20/2017 Low-Income Seniors Stevenson House 7/6/2017 Neighborhood Representatives Mitchell Park Community Center 7/11/2017 Source: Actionable Insights, LLC, 2017, unpublished data. Methodology for Collection and Interpretation of Community Input Each key informant interview was recorded for internal use only, with permission from the interviewees, and transcribed for use as a stand-alone data source. When all interviews and focus groups had been conducted, Actionable Insights used qualitative research software tools5 to analyze the information and tabulate all of the health needs that were identified as 5 Dedoose Version 7.0.23, web application for managing, analyzing, and presenting qualitative and mixed method research data (2016). Los Angeles, CA: SocioCultural Research Consultants, LLC (www.dedoose.com). PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 10 February 1, 2018 important and pressing to the community. The key findings were combined with the statistical data. This combined articulation was then used to assess community health priorities. Data Limitations and Information Gaps Despite the substantial amount of information collected, data were not available for every single aspect of each community health need, keeping us from being able to fully assess the identified needs. While this is typical for community health needs assessments (CHNAs), it is still important to acknowledge here. Limited data also constrained our understanding of the needs of special populations, including the undocumented, the LGBTQI population, and monolingual speakers of certain languages. Related to this issue, data for these populations can be statistically unstable due to the small numbers or small response rates of community members from these populations, and, as such, are of limited utility in assessing community health needs. Again, these concerns are typical for CHNAs, and are a perennial issue for public health departments and others whose work involves research into community health needs. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 11 February 1, 2018 Predictive Analysis Methods Summary This is a summary of the data and methods used in the predictive analysis. For a detailed description of the data and methods, as well as charts displaying the 2015/2016 data and 2030 predictions, please see Appendices 2-9. Data Used for Predictive Analysis Actionable Insights obtained from PAFD two datasets containing data from the years 2015 and 2016. The dataset from the computer-aided dispatch records system (CAD) contained all calls, by type (e.g., fire, medical emergency, false alarm), and by city for the City of Palo Alto and for Stanford. The dataset from the emergency medical records management system (ESO) contained all patients by age, gender, incident ZIP Code, and primary impression (e.g., diabetic symptoms, respiratory distress, cardiac rhythm disturbance). Both datasets contained information on date and time of day of each incident. For the sake of increased data stability, we generated averages of each call type in the CAD dataset across the two years, and of patient volume, by age, in the ESO dataset across the two years. These averages formed the basis for the year 2030 predictions. City of Palo Alto The City of Palo Alto provided six different growth scenarios for the year 2030 based on 2014 resident, jobs, and housing data, which were developed as part of the Comprehensive Plan. These population projection scenarios were provided to Actionable Insights by Hillary Gitelman, Director, Planning & Community Environment Department, City of Palo Alto, via email on December 21, 2016. In her email, Ms. Gitelman explained: Our EIR [Environmental Impact Report] planning scenarios look out to the year 2030 and are testing the impacts of various policy changes which would presumably yield different amounts of job and population growth. Our Scenario 2 reflects the population in 2030 that we anticipate if Palo Alto keeps building housing at the same rate it has over time and if we take steps to temper the rate of job growth. Our Scenario 4 reflects ABAG projections of jobs and housing. Table 3, on the next page, provides these scenarios. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 12 February 1, 2018 Table 3, Summary of EIR Scenarios: Population and Employment Parameters(1) Population/ Housing Non-Res Sq. Ft. (2) Jobs Jobs/Housing Balance (3) BASE FIGURES (2014) 65,685/28,545 27.0M 95,460 Jobs/Employed Residents Ratio of 3.03 Scenarios: Net Change 2015-2030 (City of Palo Alto Only) Jobs/Housing Balance (3) Population/ Housing Non-Res Sq. Ft. (2) Jobs 1. Business as Usual 6,600/2,720 3.3M 15,480 Jobs/Employed Residents Ratio of 3.20 2. Slowing Growth 6,600/2,720 3.0M 9,850 Jobs/Employed Residents Ratio of 3.04 3. Housing Tested I 8,435/3,545 3.5M 12,755 Jobs/Employed Residents Ratio of 3.03 4. Sustainability Tested I 10,455/4,420 4.0M 15,480 Jobs/Employed Residents Ratio of 3.04 5. Sustainability Tested II 8,435/3,546 2.4M 8,868 Jobs/Employed Residents Ratio of 2.93 6. Housing Tested II 14,080/6,000 2.4M 8,868 Jobs/Employed Residents Ratio of 2.71 Notes: (1) The scenarios also include different ideas for zoning/implementation actions, transportation investments, and sustainability measures. (2) This number includes 1.3M sq. ft. that has already been approved at the Stanford Medical Center. The balance of the new nonresidential square footage would be located in areas both inside and outside of the “monitored areas” referenced in Policy L-8 and Map L-6 in the Comp Plan and in areas both inside and outside of the area subject to the interim annual limit of 50,000 square feet new office/R&D space. (3) The number of employed residents in 2030 is estimated at approximately 48% of total population based on ABAG Projections 2013. The ratio of jobs to employed residents in this column assumes a 2014 base of 65,685 people [i.e., total residents] and 95,460 jobs. Sources: Hilary Gitelman, City of Palo Alto Planning & Community Environment department, personal communication, December 21, 2016; Roland Rivera, City of Palo Alto Planning & Community Environment department, personal communication, November 20, 2017. To align our 2030 predictions with these scenarios, Actionable Insights obtained additional 2014 population data, including resident age breakdowns by city, county, and state from the U.S. Census Bureau (2014). PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 13 February 1, 2018 We also obtained daytime visitor estimates for the City of Palo Alto from its Chamber of Commerce, and calculated nighttime visitor estimates for the City of Palo Alto based on reported hotel occupancy rates and related taxes for 2014. Additionally, we acquired U.S. projections of year 2030 population by age from the U.S. Census (Colby & Ortman, 2015). Year 2030 age projections for the populations of Bay Area counties and the state were obtained from the California Department of Finance (CA DOF, 2017).6 Finally, we estimated daytime population figures for the City of Palo Alto based in part on a 2013 City of Palo Alto transportation survey that reported percentages of out-commuters and in-commuters from various nearby counties. Further explanation of these figures, including population proportions by age for 2014 and 2030, are available in Appendix 5. Stanford Stanford University (2016) provided 2015 data, 2018 projections, 2020 projections, and a growth scenario for 2035, both for residents and for its daytime population, in its 2018 General Use Permit Application or “GUP.” The GUP, published by Stanford University on November 21, 2016, provides estimates of the resident population (Tab 5) and of the daytime population (Tab 8), including both workers and students, for the years 2015, 2018, 2020, and 2035 (see Appendix 5). However, the GUP data and estimates are not provided at a granular-enough level to determine population ages, despite requests for the same.7 Use of age groups from other geographies (e.g., San Mateo County, Santa Clara County, the City of Palo Alto) as proxies for Stanford’s daytime or nighttime populations are inappropriate because Stanford is not like these other areas; it has a much larger young-adult population due to its university. After considerable effort by Actionable Insights and discussions with PAFD, it was decided that Stanford’s daytime population would not be estimated by age group. For both current and 2030 Stanford nighttime population projections, we used the U.S. Census Bureau’s estimates of Stanford’s 2014 resident (nighttime) population and age proportions.8 We obtained daytime visitor estimates for Stanford from its Visitor’s Center. Further explanation of these figures, including population proportions by age for 2014 and 2030, are available in Appendix 5. 6 Note that we were unable to obtain age projections by city, and were advised that no agency or organization provides such projections (Stephen Levy, personal communication, December 14, 2016). 7 Note that a request for age-related information about the Stanford daytime population was made to the authors of the GUP. The request netted some additional information from Stanford University’s Human Resources department, but the information provided was still not enough to allow us to determine age groups for Stanford’s daytime population. 8 We used 2014 rather than 2015 data for Stanford because all City of Palo Alto “current” data and the six 2030 scenarios are also based on 2014 figures; this allows for comparisons between Stanford and the City of Palo Alto if desired. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 14 February 1, 2018 Analytical Methods Used for Predictive Analysis After consultation with several statistical experts, it was determined that the most appropriate method for predicting future call volume/type and medical emergency patient volume by age group, gender, incident location, and/or primary impression was a modified life table approach. This is a population ecology approach used to describe a current population and predict future growth or decline (see, for example, Begon, Harper, & Townsend, 1990; Begon, Mortimer, & Thompson, 1996; Krebs, 1985). The U.S. Centers for Disease Control calls this a period life table approach (Arias, 2012). We used a snapshot of the years 2015/2016 and assumed that the cohort of individuals alive at that time in each city, plus the expected increase in individuals in each city over the next 15 years, would be subject (proportionally) to the age-specific Medical Emergency (ME) call/patient volumes, other call types, incident locations, and primary impressions that prevailed for the actual populations in 2015/2016. The greatest factor to affect construction of period life tables for 2030 is the expected increase in the older population in the City of Palo Alto. As one might expect, the 2015/2016 ESO data show that older individuals are more likely than younger individuals to be ME patients in the combined Palo Alto/Stanford jurisdiction. After predicting expected 2030 ME call/patient volume by age group, as well as by location and primary impression, we adjusted the predictions based on CA DOF (2017) age projections. More information about this process may be found in Appendices 6-9. We provide separate age- adjusted predictions for the City of Palo Alto (daytime and nighttime) and for Stanford (nighttime only). Another factor one would expect to affect construction of period life tables for 2030 is gender. While gender is associated with ME patient volume overall (with females somewhat more likely to be ME patients than males), the gender composition of the two cities is not expected to change significantly within the next 15 years.9 Thus, we did not take gender into account in generating the period life tables predicting ME call/patient volume, call types, locations, or primary impressions in 2030. Although it is possible that the proportions of the population of the various ZIP Code areas in the combined Palo Alto/Stanford jurisdiction may change within the next 15 years, no data were available to address population change in such a fine-grained way (i.e., by ZIP Code). Therefore, we did not predict ME patient volume in 2030 by ZIP Code of interaction, either overall or by category of primary impression. 9 Note that while there were no data projecting gender for sub-county-level areas, the projections of gender by county (CA DOF, 2017) suggest that the expected change in gender proportions between now and 2030 is not statistically significant in either San Mateo County or Santa Clara County (Chi-square = 2.00, p > 0.05). Therefore, we do not include gender as a factor in our predictions. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 15 February 1, 2018 Finally, we provide ME call/patient volume and call type projections both overall and separately for daytime (City of Palo Alto only) and nighttime (both City of Palo Alto and Stanford, separately). The nighttime population is equivalent to the resident population plus any estimates of overnight visitors. As suggested by Bhadur (2007), daytime and nighttime populations are fundamentally different, with daytime population including workers, tourists, and business travelers, as well as the residual nighttime resident population that does not leave the area during the day. For PAFD’s purposes, it is important to predict year 2030 call/patient volume and call type, as well as ME patients by location and primary impression, based on these two very different daytime and nighttime populations. Further information on how these populations were constructed and the predictions based on them are available in Appendix 5. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 16 February 1, 2018 Palo Alto/Stanford Area Geography and Demographics The Palo Alto/Stanford area is the focus of the study. Places in the City of Palo Alto are commonly referred to as north and south, with Oregon Expressway/Page Mill Road serving as the dividing line between the two. As shown on the map in Figure 1, the Palo Alto/Stanford area has many distinct neighborhoods. Figure 1, Map of the Cities of Palo Alto and Stanford Source: Amber Cameron, City of Palo Alto Public Safety Department, personal communication, October 4, 2017. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 17 February 1, 2018 The Santa Clara County Public Health Department provides health data for the City of Palo Alto, City of Stanford, and for certain local neighborhoods, which are grouped into five areas to facilitate data analysis in this study:10,11 Barron Park: Includes Green Acres. Downtown North: Includes Crescent Park. Midtown North: South of Oregon Expressway and northeast of Middlefield Road. Includes Palo Verde and Charleston Gardens. Midtown South: South of Oregon Expressway and southwest of Middlefield Road. Includes Ventura and Charleston Meadow (Fairmeadow). Professorville: Includes Old Palo Alto, Duveneck, and St. Francis. City of Palo Alto The City of Palo Alto is the northernmost of the 15 cities in Santa Clara County. The resident population of the City of Palo Alto is approximately 65,700 (U.S. Census Bureau, 2014). However, during the day there is a large influx of workers into the area, as well as approximately 2,400 daily and/or overnight visitors, which together increase the daytime population of the combined area to approximately 132,000 (double the resident population). See Figure 2 for a comparison of resident and non-resident proportions during the daytime. Figure 2, City of Palo Alto Daytime Population by Resident Status Sources: Dawn Billman, Palo Alto Chamber of Commerce, personal communication, April 26, 2017; Hilary Gitelman, City of Palo Alto Planning & Community Environment department, personal communication, December 21, 2016; Thorp, 2014; U.S. Census Bureau, 2014. See Appendix 5 for more details. 10 Visit www.sccgov.org/sites/sccphd/en-us/Partners/Data/Pages/Palo-Alto.aspx for more information on the Santa Clara County Public Health Department’s methodology. 11 Palo Alto Hills was not included in the Palo Alto neighborhood profiles published by Santa Clara County Public Health Department. Non-resident population, 50% Resident population, 50% PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 18 February 1, 2018 Ethnicity, Origin, and Language Compared to Santa Clara County, all City of Palo Alto neighborhoods have a higher proportion of White residents (between 50% and 70% compared to 35% in the county overall). Both Midtown areas (North and South) have higher proportions of Asian residents than the county overall, at 38% and 33% respectively, compared to 32% in the county. (See Figure 3 for chart.) Figure 3, Ethnicity by Geography, Santa Clara County versus City of Palo Alto Neighborhoods Source: SCCPHD, 2016. Between 23% and 37% of the residents of each neighborhood are foreign-born (37% in the county overall), and one quarter or more speak a language other than English at home (52% in the county overall). Those speaking a language other than English at home are most likely to live in Midtown neighborhoods (where the rate is 48% in the north, and 43% in the south). While these data do not specify whether these residents also speak English, it is worth noting that in Santa Clara County over half of those who speak one of the county’s top seven non- English languages at home also speak English “very well” (U.S. Census Bureau, 2015). Age The City of Palo Alto has a higher proportion of adults aged 65 and older than the county overall (17% compared to 12% in the county). The proportion of children in the City of Palo Alto under age five is smaller than the county overall (5% compared to the county at 7%). See Figure 4, below. 0% 20% 40% 60% 80% 100% Santa Clara County Barron Park Downtown North Midtown North Midtown South Professorville Black/Af-Am Latino Asian White PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 19 February 1, 2018 Figure 4, Population by Geography and Age Range, Santa Clara County versus City of Palo Alto Santa Clara County City of Palo Alto Source: U.S. Census Bureau, 2014. The neighborhood with the highest proportion of older adults is Downtown North, where one in five (21%) is over age 65 (SCCPHD, 2016). See Appendix 11 for additional neighborhood data. Stanford The City of Stanford is bounded by the City of Palo Alto on three sides; its remaining boundary, to the southwest, is Junipero Serra Boulevard (show on the map in Figure 1 as the road immediately southwest of the small lake depicted on the Stanford University campus). The resident population of Stanford is estimated at approximately 13,500. However, during the day there is a large influx of workers into the area, as well as approximately 370 daily visitors, which together increase the daytime population of Stanford to nearly 40,000. See Figure 5 for a comparison of resident and non-resident proportions during the daytime. 0-4 5-9 10-14 15-19 20-24 25-34 35-44 45-54 55-64 65+0-4 5-9 10-14 15-19 20-24 25-34 35-44 45-54 55-64 65+ PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 20 February 1, 2018 Figure 5, City of Stanford Daytime Population by Resident Status Sources: D.J. Dull-MacKenzie, Stanford University Visitors Center, personal communication, April 26, 2017; Stanford University, 2016; U.S. Census Bureau, 2014. Note: No estimate is available for overnight visitors to the City of Stanford. See Appendix 5 for more details. Ethnicity, Origin, and Language Just over one quarter of Stanford residents are foreign-born (26%) compared to 37% in the county overall. More than one third (35%) speak a language other than English at home, compared to 52% of county residents overall. While these data do not specify whether these residents also speak English, as mentioned previously, in Santa Clara County over half of those who speak one of the county’s top seven non-English languages at home also speak English “very well” (U.S. Census Bureau, 2015). Compared to Santa Clara County, Stanford has a higher proportion of White residents (44% compared to 35% in the county overall) and Black residents (5% compared to 2% in the county overall). Stanford has a lower proportion of Latino residents than Santa Clara County (11% compared to 27% in the county overall). See Figure 6. Figure 6, Ethnicity by Geography Source: SCCPHD, 2016. Non- resident population, 66% Resident population, 34% 0% 20% 40% 60% 80% 100% Santa Clara County Stanford Black/Af-Am Latino Asian White PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 21 February 1, 2018 Age Due to its university, Stanford has much higher proportions of young adults aged 15-19 and 20- 24 than the county overall (respectively, 29% and 35% compared to 6% for each age group in the county overall). The proportion of children residents in Stanford under age five is smaller than the county overall (3% compared to the county at 7%) and the older adult resident population (age 65+) of Stanford is also smaller than the county (5% compared to 12% in the county overall). See Figure 7, below. Figure 7, Population by Geography and Age Range Santa Clara County Stanford Source: U.S. Census Bureau, 2014. 0-4 5-9 10-14 15-19 20-24 25-34 35-44 45-54 55-64 65+ 0-4 5-9 10-14 15-19 20-24 25-34 35-44 45-54 55-64 65+ PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 22 February 1, 2018 Community Health Needs Assessment (CHNA) Research Questions The PAFD’s CHNA focused on three main questions: 1. What are the pressing health needs in the Palo Alto/Stanford area? 2. Are all Palo Alto/Stanford area residents benefiting equally from available health care services? 3. What suggestions does the community have for improving health in the Palo Alto/Stanford area? We review each of these questions in the sections below. Pressing Health Needs Actionable Insights reviewed statistical data about health needs in the Palo Alto/Stanford area and combined this with the community feedback (six key informant interviews and three focus groups). Note that experts and community representatives were asked to comment about the entire Palo Alto/Stanford area, and thus we present the CHNA data for the entire area, blended rather than separated out by city. The list below was derived by comparing statistical data to benchmarks (county statistics or statewide statistics) and by analyzing the community feedback for common themes. (Please see Appendices 10 and 11 for detailed statistical data.) The following pressing health needs emerged from this analysis (in alphabetical order): Behavioral health (mental health and substance abuse) Health care access and delivery, including equity and cultural competency Housing Older adult health Transportation and traffic Youth/young adult health The PAFD did not engage in ranking the health needs; the goal was to bring together both available statistics and community perceptions to identify the set of most pressing community health needs. Behavioral Health Behavioral health is a category of health needs that includes mental health and substance use. This umbrella designation is often used in health settings, including public health departments, since mental health and substance use issues (including use of illegal drugs, misuse of PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 23 February 1, 2018 prescription drugs, excessive use of alcohol, and the use of tobacco) co-occur so frequently.12 While data on illegal drug use in the Palo Alto/Stanford area are lacking, the data do show that Palo Alto/Stanford area residents spend 18% of their household expenditures on alcohol compared to 13% in California households overall (Community Commons, 2017). Substance use was mentioned by four of six key informants, and excessive alcohol use was mentioned specifically by those who serve the students of Stanford University. Further data on accessibility of alcohol and tobacco are found in the Housing section below. Mental health was discussed in all focus groups and in all but one key informant interview. Many concerns related to mental health were discussed, including the prevalence of mental health issues within the homeless population, youth suicide, social isolation, and PTSD among veterans. The Veterans Affairs (VA) representative also mentioned that veterans with traumatic brain disorders (TBIs) may also have mental health issues, and sometimes TBI symptoms are mistaken for mental health problems. Also, some mentioned the difficulty that first responders face when addressing medical emergencies where there are also mental health concerns. The third-highest type of primary impression among patients whose incidents occurred in the Stanford area is behavioral health, representing over 9% of Stanford’s annual Medical Emergency patient volume on average (see Table 6 on page 54). “One thing that is weighing heavily on me these days, mostly because of the criticism that we’re taking, really is more of around mental health. Are we, as a community, as a nation, addressing mental health issues appropriately?” – Key Informant Regarding proximity to tobacco retail outlets, all of the City of Palo Alto’s neighborhoods have a lower density of those establishments compared to the county (Santa Clara County Public Health Department [SCCPHD], 2016). However, Downtown North, Midtown North, and Midtown South have higher densities of alcohol retail outlets than the City of Palo Alto overall (SCCPHD, 2016). In Downtown North and Midtown North, the number of establishments that sell alcohol per square mile exceeds that of the county overall at 5.1 and 3.7 respectively (compared to 2.7 in the county) (SCCPHD, 2016). Excessive alcohol use can drive the rate of violent crime (National Institute on Alcohol Abuse and Alcoholism, 1997). We note that Downtown North has the highest rate of violent crimes within one mile – 11.6 per 100,000 – among the City of Palo Alto’s neighborhoods, which is 28% lower than the county overall (at 16.0), but 84% higher than the City of Palo Alto overall (at 6.3) (SCCPHD, 2016). 12 See SAMHSA website for more information: www.samhsa.gov/disorders. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 24 February 1, 2018 Health Care Access and Delivery Health care access was mentioned in every focus group and key informant interview. Nearly two thirds (65%) of Palo Alto residents say that the availability of affordable, quality health care is good or excellent, and 74% say that the availability of preventative health care services is good or excellent (Palo Alto City Auditor, 2016). Indeed, 2012 data indicate that in most neighborhoods, only between 4% and 10% of adults are uninsured (SCCPHD, 2016). The highest estimates of uninsured are in Barron Park (13%-15%) and Midtown South (10%-13%), which are still better than the national baseline of 58% (SCCPHD, 2016). Indeed, many community members who participated in this study perceive that most Palo Alto/Stanford area residents have health insurance and the means to afford health care.13 Despite the low proportions of uninsured, the community is concerned that some health care is unaffordable, especially for those in the middle class. While many programs exist for those with low incomes, the community feels that many in the middle class who do not qualify for Medi- Cal or other subsidies may struggle to afford health care. This may be particularly true for seniors in the City of Palo Alto who own homes and do not qualify for Medi-Cal. Affording in- home services may be difficult for those who want to remain in their homes (a significant asset) but have limited incomes. It is also worth noting that the Palo Alto/Stanford area has fewer Federally Qualified Health Centers than the county (1.3 compared to 2.4 per 100,000) (Community Commons, 2017). Regarding mental health care access, fewer than half of the City of Palo Alto’s residents (46%) say there is good availability of affordable, quality mental health care, worse than in 2014 when 63% said so (Palo Alto City Auditor, 2016). Community input validates this finding. One participant explained that traditional mental health office hours (ending at 5:00 p.m.) do not serve the student community well. This limited access puts the burden of addressing mental health crises on the police department or emergency paramedics. “In the U.S. . . . insurance companies have increasingly restricted access to psychiatrists, psychologists, and counseling as part of their health care plans. And what it means is psychiatrists and psychologists have increasingly gone to pay out-of-pocket plans. And so, it restricts who can access those services to just the wealthy.” – Key Informant Our discussion of equity and cultural competency in health care access and delivery may be found in the next section, since focus groups and key informants were asked a distinct and separate question about this topic. 13 We appreciate a reader comment noting that availability and the ability to access are not the same. Further research is suggested to determine whether residents of neighborhoods with relatively higher proportions of uninsured receive the necessary information and resources to access affordable and high-quality health care. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 25 February 1, 2018 Housing Quality and type of housing are important determinants of health, and substandard housing is a major public health issue. Residents of the City of Palo Alto and Stanford are more likely than county residents overall to live in multi-unit housing, with 38% in the City of Palo Alto and 90% in Stanford, compared to 33% in the county (SCCPHD, 2016). However, a smaller proportion in the combined Palo Alto/Stanford area live in overcrowded households (3% or less compared to 8% in the county) (SCCPHD, 2016). Housing in the county is extremely expensive, as evidenced by 38% of the population spending 30% or more of their income on housing. In the Palo Alto/Stanford area, this applies to 35% of the population. See Figure 8. Figure 8, Proportion of Population Spending 30% or More of Income on Housing, 2011-2015 Source: SCCPHD, 2016. Older Adult Health With older adult populations making up a disproportionate share of patients, it was important to consult seniors and professionals who serve seniors for this study. Many concerns were expressed, including concerns about chronic conditions. In the county, 8% have been diagnosed with diabetes, and 10% have been diagnosed with pre-diabetes (SCCPHD, 2014a). The prevalence among those aged 65 and older is highest, at 18% with diabetes and 16% with pre-diabetes (SCCPHD, 2014a). Heart disease prevalence among adults in the county (5.3%) overall is better than the state (6.3%), but worse among Latino adults in the county (5.0%) (University of California Center for Health Policy Research, 2012). It is estimated that 27% of the county’s residents have been diagnosed with high blood pressure (SCCPHD, 2014c). (Data on diabetes prevalence and high blood pressure are not available for Palo Alto/Stanford.) It should be noted that in Palo Alto/Stanford the rates of overweight/obesity, stroke mortality, heart 38%35% 0% 10% 20% 30% 40% 50% Santa Clara County Palo Alto/Stanford Area PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 26 February 1, 2018 disease mortality, and diabetes hospitalizations are favorable compared to the county, but heart disease prevalence is essentially the same compared to the county (both 5.3) (SCCPHD, 2016; University of California Center for Health Policy Research, 2012). However, Alzheimer’s disease mortality data do indicate reason for concern (discussed further below; see Figure 9). “I think senior care is a big problem in a lot of areas. If you’re wealthy, you can look at the fee or whatever, and you get great care. And there are a couple of very good chronic care facilities in Palo Alto, but there’s still lots of [seniors] that have problems accessing care. . . So, I think that population is relatively underserved.” – Key Informant Figure 9, Alzheimer’s Disease Mortality Rate Source: SCCPHD, 2016. Note: See Demographics section for neighborhood definitions. Dementia and cognitive functioning were also big concerns. The mortality rate for Alzheimer’s disease in the City of Palo Alto is 32.4 per 100,000, similar to that of the county overall (at 34.6 per 100,000) (SCCPHD, 2016). In Midtown South, the rate is 40.0 per 100,000, which is 15% higher than the county rate (SCCPHD, 2016). While this rate is not age-adjusted, the proportion of residents aged 65 and older in Midtown South is lower than in most other neighborhoods (15%, compared to 17% in the City of Palo Alto overall) (SCCPHD, 2016) , which may indicate that the likelihood of dying from Alzheimer’s disease in Midtown South is higher than in other neighborhoods (SCCPHD, 2016). Further research is required to better assess this issue. 14 Regarding senior falls, county data show that seniors in Santa Clara County are less likely to be hospitalized but are more likely to die from a fall than California seniors overall (county rate of 14 This rate may be related to skilled nursing facilities and/or senior home care facilities in the area. Mapping these various facilities was beyond the scope of this assessment. 34.6 32.4 34.0 31.9 28.7 40.0 25.7 0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 45.0 Santa Clara County City of Palo Alto Barron Park Downtown North Midtown North Midtown South Professorville PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 27 February 1, 2018 55.4 deaths per 100,000 people compared to state rate of 36.1 per 100,000, not age-adjusted) (California Department of Public Health EpiCenter [EPIC], 2013). Seniors also talked about social isolation and the need for a community where people look out for one another. “In my neighborhood, we are so separated because we have larger lots and actually you don’t see them. Most of the time, you don’t see the neighbors except for an annual function that we have.” – Focus Group Participant Transportation and Traffic Transportation and traffic are public health hazards that can affect individual decisions about health, and can compromise community members’ health through exposure to pollution and other harmful elements. Transportation and traffic were mentioned in all three focus groups, and by two out of six key informants. Residents are concerned with the amount of traffic causing stress, traffic accidents, and threatening the safety of pedestrians and bicyclists. As described earlier, two thirds (64%) of the City of Palo Alto’s daytime population are non- residents and thus commute to the city (Actionable Insights, unpublished data, 2017). We also know that the most common mode of transportation to work among the City of Palo Alto’s residents is driving alone (65%) (SCCPHD, 2016), which contributes to the amount of traffic experienced in the community. The vast majority of Palo Alto/Stanford area residents live in close proximity to public transportation (within a half-mile of a regional bus or train and within a quarter-mile of any bus/light rail) (SCCPHD, 2016). However, in the Midtown North neighborhood, this proportion is only 59%, compared to 85% in the City of Palo Alto overall (SCCPHD, 2016). Twelve percent of the City of Palo Alto’s residents carpool to work or take public transportation most of the time, which is similar to the county (13%) (SCCPHD, 2016). But there has been a substantial decrease in the proportion of City of Palo Alto residents who say that public transportation is easy to use, dropping from 60% in 2006 to 28% in 2016 (Palo Alto City Auditor, 2016). A concern of study participants may be related to the large influx of daytime workers in the Palo Alto/Stanford area; some participants believe that car commuters are using neighborhood streets to get to work. Community members in the study (and those in the countywide study, Stanford Health Care, 2016) expressed concern about the increased risk of traffic and pedestrian accidents in their neighborhoods, as well as the increased noise from traffic. Airplane noise was also cited as a cause of noise pollution and stress. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 28 February 1, 2018 Youth and Young Adult Health Referring back to Figure 7, nearly two thirds of the Stanford population (65%) are adults aged 18-24, due in part to the undergraduate and graduate student populations at Stanford University. Youth and young adult health was mentioned by several study participants, including three key informants. Mental health (and specifically suicide) is a problem that the community has been very aware of since 2009, and the issue has been the subject of a CDC study (Garcia-Williams et al., 2016) and numerous media articles across the country (Lee, 2017; Rosin, 2015). According to the Santa Clara County Public Health Department’s 2016 report, the suicide rate among youth and young adults 10 to 24 years old is highest in the City of Palo Alto, at 14.1 per 100,000, more than double the county rate of 5.4 deaths of those aged 10-24 (Garcia-Williams et al., 2016). Project Safety Net is a collaborative that was formed in 2009 in response to this crisis, and includes the Palo Alto Unified School District and the Santa Clara County Public Health Department (City of Palo Alto, 2017a). Indeed, many new programs related to suicide prevention have been instituted in the community since 2009, which were known to study participants. Some community members indicated that youth and young adults are well aware of resources for mental health crisis supports. However, mental health was not the only health need mentioned in the context of youth and young adult health. Participants who serve Stanford University students also discussed the risk of injury and death due to alcohol consumption. There was concern among participants that the problem of alcohol consumption can mask more complex mental health issues. For example, students with mental health problems who have consumed excessive alcohol may be treated for drunkenness but in fact may also need help for underlying mental health issues. Likewise, a student who calls 911 because they were injured while intoxicated may receive treatment for the injury without being assessed for alcohol dependence. For this reason, study participants felt a holistic approach to treating this population is needed. Pedestrian/bicycle safety is another need that youth and young adults have expressed to community providers that serve them, as described in the quote below. “A lot of the kids in this area ride their bikes to school, so they talk about safety a lot. They also talk about . . . getting to school – and the high levels of traffic that there are in the mornings that impede them from getting to school on time.” – Focus Group Participant Equity and Cultural Competency One of the primary CHNA research questions related to equity: Are all residents of the Palo Alto/Stanford area benefiting equally from available health care services and other things that make residents healthy? Regarding equity, community members expressed concern that not PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 29 February 1, 2018 everyone has the same access to quality health care within the area and within the county because not all workers can afford to live in the area due to the high cost of living. Specific local populations identified by community members as lacking access to health care were those experiencing homelessness, those who are low-income, and undocumented immigrants who do not have access to subsidized health insurance under the Affordable Care Act. Cultural competence is the ability to interact effectively with people of different cultures, which is key to effective health care access and delivery. A lack of cultural competency contributes to health disparities, resulting in behaviors, attitudes, and policies that undermine the goal of providing quality health care and emergency response. PAFD asked key informants whether cultural competency was an issue in the Palo Alto/Stanford area. Health care professionals who were consulted explained that cultural competency is addressed adequately in health systems such as the VA Palo Alto Health Care System, Stanford Health Care, and Stanford University’s Vaden Health Center, and that there is a recent focus on building competencies to serve the LGBTQI population. For example, a key informant explained that while gynecologists are understandably trained to address pregnancy with patients, asking lesbian patients about their risk for pregnancy can cause these patients to feel misunderstood and/or disrespected. Professionals who participated in a focus group also said that cultural competency related to gender and sexual orientation is important so that everyone feels welcome by health care professionals. One key informant who represented the Muslim population also indicated that it is important for first responders to understand cultural norms, and that there are some misconceptions about treating Muslims. For example, she explained that while Muslims generally prefer a clinician of the same sex, in emergencies one should not hold back providing care if only a male EMT is available to treat a female patient. To Muslims and many others, maintaining modesty (with regards to exposing one’s body) is important; when clinicians cover patients or keep onlookers away, they preserve the dignity of their patients. In addition, this same key informant indicated that because many people in the Palo Alto/Stanford area are vegetarian, hospitals should take extra care to offer foods that are made without any animal products (e.g. gelatin, broth). While cultural competency regarding race/ethnicity was not mentioned by other study participants, additional information provided by the African Ancestry Report indicated that Africans/African Americans of different national origins, religions, and sexual orientations in the county commonly experience mistreatment and a lack of cultural competence among health care providers (SCCPHD, 2014b). For the veteran population, gender equity was discussed by the expert interviewed. While men are estimated to make up 91% of veterans in the U.S., the proportion of female veterans is rising over time (U.S. Department of Veterans Affairs, 2017). The Veterans Affairs (VA) PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 30 February 1, 2018 representative interviewed said, “If you’re not thinking about women as veterans, then you may miss out on a host of issues that women may face.” The representative suggested that women combat veterans who choose to go to non-VA providers are particularly at risk in this regard, as non-VA providers may be less likely to know about or consider the health implications of their patient being a combat veteran. It is important for residents to be able to understand information, warnings, and directions given by health care and emergency medical personnel to their clients and those who are caring for their clients. Language capacity is also important for delivery of services because, as expressed by a study key informant, when professional interpretation is lacking, patients rely on family members to translate, and those family members may “introduce their own cultural sentiments.” This creates uncertainty for health care providers about whether everything they are saying is being shared with the patient. Such uncertainty is of particular concern in an emergency situation, when the impact of communication barriers or breakdowns could be especially dire (Pressman & Schneider, 2009). Study participants indicated that Spanish and Mandarin are common languages where interpretation and translation are needed for health care/emergency services in the Palo Alto/Stanford area. Additional Community Concerns In response to a question about what changes they predict are coming in the future, the key informants expressed concern about possible changes in health insurance, namely that changes to the Affordable Care Act could result in many Palo Alto/Stanford area residents losing health insurance. Key informants and residents alike also reported that they anticipate technology playing a larger role in health care. Examples include innovations in wearable technology, laboratory testing, and telemedicine. Community members cautioned, however, that telemedicine (such as video visits or communicating with doctors using messaging) is not a mode that is accessible for all older adults, and that technology cannot always take the place of face-to-face visits with health care professionals. Residents’ and experts’ feedback on how to improve the health of the community may be found in the Recommendations chapter of this report. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 31 February 1, 2018 PAFD Calls and Medical Emergency Patients: City of Palo Alto, 2015 and 2016 The data provided by PAFD contained all unique 2015 and 2016 calls from PAFD’s internal CAD and ESO Solutions medical emergency services data systems. The CAD dataset includes all of the various calls to PAFD for dispatch (e.g., fire, medical emergency, false alarm), which were restricted for the purposes of these analyses to the City of Palo Alto (and, in a later section of this report, Stanford). Note that CAD call types are the initial presumption of incident type, based on what dispatchers are told, not the incident type verified by PAFD personnel once they are at the scene. The ESO dataset contains only medical emergency patients for whom PAFD was dispatched, but includes incident ZIP Code, patient age, gender, and primary impression. A “primary impression” is the main problem, condition, or symptom that brought about the encounter between the patient and EMS personnel. This section briefly describes the current data, with a special focus on medical emergency calls/patients. PAFD CAD Calls Originating in the City of Palo Alto Current Call Volume and Type Of all calls for service to PAFD during 2015 and 2016, an average of approximately 85% per year originated in the City of Palo Alto (84.2% in 2015 and 85.7% in 2016). Based on the CAD data provided by PAFD for calls originating only in the City of Palo Alto, in 2015 and 2016 there were a total of 14,037 calls for service to PAFD (6,887 in 2015 and 7,150 in 2016).15 See Figure 10 for a chart of the percentage of calls by call type originating in the City of Palo Alto. Appendix 6 shows the number of unique calls of each type by year. 15 Note that these figures do not tie to PAFD’s total annual costs because they represent only calls from individuals in the City of Palo Alto, not in all cities. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 32 February 1, 2018 Figure 10, PAFD Calls for Service Originating in the City of Palo Alto, 2015/2016 Average Source: Amber Cameron, City of Palo Alto Public Safety Department, personal communication containing PAFD CAD unduplicated data (2015/2016), January 24, 2017 (“City of Palo Alto PSD, 2017: CAD unduplicated data 2015/2016”). Note: “Service” calls are, for example, assisting with lock-outs, water leaks, police assists. * “All Other” is comprised of all call types that by themselves represent less than 1.5% of calls from the City of Palo Alto to PAFD.16 16 The categories of call included in “All Other” are: Airplane Emergency, Auto Aid, Full First Alarm, Gas, Hazardous Materials, Mutual Aid, Second Alarm, 1056 (suicide), Suspicious Circumstances, Technical Rescue, Train, Utilities, Vegetation Fire, Welfare Check, and Wires. Taken all together, these represent less than 3% of calls from the City of Palo Alto to PAFD. Fire Smoke All Other * Accident Service False Alarm Medical Emergency 0% 10% 20% 30% 40% 50% 60% 70% 80% PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 33 February 1, 2018 Current Medical Emergency Calls by Time of Day Approximately 70% of all calls for service in Palo Alto each year were classified as medical emergency (70.7% in 2015 and 69.7% in 2016). More calls came in during the day than at night. See Figure 11. Figure 11, PAFD Calls by Type and Time of Day, 2015/2016 Average, City of Palo Alto Source: City of Palo Alto PSD, 2017: CAD unduplicated data 2015/2016. PAFD Medical Emergency Patient Volume Originating in the City of Palo Alto Based on the ESO data provided by PAFD, there were a total of 8,837 medical emergency patients during 2015 and 2016 (4,446 in 2015 and 4,391 in 2016). Note that an individual may be a patient more than once in a given year, and each time a patient presents, s/he requires PAFD’s attention. Actionable Insights was asked to analyze the ESO data by age, gender, ZIP Code, and primary impression, and to provide PAFD with predictions about future medical emergency patients in the City of Palo Alto (and, in a later section of this report, in the Stanford area) based on these data. Current data are found in this section, while predictions may be found in the next section. 3,342 1,584 1,434 659 0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 4,500 5,000 Palo Alto Daytime Palo Alto Nighttime All Others Medical Emergency PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 34 February 1, 2018 Current Medical Emergency Patient Volume by Age Almost all (98.1%) cases had data on age. Based on a preliminary analysis of the ESO data,17 we found the relationship of the number of patients to patient age to be positive (1.033) and statistically significant (p < 0.001). That is, there are more older individuals represented in the ESO dataset than younger individuals, to a statistically significant degree. This indicates that older individuals are more likely than younger individuals to have a medical emergency in the combined Palo Alto/Stanford jurisdiction. We converted age into categories used by the U.S. Census Bureau’s American Community Survey (ACS) and the CA DOF, the better to compare statistics on current age of patients to predicted age of future patients. Note that certain ACS/DOF age groups contain fewer years (e.g., 60-64) than others (e.g., 65-74). For ease of viewing in the various charts below, the three lowest age categories have been merged; the 15-19 and 20-24 age categories have been merged, and the 55-59 and 60-64 age categories have been merged. (Data by original age categories are available upon request.) As shown in Figure 12, there tend to be more medical emergencies as individuals age. Figure 12, Current Medical Emergency Patient Volume by Age Category, 2015/2016 Average Counts and Percentages, City of Palo Alto Source: Amber Cameron, City of Palo Alto Public Safety Department, personal communication containing PAFD ESO unduplicated data (2015/2016), July 24, 2017 (“City of Palo Alto PSD, 2017: ESO unduplicated data 2015/2016”). Notes: City was determined by ZIP Codes. Data for the City of Palo Alto are shown only for the 98% of cases where age of patient was available. Data for Stanford are shown in the chart in the Stanford section; data for other cities (2% of the cases used to generate the average counts and percentages) are not shown. 17 We used a simple linear regression in which we predicted number of medical emergency patients by patient age in combined years 2015 and 2016. 0% 5% 10% 15% 20% 25% 0 200 400 600 800 1,000 1,200 Age 0-14 Age 15-24 Age 25-34 Age 35-44 Age 45-54 Age 55-64 Age 65-74 Age 75-84 Age 85+ Av e r a g e N u m b e r o f P a t i e n t s ( P a t i e n t Vo l u m e ) Palo Alto # Palo Alto % PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 35 February 1, 2018 Comparatively, a much greater proportion of older adults are being served by PAFD for medical emergencies compared to the proportion of older adults in the resident population. In Figure 13, we compare the average current medical emergency patient volume in the City of Palo Alto to its residential population, by the proportion of individuals of each age group represented among patients and residents. Figure 13, Current Medical Emergency Patient Volume by Age Category, 2015/2016 Average Percentages, Compared to Residential Population by Age Category, City of Palo Alto Sources: City of Palo Alto PSD, 2017: ESO unduplicated data 2015/2016; U.S. Census Bureau, 2014. Notes: For purposes of medical emergency data, city was determined by ZIP Codes. Data for Stanford are shown in the chart in the Stanford section; data for other cities (216 of the 10,943 cases used to generate the percentages) are not shown. One can see that in Palo Alto, adults aged 55 and older are medical emergency patients at higher rates than they are present in the residential population, with medical emergency patients who are aged 85 and older representing more than seven times their proportion in the residential population. Current Medical Emergency Patient Volume by Gender Of the patients originating in the City of Palo Alto for whom gender was reported (97.1%), slightly more than half (52.3%) were female, on average. Current Medical Emergency Patient Volume by Incident Location (ZIP Code) Virtually all cases included data on incident ZIP Code. The vast majority of medical emergency patients (98%) identified in the ESO dataset were assisted in the cities of Palo Alto or Stanford (see Figure 14; map does not show the approximately 8% of cases associated with ZIP Codes 4% 7% 9%8% 10% 14% 12%13% 22% 19% 10% 13%14%15% 12% 9% 5% 3% 0% 5% 10% 15% 20% 25% 30% Age 0-14 Age 15-24 Age 25-34 Age 35-44 Age 45-54 Age 55-64 Age 65-74 Age 75-84 Age 85+ Palo Alto Medical Emergency Patient Volume Palo Alto Resident Population PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 36 February 1, 2018 representing P.O. boxes). Most patients (an average of 43%) were found downtown, in the 94301 ZIP Code. A further breakdown of patient volume by ZIP Code, by year, may be found in Appendix 7. Figure 14, Map of Current Medical Emergency Patient Volume by Incident ZIP Code, 2015/2016 Average Sources: Map from DataSF.org, 2017; data from City of Palo Alto PSD, 2017: ESO unduplicated data 2015/2016. Note: Map does not show P.O. box ZIP Codes 94302 (6.4%) and 94309 (1.5%). 2,360 (43%) 546 (10%) 696 (13%) 862 (16%) 467 (9%) PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 37 February 1, 2018 Current Medical Emergency Patient Volume by Type of Illness/Issue Medical emergency patients called 911 with a wide variety of symptoms. Table 4 offers categories (higher-level groupings) of primary impressions by year. We remind the reader that some patients are represented more than once in the ESO data, since they may have called and been seen by PAFD more than once. The primary impressions associated with the greatest proportions of medical emergency patients in Palo Alto appear to be altered level of consciousness (ALOC)18, followed by injury and cardiac-related issues. Table 4, All Medical Emergency (ME) Patients, Categories of Primary Impressions, 2015/2016 Average, in Order of Frequency for the City of Palo Alto Primary Impression Category Annual Avg. % (#) of ME Patients in Palo Alto, 2015/2016 Altered Level of Consciousness 29.2% (1,290) Injury (Traumatic or Otherwise) or Hemorrhage 23.2% (1,026) Cardiac-Related 10.6% (469) Non-Specific Pain 8.8% (389) Abdominal Discomfort 7.4% (328) Respiratory-Related 6.6% (291) Behavioral Health 5.0% (220) All Other* 4.5% (198) No Complaints or Injury/Illness Noted 2.6% (116) Data Missing 2.2% (96) TOTAL 4,423 Source: City of Palo Alto PSD, 2017: ESO unduplicated data 2015/2016. Note: Percentages may not add to exactly 100% due to rounding. *“All other” comprises other impression types that represented < 1% of calls both years. 18 Note that this last category is based on a patient’s actual level of wakefulness and/or disorientation (Tindall, 1990), and does not refer to drug-induced hallucinatory experience. However, based on a scan of the data, some small number (<5%) of cases identified as ALOC may be related to alcohol or drug use. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 38 February 1, 2018 Appendix 9 provides patient volume by primary impression, separately for day versus night. Based on Appendix 9, medical emergency patients are seen more often during the day than at night in the City of Palo Alto. More specifically, ALOC, injury/hemorrhage, and (to a lesser extent) cardiac-related issues occur significantly more often (p < 0.05, based on a one-way ANOVA analysis) during the daytime, while behavioral health issues, abdominal discomfort, and non-specific pain occur significantly more often at night, statistically speaking (p < 0.05). For reference, Appendix 3 lists the total number and percentage of patient cases by all primary impressions by year, and includes the category into which each primary impression was placed. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 39 February 1, 2018 Predictive Analysis Findings: City of Palo Alto, Year 2030 Based on our analysis of current call volume and calls by type, we have predicted year 2030 call volume and calls by type for the City of Palo Alto (and for Stanford, separately, in the next section).19 As described briefly in the Predictive Analysis Methods Summary section at the beginning of this report, we used a modified life table approach, which is generally used to describe a current population and predict future growth or decline. We assume that the 2030 cohort will be subject, proportionally, to the age-specific medical emergency call/patient volumes, other call type volumes, incident locations, and primary impressions that prevailed for the 2015/2016 cohort. We adjusted medical emergency call/patient volume predictions by age based on CA DOF (2017) projections. We predicted call/patient volume and calls by type for day and night separately, using population figures based on CA DOF figures, City of Palo Alto scenarios, and City of Palo Alto Transportation Survey results (2013). A full description of this process may be found in Appendix 4. We provide predictions for all call types that represented at least 1% of calls in the 2015 or 2016 CAD datasets. The remainder are grouped into a call type category labeled “Other.” We used ESO data (described in Predictive Analysis Methods Summary section of this report) to adjust predictions for the medical emergency call type based on medical emergency patient volume and expected age-related shifts in the demographics for the City of Palo Alto (and for Stanford, separately, in the next section) between now and 2030.20 This process is also further described in Appendix 5. The overall 2030 predictions for the City of Palo Alto for all call types and for patient volumes take into account projected daytime and nighttime visitors, as well as larger daytime populations based on net commuters, again based on CA DOF figures, City of Palo Alto scenarios, and City of Palo Alto Transportation Survey results.21 Further information on this process may be found in Appendix 5. Projected Population Overall and by Time of Day As shown earlier in Table 3, the City of Palo Alto has a current (2014) residential population of 65,685 people, which is projected to rise to between 72,285 and 79,765 by the year 2030 (see scenarios from Hilary Gitelman, Planning & Community Environment Department, City of Palo Alto, personal communication, December 21, 2016). Nighttime population is the same as what 19 Due to the lack of ZIP Codes in the CAD data, we were not able to estimate call volume by ZIP Code, PAFD district, or other sub-city-level areas (e.g., neighborhood). 20 Note that while there were no data projecting gender for sub-county-level areas, the projections of gender by county (CA DOF, 2017) suggest that the projected change in gender proportions between now and 2030 is not statistically significant in either San Mateo County or Santa Clara County (Chi-square = 2.00, p > 0.05). Therefore, we do not include gender as a factor in our predictions. 21 However, figures and tables in this report that show data by age group generally do not include visitors (see text associated with various figures/tables for explanation). PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 40 February 1, 2018 the U.S. Census would consider the residential population. Daytime population is a net total of nighttime residents who stay, residents who “out-commute” to somewhere besides the City of Palo Alto, and individuals from other cities who “in-commute” to the City of Palo Alto. Based on Table 3, the City of Palo Alto’s year 2014 daytime population (not including visitors) was estimated to be 129,616 people, and projected to rise to between 142,841 and 150,587 by the year 2030. Figures 15 and 16 show the recent (2014) and projected 2030 population for the City of Palo Alto, comparing daytime (Figure 15) and nighttime (Figure 16) figures by age. The charts show the lowest and highest 2030 population projections based on the six scenarios provided by the City of Palo Alto (see Table 3); note that the lowest and highest scenarios are different for the daytime population (Scenarios 5 and 4) than for the nighttime population (Scenarios 2 and 6). Visitors were not included in Figures 15 and 16 because the data available for estimating visitors’ ages in 2030 grouped age into four broad categories rather than the 13 categories available for estimating residents’ and workers’ ages; however, visitors are very small proportions of the total populations. Figure 15, Current and Year 2030 Projected Populations by Age Category, Daytime, City of Palo Alto Sources: Hilary Gitelman, City of Palo Alto Planning & Community Environment department, personal communication, December 21, 2016; U.S. Census Bureau, 2014. Note: Excludes visitors to increase age category granularity (see text for further explanation). In both Scenario 4 (the highest 2030 daytime population projection) and Scenario 5 (the lowest 2030 daytime population projection) the projection is higher than its 2014 comparison for all age categories except age 0-14 and 65-74. - 5,000 10,000 15,000 20,000 25,000 30,000 Age 0-14 Age 15-24 Age 25-34 Age 35-44 Age 45-54 Age 55-64 Age 65-74 Age 75-84 Age 85+ 2014 Daytime (Residential & Commuter) Pop. 2030 Projected Daytime Pop. - Sc. 5 2030 Projected Daytime Pop. - Sc. 4 PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 41 February 1, 2018 Figure 16, Current and Year 2030 Projected Populations by Age Category, Nighttime, City of Palo Alto Sources: Hilary Gitelman, City of Palo Alto Planning & Community Environment department, personal communication, December 21, 2016; U.S. Census Bureau, 2014. Note: Excludes visitors to increase age category granularity (see text for further explanation). Scenario 2 (the lowest year 2030 nighttime population projection) is higher than its year 2014 comparison for all age categories except ages 0-14, 25-34, and 35-44. Scenario 6 (the highest 2030 nighttime population projection) is higher than its 2014 comparison for all age categories except ages 0-14 and 25-34. Overall, based on these scenarios, it appears that the absolute numbers of residents in younger age categories (particularly ages 0-14, 25-34, and 35-44, but not age 15-2422) are generally projected to be similar or to decrease in 2030 compared to 2014, while the numbers of residents in older age categories (particularly ages 55 and older23) are generally projected to increase in 2030 compared to 2014. 22 It may be that the 15-24 age group is not predicted to decrease in part because some of Stanford University’s students live in Palo Alto and the university’s student population is expected to increase over time (Stanford University, 2016). 23 Since the workforce is overwhelmingly made up of individuals between ages 20-64 (Bureau of Labor Statistics, 2016), even though labor force participation among those aged 65 and older has been increasing due to longer life spans and financial obligations (Kromer & Howard, 2013), we remained conservative in our predictions and used age groups 20-64 to represent the working population in the City of Palo Alto. It may well be that there are more individuals aged 65-74 commuting into the City of Palo Alto than our predictions indicate. - 5,000 10,000 15,000 20,000 25,000 30,000 Age 0-14 Age 15-24 Age 25-34 Age 35-44 Age 45-54 Age 55-64 Age 65-74 Age 75-84 Age 85+ 2014 Nighttime (Residential) Pop. 2030 Projected Nighttime (Residential) Pop. - Sc. 2 2030 Projected Nighttime (Residential) Pop. - Sc. 6 PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 42 February 1, 2018 Predicted Medical Emergency Call Volume and Other Call Types by Time of Day For the purposes of our predictions, daytime is considered to be 8:00 a.m. to 7:59 p.m., and nighttime is considered to be 8:00 p.m. to 7:59 a.m. (Amber Cameron, City of Palo Alto Public Safety Department, personal communication, March 18, 2017). The time indicator used was the time of dispatch. Table 9 in Appendix 6 shows the average annual 2015/2016 call volume and the age-adjusted 2030 predicted call volume, by call type and time of day, for the City of Palo Alto. Table 9 also takes into account the six different City of Palo Alto scenarios. See Table 3 earlier in this report for the six scenarios. It is clear from the tables generated by the data that the total number of calls will increase between now and 2030 for all call types, and the volume of patients will increase between now and 2030 for medical emergencies. Figure 17 below shows the predicted change from 2015/2016 to 2030 for medical emergency calls and other major call types, for the City of Palo Alto. The grey portion of the bars shows the current (2015/2016 average) number of calls by type, the light green portion of the bars illustrates the lowest predicted increase, and the dark green portion of the bars indicates the highest predicted increase based on the six City of Palo Alto scenarios. Proportionally, it is predicted that CAD medical emergency calls will increase between 25% and 34% (depending on the growth scenario), while other types of calls will increase between 10% and 16% each, depending on the scenario.24 Figure 17, Predicted Change in Average Annual Call Volume from 2015/2016 to 2030, Major Call Types, City of Palo Alto Source: City of Palo Alto PSD, 2017: CAD unduplicated data 2015/2016. Notes: Medical emergency calls are age- adjusted. “All other” comprises calls that make up less than 1.5% each in the unduplicated 2015/2016 CAD dataset (Gas, HazMat, Mutual Aid, etc.). 24 Since ESO medical emergency patient volume predicted increases are within one percentage point of CAD medical emergency call volume predicted increases, we do not display them separately here. They may be found in Appendix 7. 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 Fire Smoke All Other* Accident Service False Alarm Medical Emergency Base (2015/2016 Average) Minimum Predicted Increase Maximum Predicted Increase PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 43 February 1, 2018 Figure 18 below shows the predicted change from 2015/2016 to 2030 for Medical Emergency calls and other major call types, for the City of Palo Alto during the daytime hours. The grey portion of the bars shows the current (2015/2016 average) number of calls by type, the light orange portion of the bars illustrates the lowest predicted increase, and the dark orange portion of the bars indicates the highest predicted increase based on the six City of Palo Alto scenarios. It is predicted that medical emergency calls will increase between 24% and 32% during the daytime (depending on the growth scenario), while other types of calls will increase between 9% and 16% each during the daytime, depending on the scenario. Figure 18, Predicted Change in Average Annual Call Volume from 2015/2016 to 2030, Major Call Types, Daytime, City of Palo Alto Source: City of Palo Alto PSD, 2017: CAD unduplicated data 2015/2016.Notes: Medical emergency calls are age- adjusted. “All other” comprises calls that make up less than 1.5% each in the unduplicated 2015/2016 CAD dataset (Gas, HazMat, Mutual Aid, etc.). 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 Fire Smoke All Other* Accident Service False Alarm Medical Emergency Base (2015/2016 Average) Minimum Predicted Increase Maximum Predicted Increase PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 44 February 1, 2018 Figure 19 below shows the predicted change from 2015/2016 to 2030 for medical emergency calls and other major call types, for the City of Palo Alto during the nighttime hours. The grey portion of the bars shows the current (2015/2016 average) number of calls by type, the light blue portion of the bars illustrates the lowest predicted increase, and the darker blue portion of the bars indicates the highest predicted increase based on the six City of Palo Alto scenarios. For the nighttime, it is predicted that medical emergency calls will increase between 26% and 39% (depending on the growth scenario), while other types of calls will increase between 8% and 22% each at night, depending on the scenario. Figure 19, Predicted Change in Average Annual Call Volume from 2015/2016 to 2030, Major Call Types, Nighttime, City of Palo Alto Source: City of Palo Alto PSD, 2017: CAD unduplicated data 2015/2016.Notes: Medical emergency calls are age- adjusted. “All other” comprises calls that make up less than 1.5% each in the unduplicated 2015/2016 CAD dataset (Gas, HazMat, Mutual Aid, etc.). Although the absolute numbers of increased calls are smaller for nighttime than for daytime, the maximum proportional increases of medical emergency and other calls at night (39% and 22%, respectively) will be somewhat greater than the maximum proportional increases of medical emergency and other calls during the day (32% and 16%, respectively). Predicted Medical Emergency Patient Volume Originating in the City of Palo Alto Predicted Medical Emergency Patient Volume Overall and by Age Again, we converted age into categories used by the ACS and DOF, the better to compare statistics on current age of patients to predicted age of patients. For the sake of simplicity, in Figure 20 we merged certain age categories and show current and projected data on the City of Palo Alto and the two scenarios that generated the lowest (Scenario 2) and highest (Scenario 6) predictions. Visitors were not included in Figure 20 because the data available for estimating visitors’ ages in 2030 grouped age into four broad categories rather than the 13 categories 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 Fire Smoke All Other* Accident Service False Alarm Medical Emergency Base (2015/2016 Average) Minimum Predicted Increase Maximum Predicted Increase PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 45 February 1, 2018 available for estimating residents’ and workers’ ages; however, visitors are very small proportions of the total populations.25 Figure 20, Current and Year 2030 Predicted Age-Adjusted Medical Emergency Patient Volume by Age Category, City of Palo Alto Source: City of Palo Alto PSD, 2017: ESO unduplicated data 2015/2016. Note: Excludes visitors to increase age category granularity (see text for further explanation). The 2030 City of Palo Alto predictions show overall increases of 24% to 33% in medical emergency patient volume (depending on the scenario) compared to current medical emergency patient volume (see Appendix 6). Over two thirds (68%) of the City of Palo Alto’s predicted 2030 medical emergency patients will be aged 55+. The City of Palo Alto will experience a statistically significant change in age between current and 2030 medical emergency patients.26 The data suggest that the City of Palo Alto’s medical emergency patients in 2030 will skew significantly towards older adults, from a mean patient age of 55-64 to a slightly older demographic, with the modal, i.e., most common, patient age category in 2030 being those aged 85+ in all six scenarios. In other words, the City of Palo Alto will have more older medical emergency patients in 2030, both in terms of absolute numbers (Figure 20) and a skew towards older adult patients. This is the major issue for which the City of Palo Alto and PAFD will need to prepare. Predicted Medical Emergency Patient Volume by Time of Day Based on ESO data, there will be very little change in the City of Palo Alto’s proportion of medical emergency patients by time of day. In 2015/2016, the average proportion of daytime medical emergency patients was 68.6%, while in 2030 the proportion of daytime medical emergency patient volume originating in the City of Palo Alto is predicted to be between 67.3% (Scenario 6) – slightly lower than the current proportion – and 68.9% (Scenario 1) – slightly 25 Including daytime and overnight visitors would only increase the predicted numbers of patients by an average of 39 in total, across all age groups – a very small proportion of the total number of predicted interactions overall. 26 An increase of 0.30 in Scenario 2 and 0.34 in Scenario 6, approximately one third of an age category using the merged age categories in the chart above, p < 0.001 in both scenarios, F = 37.288 for Scenario 2 and 47.671 for Scenario 6. - 200 400 600 800 1,000 1,200 1,400 Age 0-14 Age 15-24 Age 25-34 Age 35-44 Age 45-54 Age 55-64 Age 65-74 Age 75-84 Age 85+ 2015/2016 Avg. # PA ME Patients 2030 Predicted # PA ME Patients-Sc. 2 2030 Predicted # PA ME Patients-Sc. 6 PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 46 February 1, 2018 higher than the current proportion. In none of the scenarios is the proportional change from 2015/2016 to 2030 statistically significant. This means that PAFD will continue to see a much higher call volume from the City of Palo Alto during the daytime than at night. Predicted Medical Emergency Patient Volume by Type of Illness/Issue In 2015/2016, the primary impressions that were associated with the greatest proportions of medical emergency patients in the City of Palo Alto were altered level of consciousness,27 followed by injury and cardiac-related issues (see Table 4, earlier in this report). As shown in Table 5, the order of the top primary impressions remains the same in the predictions for year 2030. Table 5, Predicted Year 2030 Medical Emergency Patients, Categories of Primary Impressions, by City, in Order of Frequency, City of Palo Alto Primary Impression Category % (#) Scen. 2 (Lowest) PA, 2030 % (#) Scen. 6 (Highest) PA, 2030 Altered Level of Consciousness 30.4% (1,625) 30.5% (1,748) Injury/Hemorrhage 23.2% (1,239) 23.1% (1,328) Cardiac-Related 11.4% (607) 11.4% (654) Non-Specific Pain 8.9% (478) 8.9% (513) Abdominal Discomfort 7.6% (404) 7.5% (433) Respiratory-Related 6.9% (371) 7.0% (402) Behavioral Health 4.6% (246) 4.5% (259) All Other* 4.5% (238) 4.4% (255) No Complaints 2.6% (137) 2.6% (147) TOTAL 5,345 5,739 Source: Actionable Insights, LLC, 2017, unpublished data, based on City of Palo Alto PSD, 2017: ESO unduplicated data 2015/2016. Note: Percentages may not add to exactly 100% due to rounding. *“All other” comprises other primary impression types that represented < 1% of calls in both years. 27 Altered level of consciousness is based on “a patient’s actual level of wakefulness and/or disorientation” (Tindall, 1990) and has a variety of causes, including dehydration, hypoglycemia, and shock, but is not intended to refer to symptoms that are directly related to drug or alcohol use. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 47 February 1, 2018 However, it is also important for PAFD’s future planning to review the primary impressions that will increase the most between 2015/2016 and the year 2030. In absolute numbers, predictions suggest that the greatest increase in the City of Palo Alto will be among ALOC patients (ranging from an increase of 335, or 26%, in Scenario 2 to an increase of 458, or 36%, in Scenario 6). However, proportionally, the greatest increases in the City of Palo Alto will be among cardiac patients28 and respiratory patients.29 See Figures 21 and 22 for a representation of City of Palo Alto increases in absolute numbers by primary impression, for daytime and nighttime separately.30 See Appendix 9 for total predicted changes in primary impression categories, by city and time of day. Figure 21, Comparison of Actual Average Annual Patient Volume versus Predicted Change in Average Annual Patient Volume from 2015/2016 to 2030, by Primary Impression, City of Palo Alto, Daytime Source: Actionable Insights, LLC, 2017, unpublished data, and City of Palo Alto PSD, 2017: ESO unduplicated data 2015/2016. 28 Ranging from an increase of 30%, or 139 more patients, in Scenario 2 to an increase of 40%, or 186 more patients, in Scenario 6. 29 Ranging from an increase of 28%, or 81 more patients, in Scenario 2 to an increase of 39%, or 112 more patients, in Scenario 6. 30 Note that Scenario 4, not Scenario 6, has the highest increases for daytime in the City of Palo Alto. - 250 500 750 1,000 1,250 Abdominal Altered Level of Consciousness Behavioral Health Cardiac Injury/ Hemorrhage Non-Specific Pain Respiratory 2015/2016 Average 2030 Lowest # Increase (Sc. 2)2030 Highest # Increase (Sc. 4) PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 48 February 1, 2018 Figure 22, Comparison of Actual Average Annual Patient Volume versus Predicted Change in Average Annual Patient Volume from 2015/2016 to 2030, by Primary Impression, City of Palo Alto, Nighttime Source: Actionable Insights, LLC, 2017, unpublished data, and City of Palo Alto PSD, 2017: ESO unduplicated data 2015/2016. We conducted a binary logistic regression analysis to determine which variables, if any, were significant in determining the increase in patient volume between 2015/2016 and 2030 (note that we used Scenario 6 for the City of Palo Alto’s 2030 figures). Regression analysis is a statistical method to explain the relationship between a dependent variable (in this case, medical emergency patient volume in the year 2015/2016 compared to the year 2030) and various independent or “explanatory” variables (for this analysis, patient age categories, daytime versus nighttime dispatch, and what kind of primary impression the patient presents). Binary logistic regression is a special form of regression analysis used when the dependent variable has only two values (in this case, total patient volume for the year 2015/2016 and predicted total patient volume for the year 2030). We tried a variety of regression equations, changing how we represented primary impression as a variable in the models. No matter how we constructed the primary impression variable(s), only patient age was significant in the binary logistic regression analyses for the City of Palo Alto; primary impressions were not statistically significant indicators in any model. As described earlier in this section on 2030 predictions for PAFD, the main issue for which the City of Palo Alto and PAFD will need to prepare is the expected increase in older medical emergency patients in 2030. - 250 500 750 1,000 1,250 Abdominal Altered Level of Consciousness Behavioral Health Cardiac Injury/ Hemorrhage Non-Specific Pain Respiratory 2015/2016 Average 2030 Lowest # Increase (Sc. 2)2030 Highest # Increase (Sc. 6) PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 49 February 1, 2018 PAFD Calls and Medical Emergency Patients: Stanford, 2015 and 2016 As described in the City of Palo Alto section, the PAFD data contained all unique 2015 and 2016 calls from PAFD’s internal CAD and ESO medical emergency services data systems. The CAD dataset includes all types of calls to PAFD for dispatch (e.g., fire, medical emergency, false alarm), which were restricted for the purposes of these analyses to Stanford (and, in the previous section, the City of Palo Alto). Note that CAD call types are the initial presumption of incident type, based on what dispatchers are told, not the incident type verified by PAFD personnel once they are at the scene. The ESO dataset contains only medical emergency patients for whom PAFD was dispatched, but includes incident ZIP Code, patient age, gender, and primary impression. A “primary impression” is the main problem, condition, or symptom that brought about the encounter between the patient and EMS personnel. This section briefly describes the current data for Stanford, with a special focus on medical emergency calls/patients. PAFD CAD Calls Originating in the Stanford Area Current Call Volume and Type Of all calls for service to PAFD during 2015 and 2016, an average of approximately 15% per year originated in Stanford (15.8% in 2015 and 14.3% in 2016). Based on the CAD data provided by PAFD for calls originating only from Stanford, in 2015 and 2016 there were a total of 2,482 calls for service to PAFD (1,289 in 2015 and 1,193 in 2016). See Figure 23 on the next page for a chart of the percentage of calls by call type originating from Stanford. Appendix 6 shows how many unique calls of each type by city, on average. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 50 February 1, 2018 Figure 23, PAFD Calls for Service Originating from Stanford, 2015/2016 Average Source: City of Palo Alto PSD, 2017: CAD unduplicated data 2015/2016. Note: “Service” calls are, for example, assisting with lock-outs, water leaks, police assists. * “All Other” is comprised of all call types that by themselves represent less than 1.5% of calls from Stanford to PAFD.31 Current Medical Emergency Calls by Time of Day Approximately 50% of all calls for service in Stanford each year were classified as medical emergency (50.2% in 2015 and 50.1% in 2016). More calls came in during the day than at night. 31 The categories of call included in “All Other” for Stanford are: Auto Aid, Gas, Hazardous Materials, Mutual Aid, Second Alarm, Vegetation Fire, and Wires. Taken all together, these represent less than 3% of calls from Stanford to PAFD. Smoke remains a separate category so this chart can be comparable to Figure 10 (City of Palo Alto calls by type). Smoke Fire Accident All Other * Service False Alarm Medical Emergency 0% 10% 20% 30% 40% 50% 60% PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 51 February 1, 2018 See Figure 24. It is important to note that the total difference in volume between daytime and nighttime calls on Stanford is much smaller than that of the City of Palo Alto. Figure 24, PAFD Calls by Type and Time of Day, 2015/2016 Average, Stanford Source: City of Palo Alto PSD, 2017: CAD unduplicated data 2015/2016. PAFD Medical Emergency Patient Volume Originating in the Stanford Area Based on the ESO data provided by PAFD for the Stanford area, there were a total of 1,890 medical emergency patients during 2015 and 2016 (916 in 2015 and 974 in 2016). Note that an individual may be a patient more than once in a given year, and each time a patient presents, s/he requires PAFD’s attention. Actionable Insights was asked to analyze the ESO data by age, gender, ZIP Code, and primary impression, and to provide PAFD with predictions about future medical emergency patients in the Stanford area based on these data. Current data are found in this section, while predictions may be found in the next section. 370 253 399 220 0 200 400 600 800 1,000 Stanford Daytime Stanford Nighttime All Others Medical Emergency PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 52 February 1, 2018 Current Medical Emergency Patient Volume by Age As described in the section on the City of Palo Alto, based on a preliminary analysis of the ESO data,32 we find that the relationship of the number of patients to patient age is positive (1.033) and statistically significant (p < 0.001). That is, there are more older individuals represented in the ESO dataset than younger individuals, to a statistically significant degree. This indicates that older individuals are more likely than younger individuals to have a medical emergency in the combined Palo Alto/Stanford jurisdiction. We converted age into categories used by the U.S. Census Bureau’s American Community Survey (ACS) and the CA DOF, the better to compare statistics on current age of patients to predicted age of future patients. Note that certain ACS/DOF age groups contain fewer years (e.g., 60-64) than others (e.g., 65-74). For ease of viewing in the various charts below, the three lowest age categories have been merged; the 15-19 and 20-24 age categories have been merged, and the 55-59 and 60-64 age categories have been merged. (Data by original age categories are available upon request.) As shown in Figure 25, aside from the spike for young adults ages 15-24 (which is appropriate given the demographics of the Stanford area), there tend to be more medical emergencies as individuals age. Figure 25, Current Medical Emergency Patient Volume by Age Category, 2015/2016 Average Counts and Percentages, Stanford Source: City of Palo Alto PSD, 2017: ESO unduplicated data 2015/2016. Notes: City was determined by ZIP Codes. Data for Stanford are shown only for the 99% of cases where age of patient was available. Data for the City of Palo Alto are shown in the chart in the Palo Alto section; data for other cities (2% of the cases used to generate the average counts and percentages) are not shown. 32 We used a simple linear regression in which we predicted number of medical emergency patients by patient age in combined years 2015 and 2016. 0% 5% 10% 15% 20% 25% 30% 0 50 100 150 200 250 300 350 400 450 500 Age 0-14 Age 15-24 Age 25-34 Age 35-44 Age 45-54 Age 55-64 Age 65-74 Age 75-84 Age 85+ Av e r a g e N u m b e r o f P a t i e n t s ( P a t i e n t Vo l u m e ) Stanford # Stanford % PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 53 February 1, 2018 Comparatively, a much greater proportion of older adults are being served by PAFD for medical emergencies compared to the proportion of older adults in the resident population. In Figure 26, we compare the average current medical emergency patient volume for Stanford to its residential population, by the proportion of individuals of each age group represented among patients and residents. Figure 26, Current Medical Emergency Patient Volume by Age Category, 2015/2016 Average Percentages, Compared to Residential Population by Age Category, Stanford Sources: City of Palo Alto PSD, 2017: ESO unduplicated data 2015/2016; U.S. Census Bureau, 2014. Notes: For purposes of medical emergency data, city was determined by ZIP Codes. Data for the City of Palo Alto are shown in the chart in the Palo Alto section; data for other cities (216 of the 10,943 cases used to generate the percentages) are not shown. At Stanford, adults aged 35 and older are medical emergency patients at higher rates than they are present in the residential population, with medical emergency patients who are aged 85 and older being represented at 15 times their proportion in the residential population. Current Medical Emergency Patient Volume by Gender Of the patients originating in the Stanford area for whom gender was reported (98.4%), slightly more than half (52.1%) were female, on average. Current Medical Emergency Patient Volume by Incident Location (ZIP Code) Virtually all cases included data on incident ZIP Code. The vast majority of medical emergency patients (98%) identified in the ESO dataset were assisted in the cities of Palo Alto or Stanford 4% 28% 13% 7%8%10%7%8% 15% 6% 64% 15% 4%3%4%3%1%1% 0% 15% 30% 45% 60% 75% Age 0-14 Age 15-24 Age 25-34 Age 35-44 Age 45-54 Age 55-64 Age 65-74 Age 75-84 Age 85+ Stanford Medical Emergency Patient Volume Stanford Resident Population PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 54 February 1, 2018 (see map in the City of Palo section, Figure 14; map does not show the approximately 8% of cases associated with ZIP Codes representing P.O. boxes). An average of 16% of all patients were found in the 94305 (Stanford) ZIP Code. A further breakdown of patient volume by ZIP Code, by year, may be found in Appendix 7. Current Medical Emergency Patient Volume by Type of Illness/Issue Medical emergency patients called 911 with a wide variety of symptoms. Table 6 offers categories (higher-level groupings) of primary impressions by year. We remind the reader that some patients are represented more than once in the ESO data, since they may have called and been seen by PAFD more than once. Table 6, All Medical Emergency (ME) Patients, Categories of Primary Impressions, 2015/2016 Average, in Order of Frequency for Stanford Primary Impression Category Annual Avg. % (#) of ME Patients in Stanford, 2015/2016 Altered Level of Consciousness 28.9% (273) Injury (Traumatic or Otherwise) or Hemorrhage 28.0% (265) Behavioral Health 9.4% (89) Non-Specific Pain 7.3% (69) Abdominal Discomfort 7.0% (66) Cardiac-Related 6.1% (58) Respiratory-Related 4.8% (45) All Other* 4.0% (38) No Complaints or Injury/Illness Noted 3.3% (31) Data Missing 1.2% (11) TOTAL 945 Source: City of Palo Alto PSD, 2017: ESO unduplicated data 2015/2016. Note: Percentages may not add to exactly 100% due to rounding. *“All other” comprises other impression types that represented < 1% of calls both years. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 55 February 1, 2018 The primary impressions associated with the greatest proportions of medical emergency patients in the Stanford area appear to be ALOC33, followed by injury and behavioral health issues. Appendix 9 provides patient volume by primary impression, separately for day versus night. Based on Appendix 9, in the Stanford area all primary impressions except behavioral health issues occur more often during the day than at night. Behavioral health issues occur significantly more often at night than any other primary impression, statistically speaking.34 For reference, Appendix 3 lists the total number and percentage of patient cases by all primary impressions by year, and includes the category into which each primary impression was placed. 33 Note that this last category is based on a patient’s actual level of wakefulness and/or disorientation (Tindall, 1990), and does not refer to drug-induced hallucinatory experience. However, based on a scan of the data, some small number (<5%) of cases identified as ALOC may be related to alcohol or drug use. 34 Depending on the primary impression to which behavioral health is being compared, these differences range from 28% more often (compared to abdominal discomfort) to 39% more often (compared to injury/hemorrhage). PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 56 February 1, 2018 Predictive Analysis Findings: Stanford, Year 2030 Based on our analysis of current call volume and calls by type, we have predicted year 2030 call volume and calls by type for Stanford (and for the City of Palo Alto, separately, in the previous section).35 As described briefly in the Predictive Analysis Methods Summary section at the beginning of this report, we used a modified life table approach, which is generally used to describe a current population and predict future growth or decline. Stanford’s daytime population could not be broken down by age, as data on age ranges of the Stanford workforce were not available to us at a granular-enough level. Therefore, we adjusted nighttime (but not daytime) medical emergency call/patient volume predictions by age based on CA DOF (2017) projections (see Appendix 5 for details).36 We predicted call/patient volume and calls by type for day and night separately, using population figures based on CA DOF figures, Stanford University (2016) GUP projections, and City of Palo Alto Transportation Survey results (2013). A full description of this process may be found in Appendix 4. We provide predictions for all call types that represented at least 1% of calls in the 2015 or 2016 CAD datasets. The remainder are grouped into a call type category labeled “Other.” We used ESO data (described in the Predictive Analysis Methods Summary section at the beginning of this report) to adjust predictions for the medical emergency call type based on medical emergency patient volume and (for the nighttime population only) expected age- related shifts in the demographics for Stanford between now and 2030. This process is also further described in Appendix 5. The 2030 predictions for Stanford for all call types and for patient volumes take into account projected daytime and nighttime visitors, as well as larger daytime populations based on net commuters, again based on CA DOF figures, Stanford University GUP projections, and City of Palo Alto Transportation Survey results. Further information on this process may be found in Appendix 5. Projected Population The Stanford area has a current (2014) residential population of 13,506 people (U.S. Census Bureau, 2014), which is projected to rise to 20,296 by the year 2030 (calculated based on the GUP, Stanford University, 2016). The “nighttime” population we discuss in this section is the same as what the U.S. Census would consider the residential population. 35 Due to the lack of ZIP Codes in the CAD data, we were not able to estimate call volume by ZIP Code, PAFD district, or other sub-city-level areas (e.g., neighborhood). 36 Note that while there were no data projecting gender for sub-county-level areas, the projections of gender by county (CA DOF, 2017) suggest that the projected change in gender proportions between now and 2030 is not statistically significant in either San Mateo County or Santa Clara County (Chi-square = 2.00, p > 0.05). Therefore, we do not include gender as a factor in our predictions. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 57 February 1, 2018 Daytime population is a net total of nighttime residents who stay, residents who “out- commute” to somewhere besides Stanford, and individuals from other cities who “in- commute” to Stanford. Stanford’s 2014 daytime population was estimated to be 38,327 people, and projected to rise to 49,695 (see Appendix 5 for further information on the calculation of projected populations). As mentioned previously, Stanford’s daytime population could not be broken down by age, as data on age ranges of the Stanford workforce were not available to us at a granular-enough level. Figure 27 shows the recent (2014) and projected 2030 residential (nighttime) population for Stanford, by age range. Note that visitors were not included in the chart because the data available for estimating visitors’ ages in 2030 grouped age into four broad categories rather than the 13 categories available for estimating residents’ ages; however, visitors are very small proportions of the total population. Figure 27, Current and Year 2030 Projected Residential Populations by Age Category, Stanford Source: Stanford University, 2016; U.S. Census Bureau, 2014. Note: Excludes visitors to increase age category granularity (see text for further explanation). Predicted Medical Emergency Call Volume and Other Call Types For the purposes of our predictions, daytime is considered to be 8:00 a.m. to 7:59 p.m., and nighttime is considered to be 8:00 p.m. to 7:59 a.m. (Amber Cameron, personal communication, March 18, 2017). The time indicator used was the time of dispatch. Appendix 6 contains tables showing the average annual 2015/2016 call volume and the age-adjusted 2030 predicted call volume, by call type and time of day (not age-adjusted for Stanford daytime predicted 2030 calls). As mentioned previously, Stanford’s daytime population could not be - 1,000 2,000 3,000 4,000 5,000 6,000 7,000 Age 0-14 Age 15-24 Age 25-34 Age 35-44 Age 45-54 Age 55-64 Age 65-74 Age 75-84 Age 85+ 2014 Nighttime (Residential) Pop. 2030 Projected Nighttime (Residential) Pop. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 58 February 1, 2018 broken down by age, as data on age ranges of the Stanford workforce were not available to us at a granular-enough level. This means that only nighttime, not daytime, CAD medical emergency call predictions for Stanford are age-adjusted. Total average nighttime Stanford call volume in 2015/2016 was 473, and predicted to be 710 (age-adjusted) in 2030. Total average daytime Stanford call volume in 2015/2016 was 769, and predicted to be 973 (not age-adjusted) in 2030. It is clear from these data that the total number of calls will increase between now and 2030 for all call types. Figure 28 below shows the predicted change from 2015/2016 to 2030 for medical emergency calls and other major call types, for Stanford. The grey portion of the bars shows the current (2015/2016 average) number of calls by type and the red portion of the bars shows the predicted increase. Proportionally, it is predicted that medical emergency calls will increase by 36%.37 Figure 28, Predicted Change in Average Annual Call Volume from 2015/2016 to 2030, Major Call Types, Stanford Source: City of Palo Alto PSD, 2017: CAD unduplicated data 2015/2016. Notes: Nighttime medical emergency calls are age-adjusted, daytime medical emergency calls are not (see text for explanation). “All other” comprises calls that make up less than 1.5% each in the unduplicated 2015/2016 CAD dataset (Gas, HazMat, Mutual Aid, etc.). 37 Since ESO medical emergency residential patient volume predicted increases are within one to two percentage points of CAD medical emergency residential call volume predicted increases, we do not display them separately here. They may be found in Appendix 7. 0 200 400 600 800 1,000 Smoke Fire Accident All Other* Service False Alarm Medical Emergency Base (2015/2016 Average) Predicted Increase PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 59 February 1, 2018 Figure 29 below shows the predicted change from 2015/2016 to 2030 for medical emergency calls and other major call types, for Stanford during the daytime hours. Again, we note that these figures are not age-adjusted due to the lack of granularity in age-range information for Stanford employees. The grey portion of the bars shows the current (2015/2016 average) number of calls by type and the yellow portion of the bars indicates the predicted increase. It is predicted that medical emergency calls will increase by approximately 27% during the daytime. Figure 29, Predicted Change in Average Annual Call Volume from 2015/2016 to 2030, Major Call Types, Daytime, Stanford Source: City of Palo Alto PSD, 2017: CAD unduplicated data 2015/2016. Notes: Daytime medical emergency calls are not age-adjusted (see text for explanation). “All other” comprises calls that make up less than 1.5% each in the unduplicated 2015/2016 CAD dataset (Gas, HazMat, Mutual Aid, etc.). 0 200 400 600 800 1,000 Smoke Fire Accident All Other* Service False Alarm Medical Emergency Base (2015/2016 Average) Predicted Increase PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 60 February 1, 2018 Figure 30 below shows the predicted change from 2015/2016 to 2030 for medical emergency calls and other major call types, for Stanford during the nighttime hours. The grey portion of the bars shows the current (2015/2016 average) number of calls by type and the dark blue portion of the bars indicates the predicted increase. It is predicted that medical emergency calls will increase by 50% during the nighttime. Figure 30, Predicted Change in Average Annual Call Volume from 2015/2016 to 2030, Major Call Types, Nighttime, Stanford Source: City of Palo Alto PSD, 2017: CAD unduplicated data 2015/2016. Notes: Nighttime medical emergency calls are age-adjusted, daytime medical emergency calls are not (see text for explanation). “All other” comprises calls that make up less than 1.5% each in the unduplicated 2015/2016 CAD dataset (Gas, HazMat, Mutual Aid, etc.). Although the absolute numbers of increased calls are smaller for Stanford during the nighttime than for the daytime, the proportional increases of medical emergency and other calls at night (50%) will be much greater than the proportional increases of medical emergency and other calls during the day (27%). These predictions, taken with the current baseline data, could suggest that nighttime call volume at Stanford may increase faster than daytime call volume over time, resulting in a closer balance between day and night call volume. That is, based on this analysis (which we remind the reader is limited by the fact that Stanford daytime medical emergency call predictions could not be age-adjusted), nighttime calls to PAFD from Stanford are predicted to drop from 62% to 58% of total Stanford call volume by 2030. 0 200 400 600 800 1,000 Smoke Fire Accident All Other* Service False Alarm Medical Emergency Base (2015/2016 Average) Predicted Increase PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 61 February 1, 2018 Predicted Medical Emergency Patient Volume Originating in the Stanford Area Predicted Medical Emergency Patient Volume Overall and by Age Overall, there were an average of 933 medical emergency patients annually from Stanford in 2015/2016, with 600 during the daytime and 333 at night. It is predicted that there will be 1,213 medical emergency patients from Stanford in 2030, with 713 predicted during the daytime (not age-adjusted) and 500 predicted at night (age-adjusted). Below, we review the nighttime (residential) predictions by age. Again, we converted age into categories used by the ACS and DOF, the better to compare statistics on current age of patients to predicted age of patients. For the sake of simplicity, in Figure 31 we merged certain age categories. The figure shows current and projected data on Stanford’s residential (i.e., nighttime) population. As mentioned previously, granular-enough data are not available for Stanford’s daytime population by age range. Visitors were not included in the figure because the data available for estimating visitors’ ages in 2030 grouped age into four broad categories rather than the 13 categories available for estimating residents’ and workers’ ages; however, visitors are very small proportions of the total population.38 Figure 31, Current and Year 2030 Predicted Age-Adjusted Residential Medical Emergency Patient Volume by Age Category, Stanford Source: City of Palo Alto PSD, 2017: ESO unduplicated data 2015/2016. Note: Excludes visitors to increase age category granularity (see text for further explanation). 38 Including daytime and overnight visitors would only increase the predicted numbers of patients by an average of 39, across all age groups – a very small proportion of the total number of predicted interactions overall. - 20 40 60 80 100 120 140 160 180 200 Age 0-14 Age 15-24 Age 25-34 Age 35-44 Age 45-54 Age 55-64 Age 65-74 Age 75-84 Age 85+ 2015-16 Avg. # Stanford ME Patients 2030 Predicted # Stanford ME Patients PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 62 February 1, 2018 As with the 2030 Palo Alto predictions, the 2030 Stanford predictions show increases in medical emergency patient volume compared to current medical emergency patient volume at approximately 31% overall (see Appendix 6). However, less than a third (32%) of Stanford’s predicted 2030 residential medical emergency patients will be age 55+, while over 40% will be age 15-24. Predicted Medical Emergency Patient Volume by Type of Illness/Issue In 2015/2016, the primary impressions that were associated with the greatest proportions of medical emergency patients in Stanford were altered level of consciousness,39 followed by injury and behavioral health issues. The order of the top primary impressions for Stanford (see Table 7) remains the same in the predictions for year 2030, although we note that the portion of primary impression predictions that relate to the daytime population are not age-adjusted due, as discussed previously, to a lack of granular-enough data on Stanford’s employees’ age ranges. Table 7, Predicted Year 2030 Medical Emergency Patients, Categories of Primary Impressions, in Order of Frequency Primary Impression Category % (#) Stanford, 2030 Altered Level of Consciousness 29.2% (367) Injury/Hemorrhage 28.0% (352) Behavioral Health 10.1% (127) Non-Specific Pain 7.4% (93) Abdominal Discomfort 7.1% (90) Cardiac-Related 6.2% (78) Respiratory-Related 4.8% (61) All Other* 4.1% (51) No Complaints 3.2% (40) TOTAL 1,259 Source: Actionable Insights, LLC, 2017, unpublished data, based City of Palo Alto PSD, 2017: ESO unduplicated data 2015/2016. Notes: Not age-adjusted. Percentages may not add to exactly 100% due to rounding. *“All other” comprises other primary impression types that represented < 1% of calls in both years. 39 Altered level of consciousness is based on “a patient’s actual level of wakefulness and/or disorientation” (Tindall, 1990) and has a variety of causes, including dehydration, hypoglycemia, and shock, but is not intended to refer to symptoms that are directly related to drug or alcohol use. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 63 February 1, 2018 However, it is also important for PAFD’s future planning to review the primary impressions that will increase the most between 2015/2016 and the year 2030. Stanford predictions suggest that the greatest increase in absolute numbers will be among ALOC patients40 and injury patients.41 However, the greatest proportional increases will be among behavioral health patients (43%, or an increase of 38 more patients). See Figures 32 and 32 for a representation of Stanford increases in absolute numbers by primary impression, for daytime and nighttime separately. See Appendix 9 for total predicted changes in primary impression categories, by city and time of day. Figure 32, Comparison of Actual versus Predicted Absolute Change in Average Annual Patient Volume from 2015/2016 to 2030, by Primary Impression, Stanford, Daytime Source: Actionable Insights, LLC, 2017, unpublished data, and City of Palo Alto PSD, 2017: ESO unduplicated data 2015/2016. Note: Not age-adjusted. 40 An increase of 94 more patients, or 34%. 41 An increase of 87 more patients, or 33%. - 100 200 300 Abdominal Altered Level of Consciousness Behavioral Health Cardiac Injury/ Hemorrhage Non-Specific Pain Respiratory 2015/2016 Average 2030 # Increase PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 64 February 1, 2018 Figure 33, Comparison of Actual versus Predicted Absolute Change in Average Annual Patient Volume from 2015/2016 to 2030, by Primary Impression, Stanford, Nighttime Source: Actionable Insights, LLC, 2017, unpublished data, and City of Palo Alto PSD, 2017: ESO unduplicated data 2015/2016. As described in the section predicting the City of Palo Alto’s primary impressions, age was the predominant factor in the increase in medical emergency patient volume between 2015/2016 and 2030. Because we did not have projections of age for Stanford’s daytime population, we did not conduct such an analysis for Stanford. - 100 200 300 Abdominal Altered Level of Consciousness Behavioral Health Cardiac Injury/ Hemorrhage Non-Specific Pain Respiratory 2015/2016 Average 2030 # Increase PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 65 February 1, 2018 Recommendations & Next Steps Addressing the Expected Increase in EMS Calls Increases in fire department EMS calls are a national issue. According to the National Fire Protection Association’s data, the proportion of total fire department calls nationwide that were for medical aid increased from 54% in 1986 to 64% in 2015 (National Fire Protection Association, 2016). Fire departments nationwide have reported this trend of increasing EMS calls (see, for example, Cooper, 2016; Keisling, 2015; Los Angeles Fire Department, 2015; Luthern 2014). Fire departments with ambulance services are also having to address non- emergency use of 911 dispatch. The National Conference of State Legislatures (NCSL) (n.d.) identified research that suggests non-emergencies account for between 10% and 40% of EMS responses, and other researchers have suggested that successful reduction of inappropriate ambulance use could consequently reduce the load on emergency departments (EDs) by 11% (Patton & Thakore, 2012). The traditional model to address expected increases in EMS calls would be to add more 24-hour ambulances. However, this is a costly method of addressing increases in call volume. Other jurisdictions have turned to alternate models, including various forms of community paramedicine and mobile integrated health care practice approaches (CP/MIHP). We review these approaches below, with reference to Actionable Insights’ literature review conducted on behalf of PAFD (unpublished, 2017). In 2015 there were over 100 CP/MIHP programs nationwide, and they are expanding rapidly. This presents PAFD with a timely opportunity to leverage the knowledge gains from these other programs. Successful Strategies to Reduce EMS Calls and/or Reliance on Emergency Transport The following are four primary evidence-based programs and strategies that are applicable to the situation facing the Palo Alto/Stanford community. 1. Community paramedicine and mobile integrated health care practice (CP/MIHP) approaches. The CP/MIHP approach is broadly characterized as a model in which a community has greater access to 24/7 non-emergency healthcare, including health education, outreach, monitoring, and prevention, brought to the patient by inter-professional teams42 who are integrated into and/or partnered with local or regional healthcare systems and under appropriate medical direction (HRSA, 2012, Myers et al., 2012, NCSL, n.d.). This could include any or all of the following (NCSL, n.d., Wang, 2011, WECAD, 2011): Conducting assessments and/or screenings (e.g., for congestive heart failure or 42 These may include paramedics, emergency medical technicians (EMTs), nurse practitioners (NPs), registered nurses (RNs), physician assistants (PAs), and other allied health professionals. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 66 February 1, 2018 diabetes) Providing treatment (such as wound care) Engaging in prevention (like health education, heat surveillance, vaccinations, fall assessment, helmet fitting, and car seat installation) Making referrals (that is, to other medical providers or social services) including addressing mental health, housing, and other needs (also known as “patient navigation”) Regularly checking in with high-risk patients – 911 or ED super-users – to assist with chronic disease management, medication compliance, and/or insurance enrollment. CP/MIHP programs look different in different communities because they are uniquely responsive to gaps in assessed community needs, and no two communities’ needs are the same (Myers et al., 2012) However, all CP/MIHP programs involve partnerships with existing healthcare organizations/infrastructure, address gaps in service, enable providers to exercise the full scope of their practice, and focus on patient outcomes to improve individual and community health. The focus on patient outcomes is particularly important. One reporter described CP/MIHP as “a redefining of what EMS is, emphasizing measuring patient outcomes over processes like response times” (Goodwin, 2013). Recently, certain researchers (Choi et al., 2016) noted that while there are relatively few studies assessing CP/MIHP program safety, efficacy, and cost- benefit ratios, data from existing programs suggest that these approaches may have impacts such as reduced health disparities; reduced Examples of EMS CP/ MIHP Program Results MedStar, Ft. Worth, TX (AHRQ, 2016): Frequent 911 users enrolled in program (incl. home visits and phone support). Total of 300 program “graduates.” ▪ 911 call volume from “graduates” down by 61% in the 12 months post-program ▪ Transport costs reduced from pre- to post-program by $4.98M ($16,600 per “graduate” in 1 year) ▪ Freed up 14,000 bed hours at local EDs over 5 years CARES, Colorado Springs, CO (Zavadsky et al., 2015): Frequent 911 users enrolled in program (incl. home visits, health ed, navigation to resources). Total of 500 patients. ▪ 911 call volume from two-thirds of patients down 50% in first 2 years ▪ Hospital readmissions down 75% in first 3 months post-program, led to $150K savings in Medicaid claims REMSA, Reno, NV (Gerber, 2015): Post-discharge congestive heart failure patients enrolled in program (incl. 1 hospital visit, follow-up home visits, screenings, assessments, services). Total of 444 patients. ▪ Avoided 28 readmissions and 97 ED visits ▪ First year total savings of $1.6M PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 67 February 1, 2018 hospital readmission rates for congestive heart failure, chronic obstructive pulmonary disease, and pneumonia; increased immunization rates; injury prevention; and reduced mortality rates (Choi et al., 2016, Myers et al., 2012, Wang, 2011, WECAD, 2011). Another important element of CP/MIHP is the reduction in healthcare costs, both directly and via reduction in ED transports and/or hospital readmissions. Krumperman (2010) has indicated that part of the value of CP/MIHP is its ability to provide cost- effective care to patients for whom services could be provided in a non-ED setting. Several writers have described the cost savings that may be found in reducing unnecessary ED transports and potential hospital readmissions (Coffman & LaFrance, 2016, Tadros et al., 2012, and Widiatmoko et al., 2008). 2. Use of nurses and/or physician assistants (PAs) in EMS and/or triage; can also be considered part of a CP/MIHP approach. Dr. Eric Beck, the medical director for the City of Chicago EMS System and the Chicago Fire Department, has been quoted as saying, “EMS needs to rethink its basic mission of being about transportation and instead be about providing care in the most effective way for the patient. …That could be community paramedicine. It could be by integrating nurse triage into dispatch, or using telemedicine to enable patients to be treated at home without having to transport” (Goodwin, 2013, emphasis ours). Walsh and Little’s (2001) small study of a nurse practitioner (NP) working in a paramedic role found that although the NP took more time at the scene to diagnose and treat a patient than a conventional EMS team would have, up to one third of calls attended to by the NP avoided ED transport. Widiatmoko et al. (2008), in describing a pilot study of nurse/paramedic partners dispatched to visit non-urgent emergency calls, suggested that the costs of such partnerships to the national Ambulance Services would be more than made up for by the savings in secondary care costs. There have been similar findings in studies on nurse telephone consultation (also known as telephone triage). A UK study found after-hours nurse telephone consultation was “at least as safe” as existing after-hours options and reduced both home visits by general practitioners and ED visits outside of normal business hours (Lattimer et al., 1998). The companion study found that there were moderate net annual cost savings of £13,185, the equivalent of about $33,000 in 2018 dollars (Lattimer et al., 2000).43 The researchers stated, “The greatest impact on the results of the cost analysis was generated by costs for emergency hospital admissions, a secondary analysis of admission data for this trial 43 Currency conversion from British pounds to U.S. dollars for historical period January 1997-January 1998 via fxtop.com. Calculation for inflation to 2018 dollars via www.in2013dollars.com. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 68 February 1, 2018 having shown that the intervention saved short stays (1-3 days) in hospital” (Lattimer et al., 2000:1056). They noted that this savings excludes the reduced costs of physicians not having to attend at the ED (Lattimer et al., 2000). There are a limited number of studies on the use of physician assistants (PAs) in EMS and/or triage (Bloemhoff et al., 2016, Callard, 2012, Nestler et al., 2012). ED-related studies found improvement in patient wait times, and one reported reduced stress on medical staff (Callard, 2012). A study of nearly 1,000 patients in the Netherlands, which compared results of the use of PAs versus nurses as solo EMS providers in ambulances, found that while PAs consulted twice as often with medical specialists than did nurses, PAs also referred a significantly smaller proportion of patients to general practitioners or the ED compared to nurses (50% versus 73%), with no apparent difference in 72-hour follow-up (Bloemhoff et al., 2016). This research suggests that use of PAs can reduce unnecessary hospital admissions and related costs (Bloemhoff et al., 2016). 3. Telemedicine and in-home visit preventive care models; can also be considered part of a CP/MIHP approach. Some studies have shown that home visits (face-to-face care at home) by community health workers, community paramedics, and/or personal support workers result in good patient outcomes and reduce the volume of 911 calls, ambulance transports, and ED visits (see, for example, Findley et al., 2014 and Ruest et al., 2012). Qualitative studies available on community paramedic home visits have described improved access to and use of healthcare services by patients, improved disease management, improved safety, and increased patient satisfaction and/or peace of mind (see, for example, Do et al., 2016 and Pennell et al., 2016). A meta-analysis of studies comparing in-home care to no home care found that in-home care delivered by nurses led to improved outcomes in patients with various levels of heart disease (Health Quality Ontario, 2013). This meta-analysis also found that home visits from physical or occupational therapists with the goal of modifying tasks and the home environment improved the ability of adults with chronic disease to engage in activities of daily living (Health Quality Ontario, 2013). With regard to end-of-life, researchers found that home visits by nurses or by community-based palliative care teams were effective in lowering levels of patient distress, reducing ED visits, and allowing more individuals to die at home rather than in the hospital (see, for example, McCorkle et al., 1989 and Seow et al., 2014). As mentioned briefly in the previous section, the use of telemedicine may be an effective way to provide patient care. There are several categories of telemedicine, including remote monitoring, “store-and-forward” technology (data transmission to a remote clinician followed by a later report or consultation), and real-time consultation (Flodgren PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 69 February 1, 2018 et al., 2015). A meta-analysis of 93 studies of interactive telemedicine concluded that face-to-face and telephone delivery of care in managing congestive heart failure and diabetes have similar outcomes, with perhaps greater decreases in high blood pressure and better control of blood glucose for telemedicine patients compared to those receiving face-to-face care (Flodgren et al., 2015). Research on the use of telemedicine for those who suffer chronic wounds found that advice-giving via telemedicine correlated with significantly increased healing of wounds in comparison to a best-practice conventional approach (Zarchi et al., 2015). A study of the implementation of a telemedicine service at a nursing home found fewer hospitalizations of nursing home residents (associated with reduced health complications and lower mortality), and consequent Medicare savings (Grabowski and O’Malley, 2014). Telemedicine was also used to good effect in palliative care for advanced cancer patients by improving healthcare access, reducing EMS use, and improving symptom assessment/control (Hennemann-Krause et al., 2015). 4. Outreach and education efforts also complement a prevention approach. A review of research shows that lay health educators (such as promotores) have been found to be effective in improving community members’ health status (Ayala, 2010). Reductions in hospital expenditures for diabetes and hypertension were identified in areas that had lay health educators (Boone, 2002). In a Denver community served by community health workers, economic savings were realized in health care costs, including urgent care (Whitley et al., 2006). The community consulted for this study suggested that the community lacks awareness of the appropriate use of the emergency departments and alternatives, including urgent care clinics. “Well, I think that a lot of people don’t know about urgent care and express care. I mean we deescalate so many things to express care clinics rather than the emergency room. I deescalated one just the other day for a gentleman. His sister couldn’t get into primary care. He had something going on with his neck. He was recently moved here from out of state just to be by Stanford. And they had a little bit of wait to go into primary care and they were gonna come back to the emergency room. And I told them about this express care clinic and it worked out great.” – Key Informant PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 70 February 1, 2018 Based on our research, community paramedicine and integrated health care practice models currently appear to be the most successful and innovative approaches available to fire departments with ambulance services. As described, these models can be implemented in a wide variety of ways, including telemedicine, home visits, arrangements for direct transport/diversion, community clinic extensions, nurse/PA triage, etc. Below, we describe some of the challenges to and best practices for implementing a CP/MIHP approach. Implementation Challenges and Best Practices in CP/MIHP The NCSL (n.d.) has identified several challenges to implementation of a CP/MIHP program, including financial, regulatory, and workforce-related issues. 1. Solutions to financial (i.e., reimbursement) challenges: It is suggested that financial issues (that is, reimbursement) can be worked out on the state or community level with public and/or private insurers (NCSL, n.d.). There are a variety of examples in the literature, including obtaining added funding via non-EMS work (HRSA, 2012), adapting the payment models of accountable care organizations (ACOs) to CP/MIHP programs (Kizer et al., 2013), and aligning CP/MIHP program financial interests with the financial interests of their partners by “linking sustainable program funding to savings accruing to system payers from reduced re-admission rates” (MIHP Resource Center, 2014). Goodwin (2013) observes that the latter requires the CP/MIHP program to share both the savings and the risk. With regard to other sustainable funding models, there are various subscription programs (City of Arcadia, 2015, City of Fountain Valley, 2010, City of Huntington Beach, 2017, Farkas, 2016), some of which include sliding scale fees for low- income households, and there are financial models based on property taxes or other municipal funds (Pearson & Shaler, 2015, Snohomish County Fire District 1, 2017, Zavadsky et al., 2015). 2. Solutions to regulatory challenges: With regard to regulatory barriers, the NCSL (n.d.) has noted that some states are working to define CPs as a class and to overcome regulatory barriers for CP programs. The Western Eagle County Ambulance District (WECAD) in Colorado has recommended that CP/MIHP program developers talk with their state’s EMS regulatory body to address any issues that may impact the operation of such a program. While WECAD began their program at the grassroots level by partnering with their area public health department, they have stated clearly that the program was “in cooperation with the state EMS Office” (WECAD, 2011:6). PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 71 February 1, 2018 3. Solutions to workforce-related challenges: With regard to workforce concerns, the NCSL (n.d.) has indicated that health professionals (like nurses or home healthcare associations) have occasionally opposed CP/MIHP programs because of concerns about encroachment on their scope of practice. Wang (2011:10) stated, “Rapid implementation of CP without identification of expanded roles causes resistance from other health care professionals.” Wang (2011) also noted that past CP/MIHP program failure has occurred in part simply because of a lack of relationships with community residents, from whom program support might otherwise be found. The literature has suggested that continued engagement of providers and other community members throughout the development and introduction of CP/MIHP programs is essential for role clarity, to address various concerns among providers and the community, and to plan for optimal collaboration among all providers (NCSL, n.d.). EMS agencies that already have CP/MIHP programs in place have recommended that program developers involve stakeholders from the very start to assure them the program is intended for collaboration rather than competition (Zavadsky et al., 2015). It is noted that “[i]n well developed, mature CP programs, the community paramedic can be the eyes and ears of primary and emergency care physicians and an extension to their practices” (NOSORH, 2010). The literature has emphasized that CP/MIHP programs expand the role of EMS practitioners but do not change their scope of practice. The International Roundtable of Community Paramedicine (2011) stated that community paramedics work “in non-traditional roles using existing skills… [and] will provide services through unique models of delivery and enhanced protocols through an integrated collaborative network with other health care providers.” Goodwin (2013) described how EMTs and paramedics working in CP/MIHP have an expanded role rather than an expanded scope; they simply apply skills for which they are already licensed in order to help community members manage chronic disease, navigate through the healthcare system, and prevent ED visits via prevention campaigns and/or public education. 4. Best practices: Engage in external quality control (Wang, 2011) and address quality assurance via medical direction and program evaluation (WECAD, 2011); for the latter, integration with state or area electronic health information exchanges is crucial in order to measure patient outcomes, experience, and the program’s impact on patient costs (Zavadsky et al., 2015). Use WECAD (2011) handbook with recommended steps for a grassroots approach to implementing a CP/MIHP program: a. Plan to plan (learn about CP/MIHP) b. Assess program feasibility and engage key partners (like medical providers) PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 72 February 1, 2018 c. Determine how to provide medical direction d. Assess community health needs e. Determine the scope of the program (services, staffing, budget/funding needs) f. Engage the community (approach key stakeholders with your community health needs assessment, program vision statement, and fact sheet with problem/ opportunity statement and FAQ) g. Develop policies and procedures h. Plan and implement training i. Develop an evaluation plan j. Begin operations k. Evaluate the pilot phase Follow broad principles developed by an MIHP working group to guide CP/MIHP program developers (Goodwin, 2013): a. Assess community needs, remain value-focused, and feature a competency- and evidence-based practice that ensures continual education, 24-hour community access, and ongoing performance improvement. b. Ensure community partnership with active medical direction. c. Deliver improved access to care and health equity for populations served through 24-hour care availability. d. Focus on patient-centered navigation and offer community-centered care by integrating existing infrastructures and resources and bringing care to patients through technology, communications, and health information exchange. e. Use evidence-based practice, incorporating multidisciplinary and inter- professional teams through which providers utilize their full scope of practice. Focus on integration with current health care and social services systems to develop a robust network of resources (Zavadsky et al., 2015). Addressing Community Health Needs Our primary research for this study indicates that the need for preventative services in the Palo Alto/Stanford area is supported by the community. For example, residents who participated in the neighborhood association focus group stressed that education is needed in the areas of emergency preparedness, especially for those who do not live in close proximity to a fire station. In response to a question about solutions to the overuse of emergency departments, some suggested that residents need to know what constitutes a medical emergency, and also where to call or go for non-emergencies. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 73 February 1, 2018 Some residents also indicated support for community integrated health care; in one focus group, the firehouse clinic operating in Hayward, CA, was suggested as a model (City of Hayward, 2015). The community envisioned it as “checking in” on those who otherwise would not go to a clinic. For example, seniors with whom we talked appreciated having access to a nurse practitioner at their senior housing facility several times per week. They observed that they checked their blood pressure and vision more often when this service was available. Also, the need to reach isolated and homebound community members was mentioned in more than one group. Health education was mentioned multiple times. Community key informants and residents suggested subjects such as youth mental health (for parents), adult caregiving, stroke risk and prevention, and falls prevention. Cultural competency and capacity of first responders were mentioned several times by community members. For example, one key informant recommended sending mental health specialists out with first responders (such as the Santa Clara County Behavioral Health pilot program). As mentioned earlier, training first responders on how to serve people of diverse cultures, gender identities, and languages is also important. In addition to the services that can be provided through an integrated community health care model, the community expressed the need for more services that are traditionally provided by other organizations such as nonprofit organizations or hospitals: Home health care for seniors, including those of the middle class who can’t afford it Phone buddy program for seniors Discharge planning Geriatric care managers Social workers to serve those with depression and handling life transitions (e.g., moving to senior housing) Telemedicine (e.g., using photographs to communicate with doctors) Sobering station with the capacity to screen for mental health problems and substance abuse Mental health services during non-traditional hours PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 74 February 1, 2018 Next Steps in Addressing Future Needs and Capacity To “engineer out” unplanned emergencies (i.e., EMS incidents) and understand what it will take to work together as a community to help address these risk factors and help prevent emergencies from occurring, Actionable Insights recommends a three-pronged approach: 1. PAFD conducts an internal resource assessment to assess its capacity to implement the recommended EMS strategies described above. Such an assessment, combined with this study of current needs, will serve as the foundation of a plan for a community integrated health model that best fits Palo Alto/Stanford. 2. PAFD convenes internal City of Palo Alto stakeholders and external stakeholders with a focus on prevention efforts in Palo Alto/Stanford. This includes discussion of those risk factors that have been identified through the work underlying this report. These discussions will include preliminary identification of community partnerships that might best help to minimize the need for emergency response by preventing emergencies from happening. Actionable Insights will meet with PAFD to review the results of the discussions. 3. Actionable Insights will then assist PAFD in developing a collective impact plan to prevent medical emergencies. A collective impact plan provides structure to collaborative efforts and recognizes the contribution of many different partners towards the same goals. This effort will bring stakeholders from the City of Palo Alto and other organizations together to plan how they will make collective impact in the years to come. Ideally, the collective impact group can also inform and even shape the integrated health model that PAFD plans to adopt. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 75 February 1, 2018 Appendix 1 Sources Agency for Healthcare Research and Quality (AHRQ). (2016). Trained paramedics provide ongoing support to frequent 911 callers, reducing use of ambulance and emergency department services. Service Delivery Innovation Profile, February 21, 2016. Accessed February 5, 2017: http://www.innovations.ahrq.gov/content.aspx?id=3343. Arias, E. (2012). United States life tables, 2008. National Vital Statistics Reports, 61(3), September 24, 2012. Accessed June 6, 2017: https://www.cdc.gov/nchs/data/nvsr/nvsr61/nvsr61_03.pdf. Ayala, G. X. et al. (2010). Outcome effectiveness of the lay health advisor model among Latinos in the United States: an examination by role. Health Education Research, 25(5), 815–840, https://doi.org/10.1093/her/cyq035. Bay Area Census. (2010). City of Palo Alto. Retrieved from: http://www.bayareacensus.ca.gov/cities/PaloAlto.htm. Begon, M., Harper, J. L., & Townsend, C. R. (1990). Ecology: individuals, populations and communities. 2nd ed. Blackwell Scientific Publications: Cambridge, MA. Begon, M., Mortimer, M., & Thompson, D. J. (1996). Population ecology: a unified study of animals and plants. 3rd ed. Blackwell Scientific Publications: Cambridge, MA. Bhadur, B. (2007). Population distribution during the day. Encyclopedia of GIS. Shekhar, S., & Xiong, H., eds. Springer Science+Business Media, LLC: New York. Bloemhoff, A., Schoonhoven, L., de Kreek, A. J. L., van Grunsven, P. M., Laurant, M. G. H., & Berben, S. A. A. (2016). Solo emergency care by a physician assistant versus an ambulance nurse: a cross-sectional document study. Scandinavian Journal of Trauma, Resuscitation, and Emergency Medicine, 24, 86. Boone, K. (2002). Lay health advisors reduce the cost of hospital expenditures for hypertension and diabetes. The 130th Annual Meeting of APHA. Presented November 12, 2002. Retrieved from: https://apha.confex.com/apha/130am/techprogram/paper_46783.htm. Bureau of Labor Statistics. (2016). Household data annual averages: 11b. Employed persons by detailed occupation and age. Accessed June 8, 2017: https://www.bls.gov/cps/cpsaat11b.pdf. California Department of Finance (DOF). (2017). Projections, February 2017. Accessed May 20, 2017: http://www.dof.ca.gov/Forecasting/Demographics/projections/. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 76 February 1, 2018 California Department of Public Health EpiCenter (EPIC). (2013). CDPH Vital statistics death master files, CA Office of Statewide Health Planning and Development (inpatient discharge data and emergency department data). Retrieved from: http://epicenter.cdph.ca.gov. Callard, J. (2012). Physician assistant provider in triage. Urgent Matters, The George Washington University School of Medicine & Health Sciences. Accessed January 25, 2017: https://smhs.gwu.edu/urgentmatters/sites/urgentmatters/files/AAPAFeaturedTool2012.pdf. Centers for Disease Control and Prevention. (2016). Promotores de salud / community health workers. Accessed December 15, 2017: https://www.cdc.gov/minorityhealth/promotores/index.html. Choi, B. Y., Blumberg, C., & Williams, K. (2016). Mobile integrated health care and community paramedicine: an emerging emergency medical services concept. Annals of Emergency Medicine, 67(3), 361-366. City of Arcadia. (2015). Paramedic Membership Program. Accessed July 23, 2017: https://www.arcadiaca.gov/government/city-departments/fire-department/paramedic- membership-program . See also Questions & Answers for Business Members (https://www.arcadiaca.gov/home/showdocument?id=1568). City of Fountain Valley. (2010). Fees & Subscription Program. Accessed July 23, 2017: http://www.fountainvalley.org/258/Fees-Subscription-Program. See also How the Paramedic Membership Program Works (http://www.fountainvalley.org/DocumentCenter/Home/View/1409) and Fountain Valley Fire Department Paramedic Membership Program (http://www.fountainvalley.org/DocumentCenter/Home/View/277). City of Hayward. (2015). Hayward’s trailblazing “firehouse clinic” set to open. Retrieved from: https://www.hayward-ca.gov/discover/news/oct15/hayward%E2%80%99s-trailblazing- %E2%80%9Cfirehouse-clinic%E2%80%9D-set-open. City of Huntington Beach. (2017). FireMed Paramedic Subscription Program. Accessed July 23, 2017: http://www.huntingtonbeachca.gov/government/departments/fire/firemed/. See also FireMed FAQ (http://www.huntingtonbeachca.gov/government/departments/fire/FireMed/FireMedFAQ.cfm ) and FireMed Fact Sheet (http://www.huntingtonbeachca.gov/files/users/fire/firemed_fact_sheet_jun_2005.pdf). City of Palo Alto. (2013). City of Palo Alto city council staff report #4322, transportation survey 2013, dated December 9, 2013. Accessed April 21, 2017: https://www.cityofpaloalto.org/civicax/filebank/documents/38089. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 77 February 1, 2018 City of Palo Alto. (2014). Demographic analysis draft. City of Palo Alto Parks, Trails, Open Space & Recreation, dated September 2014. Accessed May 30, 2017: http://www.paloaltoparksplan.org/pdf/PTOSR-Demographic-Analysis-September2014.pdf. City of Palo Alto. (2017a). Project safety net. Retrieved from: www.psnpaloalto.com/psn- plan/implementation/. City of Palo Alto. (2017b). FY 2016 Performance Report, The National Citizen SurveyTM, and Citizen Centric Report: Attachment B, The 2016 National Citizen SurveyTM. Retrieved from: https://www.cityofpaloalto.org/civicax/filebank/documents/55618. Coffman, J. & LaFrance, A. (2016). Mobile Integrated Health Care – Community Paramedicine: A Resource for Community-Dwelling People at Risk for Needing Long-Term Care. San Francisco, CA: UCSF Health Workforce Research Center. Accessed May 24, 2017: http://healthworkforce.ucsf.edu/sites/healthworkforce.ucsf.edu/files/REPORT_2016_Mobile_I ntegrated_HC_Paramedicine.pdf. Colby, S.L. & Ortman, J.M. (2015). Projections of the size and composition of the U.S. population: 2014 to 2060. U.S. Census Bureau, Population Estimates and Projections, Current Population Reports, issued March 2015, P25-1143. Accessed April 25, 2017: https://www.census.gov/content/dam/Census/library/publications/2015/demo/p25-1143.pdf. Community Commons (2017). Database. Retrieved from: https://www.communitycommons.org/. Cooper, M. (2016). Ohio fire, EMS department struggling with 911 call volume. Springfield News-Sun, March 30, 2016. Accessed January 25, 2017: https://www.ems1.com/ems- management/articles/76601048-Ohio-fire-EMS-department-struggling-with-911-call-volume/. CTV (2011). Community paramedics helping those on the fringes. CTV News, Monday, January 10, 2011. Accessed January 21, 2017: http://www.ctvnews.ca/community-paramedics-helping- those-on-the-fringes-1.594246. DataSF. (2017). Bay Area ZIP Codes. DataSF.org. Accessed June 6, 2017: https://data.sfgov.org/Geographic-Locations-and-Boundaries/Bay-Area-ZIP-Codes/u5j3- svi6/data. Do, L., Flores-Keown, B., Vu, A., & Warholak, T. (2016). Qualitative evaluation of the Rio Rico Fire Department Community Integrated Paramedic Program. Accessed January 30, 2017: http://arizona.openrepository.com/arizona/bitstream/10150/613964/2/Do,Flores- Keown,Vu2016.pdf. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 78 February 1, 2018 Farkas, R. (2016). Annual fee offers unlimited Shaler EMS ambulance trips. Tribune-Review. Accessed July 23, 2017: http://triblive.com/news/neighborhoods/northhills/10158151-74/ems- ambulance-shaler. Findley, S., Matos, S., Hicks, A., Chang, J., & Reich, D. (2014). Community health worker integration into the health care team accomplishes the triple aim in a patient-centered medical home: a Bronx tale. The Journal of Ambulatory Care Management, 37(1), 82-91. Flodgren, G., Rachas, A., Farmer, A. J., Inzitari, M., & Shepperd, S. (2015). Interactive telemedicine: effects on professional practice and health care outcomes. The Cochrane Library. Accessed January 30, 2017: https://www.researchgate.net/profile/Gerd_Flodgren/publication/281588584_Interactive_tele medicine_effects_on_professional_practice_and_health_care_outcomes/links/57ac28ec08ae0 932c9725445.pdf Garcia-Williams, A. et al. (2016). Epi-Aid 2016-018: Undetermined risk factors for suicide among youth, ages 10–24 — Santa Clara County, CA. Report prepared for Santa Clara County Public Health Department. Retrieved from: http://www.sccgov.org. Gerber, M. (2015). How an EMS System in Reno, Nev., Watches Over Post-Discharge Patients to Avoid Costly Readmissions. Journal of Emergency Medical Services. March 23, 2015. Accessed May 24, 2017: http://www.jems.com/articles/print/volume-40/issue-3/features/how-ems- system-reno-nev-watches-over-pos.html. Goodwin, J. (2013). Finding a new seat at the healthcare table. EMS World, 42(7). Accessed January 23, 2017: http://www.emsworld.com/article/10957645/community-paramedicine- mobile-integrated-healthcare. Governing.com. (2017). Median age for U.S. counties map, statistics. Accessed December 7, 2017: http://www.governing.com/gov-data/census/median-age-county-population-map.html. Grabowski, D. C., & O’Malley, A. J. (2014). Use of telemedicine can reduce hospitalizations of nursing home residents and generate savings for Medicare. Health Affairs, 33(2), 244-250. Health Quality Ontario (2013). In-home care for optimizing chronic disease management in the community: an evidence-based analysis. Ontario Health Technology Assessment Series, 13(5), 1. Health Resources and Services Administration (HRSA) (2012). Community Paramedicine Evaluation Tool. U.S. Department of Health and Human Services, Health Resources and Services Administration, Office of Rural Health Policy. Accessed January 21, 2017: https://www.hrsa.gov/ruralhealth/pdf/paramedicevaltool.pdf. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 79 February 1, 2018 Hennemann-Krause, L., Lopes, A. J., Araújo, J. A., Petersen, E. M., & Nunes, R. A. (2015). The assessment of telemedicine to support outpatient palliative care in advanced cancer. Palliative and Supportive Care, 13(04), 1025-1030. International Roundtable on Community Paramedicine (IRCP) (2017). IRCP vision and mission statements. Accessed January 21, 2017: http://ircp.info/About-Us. Keisling, P. (2015). Why we need to take the “fire” out of “fire department.” Governing, July 1, 2015. Accessed January 25, 2017: http://www.governing.com/columns/smart-mgmt/col-fire- departments-rethink-delivery-emergency-medical-services.html. Kizer, K., Shore, K., & Moulin, A. (2013). Community paramedicine: a promising model for integrating emergency and primary care. University of California eScholarship. UC Davis Previously Published Works. Accessed January 30, 2017: http://escholarship.org/uc/item/8jq9c187. Krebs, C.J. (1985). Ecology: the experimental analysis of distribution and abundance. 3rd ed. Harper and Row: New York, NY. Kromer, B., & Howard, D. (2013). Labor force participation and work status of people 65 years and older. US Department of Commerce, Economics and Statistics Administration, US Census Bureau. Accessed December 21, 2017: https://www.census.gov/prod/2013pubs/acsbr11- 09.pdf. Krumperman, K. (2010). History of Community Paramedicine. Journal of Emergency Medical Services. January 22, 2010. Accessed February 5, 2017: http://www.jems.com/articles/2010/06/history-community-paramedicine.html. Latham, L. P., & Ackroyd-Stolarz, S. (2014). Emergency department utilization by older adults: A descriptive study. Canadian Geriatrics Journal, 17(4), 118–125. http://doi.org/10.5770/cgj.17.108 Lattimer, V., George, S., Thompson, F., Thomas, E., Mullee, M., Turnbull, J., Smith, H., Moore, M., Bond, H., & Glasper, A. (1998). Safety and effectiveness of nurse telephone consultation in out of hours primary care: randomised controlled trial. BMJ, 317(7165), 1054-1059. Lattimer, V., Sassi, F., George, S., Moore, M., Turnbull, J., Mullee, M., & Smith, H. (2000). Cost analysis of nurse telephone consultation in out of hours primary care: evidence from a randomised controlled trial. BMJ, 320(7241), 1053-1057. Lee, J. (2017, March 3). CDC report: Youth suicide rates in Santa Clara County highest in Palo Alto, Morgan Hill. The Mercury News. Retrieved from: http://www.mercurynews.com. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 80 February 1, 2018 Los Angeles Fire Department (LAFD). (2015). EMS calls increase nearly 13 percent from 2013- 2014. County of Los Angeles Fire Department. March 23, 2015. Accessed January 25, 2017: https://www.fire.lacounty.gov/ems-calls-increase-nearly-13-percent-2013-2014/. Luthern, A. (2014). As medical calls flow in, fire department paramedics respond. Milwaukee Wisconsin Journal Sentinel, September 24, 2014. Accessed January 25, 2017: http://archive.jsonline.com/news/milwaukee/as-medical-calls-flow-in-fire-department- paramedics-respond-b99332685z1-276869941.html. McCorkle, R., Benoliel, J.Q., Donaldson, G., Georgiadou, F., Moinpour, C., Goodell, B. (1989). A randomized clinical trial of home nursing care for lung cancer patients. Cancer, 64: 1375-1382. McKenzie, B., Koerber, W., Fields, A., Benetsky, M., & Rapino, M. (2013). Commuter-adjusted population estimates: ACS 2006-2010. Accessed December 3, 2017: https://www.census.gov/content/dam/Census/library/working- papers/2013/acs/2013_McKenzie_02.pdf. MIHP Resource Center. (2014, September). AMR “PRIME Medic” CHF re-admission reduction program. Accessed January 30, 2017: http://mihpresources.com/programs/amr-prime- medicchf-re-admission-reduction-program/. Myers, B., Racht, E., Tan, D., & White, L. (2012). Mobile integrated healthcare practice: a healthcare delivery strategy to improve access, outcomes, and value. Accessed January 21, 2017: http://media.cygnus.com/files/cygnus/document/EMSR/2013/DEC/medtronic- download-12-9_11273203.pdf. National Conference of State Legislatures (NCSL). (n.d.). Beyond 911: State and community strategies for expanding the primary care role of first responders. Accessed December 29, 2016: http://www.ncsl.org/research/health/expanding-the-primary-care-role-of-first-responder.aspx. National Fire Protection Association (NFPA). (2016). News & research: Fire department calls. Updated September 2016. Accessed January 25, 2017: http://www.nfpa.org/news-and- research/fire-statistics-and-reports/fire-statistics/the-fire-service/fire-department-calls/fire- department-calls. National Institute on Alcohol Abuse and Alcoholism. (1997). Alcohol alert. Retrieved from: https://pubs.niaaa.nih.gov/publications/aa38.htm. National Organization of State Offices of Rural Health (NOSORH). (2010). Discussion paper on development of community paramedic programs. State Perspectives, December 2010, Joint Committee on Rural Emergency Care, National Association of State Emergency Medical Services Officials, National Organization of State Offices of Rural Health. Accessed January 21, 2017: https://nosorh.org/wp-content/uploads/2013/12/State-Perspectives-Discussion-Paper-on- Development-of-Community-Paramedic-Programs.pdf. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 81 February 1, 2018 Nestler, D. M., Fratzke, A. R., Church, C. J., Scanlan-Hanson, L., Sadosty, A. T., Halasy, M. P., Finley, J.L., Boggust, A., Hess, E. P. (2012). Effect of a physician assistant as triage liaison provider on patient throughput in an academic emergency department.” Academic Emergency Medicine: Official Journal of the Society for Academic Emergency Medicine, 19(11), 1235–1241. Patton, G. G., & Thakore, S. (2012). Reducing inappropriate emergency department attendances—a review of ambulance service attendances at a regional teaching hospital in Scotland. Emergency Medicine Journal, 30(6), 459-461. Pearson, K., & Shaler, G. (2015). Maine EMS Community Paramedicine Pilot Program Evaluation. Accessed July 23, 2017: https://www1.maine.gov/ems/documents/cp_muskie_report.pdf. Pennel, C. L., Tamayo, L., Wells, R., & Sunbury, T. (2016). Emergency medical service-based care coordination for three rural communities. Journal of Health Care for the Poor and Underserved, 27(4), 159-180. Pressman, H., & Schneider, J. (2009, August 17). Overcoming communication barriers in emergency situations. Patient Provider Communications. Accessed December 21, 2017: http://www.patientprovidercommunication.org/article_1.htm. Rosin, H. (2015, December). The Silicon Valley suicides: Why are so many kids with bright prospects killing themselves in Palo Alto? The Atlantic. Retrieved from: http://www.theatlantic.com. Ruest, M., Stitchman, A., & Day, C. (2012). Evaluating the impact of 911 calls by an in-home programme with a multidisciplinary team. International Paramedic Practice, 1(4), 125-132. Santa Clara County Public Health Department (SCCPHD). (2014a). 2013-14 behavioral risk factor survey (BRFS) health status quick facts. Retrieved from: http://www.sccgov.org. Santa Clara County Public Health Department (SCCPHD). (2014b). African & African ancestry. Retrieved from: https://www.sccgov.org/sites/opa/nr/Documents/AFRICAN%20ANCESTRY%20REPORT.pdf. Santa Clara County Public Health Department (SCCPHD). (2014c). Partners for health Santa Clara County 2015-2020: Community health assessment Community health improvement plan. Retrieved from: https://www.sccgov.org/sites/phd/collab/chip/Documents/cha-chip/cha- chip.pdf. Santa Clara County Public Health Department (SCCPHD). (2016). Neighborhood profiles. Retrieved from: https://www.sccgov.org/sites/sccphd/en-us/Partners/Data/Pages/San- Jose.aspx. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 82 February 1, 2018 Seow, H., Brazil, K., Sussman, J., Pereira, J., Marshall, D., Austin, P. C., Husain, A., Rangrej, J., & Barbera, L. (2014). Impact of community based, specialist palliative care teams on hospitalisations and emergency department visits late in life and hospital deaths: a pooled analysis. BMJ, 348, g3496. Snohomish Co. Fire District 1. (2017). Community Paramedic. Accessed July 23, 2017: http://www.firedistrict1.org/our-services/ems/community-paramedic. Stanford Health Care. (2016.) Community Health Needs Assessment. Retrieved from: https://stanfordhealthcare.org/content/dam/SHC/about-us/public-services-and-community- partnerships/docs/FINAL-SHC-2016-CHNA-with-attachments.pdf. Stanford University. (2016). Stanford University 2018 general use permit application, transmitted November 21, 2016. Accessed December 13, 2016: https://gup.stanford.edu/the- project/2018-gup-application. Tadros, A. S., Castillo, E. M., Chan, T. C., Jensen, A. M., Patel, E., Watts, K., & Dunford, J. V. (2012). Effects of an emergency medical services-based resource access program on frequent users of health services. Prehospital Emergency Care, 16(4), 541-547. Thorp, V. (2014). Palo Alto finances – Where does the money come from and where does it go? Palo Alto Pulse, November 1, 2014. Accessed April 26, 2017: http://www.paloaltopulse.com/2014/11/01/pulse-asks-city-of-palo-alto-finances-where-does- the-money-come-from-and-where-does-it-go/. Tindall, S.C. (1990). Level of consciousness. In Clinical Methods: The History, Physical, and Laboratory Examinations. 3rd edition, Ch. 57. Walker, H.K., Hall, W.D., & Hurst, J.W., editors. Butterworth Publishers: Boston, MA. Accessed May 26, 2017: https://www.ncbi.nlm.nih.gov/books/NBK380/. U.S. Census Bureau. (2013). Commuting (Journey to Work). Retrieved from: https://www.census.gov/hhes/commuting/data/calculations.html. U.S. Census Bureau. (2014). Selected demographic estimates, 2010-2014 American Community Survey 5-year estimates. Retrieved from: https://factfinder.census.gov/faces/nav/jsf/pages/index.xhtml. U.S. Census Bureau. (2015). Selected demographic estimates, 2011-2015 American Community Survey 5-year estimates. Retrieved from: https://factfinder.census.gov/faces/nav/jsf/pages/index.xhtml. U.S. Department of Veterans Affairs. (2017). National center for veterans analysis and statistics [database]. Retrieved from: https://www.va.gov/vetdata/. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 83 February 1, 2018 University of California Center for Health Policy Research. 2012. California Health interview survey (CHIS). 2011-12. Walsh, M., & Little, S. (2001). Study of a nurse practitioner working in a paramedic role. Emergency Nurse: The Journal of the RCN Accident and Emergency Nursing Association, 9(6), 11-14. Wang, H. (2011). Community paramedicine. Accessed January 21, 2017: http://communityparamedic.ca/site/media/download_gallery/comm%20paramed.pdf. Western Eagle County Ambulance District (WECAD) (2011). Community Paramedic Program Handbook. Western Eagle County Health Services District, Version 1.2 (Fall 2011). Retrieved from: http://communityparamedic.org/Program-Handbook. Whitley, E. M. & Everhart, R. M. & Wright, R. A. (2006). Measuring return on investment of outreach by Community Health Workers. Journal of Health Care for the Poor and Underserved, 17(1), 6-15. Project MUSE, doi:10.1353/hpu.2006.0015. Widiatmoko, D., Machen, I., Dickinson, A., Williams, J., & Kendall, S. (2008). Developing a new response to non-urgent emergency calls: Evaluation of a nurse and paramedic partnership intervention. Primary Health Care Research & Development, 9(3), 183-190. Zarchi, K., Haugaard, V. B., Dufour, D. N., & Jemec, G. B. (2015). Expert advice provided through telemedicine improves healing of chronic wounds: prospective cluster controlled study. The Journal of investigative dermatology, 135(3), 895. Zavadsky, M., Hagen, T., Hinchey, P., McGinnis, K., Bourn, S., & Myers, B. (2015). Mobile Integrated Healthcare and Community Paramedicine (MIH-CP). National Association of Emergency Medical Technicians. Accessed January 30, 2017: https://www.naemt.org/docs/default-source/community-paramedicine/naemt-mih-cp- report.pdf?sfvrsn=4. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 84 February 1, 2018 Appendix 2 Methods After consultation with several statistical experts (Kristi Kelly, Ph.D., personal communication, May 10, 2016; Professor Stefanie Bailey Möllborn, Ph.D., personal communication, November 28, 2016; Christina Branom, Ph.D., personal communication, February 17, 2017), Actionable Insights determined that the most appropriate method for predicting future unit call volume/type and proportion of emergency medical interactions by age group, gender, incident location, and/or primary impression was a modified life table approach. This is a population ecology approach used to describe a current population and predict future growth or decline (see, for example, Begon et al. 1990; Begon et al. 1996; Krebs, 1985). The U.S. Centers for Disease Control calls this a period life table approach (Arias, 2012). We use a snapshot of the years 2015/2016 and assume that the cohort of individuals alive at that time in the combined Palo Alto/Stanford jurisdiction, plus the expected increase in individuals in the jurisdiction over the next 15 years, will be subject (proportionally) to the age-specific call/patient volumes, call types, locations, and primary impressions that prevailed for the actual population in 2015/2016. The greatest factor that would affect construction of period life tables for 2030 is the expected increase in the older population in the City of Palo Alto. Based on a preliminary analysis of the ESO data, a simple linear regression in which we predict number of medical emergency patients by patient age in combined years 2015 and 2016 shows that the relationship of the number of patients by patient age is positive (1.033) and statistically significant (p < 0.001). That is, there are more older individuals represented in the ESO dataset than younger individuals, to a statistically significant degree. This supports the hypothesis that older individuals are more likely than younger individuals to have a medical emergency in the combined Palo Alto/Stanford jurisdiction. After predicting expected 2030 call volume by call type, medical emergency call/patient volume by age group, and patient volume by location and primary impression, we adjusted medical emergency predictions based on CA DOF (2017) age projections. More information about this process may be found in Appendices 5 and 6. We provide separate predictions for the City of Palo Alto and for Stanford.44 Another factor one would expect to affect construction of period life tables for 2030 is gender. While gender is associated with number of medical emergency (ME) patients overall (with females somewhat more likely to be an ME patient than males), the gender composition of the combined Palo Alto/Stanford jurisdiction is not expected to change significantly within the next 15 years (see footnote 9). Thus, we did not take gender into account in generating the period 44 Note that we do not apply these age adjustments to Stanford’s daytime population due to the lack of granularity in age-range data for Stanford employees. See Appendix 5 for more information. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 85 February 1, 2018 life tables predicting call type, call/patient volume, or call/patient volume by location or primary impression in 2030. Although it is possible that the proportions of the population of the various ZIP Code areas in the combined Palo Alto/Stanford jurisdiction may change within the next 15 years, no data were available to address population change in such a fine-grained way (i.e., by ZIP Code). Therefore, we did not develop period life tables predicting patient volume in 2030 by ZIP Code of interaction. Finally, we provide call/patient volume and call type projections both overall, and separately for daytime and nighttime. As suggested by Bhadur (2007), daytime and nighttime populations are fundamentally different, with daytime population including workers, tourists, and business travelers, as well as the residual nighttime resident population that does not leave the area during the day. For PAFD’s purposes, it is important to predict year 2030 call/patient volume and call type, as well as ME patient volume by location and primary impression, based on these two very different daytime and nighttime populations. Again, we provide separate predictions for the City of Palo Alto and for Stanford. Further information on how these populations were constructed is available in Appendix 6. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 86 February 1, 2018 Appendix 3 Mapping of Primary Impressions to Categories Table 8, Total Medical Emergency Patient Volume by Primary Impression by Year: Percentage and Number45 Primary Impression Category # of Patients in 2015 # of Patients in 2016 Abdominal Pain Abdominal Discomfort < 1% 1.1% (59) Abdominal Pain/Discomfort Abdominal Discomfort < 1% None Abdominal Pain/Problems Abdominal Discomfort 6.4% (342) 5.6% (301) Acute Appendicitis Abdominal Discomfort None < 1% Acute Bronchitis Respiratory- Related Issues None < 1% Acute Respiratory Distress Respiratory- Related Issues < 1% < 1% Acute Respiratory Distress (Dyspnea) Respiratory- Related Issues None < 1% Airway Obstruction Respiratory- Related Issues < 1% < 1% Alcohol Behavioral Health Issues < 1% None Alcohol Use Behavioral Health Issues < 1% < 1% Allergic Reaction All Other 1.7% (89) 1.8% (100) Altered Level of Consciousness Altered Level of Consciousness 8.3% (443) 6.8% (364) Altered Mental Status Altered Level of Consciousness < 1% < 1% Asthma Respiratory- Related Issues < 1% < 1% Back Pain Non-Specific Pain < 1% < 1% Behavioral/Psychiatric Disorder Behavioral Health Issues 1.9% (103) 2.5% (135) 45 City of Palo Alto and Stanford only. A total of 234 cases across the two years (<3%) are from other ZIP Codes. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 87 February 1, 2018 Bowel Obstruction Abdominal Discomfort < 1% < 1% Cardiac Arrest Cardiac-Related Issues < 1% < 1% Cardiac Arrhythmia/Dysrhythmia Cardiac-Related Issues < 1% < 1% Cardiac-Related Issues Cardiac-Related Issues < 1% < 1% Cardiac Rhythm Disturbance Cardiac-Related Issues 1.8% (99) 1.4% (74) Chest Pain/Discomfort Cardiac-Related Issues 5.7% (309) 5.6% (301) Chest Pain, Other (Non-Cardiac) Non-Specific Pain None < 1% CHF (Congestive Heart Failure) Cardiac-Related Issues < 1% < 1% Common Cold Respiratory- Related Issues < 1% < 1% Concussion Injury/ Hemorrhage None < 1% Confusion/Delirium Altered Level of Consciousness None < 1% Constipation Abdominal Discomfort None < 1% Contact with Venomous Animal All Other < 1% None Dehydration Altered Level of Consciousness < 1% < 1% Diabetic Hyperglycemia Altered Level of Consciousness < 1% < 1% Diabetic Hypoglycemia Altered Level of Consciousness < 1% < 1% Diabetic Symptoms Altered Level of Consciousness < 1% < 1% Diarrhea Abdominal Discomfort < 1% < 1% Electrocution All Other < 1% < 1% Epistaxis Injury/ Hemorrhage < 1% < 1% Esophageal Obstruction Respiratory- Related Issues None < 1% Eye Pain Non-Specific Pain None < 1% PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 88 February 1, 2018 Fatigue Altered Level of Consciousness None < 1% Fever All Other < 1% < 1% Foreign Body in Alimentary Tract Abdominal Discomfort None < 1% Foreign Body in Ear All Other None < 1% Foreign Body in Respiratory Tract Respiratory- Related Issues None < 1% G.I. Bleed Injury/ Hemorrhage < 1% < 1% Gastro-esophageal Reflux Disease (GERD) Abdominal Discomfort None < 1% Gastrointestinal Hemorrhage Injury/ Hemorrhage < 1% < 1% Generalized Weakness Altered Level of Consciousness 7.3% (390) 8.5% (458) Hallucinogen-Related Disorders Behavioral Health Issues None < 1% Headache Non-Specific Pain 1.0% (53) 1.0% (51) Hemorrhage Injury/ Hemorrhage < 1% < 1% Hypertension Cardiac-Related Issues < 1% 1.2% (64) Hyperthermia Altered Level of Consciousness < 1% < 1% Hyperventilation Respiratory- Related Issues None < 1% Hypotension Cardiac-Related Issues < 1% < 1% Hypothermia Altered Level of Consciousness < 1% < 1% Hypovolemia Cardiac-Related Issues None < 1% Hypovolemia/Shock Cardiac-Related Issues < 1% < 1% Influenza Respiratory- Related Issues < 1% < 1% Inhalation Injury (Toxic Gas) Respiratory- Related Issues < 1% < 1% Injury Injury/ Hemorrhage 2.1% (112) 3.1% (169) PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 89 February 1, 2018 Injury of Head Injury/ Hemorrhage None < 1% Injury of Lower Back Injury/ Hemorrhage < 1% < 1% Injury of Pelvis Injury/ Hemorrhage None < 1% Injury of Thorax (Upper Chest) Injury/ Hemorrhage None < 1% Malaise Altered Level of Consciousness < 1% < 1% Mental Disorder Behavioral Health Issues < 1% < 1% Migraine Non-Specific Pain < 1% < 1% Nausea Abdominal Discomfort < 1% < 1% Obvious Death All Other < 1% < 1% Opioid-Related Disorders Behavioral Health Issues None < 1% Other Stimulant-Related Disorders Behavioral Health Issues None < 1% Pain (Non-Traumatic) Non-Specific Pain 7.4% (392) 6.7% (360) Pelvic and Perineal Pain Non-Specific Pain < 1% < 1% Pneumothorax Respiratory- Related Issues None < 1% Poisoning/Drug Ingestion Behavioral Health Issues 1.3% (68) 1.2% (64) Pregnancy/OB Delivery All Other < 1% < 1% Pregnancy-Related Conditions All Other None < 1% Pulmonary Edema, Acute Respiratory- Related Issues None < 1% Pulmonary Embolism Respiratory- Related Issues None < 1% Renal Failure All Other < 1% < 1% Respiratory Arrest Respiratory- Related Issues < 1% < 1% Respiratory Distress Respiratory- Related Issues 4.6% (246) 4.3% (230) PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 90 February 1, 2018 Respiratory Failure Respiratory- Related Issues None < 1% Respiratory-Related Issues Respiratory- Related Issues < 1% < 1% Seizure(s) Altered Level of Consciousness 2.3% (126) 2.6% (145) Seizures Without Status Epilepticus Altered Level of Consciousness None < 1% Sepsis/Septicemia Altered Level of Consciousness < 1% < 1% Sexual Assault/Rape All Other None < 1% Smoke Inhalation Respiratory- Related Issues None < 1% Stings/Venomous Bites All Other < 1% < 1% Stroke Altered Level of Consciousness < 1% < 1% Stroke/Cerebrovascular Accident (CVA) Altered Level of Consciousness 2.3% (126) 2.1% (110) Substance/Drug Abuse Behavioral Health Issues 1.2% (66) < 1% Syncope/Fainting Altered Level of Consciousness 6.1% (327) 5.9% (316) Transient Cerebral Ischemic Attack (TIA) Altered Level of Consciousness < 1% < 1% Trauma Injury/ Hemorrhage < 1% None Traumatic Circulatory Arrest Cardiac-Related Issues < 1% < 1% Traumatic Injury Injury/ Hemorrhage 21.4% (1,147) 17.9% (962) Vaginal Hemorrhage Injury/ Hemorrhage < 1% < 1% Vertigo Altered Level of Consciousness < 1% < 1% Visual Disturbance All Other < 1% < 1% Vomiting Abdominal Discomfort < 1% < 1% No Complaints or Injury/Illness Noted No Complaints 2.9% (154) 2.5% (140) “Everything Else” All Other 1.7% (89) 1.2% (67) PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 91 February 1, 2018 Missing Missing 1.7% (92) 2.3% (123) TOTAL 5,362 5,365 Source: City of Palo Alto PSD, 2017: ESO unduplicated data 2015/2016. Notes: Primary impressions with more than zero but less than 50 patients are marked as “< 1%” to preserve patient confidentiality. Rounding of averages or predictions may cause minor differences in totals. Primary impression categories developed in consultation with Mary Vizzi, R.N. (ret.), personal communication, May 27, 2017, August 3, 2017). PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 92 February 1, 2018 Appendix 4 Call Volume and Call Type Predictions Call volume and call type predictions for 2030 are based on calendar years 2015 and 2016 data from the computer aided dispatch records system (CAD). We averaged the call counts across years 2015 and 2016 to generate an annual average before making the prediction calculations, in order to base them on more stable figures. Data are broken into daytime and nighttime calls by city; based on discussions with PAFD’s Strategic Operations Manager, daytime is considered to be 8:00 a.m. to 7:59 p.m., and nighttime is considered to be 8:00 p.m. to 7:59 a.m. (Amber Cameron, personal communication, March 18, 2017). The time indicator used was the time of dispatch. Each call type/call volume prediction is based on the proportion of such calls in 2015/2016 to the total daytime or nighttime population in that city in 2014 (see Appendix 5 for a further description of the calculation of current daytime and nighttime city populations and the rationale for using 2014 population with 2015/2016 calls), and increased proportionate to the projected 2030 total daytime or nighttime population for each city. Call type/volume predictions for the City of Palo Alto are made for each of the six 2030 scenarios (see main text, Table 3), as well as for Stanford alone (not applying the six City of Palo Alto scenarios). In order to show the difference between the CAD data and the ESO data, medical emergency call type predictions for 2030 were generated for both datasets. Note that we used ESO data for City of Palo Alto and Stanford ZIP Codes only, in order to align more closely with CAD data (which were only provided for the two cities). The advantage of the ESO data was that patients’ ages were made available in the ESO dataset, and age is considered a primary factor in predicting future medical emergency patient volume. As with the CAD data, we averaged annual patient counts in the ESO dataset across years 2015 and 2016 to obtain more stable figures. As described elsewhere in this report, we assigned age to one of 13 groups, to align with U.S. Census Bureau ACS age categories. To generate the medical emergency patient volume predictions, we first generated from the ESO dataset average patient counts for each city by age group and by daytime versus nighttime. Then, for daytime and nighttime populations separately, these average medical emergency patient counts were then divided by the number of individuals in the population who were of that age group, to obtain a ratio of medical emergency patients for each age group. Having the average ESO medical emergency patient counts by age, we were able to apply these to the CAD data by dividing each average ESO medical emergency age group count by the total average ESO medical emergency count and then multiplying by the total average CAD medical emergency calls to obtain estimated average counts and ratios by age group for CAD medical emergency calls. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 93 February 1, 2018 These two sets of medical emergency ratios, ESO and CAD, were then applied to the projected 2030 populations, by city, age group, and time of day, to generate predicted counts of 2030 medical emergency patients and 2030 medical emergency calls. Note that for Stanford, the age adjustments described in this paragraph were applied only to the nighttime population, as the data for the daytime population were not granular enough to determine age ranges for Stanford employees. Information on how the 2030 populations were obtained, including more information on the lack of age-range granularity for Stanford’s data, may be found in Appendix 5. For example, there were an estimated 2,066 individuals age 25-34 in the Stanford nighttime population in 2014, and an average of 41 medical emergency patients in the 2015/2016 EMS dataset for individuals age 25-34 that were found in Stanford ZIP Codes 94305 or 94309 between the hours of 8:00 p.m. and 8:00 a.m. The medical emergency patient ratio for Stanford residents aged 25-34 during the nighttime was thus 0.02 (41 divided by 2,066). PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 94 February 1, 2018 Appendix 5 Population Figures by Age – Current and Predicted Current Figures, Including Populations by Age City of Palo Alto In all cases, “current” figures are 2014 actuals or estimates, to align with the 2014 baseline figures used for the six City of Palo Alto 2030 scenarios.46 To obtain current City of Palo Alto resident figures, we began with the total 2014 base population for the City of Palo Alto (see main text, Table 3, City of Palo Alto Scenarios Table, note 3). We applied to this base number the U.S. Census Bureau American Community Survey (ACS) 2010-2014 percentages of the City of Palo Alto’s residents by age. To obtain current City of Palo Alto nightly (overnight) visitor figures (a conservative estimate of tourists and business travelers), we began with the dollar amount of Transient Occupancy Tax (i.e., hotel tax) reported in 2014 ($14 million), the 12% tax rate at the time, and the average room rate of $202 per night (Thorp, 2014). Dividing $14 million by 12% of $202, then dividing by 365, gave us an estimate of 1,582 occupied rooms per night. We used a conservative estimate of one person per room, giving us an estimated 1,582 overnight visitors in the City of Palo Alto during each 24-hour period.47 We applied to this base number the U.S. Census Bureau ACS 2010-14 percentages of U.S. residents by age.48 To obtain current City of Palo Alto daily visitor figures (i.e., shoppers and others making non- work-related day-trips to Palo Alto), we contacted the City of Palo Alto Chamber of Commerce, which estimated approximately 300,000 visitors in the year 2016 (Dawn Billman, personal conversation, March 26, 2017).49 We estimated the total population of the City of Palo Alto in 2016 based on the net increase between the 2014 and 2015 City of Palo Alto populations as provided by the U.S. Census Bureau (2014; 2015),50 then adjusted downward the estimated 46 PAFD’s CAD and ESO data systems were not implemented until March of 2014, and confidence by PAFD in the 2014 CAD and ESO data quality was not as strong as in the 2015 and 2016 data (Amber Cameron, Judy Maloney, & Kim Roderick, personal communication, December 6, 2016); thus, we received 2015/2016 data and employed averages from the 2015 and 2016 combined datasets to obtain more stable statistics related to call volume, call type, and medical emergency interactions in connection with the 2014 population figures. 47 We realize that there are informal options for overnight guests (e.g., staying with friends/relatives, using AirBnB or other services not formally subject to the hotel tax); again, this is a conservative estimate of overnight visitors to the City of Palo Alto. 48 While it is likely that some visitors are from outside the U.S., we could not estimate what proportion of total visitors these might be, and therefore applied only U.S. age percentages. Note that visitors are a very small proportion of total population numbers. 49 While we attempted to obtain shopper statistics from Stanford Shopping Center (located within the boundaries of the City of Palo Alto), the local mall management office claimed it did not track such figures and its Chicag0-based headquarters (Simon Property Group) did not respond to requests for information. 50 Note that 2016 population estimates were not yet available at the time of these analyses. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 95 February 1, 2018 2016 annual number of visitors using the total 2014 base population for the City of Palo Alto (see main text, Table 3, note 3). We applied to this base number the U.S. Census Bureau ACS 2010-14 percentages of U.S. residents by age.48 Stanford To obtain current Stanford resident figures, we employed the U.S. Census Bureau ACS 2010- 2014 estimates and applied percentages of Stanford residents by age from the same source. To obtain current Stanford daily visitor figures (i.e., tourists and others making non-work day- trips to Stanford), we contacted Stanford University’s Visitor Center, which estimated approximately 150,000 visitors in the year 2016 (D.J. Dull-MacKenzie, personal conversation, April 26, 2017). We estimated the total population of Stanford in 2016 based on one quarter of the difference between the 2015 and 2018 estimated Stanford resident populations as provided by the Stanford University 2018 General Use Permit Application (GUP) (Stanford University, 2016), then adjusted downward the estimated 2016 annual number of visitors using the U.S. Census Bureau ACS 2010-2014 population for Stanford. We applied to this base number the U.S. Census Bureau ACS 2010-14 percentages of U.S. residents by age.48,51 Year 2030 Predictions, Including Predictions by Age City of Palo Alto Nighttime Population To predict 2030 City of Palo Alto residents, we took the 2014 base population as provided in Table 3, note 3, and added to it, for each of the six scenarios, the predicted net change in population indicated in the City of Palo Alto Scenarios (Table 3). To predict 2030 City of Palo Alto overnight visitors, we took the 2014 overnight visitors estimate and adjusted it upward by applying, to the 2014 estimate, the ratios of the 2014 base population to the projected 2030 resident population figures associated with the six scenarios (2030 projected resident population figures calculated as described in the bullet above). We applied to the estimates of the six City of Palo Alto resident scenarios CA DOF projected 2030 percentages of San Mateo County residents by age (CA DOF, 2017). Although the City of Palo Alto is technically in Santa Clara County, it is just south of the San Mateo County line; we use San Mateo County 2030 projected age percentages because (a) neither the DOF nor any other agency provides age projections by city (Stephen Levy, personal communication, December 14, 2016), and (b) the latest available population age distribution of the City of Palo 51 There are no hotels within the city limits of Stanford, no record of the number of informal overnight guests on the campus/within the city limits, and no clear way to determine this number; thus, we did not estimate figures for overnight visitors to Stanford. Note that daytime visitors are a very small proportion of total population numbers, and it might be expected that nighttime visitors are similarly a small proportion compared to the total population. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 96 February 1, 2018 Alto was more like the population age distribution of San Mateo County than it was like the population age distribution of Santa Clara County (U.S. Census Bureau, 2015). For overnight visitors to the City of Palo Alto, we applied to the various 2030 estimates U.S. Census 2030 projections of percentages of U.S. residents by age (Colby & Ortman, 2015).48 City of Palo Alto nighttime population estimates by age thus comprise age-adjusted numbers of residents and overnight visitors for each of the six 2030 resident scenarios. City of Palo Alto Daytime Population To predict 2030 City of Palo Alto residents, we took the 2014 base population from Table 3, note 3, and added to it, for each of the six scenarios, the predicted net change in population indicated Table 3. To predict 2030 City of Palo Alto jobs, we took the 2014 base number of jobs as provided in Table 3, note 3, and added to it, for each of the six scenarios, the predicted net change in jobs indicated in Table 3. The figures for predicted 2030 Palo Alto employed residents in each scenario were derived by taking the predicted number of jobs and dividing them by the “Jobs/Employed Residents” ratio from Table 3. Finally, using the U.S. Census Bureau’s “Method #2” for calculating daytime population estimates (U.S. Census Bureau, 2013; see also McKenzie et al., 2013), for each 2030 City of Palo Alto scenario we then took the total predicted resident population, plus the total predicted workers working in the city52, minus the total predicted workers living in the city.53 This provided estimated daytime populations for the six 2030 scenarios but did not include visitors. To predict 2030 City of Palo Alto daytime visitors, we took the 2014 daytime visitors estimate and adjusted it upward by applying to the 2014 estimate the ratios of the base 2014 population to the projected resident population figures for the six 2030 scenarios (calculated as described above). We expect that overnight visitors to the City of Palo Alto also enjoy the City of Palo Alto during the day. Thus, the total estimated daytime populations for the six 2030 scenarios included both 52 McKenzie et al. state that “The total number of workers working in an area includes all workers who indicate a specified area as their place of work regardless of where they live” (2013:3). While we do not have the total number of predicted workers who work in the city (such projections are not available on the city level, per Stephen Levy, personal communication, December 14, 2016), we do have the total number of projected jobs in the city. For purposes of this analysis, we use total projected jobs in the city as a proxy for total predicted workers working in the city. 53 McKenzie et al. state that “The total workers living in a specified geography is defined as the number of workers who are also residents. This estimate does not reflect location of work” (2013:3); predicted workers living in the city thus equate to predicted employed residents as provided by the City of Palo Alto 2016 (see Table 3). PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 97 February 1, 2018 daytime visitors and overnight visitors to the City of Palo Alto. For daytime and overnight visitors to the City of Palo Alto, we applied to the various 2030 estimates U.S. Census Bureau 2030 projections of percentages of U.S. residents by age (Colby & Ortman, 2015).48 To predict the City of Palo Alto’s daytime population by age: We retained 39% of the predicted City of Palo Alto employed residents aged 20-64,54 using the proportion of employed respondents who reported that they lived and worked locally, based on the City of Palo Alto Transportation Survey (City of Palo Alto, 2013).55 We treated the remaining proportion of employed residents as out-commuters. We calculated in-commuters to the City of Palo Alto based on the remaining predicted number of jobs, using the proportions of the working population (aged 20-64) in the top three counties identified as City of Palo Alto in-commuters in the Transportation Survey (City of Palo Alto, 2013): Santa Clara County (39%), San Mateo County (27%), and Alameda County (15%). We attributed the rest (“Other”) to California generally (19%). We then had four sets of in-commuters for each of the six 2030 City of Palo Alto scenarios (three counties and California generally). For each set of predicted in- commuters, we applied that county’s projected 2030 percentages of residents by age for the working population (aged 20-64) only, using California state projected percentages by age for the set not representing a specific county (CA DOF, 2017 for counties and state). Therefore, the overall City of Palo Alto 2030 daytime population estimates by age for each of the six scenarios include both remaining residents and expected in-commuters, as well as daytime and overnight visitors, all age-adjusted. Stanford Nighttime Population The GUP provides, in its Tab 5.5 (reproduced below in Table 9), projections of the Stanford resident population for the years 2015 and 2035 (Stanford University, 2016). We calculated the 2030 Stanford resident population by taking the average increase per year between 2018 and 2035, and subtracting 5 years’ worth of that average from the 2035 estimates. The estimated 2030 resident population is equal to the estimated 2030 nighttime population for Stanford.51 For Stanford, dramatic change was not expected by age category for residents, given that most of the residents are students, who are generally likely to be from certain age groups. For age 54 The vast majority of the U.S. labor force is aged 20-64 (Bureau of Labor Statistics, 2016). Even though labor force participation among those aged 65 and older has been increasing due to longer life spans and financial obligations (Kromer & Howard, 2013), we remain conservative in our predictions and use age groups 20-64 to represent the working population in the City of Palo Alto. 55 We note that 39% is consistent with prior reports of Palo Alto residents who work where they live (e.g., this figure was 36% in the 2000 U.S. Census; see Bay Area Census, 2010). PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 98 February 1, 2018 ranges in both current and 2030 Stanford nighttime population projections, we used the U.S. Census Bureau’s estimates of Stanford’s 2014 resident (nighttime) population proportions by age.56 Stanford nighttime population estimates by age are thus comprised of predicted, age-adjusted 2030 residents only. Stanford Daytime Population To obtain the predicted 2030 daytime population: Daytime Resident, Student, and Worker Population: o Overall: The GUP (Stanford University, 2016) provides in its Tab 8 estimates of the daytime population, including both workers and students, for the years 2018 and 2035. These estimates are reproduced below in Table 10. We calculated the 2030 Stanford daytime population by taking the average increase per year between 2018 and 2035, and subtracting 5 years’ worth of that average from the 2035 estimates. We added graduate students’ non-student spouses and faculty/staff “other family members” in the same proportions, and added in children of graduate students (a number that is not projected to change between 2018 and 2035, based on Tab 5.5 of the GUP; see Table 9). o By age: The GUP (Stanford University, 2016) data and estimates are not provided at a granular-enough level to determine commute (i.e., daytime) population ages, despite requests for the same.57 Use of age groups from other geographies (e.g., Santa Clara County, the City of Palo Alto) as proxies for Stanford’s daytime population is inappropriate because Stanford is not like these other areas; it has a much larger young-adult population due to its university. After considerable effort and discussions with PAFD, it was decided that Stanford’s daytime population would not be estimated by age group. Because we do not have granular-enough data on the current Stanford daytime population by age, we cannot predict how many working-age adults will stay on the campus and how many will out-commute, so we treat them all as staying. This means our analysis uses a slightly more liberal estimate of Stanford’s daytime population than might otherwise be generated. Daytime Visitors: To predict 2030 Stanford daytime visitors, we took the 2014 Stanford 56 We used 2014 rather than 2015 data for Stanford because all City of Palo Alto “current” data and the six 2030 scenarios are also based on 2014 figures; this allows for comparisons between Stanford and the City of Palo Alto if desired. 57 Note that a request for age-related information about the Stanford daytime population was made to the authors of the GUP. The request netted some additional information from Stanford University’s Human Resources department, but the information provided was still not enough to allow us to determine age groups for Stanford’s daytime population. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 99 February 1, 2018 daytime visitors estimate and adjusted it upward by applying to the 2014 estimate the ration of the 2014 resident population to the projected 2030 resident population figure. Together, these figures provided the estimated 2030 daytime population for Stanford. Table 9, Stanford Projected Growth in Academic Year Residential Population Through 2035 (from Stanford GUP Tab 5.5) Affiliation Fall 2015 On- Campus Residential Population Housing Units/ Student Beds Added by Fall 2018 and Resulting Residential Population in 2018 Housing Units/ Student Beds Added by Fall 2020 and Resulting Residential Population in 2020 Housing Units/ Student Beds Added under 2018 General Use Permit and Resulting Residential Population in 2035 Undergraduates 6,401 existing beds occupied 216 beds added at Lagunita 0 beds added 1,700 beds added 6,401 undergraduate students living on campus 6,617 undergraduate students living on campus 6,617 undergraduate students living on campus 8,317 undergraduate students living on campus Graduate Students, including Ph.D.s (see note below) 4,892 existing beds occupied 200 beds added at Highland Hall 2,020 beds added at EVGR 900 beds added 5,001 graduate students living on campus plus 644 non-student spouses and 420 children 5,205 graduate students living on campus plus 660 non-student spouses and 420 children 7,265 graduate students living on campus plus 822 non-student spouses and 420 children 8,183 graduate students plus 894 non-student spouses and 420 children Postdoctoral Scholars 28 existing beds occupied 0 units added 0 units added N/A – included with faculty/staff 28 postdocs living on campus 28 postdocs living on campus 28 postdocs living on campus Faculty/Staff 937 existing faculty/staff housing units built 0 units added 0 units added 550 units added 937 faculty/staff living on campus plus 1,471 other family members 937 faculty/staff living on campus plus 1,471 other family members 937 faculty/staff living on campus plus 1,471 other family members 1,515 faculty/staff/ postdocs living on campus plus 2,335 other family members Total 14,902 15,338 17,560 21,664 Source: Reproduced from Stanford University 2018 General Use Permit (Stanford University, 2016), Tab 5, Table 3. Original source: Stanford University Land Use and Environmental Planning Office, in consultation with Stanford University Residential and Dining Enterprises. Original note: Increases in units/beds are not one-to-one increases except with regard to undergraduates. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 100 February 1, 2018 Table 10, Stanford Projected Growth in Worker Population Through 2035 (from Stanford GUP Tab 8) Population Fall 2015 Fall 2015 to Fall 2018 Fall 2018 to Fall 20201 Fall 2018 to Fall 2035 Total Increase Total Increase Total Increase Total Students Undergraduate Students 6,994 91 7,085 0 7,085 1,700 8,785 Graduate Students 9,196 332 9,528 0 9,528 1,200 10,728 Total Students 16,190 423 16,613 0 16,613 2,900 19,513 Stanford Faculty & Staff Faculty 2,959 114 3,073 0 3,073 789 3,862 Staff 8,612 373 8,985 0 8,985 2,438 11,423 Postdoctoral Scholars 2,264 139 2,403 0 2,403 961 3,364 Total Faculty/Staff 13,835 626 14,461 0 14,461 4,188 18,649 Stanford Other Workers Casual 2,080 87 2,167 0 2,167 579 2,746 Contingent 980 41 1,021 0 1,021 273 1,294 Temporary 1,390 58 1,448 0 1,448 387 1,835 Non-Employee Affiliates (Including Non-Matriculated Students) 2,636 111 2,747 0 2,747 733 3,480 Total Stanford Other Workers 7,086 297 7,383 0 7,383 1,971ǂ 9,354 Non-Stanford Workers Third Party Contractors 300 24 324 0 324 72 396 Janitorial Shift Contractors 240 19 259 0 259 57 316 Construction Contractors 1,200 0 1,200 0 1,200 0 1,200 Total Non-Stanford Workers 1,740 43 1,783 0 1,783 129 1,912 Grand Total 38,851 1,389 40,240 0 40,240 9,188ǂ 49,428 Original source: Stanford LUEP & Fehr & Peers, 2016. Original note: (1) Fall 2018 to Fall 2020 Scenario evaluates the change in VMT with the new Escondido Village Graduate Residents (2020 beds). Source: Reproduced from Stanford General Use Plan, Tab 8, Table 3 (Stanford University, 2016). Note: ǂ Total is incorrect; figures copied from original source. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 101 February 1, 2018 Appendix 6 2030 Predicted ME Call/Patient Volume, Based on CAD Calls and ESO ME Patient Volume City of Palo Alto Table 11 below shows the average 2015/16 CAD call volume by call type (in alphabetical order) for the City of Palo Alto. Note that these are based on unique (unduplicated) CAD calls, not unit calls. Table 11, Current 2015/2016 Call Volume, City of Palo Alto CAD Call Type 2015 2016 2015/2016 Average 1056 0 1 0.5 2nd Alarm 1 6 3.5 Accident 245 299 272.0 Airplane Emergency 3 6 4.5 Auto Aid 3 3 3.0 False Alarm 922 985 953.5 Fire 105 105 105.0 Full First Alarm 1 6 3.5 Gas 47 41 44.0 Hazmat 73 72 72.5 Medical Emergency 4,866 4,986 4,926.0 Mutual Aid 2 5 3.5 Service 443 490 466.5 Smoke 148 110 129.0 Suspicious Circumstance 1 0 0.5 Technical Rescue 5 1 3.0 Train-FD 2 3 2.5 Utilities 0 1 0.5 Vegetation Fire 12 11 11.5 Welfare Check 0 3 1.5 Wires 8 16 12.0 CAD Total 6,887 7,150 7,018.5 Source: City of Palo Alto PSD, 2017: CAD unduplicated data 2015/2016. Table 12 below shows the average 2015/16 CAD call volume and the 2030 predicted call volume, by major call type for the City of Palo Alto, taking the six scenarios into account (see PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 102 February 1, 2018 main text, Table 3 for the scenarios). Medical Emergency call volume predictions are age- adjusted. Note that these are based on unique (unduplicated) CAD calls, not unit calls. Table 12, Current 2015/2016 and Predicted 2030 Call Volume, City of Palo Alto CAD Call Type 2015/2016 Average 2030 Predictions Scen. 1 Scen. 2 Scen. 3 Scen. 4 Scen. 5 Scen. 6 Accident 272 309 300 307 316 301 310 False Alarm 954 1,078 1,051 1,078 1,106 1,060 1,101 Fire 105 119 116 119 122 117 121 CAD-ME 4,926 6,230 6,148 6,305 6,476 6,250 6,605 Service 467 526 514 527 541 519 540 Smoke 129 146 142 146 150 143 150 All Other* 166 190 184 188 194 184 190 CAD Total 7,022ǂ 8,598 8,455 8,670 8,905 8,574 9,017 Source: City of Palo Alto PSD, 2017: CAD unduplicated data 2015/2016. Notes: CAD ME call volume predictions are age-adjusted. *“All other” comprises call types that represented < 1% of unique calls in the unduplicated 2015/2016 CAD dataset. ǂRounding of averages may cause minor differences in totals. Table 13 below shows the average daytime 2015/2016 CAD call volume and the 2030 predicted daytime call volume, by major call type for the City of Palo Alto, taking the six scenarios into account. Again, Medical Emergency call volume predictions are age-adjusted. Table 13, Current 2015/2016 and Predicted 2030 Daytime Call Volume, City of Palo Alto CAD Call Type 2015/2016 Average 2030 Predictions Scen. 1 Scen. 2 Scen. 3 Scen. 4 Scen. 5 Scen. 6 Accident 223 255 246 252 259 246 251 False Alarm 637 729 702 720 739 702 716 Fire 70 80 77 79 81 77 78 CAD-ME 3,342 4,241 4,159 4,265 4,381 4,210 4,410 Service 296 338 326 334 343 326 332 Smoke 78 89 85 88 90 85 87 All Other* 132 151 145 149 153 145 148 CAD Total 4,778ǂ 5,883 5,740 5,887 6,046 5,791 6,022 Source: City of Palo Alto PSD, 2017: CAD unduplicated data 2015/2016. Notes: CAD ME call volume predictions are age-adjusted. *“All other” comprises call types that represented < 1% of unique calls in the unduplicated 20152016 CAD dataset. ǂRounding of averages may cause minor differences in totals. Table 14 shows the average nighttime 2015/2016 CAD call volume and the 2030 predicted nighttime call volume, by major call type for the City of Palo Alto, taking the six scenarios into account. Once again, Medical Emergency call volume predictions are age-adjusted. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 103 February 1, 2018 Table 14, Current 2015/2016 and Predicted 2030 Nighttime Call Volume, City of Palo Alto CAD Call Type 2015/2016 Average 2030 Predictions Scen. 1 Scen. 2 Scen. 3 Scen. 4 Scen. 5 Scen. 6 Accident 49 54 54 55 57 55 59 False Alarm 317 349 349 358 367 358 385 Fire 36 39 39 40 41 40 43 CAD-ME 1,584 1,989 1,989 2,040 2,095 2,040 2,195 Service 171 188 188 193 198 193 208 Smoke 52 57 57 58 60 58 63 All Other* 35 39 39 39 41 39 42 CAD Total 2,244ǂ 2,715 2,715 2,783 2,859 2,783 2,995 Source: City of Palo Alto PSD, 2017: CAD unduplicated data 2015/2016. Notes: CAD ME call volume predictions are age-adjusted. *“All other” comprises call types that represented < 1% of unique calls in the unduplicated 2015/2016 CAD dataset. ǂRounding of averages may cause minor differences in totals. Table 15 shows the average 2015/2016 ESO Medical Emergency patient volume and the 2030 predicted patient volume, for daytime compared to nighttime, for the City of Palo Alto, taking the six scenarios into account. Medical Emergency patient volume predictions are age-adjusted. Table 15, Current 2015/2016 and Predicted 2030 Medical Emergency (ME) Patient Volume, City of Palo Alto ESO Patients 2015/2016 Average 2030 Predictions Scen. 1 Scen. 2 Scen. 3 Scen. 4 Scen. 5 Scen. 6 ESO-ME - Daytime 2,971 3,736 3,663 3,757 3,859 3,707 3,883 ESO-ME - Nighttime 1,359 1,707 1,707 1,750 1,798 1,750 1,883 ESO Total 4,330 5,443 5,370 5,507 5,657 5,457 5,766 Source: City of Palo Alto PSD, 2017: ESO unduplicated data 2015/2016. Note: ESO ME patient volume predictions are age-adjusted. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 104 February 1, 2018 Stanford Table 16 below shows the average 2015/2016 CAD call volume by call type (in alphabetical order) for Stanford. Note that these are based on unique (unduplicated) CAD calls, not unit calls. Table 16, Current 2015/2016 Call Volume, Stanford CAD Call Type 2015 2016 2015/2016 Average 1056 None None None 2nd Alarm 0 1 0.5 Accident 30 33 31.5 Airplane Emergency None None None Auto Aid 1 0 0.5 False Alarm 450 409 429.5 Fire 25 22 23.5 Full First Alarm None None None Gas 18 15 16.5 Hazmat 7 8 7.5 Medical Emergency 647 598 622.5 Mutual Aid 2 1 1.5 Service 83 88 85.5 Smoke 20 11 15.5 Suspicious Circumstance None None None Technical Rescue None None None Train-FD None None None Utilities None None None Vegetation Fire 6 5 5.5 Welfare Check None None None Wires 0 2 1.0 CAD Total 1,289 1,193 1,241.0 Source: City of Palo Alto PSD, 2017: CAD unduplicated data 2015/2016. Table 17 below provides the average 2015/2016 CAD call volume and the 2030 predicted call volume, by major call type, for Stanford. Medical Emergency call volume predictions are age- adjusted for nighttime but not daytime, due to the lack of granularity of age-range data on Stanford employees. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 105 February 1, 2018 Table 17, Current 2015/2016 and Predicted 2030 Call Volume, Stanford CAD Call Type 2015/2016 Average 2030 Predictions Accident 32 41 False Alarm 430 582 Fire 24 31 CAD-ME 623 848 Service 86 115 Smoke 16 21 All Other* 33 44 CAD Total 1,244ǂ 1,682 Source: City of Palo Alto PSD, 2017: CAD unduplicated data 2015/2016. Notes: Nighttime CAD ME call volume predictions are age-adjusted, daytime predictions are not. *“All other” comprises call types that represented < 1% of unique calls in the unduplicated 2015/2016 CAD dataset. ǂRounding of averages may cause minor differences in totals. Table 18 below provides the average daytime 2015/2016 CAD call volume and the 2030 predicted daytime call volume, by major call type, for Stanford. Again, Medical Emergency call volume predictions are age-adjusted for nighttime but not daytime, as described above. Table 18, Current 2015/2016 and Predicted 2030 Daytime Call Volume, Stanford CAD Call Type 2015/2016 Average 2030 Predictions Accident 26 32 False Alarm 269 340 Fire 18 23 CAD-ME 370 468 Service 56 71 Smoke 8 9 All Other* 23 30 CAD Total 770 973 Source: City of Palo Alto PSD, 2017: CAD unduplicated data 2015/2016. Notes: Daytime CAD ME call volume predictions are not age-adjusted. *“All other” comprises call types that represented < 1% of unique calls in the unduplicated 2015/2016 CAD dataset. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 106 February 1, 2018 Table 19 below provides the average nighttime 2015/16 CAD call volume and the 2030 predicted nighttime call volume, by major call type, for Stanford. Once again, Medical Emergency call volume predictions are age-adjusted for nighttime but not daytime, as described above. Table 19, Current 2015/2016 and Predicted 2030 Nighttime Call Volume, Stanford CAD Call Type 2015/2016 Average 2030 Predictions Accident 6 9 False Alarm 161 242 Fire 6 8 CAD-ME 253 380 Service 30 44 Smoke 8 12 All Other* 10 14 CAD Total 474 709 Source: City of Palo Alto PSD, 2017: CAD unduplicated data 2015/2016. Note: Nighttime CAD ME call volume predictions are age-adjusted. *“All other” comprises call types that represented < 1% of unique calls in the unduplicated 2015/2016 CAD dataset. Table 20 shows the average 2015/16 ESO Medical Emergency patient volume and the 2030 predicted patient volume, for daytime and nighttime separately, for Stanford. Medical Emergency patient volume predictions are age-adjusted for nighttime but not daytime, due to the lack of granularity of age-range data on Stanford employees. Table 20, Current 2015/2016 and Predicted 2030 Patient Volume, Day versus Night, Stanford ESO Patients 2015/2016 Average 2030 Predictions ESO-ME - Daytime 600 713 ESO-ME - Nighttime 333 500 ESO Total 933 1,213 Source: City of Palo Alto PSD, 2017: ESO unduplicated data 2015/2016. Note: Nighttime ESO ME patient volume predictions are age-adjusted, daytime predictions are not. See report text for explanation. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 107 February 1, 2018 Appendix 7 Total Medical Emergency Patient Volume by Incident ZIP Code by Year: Percentage and Number Table 21, Total Medical Emergency Patient Volume by Incident ZIP Code by Year Incident ZIP Code # of Medical Emergency Patients in 2015 # of Medical Emergency Patients in 2016 72533 < 1% < 1% 93302 < 1% None 94022 < 1% < 1% 94028 < 1% < 1% 94030 None < 1% 94035 < 1% < 1% 94039 < 1% < 1% 94040 < 1% < 1% 94041 < 1% < 1% 94042 < 1% < 1% 94043 < 1% < 1% 94061 < 1% None 94065 None < 1% 94102 < 1% None 94204 None < 1% 94301 44.0% (2,405) 42.3% (2,315) 94302 7.3% (397) 5.5% (302) PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 108 February 1, 2018 94303 11.4% (621) 8.6% (472) 94304 9.9% (539) 15.6% (852) 94305 15.5% (847) 16.0% (878) 94306 8.9% (484) 8.2% (450) 94307 < 1% < 1% 94308 None < 1% 94309 1.3% (69) 1.8% (96) 94310 < 1% < 1% 94930 None < 1% 95050 None < 1% 95301 < 1% None 95305 < 1% None Missing < 1% < 1% TOTAL 5,465 5,478 Source: City of Palo Alto PSD, 2017: ESO unduplicated data 2015/2016. Note: Totals that are less than 1% are not represented, for the sake of individuals’ confidentiality. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 109 February 1, 2018 Appendix 8 Current 2015/2016 Average Patient Volume by Primary Impression and Age Note that the tables below contain only the categories of primary impression that are relatively specific; categories such as “Everything Else” and “No Complaints or Illness/Injury Noted” are not included in these tables. Table 22, Current 2015/2016 Average Patient Volume by Primary Impression and Age, City of Palo Alto Abdominal Pain/ Discomfort Altered Level of Consciousness Behavioral Health Issues Cardiac- Related Issues Injury/ Hemorrhage Non- Specific Pain Respiratory- Related Issues Age 0-4 <10* 16 <10 None 10 <10 14 Age 5-9 <10 <10 <10 <10 24 <10 <10 Age 10-14 <10 14 <10 <10 28 <10 <10 Age 15-19 <10 30 14 <10 50 <10 <10 Age 20-24 12 44 22 <10 62 11 <10 Age 25-34 38 89 52 17 120 31 19 Age 35-44 24 101 34 41 84 41 <10 Age 45-54 39 126 34 64 84 46 19 Age 55-59 35 92 20 33 50 50 23 Age 60-64 28 81 10 27 52 30 20 Age 65-74 34 151 18 93 96 58 40 Age 75-84 39 206 <10 78 122 40 50 Age 85+ 64 332 <10 102 244 66 80 Total 328 1,288 218 468 1,023 388 290 Source: City of Palo Alto PSD, 2017: ESO unduplicated data 2015/2016. Note: Totals that represent fewer than 10 patients are not represented, for the sake of individuals’ confidentiality PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 110 February 1, 2018 Table 23, Current 2015/2016 Average Patient Volume by Primary Impression and Age, Stanford Abdominal Pain/ Discomfort Altered Level of Consciousness Behavioral Health Issues Cardiac- Related Issues Injury/ Hemorrhage Non- Specific Pain Respiratory- Related Issues Age 0-4 <10* <10 <10 <10 <10 <10 <10 Age 5-9 None <10 None None <10 <10 <10 Age 10- 14 <10 <10 <10 <10 <10 <10 <10 Age 15- 19 <10 31 27 <10 36 <10 <10 Age 20- 24 10 44 26 <10 40 <10 <10 Age 25- 34 10 32 16 <10 36 <10 <10 Age 35- 44 <10 16 <10 <10 18 <10 <10 Age 45- 54 10 20 <10 <10 18 <10 <10 Age 55- 59 <10 16 <10 <10 12 <10 <10 Age 60- 64 <10 13 <10 <10 13 <10 <10 Age 65- 74 <10 22 <10 <10 21 <10 <10 Age 75- 84 <10 24 <10 10 24 <10 <10 Age 85+ <10 48 <10 14 36 12 11 Total 66 273 89 58 264 69 45 Source: City of Palo Alto PSD, 2017: ESO unduplicated data 2015/2016. Note: Totals that represent fewer than 10 patients are not represented, for the sake of individuals’ confidentiality. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 111 February 1, 2018 Appendix 9 Current 2015/2016 Average and Predicted 2030 Patient Volume by Primary Impression, by City and Time of Day Note that the tables below contain only the categories of Primary Impression that are relatively specific; categories such as “Everything Else” and “No Complaints or Illness/Injury Noted” are not included in these tables. Tables are ordered alphabetically by Primary Impression. Table 24, Current and Predicted Year 2030 Medical Emergency Patients, Categories of Primary Impressions, by Time of Day, City of Palo Alto Primary Impression Category City of Palo Alto (CoPA) 2015/2016 Average, Daytime Scen. 2 (Lowest) CoPA, 2030 Daytime (% incr.) Scen. 4 (Highest) CoPA, 2030 Daytime (% incr.) City of Palo Alto (CoPA) 2015/2016 Average, Nighttime Scen. 1 & 2 (Lowest) CoPA, 2030 Nighttime (% incr.) Scen. 6 (Highest) CoPA, 2030 Nighttime (% incr.) Abdominal Discomfort 201 244 (21%) 257 (28%) 127 159 (25%) 176 (39%) ALOC 940 1,181 (26%) 1,244 (32%) 350 445 (27%) 491 (40%) Behavioral Health 131 146 (11%) 154 (18%) 89 99 (11%) 110 (24%) Cardiac- Related 321 416 (30%) 438 (36%) 148 191 (29%) 211 (43%) Injury/ Hemorrhage 749 895 (19%) 943 (26%) 277 344 (24%) 379 (37%) Non-Specific Pain 238 289 (21%) 305 (28%) 151 189 (25%) 208 (38%) Respiratory- Related 185 232 (25%) 244 (32%) 106 139 (31%) 153 (44%) TOTAL 2,765 3,403 (23%) 3,585 (30%) 1,248 1,566 (25%) 1,728 (38%) Source: Actionable Insights, LLC, 2017, unpublished data, based on City of Palo Alto PSD, 2017: ESO unduplicated data 2015/201617. Note: All predictions are age-adjusted. Rounding of averages or predictions may cause minor differences in totals. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 112 February 1, 2018 Table 25, Current and Predicted Year 2030 Medical Emergency Patients, Categories of Primary Impressions, by Time of Day, Stanford Primary Impression Category Stanford 2015/2016 Average, Daytime Stanford, 2030 Daytime (% increase) Stanford 2015/2016 Average, Nighttime Stanford 2030 Nighttime (% increase) Abdominal Discomfort 41 52 (27%) 26 38 (46%) ALOC 181 228 (26%) 93 139 (49%) Behavioral Health 30 38 (27%) 60 89 (48%) Cardiac- Related 39 49 (26%) 19 29 (53%) Injury/ Hemorrhage 191 242 (27%) 74 110 (49%) Non-Specific Pain 43 54 (26%) 26 39 (50%) Respiratory- Related 31 39 (26%) 15 22 (47%) TOTAL 556 702 (26%) 313 466 (49%) Source: Actionable Insights, LLC, 2017, unpublished data, based on City of Palo Alto PSD, 2017: ESO unduplicated data 2015/2016. Note: Nighttime predictions are age-adjusted, daytime predictions are not. See report text for explanation. Rounding of averages or predictions may cause minor differences in totals. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 113 February 1, 2018 Appendix 10 Palo Alto/Stanford CHNA Data Dashboard Table 26, Data Employed for Community Health Needs Assessment: Combined Palo Alto/Stanford Area Source58 Indicator Palo Alto/ Stanford Area SC County Trend CA DEMOGRAPHICS CC Change in Total Population 8% 6% 10% SCCPHD Ethnicity: African American 2% 2% SCCPHD Ethnicity: Asian/Pacific Islander 27% 32% CC Ethnicity: Hispanic Population 8% 27% 38% SCCPHD Ethnicity: Latino 6% 27% SCCPHD Ethnicity: White 61% 35% CC Female Population 49% 50% 50% SCCPHD Foreign Language Spoken 39% 52% 58 “CC” stands for Community Commons, 2017. “SCCPHD” stands for Santa Clara County Public Health Department, 2014. “NCS” stands for The 2016 National Citizen SurveyTM (City of Palo Alto, 2017b). “EPIC” stands for California Department of Public Health EpiCenter, 2013. PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 114 February 1, 2018 Source58 Indicator Palo Alto/ Stanford Area SC County Trend CA SCCPHD Foreign-Born 32% 37% SCCPHD Household Size 2.4 2.9 SCCPHD Households with Children 35% 39% CC Male Population 51% 50% 50% CC Population in Limited English Households 5% 11% 9% CC Population with Any Disability 6% 8% 10% CC Population with Limited English Proficiency 11% 21% 19% SCCPHD Single Parent Households 5% 7% CC Population Density 3,033.1 1,447.9 246.6 SCCPHD CalFresh Households 1% 5% CC Children Eligible for Free/Reduced Price Lunch 8% 38% 58% SOCIAL & ECONOMIC FACTORS SCCPHD College Graduate 80% 47% PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 115 February 1, 2018 Source58 Indicator Palo Alto/ Stanford Area SC County Trend CA CC Economic Security - Commute Over 60 Minutes 5% 8% 11% CC Economic Security - Households with No Vehicle 9% 5% 8% SCCPHD Economic Security - Unemployment Proportion 3% 4% CC Economic Security - Unemployment Rate 4.7 3.3 5.0 CC Education - Head Start Program Facilities no data 3.4 6.3 CC Education - Less than High School Diploma (or Equivalent) 3% 13% 18% CC Education - School Enrollment Age 3-4 81% 57% 49% CC Food Security - Food Desert Population 8% 10% 14% CC Food Security - Food Insecurity Rate 12% 12% 15% CC Food Security - Population Receiving SNAP 6% 6% 11% SCCPHD Highest Ed: High School Grad 6% 15% SCCPHD Highest Ed: Some College 11% 24% CC Income Inequality no data 0.47 0.49 PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 116 February 1, 2018 Source58 Indicator Palo Alto/ Stanford Area SC County Trend CA CC Insurance - Population Receiving Medicaid 5% 17% 26% CC Insurance - Uninsured Population 3% 9% 15% CC Lack of Social or Emotional Support county only 22% 25% SCCPHD Median Household Income $ 126,771 $ 93,854 CC Poverty - Children Below 100% FPL 4% 11% 23% SCCPHD Poverty - Children Below 175% FPL 8% 33% SCCPHD Poverty - Families Below 185% FPL 8% 29% CC Poverty - Population Below 100% FPL 7% 10% 16% CC Poverty - Population Below 200% FPL 14% 23% 36% SCCPHD Received Free Food 16%* 11% CC Housing - Cost Burdened Households (by Tract) 35% 38% 44% CC Housing - Substandard Housing (by Tract) 36% 41% 47% CC Housing - Vacant Housing (by Tract) 5% 4% 8% PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 117 February 1, 2018 Source58 Indicator Palo Alto/ Stanford Area SC County Trend CA SCCPHD Lives in Multi-Unit Housing 38% 33% SCCPHD Overcrowded Households 3% 8% NATURAL ENVIRONMENT NCS Air Quality 81% 80% (PA 2006) CC Air Quality - Ozone (O3) 0.0% 0.0% 2.7% CC Air Quality - Particulate Matter 2.5 0.0% 0.0% 0.5% CC Climate and Health - Canopy Cover no data 10% 15% NCS Paths and Walking Trails 76% 74% (PA 2008) SCCPHD Proximity to Open Space (Miles) 0.24 0.29 TRANSIT NCS Carpooled 56% 53% (PA 2014) CC Commute to Work - Alone in Car 58% 76% 73% SCCPHD Commute to Work - Carpooled 7% 10% PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 118 February 1, 2018 Source58 Indicator Palo Alto/ Stanford Area SC County Trend CA SCCPHD Commute to Work - Other 23% 10% SCCPHD Commute to Work - Public Transportation 6% 4% CC Commute to Work - Walking/Biking 23% 4% 4% NCS Drive for Daily Needs 77% NCS Ease of Public Transportation 28% 60% (2006) CC Transit - Public Transit within 0.5 Miles 20.9% 4.43% 15.5% CC Transit - Road Network Density no data 5.23 2.02 NCS Used Public Transportation to Work 53% 50% (PA 2014) NCS Walked or Biked to Work 87% 85% (PA 2014) CC Walking/Biking/Skating to School 48% 48% 43% NCS Walks/Bikes for Daily Needs 21% N/A NCS Health Rating 95% N/A GENERAL HEALTH OUTCOMES PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 119 February 1, 2018 Source58 Indicator Palo Alto/ Stanford Area SC County Trend CA CC Years of Premature Death Rate suppressed 3,721 5,308 CC Poor General Health county only 14% 18% CC Preventable Hospital Events 42.9 58.5 83.2 HEALTH CARE ACCESS & DELIVERY CC Access to Primary Care 103.6 105.9 78.5 CC Federally Qualified Health Centers 1.3 1.6 2.4 NCS Health Care Availability 65% 57% (PA 2006) CC Lack of a Consistent Source of Primary Care county only 12% 14% NCS Preventive Health Services 74% 70% (PA 2008) BEHAVIORAL HEALTH CC Access to Mental Health Providers county only 272.4 280.6 SCCPHD Adult Smokers 3%* 10% CC Alcohol - Excessive Consumption county only 15% 17% CC Alcohol - Expenditures 18% suppressed 13% PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 120 February 1, 2018 Source58 Indicator Palo Alto/ Stanford Area SC County Trend CA CC Depression Among Medicare Beneficiaries 65+ county only 11% 14% CC Needing Mental Health Care county only 14% 16% CC Poor Mental Health Days county only 2.7 3.6 NCS Mental Health Care 46% 63% (PA 2014) CC Mortality - Suicide 10.7 7.9 9.8 CC Tobacco Expenditures 0.5% suppressed 1.0% CC Tobacco Usage county only 10% 13% CANCER CC Breast Cancer Incidence county only 121.5 122.1 CC Cervical Cancer Incidence* county only 5.9 7.7 CC Colon and Rectum Cancer Incidence* county only 38.7 40.0 CC Lung Cancer Incidence county only 41.3 48.0 CC Prostate Cancer Incidence county only 140.6 126.9 PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 121 February 1, 2018 Source58 Indicator Palo Alto/ Stanford Area SC County Trend CA CC Breast Cancer Screening (Mammogram) county only 62% 60% CC Cervical Cancer Screening (Pap) county only 79% 78% CC Colon and Rectum Cancer Screening (Sigmoid/Colonoscopy) county only 65% 58% CC Cancer Mortality 120.4 140.8 157.1 HEART DISEASE & STROKE CC Heart Disease Prevalence county only 5% 6% CC High Blood Pressure - Unmanaged county only 27% 30% CC Ischemic Heart Disease Mortality* county only 118.55 163.18 CC Mortality - Stroke 15.1 27.2 37.4 DIABETES & OBESITY CC Diabetes Hospitalizations 3.1 7.9 10.4 CC Diabetes Management (Hemoglobin A1c Test) 80% 85% 82% CC Diabetes Prevalence county only 8% 8% PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 122 February 1, 2018 Source58 Indicator Palo Alto/ Stanford Area SC County Trend CA CC Obesity (Adult) county only 19% 22% CC Obesity (Youth) no data 15% 19% CC Overweight (Adult) county only 33% 36% CC Overweight (Youth) no data 17% 19% HEALTHY EATING/ACTIVE LIVING SCCPHD Adult Aerobic Physical Activity 75% 58% NCS Affordable Food 59% 62% (PA 2006) SCCPHD Farmers' Markets 1.4 1.6 SCCPHD Fast Food Consumption - Adults 24% 38% SCCPHD Fast Food Proximity 1.5 2.8 NCS Fitness Opportunities 79% 78% (PA 2014) CC Food Environment - Fast Food Restaurants 80.5 78.7 74.5 CC Food Environment - Grocery Stores 17.0 19.0 21.5 PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 123 February 1, 2018 Source58 Indicator Palo Alto/ Stanford Area SC County Trend CA SCCPHD Fruit Consumption - Adults 34% 27% NCS Fruit/Vegetable Consumption 64% N/A CC Fruit/Vegetable Expenditures 14.4% suppressed 14.1% SCCPHD Grocery Stores 0.57 0.56 CC Low Fruit/Vegetable Consumption (Adult) county only 69% 72% CC Low Fruit/Vegetable Consumption (Youth) county only 60% 47% NCS Moderate Physical Activity or Better 67% N/A CC Physical Inactivity (Adult) county only 15% 17% CC Physical Inactivity (Youth) county only 25% 36% NCS Recreational Opportunities 77% 83% (PA 2006) SCCPHD Shops for Fruits/Vegetables Locally 93% 92% CC Soft Drink Expenditures 3.2% suppressed 3.6% PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 124 February 1, 2018 Source58 Indicator Palo Alto/ Stanford Area SC County Trend CA NCS Used Recreation Centers 63% 63% (PA 2006) SCCPHD Vegetable Consumption - Adults 19% 19% NCS Visit Park 93% 93% (PA 2006) INFECTIOUS DISEASE CC STD - Chlamydia county only 329.37 459.2 CC STD - HIV Hospitalizations county only 0.87 1.98 CC STD - HIV Prevalence county only 208.42 376.16 CC STD - No HIV Screening county only 64.0% 60.8% MATERNAL & INFANT HEALTH CC Breastfeeding (Any) county only 97% 93% CC Breastfeeding (Exclusive) county only 77% 65% CC Infant Mortality suppressed 3.5 5 CC Lack of Prenatal Care 0.94% no data 3.1% CC Low Birth Weight 5.7% 6.89% 6.8% PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 125 February 1, 2018 Source58 Indicator Palo Alto/ Stanford Area SC County Trend CA ORAL HEALTH CC Absence of Dental Insurance Coverage county only 36.0% 40.9% CC Dental Care - Lack of Affordability (Youth) county only 4.2% 6.3% CC Dental Care - No Recent Exam (Adult) county only 19% 31% CC Dental Care - No Recent Exam (Youth) county only 30% 19% CC Poor Dental Health county only 8% 11% RESPIRATORY HEALTH CC Asthma - Hospitalizations 4.12 6.57 8.9 CC Asthma - Prevalence county only 14% 14% CC Pneumonia Vaccinations (Age 65+) county only 67% 63% VIOLENCE & INJURY (SAFETY) SCCPHD Bicycle/Vehicle Injuries 83 636 NCS Downtown Safety 74% 67% (PA 2006) NCS Downtown Safety* 92% 91% (PA 2006) PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 126 February 1, 2018 Source58 Indicator Palo Alto/ Stanford Area SC County Trend CA CC Mortality - Homicide 3.50 2.8 5.2 CC Mortality - Motor Vehicle Accident 2.12 4.0 5.2 CC Mortality - Pedestrian Accident* 0.53 1.54 1.97 SCCPHD Motor Vehicle Collisions 459 6,669 NCS Perceived Neighborhood Safety at Night 87% 79% (2006) NCS Perceived Neighborhood Safety during Day 98% 94% (2006) SCCPHD Pedestrian/Vehicle Injuries 30 478 CC Violence - All Violent Crimes county only 262.1 425.0 CC Violence - Assault (Crime) county only 152.2 249.4 CC Violence - Rape (Crime) county only 21.1 21.0 CC Violence - Robbery (Crime) county only 86.3 149.5 EPIC Unintentional Senior Falls Hospitalizations (Crude Rate) county only 1,259.5 1,501.4 EPIC Senior Deaths due to Falls (Crude Rate) county only 55.4 36.1 PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 127 February 1, 2018 Appendix 11 Palo Alto/Stanford Neighborhood Data Dashboard Table 27, Data Employed for Community Health Needs Assessment: Cities of Palo Alto and Stanford (Separately) Neighborhood Desired Direction ↑↓ Santa Clara County City of Palo Alto Stanford Barron Park / Green Acres Downtown North / Crescent Park Midtown North / Palo Verde / Charleston Gardens Midtown South / Ventura / Charleston Meadow Professorville / Old PA / Duveneck / St Francis Population Size N/A 1,781,642 64,403 10,902 14,647 12,094 11,692 14,527 12,755 Race/Ethnicity African American N/A 2% 2% 5% 2% 2% 2% 2% 1% Asian/Pacific Islander N/A 32% 27% 31% 23% 19% 38% 33% 21% Latino N/A 27% 6% 11% 8% 5% 6% 8% 5% White N/A 35% 61% 44% 63% 70% 50% 53% 69% Foreign-Born N/A 37% 32% 29% 32% 28% 34% 37% 25% Speaks Language Other than Eng. at Home N/A 52% 39% 41% 38% 30% 45% 46% 33% Single Parent Households N/A 7% 5% 1%* 5% 4% 4%* 5% 7% Households with Children N/A 39% 35% 12% 35% 24% 44% 37% 37% Average Household Size N/A 2.90 2.41 1.9 2.5 2.1 2.7 2.5 2.5 PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 128 February 1, 2018 Neighborhood Desired Direction ↑↓ Santa Clara County City of Palo Alto Stanford Barron Park / Green Acres Downtown North / Crescent Park Midtown North / Palo Verde / Charleston Gardens Midtown South / Ventura / Charleston Meadow Professorville / Old PA / Duveneck / St Francis Income, Employment, Poverty Median Household Income ($) ↑ 93,854 126,771 33,994 121,332 119,781 156,210 119,932 156,530 Unemployed (Ages ≥ 16 Years) ↓ 4% 3% 6%* 4% 4% 7% 7% 5% Families below 185% FPL ↓ 16% 8% 29% 9% 7%* 4%* 10% 8% Children (Ages 0-17) below 185% FPL ↓ 25% 8% 33%* 12%* 3% 2%* 8%* 12%* Educational Attainment (Ages ≥25) Highest Ed Attained: Less than High School ↓ 13% 3% 2* 5 1* 2* 3* 1* Highest Ed Attained: High School Graduate Relative 15% 6% 2* 7% 6% 4% 6% 5% Highest Ed Attained: Some College/AA Relative 24% 11% 3* 10% 9% 11% 15% 11% Highest Ed Attainment: College Graduate ↑ 47% 80% 94% 79% 84% 82% 76% 84% Transportation N Vehicle-Pedestrian Injury Collisions, 10 Years ↓ 478 30 20 39 94 11 34 35 N Vehicle-Bicycle Injury Collisions, 10 Years ↓ 636 83 68 143 146 39 128 149 N Motor Vehicle Collisions, 1 Year ↓ 6669 459 15 82 83 41 55 86 Lives in Close Proximity to Public Transit ↑ 65 85 96 86 89 59 99 92 PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 129 February 1, 2018 Neighborhood Desired Direction ↑↓ Santa Clara County City of Palo Alto Stanford Barron Park / Green Acres Downtown North / Crescent Park Midtown North / Palo Verde / Charleston Gardens Midtown South / Ventura / Charleston Meadow Professorville / Old PA / Duveneck / St Francis Percentage of Residents Who Commute to Work by Mode Drove Alone ↓ 76 65 11 54 55 73 68 70 Carpooled ↑ 10 7 3 7 6 6 8 6 Public Transportation ↑ 4 6 4 8 6 3 5 5 Other N/A 10 23 82 31 33 18 19 20 Nutrition & Food Households receiving CalFresh Benefits Relative 5% 1% 1* 0* 1* 2* 1* 2* Ave. Miles to Nearest Full-Service Grocery Store ↓ 0.56 0.57 0.75 0.61 0.61 0.61 0.62 0.44 Ave. Miles to Nearest Farmers’ Market ↓ 1.60 1.37 1.00 0.73 1.91 1.77 1.18 1.32 N of Fast Food Outlets per Square Mile ↓ 2.80 1.50 1.50 2.10 7.90 1.20 2.10 1.10 Housing Household Rent >=30% of Household Income ↓ 46 40 69 41 35 36 44 44 Overcrowded Households ↓ 8 3 2* 5* 3 2* 3* 1* Lives in Multi-Unit Housing ↓ 33 38 90 37 60 30 31 24 Average Miles to Nearest Park or Open Space ↓ 0.29 0.24 0.25 0.33 0.21 0.23 0.25 0.24 N of Tobacco Retail Outlets per Square Mile ↓ 3.50 1.20 0.0 1.6 3.4 3.1 2.6 0.5 Average N of Violent Crimes within 1 Mile ↓ 16.04 6.30 1.8 3.1 11.6 1.8 5.1 9.2 N of Alcohol Retail Outlets per Square Mile ↓ 2.7 1.00 0.0 0.7 5.1 3.7 2.1 1.1 PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 130 February 1, 2018 Neighborhood Desired Direction ↑↓ Santa Clara County City of Palo Alto Stanford Barron Park / Green Acres Downtown North / Crescent Park Midtown North / Palo Verde / Charleston Gardens Midtown South / Ventura / Charleston Meadow Professorville / Old PA / Duveneck / St Francis Maternal, Child, & Infant Health Births per 1,000 People N/A 13.50 9.70 7 8 12 9 10 7 Low Birth Weight Infants ↓ 7% 6% 5% 4% 7% 7% 7% 5% Preterm Births ↓ 9% 8% 6% 6% 10% 7% 7% 7% Overweight/Obese in 1st Trimester ↓ 38% 19% 12% 20% 15% 21% 21% 17% Prenatal Care ↑ 74% 85% 83% 85% 87% 85% 82% 88% Teen Live Births ↓ 19.2 2.3 0 2 2 2 3 1 Mortality Life Expectancy ↑ 83.4 87.0 -- 87.2 86.8 87.4 86.8 87.5 Cancer Mortality Rate ↓ 140.3 111.7 -- 109.9 118.0 111.9 116.8 104.1 Heart Disease Mortality Rate ↓ 118.8 82.7 -- 86.9 84.8 72.8 84.1 81.4 Alzheimer’s Disease Mortality Rate ↓ 34.6 32.4 -- 34.0 31.9 28.7 40.0 25.7 Stroke Mortality Rate ↓ 27.8 17.3 -- -- 13.5 -- -- 20.5 Chronic Lower Respiratory Mortality Rate ↓ 25.5 16.7 -- -- 20.6 -- -- -- Unintentional Injury Mortality Rate ↓ 23.4 15.1 -- -- -- -- -- -- Diabetes Mortality Rate ↓ 23.3 7.8 -- -- -- -- -- -- Influenza and Pneumonia Mortality Rate ↓ 14.3 10.2 -- -- -- -- -- -- Hypertension Mortality Rate ↓ 14.6 9.4 -- -- 10.7 -- -- -- PAFD 2017 Report Final Draft © Actionable Insights, LLC Page 131 February 1, 2018 Neighborhood Desired Direction ↑↓ Santa Clara County City of Palo Alto Stanford Barron Park / Green Acres Downtown North / Crescent Park Midtown North / Palo Verde / Charleston Gardens Midtown South / Ventura / Charleston Meadow Professorville / Old PA / Duveneck / St Francis Other Health-Related Currently Uninsured (18-64) ↓ -- -- 4% to 10% 13% to 15% 4% to 10% 4% to 10% 10% to 13% 4% to 10% Current Smoker (18+) ↑ -- -- 3% to 8% 8% to 10% 3% to 8% 3% to 8% 8% to 10% 3% to 8% Obese (18+) ↓ -- 11% 1% to 15% 1% to 15% 1% to 15% 1% to 15% 1% to 15% 1% to 15% Overweight or Obese (12-17) ↓ -- -- 0% to 11% 11% to 17% 0% to 11% 0% to 11% 11% to 17% 0% to 11% Overweight for Age (2-11) ↓ -- -- 0% to 6% 0% to 6% 0% to 6% 0% to 6% 0% to 6% 0% to 6% Received Flu Vaccine (65+) ↓ -- -- 75% to 76% 78% to 84% 75% to 76% 78% to 84% 78% to 84% 76% to 78% Received Flu Vaccine (6m-11) ↓ -- -- 52% to 71% 52% to 71% 52% to 71% 49% to 50% 50% to 52% 50% to 52% Source: Santa Clara County Public Health Department, 2016.