Social and Behavioral Information in Electronic Health Records New Opportunities for Medicine and Public Health Ana V. Diez Roux, MD, PhD,1 Mitchell Katz, MD,2 Deidra C. Crews, MD, ScM,3,4 David Ross, ScD,5 Nancy Adler, PhD6
A
lthough linked in many ways, medicine and public health have, to a large extent, operated independently, with one emphasizing the treatment of the individual patient and the other emphasizing broader drivers of health outside the patient, prevention rather than treatment, and outcomes for populations rather than individuals. The growing emphasis in many countries toward improving outcomes has caused medicine to increasingly embrace the public health mantra that health cannot be improved without considering the many determinants that lie outside the traditional purview of the healthcare system. This has led to a growing recognition of the importance of the social and behavioral determinants of health, and a rapprochement of sorts between the fields of medicine and public health. The convergence of medicine and public health is reflected in renewed interest among healthcare providers in broadening the focus of medical practice beyond the individual patient to the communities and other contexts in which patients live. This trend has been fueled in part by financial incentives. For example, hospitals facing penalties for patients being readmitted are motivated to consider social and behavioral factors that can affect readmission, such as the ability of patients to take their medication correctly and keep follow-up appointments, as well as the living situation and family support received by patients. Similarly, providers caring for patients under capitated models have strong incentives to address the true cost drivers of health care, including factors such as From the 1Department of Epidemiology and Biostatistics, Drexel University School of Public Health, Philadelphia, Pennsylvania; 2Los Angeles Department of Health Services, Los Angeles, California; 3Division of Nephrology, Department of Medicine, Johns Hopkins University School of Medicine, Baltimore, Maryland; 4Welch Center for Prevention, Epidemiology, and Clinical Research, Johns Hopkins Medical Institutions, Baltimore, Maryland; 5Task Force for Global Health Public Health Informatics Institute, Atlanta, Georgia; and 6Departments of Psychiatry and Pediatrics and the Center for Health and Community, University of California San Francisco, San Francisco, California Address correspondence to: Ana V. Diez Roux, MD, PhD, Drexel University School of Public Health, 3215 Market Street, Philadelphia PA 19104. E-mail: [email protected]. 0749-3797/$36.00 http://dx.doi.org/10.1016/j.amepre.2015.08.027
980 Am J Prev Med 2015;49(6):980–983
obesity, sedentary lifestyle, smoking, and even social circumstances. Another manifestation of the convergence between medicine and public health is the growing use of the term “population health” in the medical field.1 Although sometimes “population health” is used in a narrow sense to refer to the health of panels of patients, in its broadest sense it encompasses the health of an entire community including people not currently in care, or not in care with a particular provider. It explicitly recognizes the key role of factors outside the healthcare system, including social and behavioral determinants, in shaping health and outcomes of clinical care.1,2 Along with improving the experience of care and lowering per capita costs, improving population health is implicitly part of the “triple aim” and the Affordable Care Act, reflecting the recognition that healthcare costs cannot be reduced without a focus on population health.3 The inclusion of social and behavioral information in electronic health records (EHRs) represents an opportunity to improve both clinical care and public health.4,5 Incentives for adoption and “meaningful use” of EHRs have expanded their use and it is now estimated that more than 70% of office-based physicians are using an EHR based on provider reports.6 With adequate attention to privacy and security, EHRs have the potential to collect and store standard measures of social and behavioral determinants, making this information available not only to clinicians but also to patients, health systems, public health organizations, and researchers. Although a growing set of measures have been added to EHRs, most additions have been made in an unsystematic way by individual health organizations. As a result, substantial variation in what is collected makes it difficult to use the information consistently and limits the opportunities for data exchange and interoperability. Prior work has identified the need to include measures of health behaviors and psychosocial factors in EHRs,7 but until now, there have been no standard recommendations for the routine inclusion of social or behavioral information. A committee of IOM, composed of social
& 2015 American Journal of Preventive Medicine
Published by Elsevier Inc.
981
Diez Roux et al / Am J Prev Med 2015;49(6):980–983
scientists, clinicians, public health practitioners, and informatics experts, has recently proposed a set of measures of social and behavioral determinants for inclusion in a standardized manner in EHRs (Table 1). The systematic inclusion of this type of information in EHRs has potential for enhancing both clinical practice and public heath, as well as enriching the connections between both fields. Uses of information on social and behavioral determinants in clinical practice include (1) enabling greater
precision in diagnoses and improved treatment (through, e.g., improved risk stratification and greater attention to patient context in treatment decisions); (2) detecting patients with social or behavioral risk factors that, if addressed, can lower disease burden and improve management and recovery; (3) allowing referrals to community and social service agencies as necessary (which could in turn improve outcomes of clinical care); and (4) improving the capacity of health systems to tailor services to the needs of the population they serve.5
Table 1. Recommended Domains and Measures Domain/measure
Measure
Frequency
Race and ethnicitya
1. What is your race? 2. Are you of Hispanic, Latino, or Spanish origin?
At entry
Education
1. What is the highest level of school you have completed? 2. What is the highest degree you earned?
At entry
Financial resource strain
1. How hard is it for you to pay for the very basics like food, housing, medical care,
Screen and follow up
Stress
1. Stress means a situation in which a person feels tense, restless, nervous, or anxious,
and heat? or is unable to sleep at night because his or her mind is troubled all the time. Do you feel this kind of stress these days?
Depression
Over the past 2 weeks, how often have you been bothered by 1. Little interest or pleasure in doing things? 2. Feeling down, depressed, or hopeless?
Physical activity
1. On average, how many days per week do you engage in moderate to strenuous
exercise (like walking fast, running, jogging, dancing, swimming, biking, or other activities that cause a light or heavy sweat)? 2. On average, how many minutes do you engage in exercise at this level? Tobacco usea
1. Have you smoked at least 100 cigarettes in your entire life?
If yes: 2. Do you now smoke cigarettes every day, some days, or not at all?
Screen and follow up Screen and follow up Screen and follow up
Screen and follow up
Alcohol usea
1. How often do you have a drink containing alcohol? 2. How many standard drinks containing alcohol do you have on a typical day? 3. How often do you have six or more drinks on one occasion?
Screen and follow up
Social connection and social isolation
1. In a typical week, how many times do you talk on the telephone with family, friends,
Screen and follow up
Intimate partner violence
1. Within the last year, have you been humiliated or emotionally abused in other ways
Residential addressa
1. What is your current address?
Verify every visit
Census tract-median income
Geocoded
Update on address change
or neighbors? 2. How often do you get together with friends or relatives? 3. How often do you attend church or religious services? 4. How often do you attend meetings of the clubs or organizations you belong to? by your partner or ex-partner? 2. Within the last year, have you been afraid of your partner or ex-partner? 3. Within the last year, have you been raped or forced to have any kind of sexual activity by your partner or ex-partner? 4. Within the last year, have you been kicked, hit, slapped, or otherwise physically hurt by your partner or ex-partner?
Screen and follow up
Source: IOM. Capturing Social and Behavioral Domains and Measures in Electronic Health Records: Phases 1 and 2. Washington, DC: National Academies Press; 2014. a Domain is already widely included in clinical practice for which systematic measurement was proposed.
December 2015
982
Diez Roux et al / Am J Prev Med 2015;49(6):980–983
for interventions and policies to improve health. Richer data on behavioral and health variations across geography and social groups, including information on health behaviors, prevalent health conditions, and outcomes of care, which is often unavailable or limited in scope in existing surveys or registries, can also be useful to public health agencies in developing and targeting interventions and in identifying potentially important policies. EHR data linked to social and behavioral determinants would also be of great value in policy evaluation, including efforts to systematically evaluate the health and health disparity consequences of policies in domains outside of health care. For example, there is growing interest in evaluating the health and policy implications of transportation policies (such as establishing bike share programs or increasing access to public transportation) or food access policies (such as subsidizing access to fruits and vegetables in poor neighborhoods). The evaluation of these policies requires capitalizing on socalled natural experiments in which policy variations over space and time are linked to appropriate timevarying and geographically varying behavioral or other health outcome data, so that the effect of the policy can be assessed. These data are difficult or impossible to collect de novo in research studies within the timeframe required but could be available from EHRs. Public health research could also benefit from systematically collected information on social and behavioral determinants in EHR. In the narrowest sense, it will allow improved investigation of the role of the social environment in modifying outcomes of clinical care—an important area that has been relatively neglected by the social determiTable 2. Examples of Public Health Uses of Social and Behavioral Data Linked to nants of health field, and that Electronic Health Records presents important opportuniSocial/behavioral factor Sample possible uses ties for productive collaboration between clinicians and public Education, financial strain, and race and Tracking of immunization rates or receipt of health researchers. More ethnicity other preventative services. Characterizing causes of late diagnosis (e.g., broadly, by dramatically of cancer) or poor response to treatment. increasing the availability of Behavioral measures (physical activity, Identifying populations to target for longitudinal health data alcohol use, smoking rates, and fruit and community-based lifestyle interventions. (including laboratory and imagvegetable consumption) Evaluation of the impact of policies (bike and ing data) linked to information walking paths, safe street policies, smoking bans, corner store initiatives). on social determinants, the availability of social and behavDepression and stress Understanding how community violence may ioral data in EHRs will facilitate affect levels of depression and reported stress. etiologic research on the drivers Evaluation of the uptake and success of of population health and the publically available mental health services. causes of health disparities. Geographic identifiers linked to Characterizing geographic disparities and This enriched EHR data environmental data identifying high-risk areas. could be mined for many purUnderstanding the role of modifiable environmental factors (e.g., traffic and asthma poses, including large-scale lonexacerbations). gitudinal investigations of
Importantly, the availability of this information will also allow research and systematic evaluation of the utility of this information in clinical practice and service delivery such that its use enhances care and related outcomes without stigmatizing patients or threatening their privacy. The systematic inclusion of social and behavioral determinants (including routine geocoding and geographic linkages) in EHRs could also have important uses in public health practice and policy. Initiatives are already underway that use EHR data for public health surveillance. In a recent study, Kaiser Permanente used EHR data to identify geographic clusters of underimmunization and vaccine refusal.8 In another example, the Macroscope Project in New York City9 uses EHRs to track conditions managed by primary care practices that are important to public health. The addition of information on social and behavioral determinants and routine geocoding could further expand and increase the informativeness of these efforts. EHR data with their broad coverage and detailed information could be an important complement to existing national surveys in monitoring the health of the nation by allowing surveillance across broader geographic areas, by social groups, and for an expanded set of outcomes. The routine geocoding of EHRs is consonant with the growing interest in enhancing the use of available data to characterize and monitor small area variations in health. Information on small area variations in health can empower communities to raise questions about the causes of these differences and allow them to advocate
www.ajpmonline.org
Diez Roux et al / Am J Prev Med 2015;49(6):980–983
environmental influences on health and outcomes of health care as well as studies of gene–environment interactions. Moreover, the availability of detailed health and healthcare data linked to social and environmental data is likely to spur the development and application of new methodologic tools appropriate for extracting meaning from “big data.”10 The increasing use of these data for research may in turn improve the quality and completeness of the EHR data. Examples of public health uses of this type of data are shown in Table 2. Of course, the use of EHR data for public health research, practice, and action presents many challenges. There are important reasons why the public health field has advocated for the collection of new data for research, surveillance, and policy evaluation. Data from health records are notorious for missing data, non-systematic data collection, and limited generalizability to the populations of interest. There is a history of attempts to use data from health records in research with mixed results.11 Populations that are not frequent users of health care or that have poor access to health care (including many disadvantaged social groups) will not be accurately represented. Additionally, important variables may not be collected. Standardization and implementation of systematic data collection will present important obstacles related to logistics and cost. For example, clinical workers may be less inclined to collect information that they do not view as immediately relevant to their practice. The records can be cumbersome to access because of the data systems used, and their use for things other than the health care of an individual patient raises a host of complex privacy issues that must be addressed. Pilot and demonstration projects will be necessary to validate the information, illustrate how it can best be used for various purposes (including improving clinical practice), and identify what improvements will be necessary before it can be utilized widely and to its full potential. Achieving the full value of social and behavioral information in the EHR requires engagement of both clinicians and the public health community. The participation of public health researchers and practitioners will ensure that methodologic challenges and data quality issues are addressed and that the incorporated and retained measures are useful not only for clinical decision making but also for true population health improvement. Public health engagement will help keep the focus on the broader social determinants and will promote the type of research that will benefit not only the individual patient but also the population as a whole. Although these EHR
December 2015
983
data may never replace other data used in public health research and action, they may provide an economical alternative of rich complementary data to many data collection efforts already underway. The incorporation of social and behavioral information in EHRs and its potential uses in both clinical practice and public health research and action presents renewed opportunities for rich and productive collaborations between medicine and public health in ways that have not been possible in the past. This type of collaboration will both enhance clinical care and improve population heath.
No financial disclosures were reported by the authors of this paper.
References 1. Sharfstein J. The strange journey of population health. Milbank Q. 2014;92(4):640–643. http://dx.doi.org/10.1111/1468-0009.12082. 2. Kindig DA, Asada Y, Booske B. A population health framework for setting national and state health goals. JAMA. 2008;299(17):2081– 2083. http://dx.doi.org/10.1001/jama.299.17.2081. 3. Roundtable on Population Health Improvement; Board on Population Health and Public Health Practice; Institute of Medicine. Populaton Health Implications of the Affordable Care Act. Washington, DC: National Academies Press, 2014. 4. IOM. Capturing Social and Behavioral Domains and Measures in Electronic Health Records: Phases 1 and 2. Washington, DC: National Academies Press, 2014. 5. Adler NE, Stead WW. Patients in context—EHR capture of social and behavioral determinants of health. N Engl J Med. 2015;372(8):698–701. http://dx.doi.org/10.1056/NEJMp1413945. 6. Hsiao CJ, Hing E, Ashman J. Trends in electronic health record system use among office-based physicians: United States, 2007-2012. Natl Health Stat Report. 2014;75:1–18. 7. Estabrooks PA, Boyle M, Emmons KM, et al. Harmonized patientreported data elements in the electronic health record: supporting meaningful use by primary care action on health behaviors and key psychosocial factors. J Am Med Inform Assoc. 2012;19(4):575–582. http://dx.doi.org/10.1136/amiajnl-2011-000576. 8. Lieu TA, Ray GT, Klein NP, Chung C, Kulldorff M. Geographic clusters in underimmunization and vaccine refusal. Pediatrics. 2015;135(2):280–289. http://dx.doi.org/10.1542/peds.2014-2715. 9. New York City Department of Health. The New York City macroscope: developing an electronic health record-based population health surveillance system. www.nyc.gov/html/doh/downloads/pdf/data/nyc-macrosummary.pdf. Published 2014. Accessed 2015. 10. Scruggs SB, Watson K, Su AI, et al. Harnessing the heart of big data. Circ Res. 2015;116(7):1115–1119. http://dx.doi.org/10.1161/CIRCRESAHA. 115.306013. 11. Dean BB, Lam J, Natoli JL, Butler Q, Aguilar D, Nordyke RJ. Review: use of electronic medical records for health outcomes research: a literature review. Med Care Res Rev. 2009;66(6):611–638. http://dx.doi. org/10.1177/1077558709332440.
A
lthough linked in many ways, medicine and public health have, to a large extent, operated independently, with one emphasizing the treatment of the individual patient and the other emphasizing broader drivers of health outside the patient, prevention rather than treatment, and outcomes for populations rather than individuals. The growing emphasis in many countries toward improving outcomes has caused medicine to increasingly embrace the public health mantra that health cannot be improved without considering the many determinants that lie outside the traditional purview of the healthcare system. This has led to a growing recognition of the importance of the social and behavioral determinants of health, and a rapprochement of sorts between the fields of medicine and public health. The convergence of medicine and public health is reflected in renewed interest among healthcare providers in broadening the focus of medical practice beyond the individual patient to the communities and other contexts in which patients live. This trend has been fueled in part by financial incentives. For example, hospitals facing penalties for patients being readmitted are motivated to consider social and behavioral factors that can affect readmission, such as the ability of patients to take their medication correctly and keep follow-up appointments, as well as the living situation and family support received by patients. Similarly, providers caring for patients under capitated models have strong incentives to address the true cost drivers of health care, including factors such as From the 1Department of Epidemiology and Biostatistics, Drexel University School of Public Health, Philadelphia, Pennsylvania; 2Los Angeles Department of Health Services, Los Angeles, California; 3Division of Nephrology, Department of Medicine, Johns Hopkins University School of Medicine, Baltimore, Maryland; 4Welch Center for Prevention, Epidemiology, and Clinical Research, Johns Hopkins Medical Institutions, Baltimore, Maryland; 5Task Force for Global Health Public Health Informatics Institute, Atlanta, Georgia; and 6Departments of Psychiatry and Pediatrics and the Center for Health and Community, University of California San Francisco, San Francisco, California Address correspondence to: Ana V. Diez Roux, MD, PhD, Drexel University School of Public Health, 3215 Market Street, Philadelphia PA 19104. E-mail: [email protected]. 0749-3797/$36.00 http://dx.doi.org/10.1016/j.amepre.2015.08.027
980 Am J Prev Med 2015;49(6):980–983
obesity, sedentary lifestyle, smoking, and even social circumstances. Another manifestation of the convergence between medicine and public health is the growing use of the term “population health” in the medical field.1 Although sometimes “population health” is used in a narrow sense to refer to the health of panels of patients, in its broadest sense it encompasses the health of an entire community including people not currently in care, or not in care with a particular provider. It explicitly recognizes the key role of factors outside the healthcare system, including social and behavioral determinants, in shaping health and outcomes of clinical care.1,2 Along with improving the experience of care and lowering per capita costs, improving population health is implicitly part of the “triple aim” and the Affordable Care Act, reflecting the recognition that healthcare costs cannot be reduced without a focus on population health.3 The inclusion of social and behavioral information in electronic health records (EHRs) represents an opportunity to improve both clinical care and public health.4,5 Incentives for adoption and “meaningful use” of EHRs have expanded their use and it is now estimated that more than 70% of office-based physicians are using an EHR based on provider reports.6 With adequate attention to privacy and security, EHRs have the potential to collect and store standard measures of social and behavioral determinants, making this information available not only to clinicians but also to patients, health systems, public health organizations, and researchers. Although a growing set of measures have been added to EHRs, most additions have been made in an unsystematic way by individual health organizations. As a result, substantial variation in what is collected makes it difficult to use the information consistently and limits the opportunities for data exchange and interoperability. Prior work has identified the need to include measures of health behaviors and psychosocial factors in EHRs,7 but until now, there have been no standard recommendations for the routine inclusion of social or behavioral information. A committee of IOM, composed of social
& 2015 American Journal of Preventive Medicine
Published by Elsevier Inc.
981
Diez Roux et al / Am J Prev Med 2015;49(6):980–983
scientists, clinicians, public health practitioners, and informatics experts, has recently proposed a set of measures of social and behavioral determinants for inclusion in a standardized manner in EHRs (Table 1). The systematic inclusion of this type of information in EHRs has potential for enhancing both clinical practice and public heath, as well as enriching the connections between both fields. Uses of information on social and behavioral determinants in clinical practice include (1) enabling greater
precision in diagnoses and improved treatment (through, e.g., improved risk stratification and greater attention to patient context in treatment decisions); (2) detecting patients with social or behavioral risk factors that, if addressed, can lower disease burden and improve management and recovery; (3) allowing referrals to community and social service agencies as necessary (which could in turn improve outcomes of clinical care); and (4) improving the capacity of health systems to tailor services to the needs of the population they serve.5
Table 1. Recommended Domains and Measures Domain/measure
Measure
Frequency
Race and ethnicitya
1. What is your race? 2. Are you of Hispanic, Latino, or Spanish origin?
At entry
Education
1. What is the highest level of school you have completed? 2. What is the highest degree you earned?
At entry
Financial resource strain
1. How hard is it for you to pay for the very basics like food, housing, medical care,
Screen and follow up
Stress
1. Stress means a situation in which a person feels tense, restless, nervous, or anxious,
and heat? or is unable to sleep at night because his or her mind is troubled all the time. Do you feel this kind of stress these days?
Depression
Over the past 2 weeks, how often have you been bothered by 1. Little interest or pleasure in doing things? 2. Feeling down, depressed, or hopeless?
Physical activity
1. On average, how many days per week do you engage in moderate to strenuous
exercise (like walking fast, running, jogging, dancing, swimming, biking, or other activities that cause a light or heavy sweat)? 2. On average, how many minutes do you engage in exercise at this level? Tobacco usea
1. Have you smoked at least 100 cigarettes in your entire life?
If yes: 2. Do you now smoke cigarettes every day, some days, or not at all?
Screen and follow up Screen and follow up Screen and follow up
Screen and follow up
Alcohol usea
1. How often do you have a drink containing alcohol? 2. How many standard drinks containing alcohol do you have on a typical day? 3. How often do you have six or more drinks on one occasion?
Screen and follow up
Social connection and social isolation
1. In a typical week, how many times do you talk on the telephone with family, friends,
Screen and follow up
Intimate partner violence
1. Within the last year, have you been humiliated or emotionally abused in other ways
Residential addressa
1. What is your current address?
Verify every visit
Census tract-median income
Geocoded
Update on address change
or neighbors? 2. How often do you get together with friends or relatives? 3. How often do you attend church or religious services? 4. How often do you attend meetings of the clubs or organizations you belong to? by your partner or ex-partner? 2. Within the last year, have you been afraid of your partner or ex-partner? 3. Within the last year, have you been raped or forced to have any kind of sexual activity by your partner or ex-partner? 4. Within the last year, have you been kicked, hit, slapped, or otherwise physically hurt by your partner or ex-partner?
Screen and follow up
Source: IOM. Capturing Social and Behavioral Domains and Measures in Electronic Health Records: Phases 1 and 2. Washington, DC: National Academies Press; 2014. a Domain is already widely included in clinical practice for which systematic measurement was proposed.
December 2015
982
Diez Roux et al / Am J Prev Med 2015;49(6):980–983
for interventions and policies to improve health. Richer data on behavioral and health variations across geography and social groups, including information on health behaviors, prevalent health conditions, and outcomes of care, which is often unavailable or limited in scope in existing surveys or registries, can also be useful to public health agencies in developing and targeting interventions and in identifying potentially important policies. EHR data linked to social and behavioral determinants would also be of great value in policy evaluation, including efforts to systematically evaluate the health and health disparity consequences of policies in domains outside of health care. For example, there is growing interest in evaluating the health and policy implications of transportation policies (such as establishing bike share programs or increasing access to public transportation) or food access policies (such as subsidizing access to fruits and vegetables in poor neighborhoods). The evaluation of these policies requires capitalizing on socalled natural experiments in which policy variations over space and time are linked to appropriate timevarying and geographically varying behavioral or other health outcome data, so that the effect of the policy can be assessed. These data are difficult or impossible to collect de novo in research studies within the timeframe required but could be available from EHRs. Public health research could also benefit from systematically collected information on social and behavioral determinants in EHR. In the narrowest sense, it will allow improved investigation of the role of the social environment in modifying outcomes of clinical care—an important area that has been relatively neglected by the social determiTable 2. Examples of Public Health Uses of Social and Behavioral Data Linked to nants of health field, and that Electronic Health Records presents important opportuniSocial/behavioral factor Sample possible uses ties for productive collaboration between clinicians and public Education, financial strain, and race and Tracking of immunization rates or receipt of health researchers. More ethnicity other preventative services. Characterizing causes of late diagnosis (e.g., broadly, by dramatically of cancer) or poor response to treatment. increasing the availability of Behavioral measures (physical activity, Identifying populations to target for longitudinal health data alcohol use, smoking rates, and fruit and community-based lifestyle interventions. (including laboratory and imagvegetable consumption) Evaluation of the impact of policies (bike and ing data) linked to information walking paths, safe street policies, smoking bans, corner store initiatives). on social determinants, the availability of social and behavDepression and stress Understanding how community violence may ioral data in EHRs will facilitate affect levels of depression and reported stress. etiologic research on the drivers Evaluation of the uptake and success of of population health and the publically available mental health services. causes of health disparities. Geographic identifiers linked to Characterizing geographic disparities and This enriched EHR data environmental data identifying high-risk areas. could be mined for many purUnderstanding the role of modifiable environmental factors (e.g., traffic and asthma poses, including large-scale lonexacerbations). gitudinal investigations of
Importantly, the availability of this information will also allow research and systematic evaluation of the utility of this information in clinical practice and service delivery such that its use enhances care and related outcomes without stigmatizing patients or threatening their privacy. The systematic inclusion of social and behavioral determinants (including routine geocoding and geographic linkages) in EHRs could also have important uses in public health practice and policy. Initiatives are already underway that use EHR data for public health surveillance. In a recent study, Kaiser Permanente used EHR data to identify geographic clusters of underimmunization and vaccine refusal.8 In another example, the Macroscope Project in New York City9 uses EHRs to track conditions managed by primary care practices that are important to public health. The addition of information on social and behavioral determinants and routine geocoding could further expand and increase the informativeness of these efforts. EHR data with their broad coverage and detailed information could be an important complement to existing national surveys in monitoring the health of the nation by allowing surveillance across broader geographic areas, by social groups, and for an expanded set of outcomes. The routine geocoding of EHRs is consonant with the growing interest in enhancing the use of available data to characterize and monitor small area variations in health. Information on small area variations in health can empower communities to raise questions about the causes of these differences and allow them to advocate
www.ajpmonline.org
Diez Roux et al / Am J Prev Med 2015;49(6):980–983
environmental influences on health and outcomes of health care as well as studies of gene–environment interactions. Moreover, the availability of detailed health and healthcare data linked to social and environmental data is likely to spur the development and application of new methodologic tools appropriate for extracting meaning from “big data.”10 The increasing use of these data for research may in turn improve the quality and completeness of the EHR data. Examples of public health uses of this type of data are shown in Table 2. Of course, the use of EHR data for public health research, practice, and action presents many challenges. There are important reasons why the public health field has advocated for the collection of new data for research, surveillance, and policy evaluation. Data from health records are notorious for missing data, non-systematic data collection, and limited generalizability to the populations of interest. There is a history of attempts to use data from health records in research with mixed results.11 Populations that are not frequent users of health care or that have poor access to health care (including many disadvantaged social groups) will not be accurately represented. Additionally, important variables may not be collected. Standardization and implementation of systematic data collection will present important obstacles related to logistics and cost. For example, clinical workers may be less inclined to collect information that they do not view as immediately relevant to their practice. The records can be cumbersome to access because of the data systems used, and their use for things other than the health care of an individual patient raises a host of complex privacy issues that must be addressed. Pilot and demonstration projects will be necessary to validate the information, illustrate how it can best be used for various purposes (including improving clinical practice), and identify what improvements will be necessary before it can be utilized widely and to its full potential. Achieving the full value of social and behavioral information in the EHR requires engagement of both clinicians and the public health community. The participation of public health researchers and practitioners will ensure that methodologic challenges and data quality issues are addressed and that the incorporated and retained measures are useful not only for clinical decision making but also for true population health improvement. Public health engagement will help keep the focus on the broader social determinants and will promote the type of research that will benefit not only the individual patient but also the population as a whole. Although these EHR
December 2015
983
data may never replace other data used in public health research and action, they may provide an economical alternative of rich complementary data to many data collection efforts already underway. The incorporation of social and behavioral information in EHRs and its potential uses in both clinical practice and public health research and action presents renewed opportunities for rich and productive collaborations between medicine and public health in ways that have not been possible in the past. This type of collaboration will both enhance clinical care and improve population heath.
No financial disclosures were reported by the authors of this paper.
References 1. Sharfstein J. The strange journey of population health. Milbank Q. 2014;92(4):640–643. http://dx.doi.org/10.1111/1468-0009.12082. 2. Kindig DA, Asada Y, Booske B. A population health framework for setting national and state health goals. JAMA. 2008;299(17):2081– 2083. http://dx.doi.org/10.1001/jama.299.17.2081. 3. Roundtable on Population Health Improvement; Board on Population Health and Public Health Practice; Institute of Medicine. Populaton Health Implications of the Affordable Care Act. Washington, DC: National Academies Press, 2014. 4. IOM. Capturing Social and Behavioral Domains and Measures in Electronic Health Records: Phases 1 and 2. Washington, DC: National Academies Press, 2014. 5. Adler NE, Stead WW. Patients in context—EHR capture of social and behavioral determinants of health. N Engl J Med. 2015;372(8):698–701. http://dx.doi.org/10.1056/NEJMp1413945. 6. Hsiao CJ, Hing E, Ashman J. Trends in electronic health record system use among office-based physicians: United States, 2007-2012. Natl Health Stat Report. 2014;75:1–18. 7. Estabrooks PA, Boyle M, Emmons KM, et al. Harmonized patientreported data elements in the electronic health record: supporting meaningful use by primary care action on health behaviors and key psychosocial factors. J Am Med Inform Assoc. 2012;19(4):575–582. http://dx.doi.org/10.1136/amiajnl-2011-000576. 8. Lieu TA, Ray GT, Klein NP, Chung C, Kulldorff M. Geographic clusters in underimmunization and vaccine refusal. Pediatrics. 2015;135(2):280–289. http://dx.doi.org/10.1542/peds.2014-2715. 9. New York City Department of Health. The New York City macroscope: developing an electronic health record-based population health surveillance system. www.nyc.gov/html/doh/downloads/pdf/data/nyc-macrosummary.pdf. Published 2014. Accessed 2015. 10. Scruggs SB, Watson K, Su AI, et al. Harnessing the heart of big data. Circ Res. 2015;116(7):1115–1119. http://dx.doi.org/10.1161/CIRCRESAHA. 115.306013. 11. Dean BB, Lam J, Natoli JL, Butler Q, Aguilar D, Nordyke RJ. Review: use of electronic medical records for health outcomes research: a literature review. Med Care Res Rev. 2009;66(6):611–638. http://dx.doi. org/10.1177/1077558709332440.
