CommunityRx: A Population Health Improvement Innovation That Connects Clinics To Communities

The Affordable Care Act shifts the value proposition in health care delivery from fees generated by delivering medical services to people with illnesses to savings generated by optimizing the physical and mental health of the population served. Physicians ¹ and other caregivers ² recognize the importance of community-based resources for the effective prevention and management of disease, but access to high-quality information about these resources is exceedingly limited. Effective population health management requires an information technology (IT) infrastructure containing high-quality data on community resources that enables health care providers and consumers to address basic, wellness, and disease management needs.

Electronic medication prescribing has established the feasibility of integrating IT infrastructure across sectors to promote population health. It produces data used by the pharmaceutical industry, insurers, and policy makers for drug development, coverage, and quality monitoring. By 2014, 70 percent of physicians and 96 percent of community pharmacies were exchanging data via a single integrated e-prescribing infrastructure. ³ An analogous IT system, similarly supported by policy incentives and a dynamic source of high-quality data on community resources, could enable cross-sector accountability for and optimization of population health.

In 2011 the Center for Medicare and Medicaid Innovation (CMMI) announced the Health Care Innovation Challenge. It called for “the most compelling new ideas to deliver better health and improved care at lower costs,” emphasizing technology-based scalable ideas with the potential to affect underserved populations and create jobs. ⁴ CommunityRx, an innovation that won a Health Care Innovation Award, adapted the Kilbridge e-prescribing model ⁵ (for details about this adaptation, see online Appendix Exhibit 1) ⁶ and an asset-based, community-engaged approach ⁷ to create an IT infrastructure that enabled clinicians to e-prescribe community resources for basic, wellness, and disease self-management needs.

The CommunityRx intervention consisted of the following three components: MAPSCorps, a youth workforce program that conducted an annual census of community resources; community health information specialists who both conducted a survey of community resources and helped participants navigate the resources; and an IT platform for community resource prescribing. The IT platform was interfaced with electronic health record (EHR) systems to generate personalized lists of community resources near patients’ homes that were delivered to patients at the point of care, known as HealtheRx (for a sample HealtheRx, see Appendix Exhibit 2). ⁶

In this article we report on the implementation of the CommunityRx system, examining its process-based outcomes. The application for a Health Care Innovation Award required innovators to project their project’s expected impact and propose a business plan for its sustainability beyond the thirty-six-month funding period. Internal and external evaluations, ⁸ assessing percentage changes in cost per beneficiary per year, are under way.

Study Data And Methods

In the first eight months of the funding period (July 2012–February 2013), we received approval from the University of Chicago Institutional Review Board and developed the workforce and IT components. CommunityRx used continuous quality improvement and rapid-cycle iteration methods. ⁹

Academic medical sites (affiliated with the University of Chicago) and community sites (federally qualified health centers) were recruited to join the CommunityRx system to ensure broad geographic coverage; wide inclusion of people eligible for Medicare, Medicaid, or both; and a diverse set of EHR platforms and practice types (for example, school-based clinics, a senior center, and emergency departments). In the ninth month, the system was launched at one federally qualified health center.

Community health information specialists located at the CommunityRx lead site at the University of Chicago and at community-based organizations across the implementation region supported participants’ navigation of resources and conducted surveys of community service providers in their assigned service areas. By the twenty-eighth month, the implementation region included sixteen contiguous ZIP codes on Chicago’s South and West Sides ( Exhibit 1 ). The region was 106 square miles and had a population of 993,812, fifty-eight percent of whom had a household income below 200 percent of the federal poverty level. It contained the primary service area of the University of Chicago’s medical center. Residents of the region who were seen at participating clinical sites were eligible to participate in CommunityRx. We defined participants as individuals for whom at least one HealtheRx was generated.

SOURCE Authors’ analysis of data for months 9–35 from the CommunityRx database and of population data for 2010–14 from the American Community Survey. NOTES “Percent receiving at least 1 service” refers to the penetration of the system, measured by the estimated percentage of the population residing in each ZIP code that received at least one HealtheRx (a personalized list of community resources near the patient’s home address) in months 9–35 of the 36-month implementation period for CommunityRx. ZIP codes with 0 percent covered were outside the implementation region. Five sites of the Near North Health Service Corporation (a federally qualified health center) located outside the implementation region adopted the CommunityRx system after a physician trained at one site to use CommunityRx “hacked” into the system to make it function at another site where he worked. This unexpected spread prompted us to expand CommunityRx prescribing functionality to all clinical sites affiliated with the Near North Health Service Corporation that provided care for patients who were residents of the implementation region. Twenty-seven of the thirty-three clinical sites were federally qualified health centers (five of the centers were school-based clinics) that were operated by five corporations. The remaining six sites were affiliated with one academic medical center (four outpatient practices, an adult emergency department [ED], and a pediatric ED). Community health information specialists were deployed at partnering community-based organizations and with the CommunityRx lead site, at the University of Chicago.

Community Resource Inventory

MAPSCorps paired local high school students with science-oriented young adults (mostly college students) to conduct an annual “feet-on-the-street” census during the summers of 2013–15 of all operating community service providers—that is, those open to the public. We have previously described the MAPSCorps protocol in detail and reported that the mapping methods yielded community resource data that were far more accurate than widely used secondary sources. ¹⁰ MAPSCorps methods have been replicated in New York City and Niagara Falls, New York, with the support of the New York State Health Foundation; and in Edgecombe and Nash Counties, North Carolina, with the support of the National Heart, Lung, and Blood Institute.

In brief, using the MapApp smartphone application, Chicago MAPSCorps youth walked block by block through the implementation region, gathering data by direct observation about each community service provider, including its name, address, phone number, and primary function and whether it was operating. MAPSCorps data were annually posted to two websites. ^11,12 Youth workers reported their own demographic characteristics and employment histories to researchers via a confidential web-based survey.

In the seventh month, we began a series of telephone surveys of the community service providers identified by MAPSCorps. These surveys, conducted primarily by community health information specialists, collected data about the resources (programs and services) available from each provider, including their eligibility requirements and cost. The surveys prioritized places that were likely to offer resources indicated for the target social and medical conditions (for a list of conditions, see Appendix Exhibit 3). ⁶ Resource data were posted online. ¹³

E-Prescribing Community Resources

The IT infrastructure linked EHR systems at the participating clinical sites to the CommunityRx system. At the end of a patient’s clinical visit, data from his or her EHR ( International Classification of Diseases , Ninth Revision [ICD-9], codes; age; home address; primary language; and other variables) were sent via an automated secure web call to the CommunityRx server. Using an algorithm that combined clinical and public health guidelines with expert opinion, the system generated a HealtheRx for the patient. The HealtheRx arrived within seconds of the secure web call (median time: 2.1 seconds) at a printer in the clinical setting and was posted to the patient’s EHR. The patient received the HealtheRx from a health care provider with a brief explanation (one or two sentences long) that included an orientation to the information on the HealtheRx, including cost and eligibility information, and information about the role of the community health information specialists and how to contact them. In twenty-nine of the thirty-three clinical sites in the CommunityRx system, the HealtheRx printed automatically with the visit summary (such a summary is required to demonstrate meaningful use of health IT); ¹⁴ in the remaining four sites, one extra click was required to print the HealtheRx.

Process measures were obtained for months 9–36 of the CommunityRx implementation period, with the exception of measures requiring EHR data that were available for analysis only through month 35. Analyses of resource referral patterns used data from months 30–35, when the CommunityRx system, including its target conditions and implementation region, was at steady state.

Participant experience was assessed in months 9–36 by a cross-sectional phone survey. About twenty participants a month were recruited for the survey through an advertisement printed on the HealtheRx that offered a monetary incentive. Provider experience was assessed in months 9–36 using an anonymous cross-sectional self-administered survey. Baseline surveys were collected before launch of the CommunityRx system, and follow-up surveys were conducted every six months after launch. Using an item from a 2011 Robert Wood Johnson Foundation survey, ¹ we assessed providers’ attitudes about patients’ social needs. This article reports provider data from eleven sites that participated in the CommunityRx system at baseline and twelve months afterward.

Analyses were conducted using Stata/SE, version 14.0.

Limitations

Our study had several limitations. First, EHR data were available for only nineteen of the thirty-three clinical sites. Data for fourteen of the federally qualified health centers (most of which launched the CommunityRx system in the last five months of the implementation period) were excluded because of acquisition and cost barriers. Excluded sites’ payer mix and patient volume, age, and income were similar to those of the included sites, but the excluded sites had a higher proportion of Hispanic patients. ¹⁵

Second, although EHR data included a unique patient identifier, confidentiality concerns and costs prohibited linking individual-level EHR data across sites. Third, self-reported youth, participant, and provider survey data may be limited by response bias. Finally, provider survey data were limited by variable participation in the survey across sites.

Study Results

Community Resource Inventory

As noted above, the implementation region covered 106 square miles. The youth in MAPSCorps walked an average of 4.6 miles per day and identified 19,589 places that provided community resources. During the implementation of the CommunityRx system, MAPSCorps created 224 paid positions for youth (83 percent identified themselves as non-Hispanic blacks, 9 percent as nonblack Hispanics, and 2 percent as Hispanic blacks; 51 percent identified themselves as female) and 64 paid positions for young adult mentors. For 53 percent of the youth, the position at MAPSCorps was their first paying job.

Four English- and Spanish-speaking community health information specialists were deployed at partnering community-based organizations, and three were deployed with the CommunityRx lead site at the University of Chicago. The specialists used MAPSCorps data to produce an inventory of community service providers that collectively provided 14,914 health-promoting resources (for descriptions of community service providers’ responses to and the resources identified by the resource inventory survey, see Appendix Exhibit 4). ⁶ Of the 19,589 community service providers identified, 5,952 (30 percent) completed a baseline survey.

E-Prescribing Community Resources

The CommunityRx IT system was integrated with the three EHR platforms (Epic, GE Centricity, and NextGen) that were used by one or more of the thirty-three clinical sites. By the twenty-eighth month, the system was generating HealtheRxs for thirty-seven target conditions (wellness for seven different age groups, twenty-three medical conditions, six mental or behavioral health conditions, and homelessness) (for a description of the target conditions and the numbers of HealtheRxs generated for each, see Appendix Exhibit 3). ⁶

More than 1,600 providers (physicians, nurses, and other staff members) were trained to deliver the HealtheRxs. At baseline, 552 of the 644 providers (86 percent) said that they wished the health care system would pay for connecting patients to resources for unmet social needs, and 119 of 642 providers (19 percent; two providers did not answer this question) were “not at all confident” in their ability to meet their patients’ unmet social needs. The proportion of “not at all confident” respondents had fallen to 15 percent at twelve months (27 out of 185 providers) (Appendix Exhibit 5). ⁶

During months 9–35, the CommunityRx system generated 253,479 total HealtheRxs (data not shown), or one or more for each of an estimated 113,295 unique individuals. Overall, the HealtheRxs provided more than eight million community resource referrals. More than 21 percent of the population living in three of the sixteen implementation ZIP codes received at least one HealtheRx ( Exhibit 1 ).

Of the 458 participants we surveyed, 71 percent found places listed on their HealtheRx that they had not known were in their community, and 19 percent reported going to a place they learned about because it was on their HealtheRx. Overall satisfaction with the HealtheRx was high: 79 percent of the respondents said that they were very satisfied, and 83 percent found the HealtheRx to be very useful.

In month nineteen, we added a survey item to assess the spread of HealtheRx information from participants to other people. Of the 374 respondents, 183 (49 percent) reported telling others about the HealtheRx. All but one of these 183 people told others something positive about the HealtheRx; the exception reported making a neutral comment about it.

When we looked at information from the nineteen clinical sites for which complete steady-state data (information for months 30–35) were available, we found that an estimated 49,655 unique participants had received a HealtheRx ( Exhibit 2 ). Of these, on average, 11,848 were referred per month to one or more of 4,646 resources at 1,935 community service providers (Appendix Exhibit 6). ⁶

Exhibit 3 shows the most commonly referred resource types. We estimated the potential strain on these resources by the average number of participants referred per month divided by the number of resources. The most strained resources were the following: classes on smoking cessation (1,001 participants referred per month per resource); pest control (497); mold assessment, removal, or both (434); warming and cooling centers (295); help paying mortgage or rent (209); and classes on weight loss (205). Resource inventory surveys were conducted throughout months 6–36 in the first ten ZIP codes in the implementation region ( Exhibit 4 and Appendix Exhibit 7). ⁶

During months 9–36, the community health information specialists received 888 requests (from fewer than 1 percent of all participants) for assistance accessing community resources (631 requests were made by phone, 193 by text, and 64 by e-mail or in person) (Appendix Exhibit 8). ⁶ In months 31–36, at sites affiliated with one federally qualified health center partner, participants could elect to receive text messages from community health information specialists. Among the 1,448 participants who received a text message from a specialist along with receipt of the HealtheRx, the engagement rate increased from 0.2 percent to 14 percent. Two crisis requests (one related to thoughts of suicide, the other to child safety) that required physician involvement were promptly resolved. Seventy-one percent of the services that participants requested were available at their medical homes.

In the final quarter of the implementation period, clinical partners were given the option to have the CommunityRx system removed from their EHRs. Every partnering organization, with the exception of one that was planning to switch to a new EHR vendor, opted to continue use. To paraphrase one internist at a federally qualified health center, “CommunityRx is now woven into the fabric of what we do.”

Discussion

In 2011 the CMMI issued its first call for applicants for Health Care Innovation Awards, described by the White House as “a call to action for technology and data innovators to team with care innovators to propose solutions that can be implemented with speed and have the opportunity to scale.” ⁴ The CommunityRx system, built in partnership with the Office of the National Coordinator’s Chicago Health Information Technology Regional Extension Center and others, became operational in nine months, and its IT infrastructure was scaled to be integrated with three EHR platforms at thirty-three diverse clinical sites.

More than 113,000 patients and 1,600 clinicians, nearly 6,000 community service providers, and hundreds of young people were engaged in the system. Participant satisfaction was consistently high. All but one clinical partner organization elected to continue operating CommunityRx after funding for the Health Care Innovation Award ended. Most CMMI interventions targeted a specific disease or high-risk subpopulation. In contrast, the CommunityRx system was designed to serve people of all ages who were seen in diverse clinical settings for primary prevention and for the management of a wide range of social and medical conditions.

We adapted the Kilbridge model for prescribing community resources, ⁵ and the activities that make up the fulfillment phase of the model provide a useful framework for analyzing this study’s findings. The fulfillment phase—the period after a participant received a HealtheRx—began with an evaluation activity to orient the patient to the HealtheRx: The clinician reviewed the printed HealtheRx with the patient, including cost and eligibility information, and highlighted the community health information specialist information. Published evidence about patients’ adherence to community resource referrals was very limited when we proposed the CommunityRx innovation. In a study of thirty-nine callers to United Way’s 211 number (used to inform people about local resources), five (12.8 percent) of the callers who were referred to community service providers reported making an appointment. ¹⁶ Based in part on this finding, we estimated that 12 percent of the participants in CommunityRx would contact a community health information specialist or a community service provider after receiving a HealtheRx.

However, the observed rate of engagement with the community health information specialists in our study was very low (less than 1 percent). The specialists hypothesized that the low rate of engagement was attributable to the comprehensiveness of information provided on the HealtheRx—they thought that the HealtheRx provided most participants with all of the information they needed to directly contact community service providers. This explanation was corroborated by the higher-than-expected rate (a self-reported 19 percent) of participants visiting at least one community service provider that they learned about because it was on the HealtheRx.

Dispensing, the second activity in the fulfillment phase of Kilbridge’s adapted model, occurred when the patient contacted or visited a community service provider. The 19 percent community resource “dispense” rate likely underestimates the impact of the CommunityRx intervention on community resource use. Most participant surveys were completed within two weeks of receiving a HealtheRx, so subsequent visits to community service providers were not captured. In addition, the rate did not include participants’ visits to community service providers listed on the HealtheRx that participants already knew about (55 percent of participants said that they had ever been to at least one of the community service providers on their HealtheRx).

Lastly, one-third of participants received two or more HealtheRxs, and a dose-response effect is possible. Text messaging appeared to multiply the effect of the intervention: We found a seventyfold increase in the rate of interaction between community health information specialists and participants (from 0.2 percent in the six months before the texting implementation to 14 percent after it).

The fulfillment phase of the CommunityRx process was complete when a community-based resource was administered (for example, a person participated in an Alcoholics Anonymous meeting or received a donation from a food pantry). Although the CommunityRx model builds on medication e-prescribing, it is important to note that the CommunityRx implementation did not digitally “close the loop” on its e-prescribing functionality. Medication e-prescribing sends an outbound message to the patient’s preferred pharmacy and sends back to the prescriber, via the EHR system, information about patient fulfillment of prescriptions. Fulfillment occurs when the patient obtains the prescription; whether the patient uses the medication is typically ascertained during the patient-provider encounter.

During the CommunityRx implementation study, we were not resourced to build an electronic interface for community service providers to receive referrals. Therefore, we could not track fulfillment beyond participants’ reports. Additional technology and research are needed to more precisely estimate the effect of the HealtheRx on participant fulfillment (especially dispensing and administration of community resources). Few empirical studies—including one randomized controlled trial of clinic-community resource referral interventions—have been published. ^17–19 We found none that describes integration with an EHR system or a closed-loop IT solution.

Our findings suggest that national cross-sector adoption of an IT infrastructure to manage clinic-based referrals to community resources could increase the visibility of community service providers and generate metadata to fill a major void in knowledge about community resources. In our study, 42 percent of participating community service providers had no website, and many requests to community health information specialists were for resources that participants did not know were available at their own medical homes.

Metadata from the CommunityRx implementation revealed variation in demand across regions and wide disparities in the supply of community resources. In the ten ZIP codes with complete resource data, 26 percent of participants and 25 percent of the total population—including 27 percent of people ages sixty-five and older and 40 percent of Hispanic people—were concentrated in the two ZIP codes with the lowest resource density (148–50 resources per 10,000 population, compared to 223–36 resources per 10,000 population in the two ZIP codes with the highest density) (calculations based on data in Exhibit 4 ).

Metadata such as these were shared with community service providers during our study and are needed by communities and policy makers to monitor demand for and equitable access to community resources. By identifying resource gaps and inequities, these data can inform health-promoting policies, community planning, community benefit investment by tax-exempt hospitals, and even entrepreneurial activities.

The CommunityRx system, especially the HealtheRx, acted like a vector, spreading information about community resources well beyond participant-provider encounters. In several ZIP codes the intervention reached more than 21 percent of the total population, rendering the system a powerful tool for communicating health-promoting information to a targeted population. With support from the National Institute on Aging, we are now studying the system’s sustainability by applying a systems science approach to quantify its community-level impact and cost-effectiveness and to project its impact on other communities.

Conclusion

This study demonstrated the feasibility of using IT and workforce innovation to bridge the gap between clinical and other health-promoting sectors in one high-poverty urban region. Building linkages between clinics and the community to address patients’ health-related social needs also generated meaningful work for youth, increased awareness of local health-promoting community service providers, and produced evidence about resource gaps that can be used to inform health-promoting community investments.

The CommunityRx innovation came to life because the Centers for Medicare and Medicaid Services, the nation’s largest health insurer, chose to test innovation and workforce development as a pathway to long-term improvement in population health. Building in part on the CommunityRx project, the CMMI Accountable Health Communities Model will invest $157 million over five years to accelerate the development of cross-sector delivery models for addressing beneficiaries’ health-related social needs. ²⁰ The Health Care Innovation Award anticipated “locally driven innovations.” ²¹ On the South Side of Chicago, whose residents were instrumental in developing and updating CommunityRx for their community’s use, we believe that strategies that succeed in creating an enduring culture of health will be those that center on whole people (instead of on diseases), engage youth, and generate value across sectors by creating meaningful, health-promoting data and jobs. ⁷

ACKNOWLEDGMENTS

CommunityRx data were presented at the Health Information Technology poster session of the AcademyHealth Annual Research Meeting, Boston, Massachusetts, June 26, 2016. This project was supported by the Centers for Medicare and Medicaid Services (CMS) (Grant No. 1C1CMS330997-01-00; Stacy Lindau, principal investigator [PI]); by the National Institute on Aging (Grant No. 1R01 AG 047869-01; Lindau, PI); and by the University of Chicago Institute for Translational Medicine (Grant No. UL1TR000430; Julian Solway, PI), a member of the National Institutes of Health Clinical and Translational Science Awards (CTSA) consortium. Under the terms of the CMS grant, the authors were expected to develop a sustainable business model that would continue and support the model tested after the grant funding ended. Lindau is the founder and owner of a social impact company NowPow, LLC. NowPow is not supported through funding from CMS or other federal funding. She is also president of the Board of Directors of MAPSCorps 501(c)3, a not-for-profit corporation. Together, these two organizations are testing for sustainability a public-private social impact model of the CommunityRx innovation. Neither the University of Chicago nor University of Chicago Medicine is endorsing or promoting any NowPow or MAPSCorps entity or its business, products, or services. The authors gratefully acknowledge Gillian Feldmeth, Chenab Navalkha, Senxi Du, and Kelsey Paradise for research assistance; Elizabeth Tung for data analysis assistance; Kellie Campbell for site leadership and assistance with implementation of CommunityRx at the University of Chicago South Shore Senior Center; Lisa Vinci for site leadership and assistance with implementation of CommunityRx at the University of Chicago Primary Care Group; and Nita K. Lee, Fred Rachman, and S. Margaret Paik for assistance in the design and implementation of CommunityRx. In addition to the many individuals on the clinical and research teams of the authors and contributors to the CommunityRx implementation, the authors also acknowledge the community health information specialists and the following community partnering organizations: Centers for New Horizons, Claretian Associates, Greater Auburn Gresham Development Corporation, and the Washington Park Consortium. The authors are grateful for the collaboration of the partnering community health centers (Alivio Medical Center, Chicago Family Health Center, Esperanza Health Centers, Friend Family Health Center, and Near North Health Service Corporation), Chicago Biomedicine Information Services at University of Chicago Medicine, the Alliance of Chicago Community Health Services (with special thanks to Fred Rachman and Andrew Hamilton), and After School Matters. The authors also acknowledge Shane Desautels, a health literacy consultant; Scott Stern, a UChicagoTech entrepreneur in residence; and Milan Makelarski and Dana Weiner, consultants. The contents of this article are solely the responsibility of the authors and do not necessarily represent the official views of the Department of Health and Human Services or any of its agencies. The contents of this article describe research that was conducted by the authors. Findings may or may not be consistent with or confirmed by the findings of the independent evaluation contractor hired by CMS to evaluate the CommunityRx system.

NOTES

• 1 Robert Wood Johnson Foundation . Health care’s blind side: the overlooked connection between social needs and good health: summary of findings from a survey of America’s physicians . [Internet]. Princeton (NJ) : RWJF ; 2011 Dec [cited 2016 Aug 30 ]. Available from: http://www.rwjf.org/content/dam/farm/reports/surveys_and_polls/2011/rwjf71795 Google Scholar
• 2 Black BS , Johnston D , Rabins PV , Morrison A , Lyketsos C , Samus QM . Unmet needs of community‐residing persons with dementia and their informal caregivers: findings from the maximizing independence at home study . J Amer Geriat Soc . 2013 ; 61 ( 12 ): 2087 – 95 . Crossref, Medline, Google Scholar
• 3 Gabriel MH , Swain M . E-prescribing trends in the United States [Internet]. Washington (DC) : Department of Health and Human Services, Office of the National Coordinator for Health Information Technology ; 2014 Jul [cited 2016 Aug 30 ]. (ONC Data Brief No. 18). Available from: https://www.healthit.gov/sites/default/files/oncdatabriefe-prescribingincreases2014.pdf Google Scholar
• 4 Chopra A . Calling all innovators—health care innovation challenge open for great ideas . WhiteHouse.gov [blog on the Internet]. 2011 Dec 7 [cited 2016 Aug 30 ]. Available from: https://www.whitehouse.gov/blog/2011/12/07/calling-all-innovators-health-care-innovation-challenge-open-great-ideas Google Scholar
• 5 Kilbridge P , Gladysheva K . E-prescribing [Internet]. Oakland (CA) : California Health Care Foundation ; 2001 Nov [cited 2016 Aug 30 ]. Available from: http://www.chcf.org/~/media/MEDIA%20LIBRARY%20Files/PDF/PDF%20E/PDF%20EPrescribing.pdf Google Scholar
• 6 To access the Appendix, click on the Appendix link in the box to the right of the article online.
• 7 Lindau ST , Makelarski JA , Chin MH , Desautels S , Johnson D , Johnson WE , et al. Building community-engaged health research and discovery infrastructure on the South Side of Chicago: science in service to community priorities . Prev Med . 2011 ; 52 ( 3–4 ): 200 – 7 . Medline, Google Scholar
• 8 RTI International . Evaluation of the Health Care Innovation Awards: community resource planning, prevention, and monitoring: second annual report [Internet]. Baltimore (MD) : Center for Medicare and Medicaid Innovation ; 2016 Mar [cited 2016 Aug 30 ]. Available from: https://downloads.cms.gov/files/cmmi/hcia-communityrppm-secondevalrpt.pdf Google Scholar
• 9 Langley GJ , Moen RD , Nolan KM , Nolan TW , Norman CL , Provost LP . The improvement guide: a practical approach to enhancing organizational performance . San Francisco (CA) : Jossey-Bass ; 2009 . Google Scholar
• 10 Makelarski JA , Lindau ST , Fabbre VD , Grogan CM , Sadhu EM , Silverstein JC , et al. Are your asset data as good as you think? Conducting a comprehensive census of built assets to improve urban population health . J Urban Health . 2012 ; 90 ( 4 ): 586 – 601 . Crossref, Google Scholar
• 11 MAPSCorps [home page on the Internet]. Chicago (IL) : MAPSCorps ; c 2016 [cited 2016 Aug 30 ]. Available from: http://www.mapscorps.org Google Scholar
• 12 MAPSCorps [Spanish version of home page on the Internet]. Chicago (IL) : MAPSCorps ; c 2016 [cited 2016 Aug 30 ]. Available from: http://www.dondeesta.org Google Scholar
• 13 HealtheRx [home page on the Internet]. Chicago (IL) : HealtheRx ; c 2013 [cited 2016 Aug 30 ]. Available from: http://www.HealtheRx.org Google Scholar
• 14 Blumenthal D , Tavenner M . The “meaningful use” regulation for electronic health records . N Engl J Med . 2010 ; 363 ( 6 ): 501 – 4 . Crossref, Medline, Google Scholar
• 15 Health Resources and Services Administration . HRSA Health Center Program [Internet]. Rockville (MD) : HRSA ; [cited 2016 Sep 13 ]. Available from: http://bphc.hrsa.gov/uds/datacenter.aspx?q=d&year=2014&state=IL#glist Google Scholar
• 16 Eddens KS , Kreuter MW . Proactive screening for health needs in United Way’s 2-1-1 information and referral service . J Soc Serv Res . 2011 : 37 ( 2 ): 113 – 23 . Crossref, Medline, Google Scholar
• 17 Garg A , Marino M , Vikani AR , Solomon BS . Addressing families’ unmet social needs within pediatric primary care: the Health Leads model . Clin Pediatr (Phila) . 2012 ; 51 ( 12 ): 1191 – 3 . Crossref, Medline, Google Scholar
• 18 Page-Reeves J , Kaufman W , Bleecker M , Norris J , McCalmont K , Ianakieva V , et al. Addressing social determinants of health in a clinic setting: the WellRx Pilot in Albuquerque, New Mexico . J Am Board Fam Med . 2016 ; 29 ( 3 ): 414 – 8 . Crossref, Medline, Google Scholar
• 19 Garg A , Toy S , Tripodis Y , Silverstein M , Freeman E . Addressing social determinants of health at well child care visits: a cluster RCT . Pediatrics . 2015 ; 135 ( 2 ): e296 – 304 . Crossref, Medline, Google Scholar
• 20 Alley DE , Asomugha CN , Conway PH , Sanghavi DM . Accountable health communities—addressing social needs through Medicare and Medicaid . N Engl J Med . 2016 ; 374 ( 1 ): 8 – 11 . Crossref, Medline, Google Scholar
• 21 Center for Medicare and Medicaid Innovation . Health Care Innovation Challenge [Internet]. Baltimore (MD) : CMMI ; 2011 Nov 14 [cited 2016 Aug 30 ]. Available from: https://innovation.cms.gov/files/x/health-care-innovation-challenge-funding-opportunity-announcement.pdf Google Scholar