INTRODUCTION
Academic medical centers in the USA disproportionately care for low-income populations.1 Although several medical organizations now recommend social determinants of health (SDH) curricula be incorporated into internal medicine (IM) residency training, programs have been slow to respond.2–4 While models screening for SDH in adult-care settings have been described,5 there have been no published trials of SDH education targeting IM residents. We implemented a pilot SDH educational intervention to improve resident knowledge, confidence, and ability to address patients’ unmet social needs.
METHODS
We conducted this study in two outpatient clinics at Montefiore Medical Center (Bronx, NY). All second-year (PGY2) IM residents (N = 38) participated in the curriculum from September 2015 through April 2016 (Fig. 1). Third-year (PGY3) IM residents served as a control group. Albert Einstein College of Medicine IRB approved this study.
Fig. 1.
Timeline of curriculum intervention and evaluation components.
The curriculum focused on SDH shown to impact health outcomes for which resources were available, including health insurance, food insecurity, housing, transportation barriers, and need for supplemental income/public benefits. Using a trigger-based approach, residents were taught to look for key features of a patient presentation, e.g., multiple missed visits or uncontrolled disease, to assist in identifying social needs. Tools, including resource sheets, a resource website, and advocacy letters, were created for addressing SDH, and available on the residency homepage. Preceptors were invited to participate in a faculty development training session on the trigger-based approach and resource tools.
Residents in the intervention group completed surveys before, immediately after, and 6 months after the intervention. The 6-month survey allowed for comparisons to controls at the same stage of residency training. Surveys assessed barriers to addressing SDH in the clinic and knowledge on underserved populations,6 confidence identifying and addressing SDH, and self-reports on the proportion of encounters where SDH were identified or addressed. Social work referrals per patient encounter for each resident were calculated using medical records data for the intervention and control groups before and after the intervention. Chi-square, t test, Fisher’s exact test, and McNemar’s test were used to analyze the data (SPSS, v.21 and Microsoft Excel).
RESULTS
Thirty-four PGY2 residents (89%) completed the baseline survey. Of those, 94% completed the post intervention survey and 76% completed surveys 6 months later. Of the 39 PGY3 control residents, 36 (92%) completed the survey. The intervention and control groups were similar except more intervention residents reported a childhood experience of difficulty “making ends meet” (24 vs. 6%, p = 0.04) and prior training (82 vs. 50%, p = 0.03) and experience working with the underserved (76 vs. 42%, p < 0.01).
At baseline, a lack of time (86%) and a lack of knowledge (60%) were the most frequently cited barriers to identifying and addressing SDH in the clinic. Post intervention, there were short-term mean gains in knowledge (16.5 vs. 19.6, p < 0.01), and persistent gains in confidence identifying and addressing SDH at 6 months (2.7 vs. 3.2, p < 0.01; Table 1). While there was no difference in knowledge between intervention and control residents (17.7 vs. 16.4, p = 0.23), confidence with SDH was greater in the intervention than the control group (3.2 vs. 2.9, p = 0.03). Behavior increased post intervention (3.0 vs. 3.3, p = 0.05) with 50% of residents reporting spending ≥ 10% more patient encounters addressing SDH 6 months post intervention. There was a 47% decline in social work referrals in the intervention group over the 16-month study period (5.8 vs. 3.1%, p = 0.01), a change not observed in the control group (3.7 vs. 3.8%, p = 0.79).
Table 1.
Intervention and Control Group Survey Results of Knowledge, Attitudes, and Self-reported Behavior
| A | B | A vs. B | C | A vs. C | D | C vs. D | |
|---|---|---|---|---|---|---|---|
| Pre-intervention mean | Post intervention mean | Pre-Post p value* |
6-month post intervention mean | p value† | Control mean | p value‡ | |
| Knowledge n = 33 Health insurance access (n = 10) |
16.5 3.6 |
19.6 4.4 |
< 0.01
< 0.01 |
17.7 4.4 |
0.15 0.04 |
16.4 3.8 |
0.23 0.13 |
| Social determinants of health (n = 10) | 5.6 | 5.9 | 0.26 | 5.7 | 0.81 | 5.1 | 0.24 |
| Racial disparities (n = 10) | 5.9 | 7.1 | < 0.01 | 5.9 | 1.00 | 5.8 | 0.87 |
| Incarceration (n = 3) | 1.5 | 2.3 | < 0.01 | 1.7 | 0.29 | 1.8 | 0.93 |
| Confidence in identifying SDH (overall mean) | 2.7 | 3.1 | < 0.01 | 3.2 | < 0.01 | 2.9 | 0.03 |
| Insurance status Out-of-pocket financial costs |
2.9 2.2 |
3.4 2.2 |
0.12 < 0.01 |
3.7 3.0 |
< 0.01
< 0.01 |
3.2 2.3 |
0.08
< 0.01 |
| Access affordable/nutritious food | 3.0 | 3.3 | 0.15 | 3.2 | 0.24 | 3.2 | 0.88 |
| Transportation to appointments | 3.3 | 3.4 | 0.85 | 3.7 | 0.03 | 3.4 | 0.21 |
| Housing conditions and stability | 3.0 | 3.2 | 0.53 | 3.4 | 0.20 | 2.8 | 0.03 |
| Eligibility for supplemental income and disability benefits | 2.0 | 3.2 | < 0.01 | 2.4 | 0.04 | 2.3 | 0.96 |
| Confidence in creating a plan to address SDH (overall mean) | 2.7 | 3.1 | < 0.01 | 3.2 | < 0.01 | 2.7 | < 0.01 |
| Insurance status Out-of-pocket financial costs |
2.8 2.7 |
3.7 2.6 |
< 0.01
0.75 |
3.5 3.0 |
< 0.01
0.24 |
2.4 2.4 |
< 0.01
0.01 |
| Access affordable/nutritious food | 2.5 | 3.1 | < 0.01 | 3.4 | < 0.01 | 2.7 | 0.01 |
| Transportation to appointments | 3.1 | 3.5 | 0.08 | 3.5 | 0.04 | 3.4 | 0.73 |
| Housing conditions and stability | 2.5 | 3.3 | < 0.01 | 3.0 | 0.04 | 2.7 | 0.14 |
| Eligibility for supplemental income and disability benefits | 2.4 | 2.4 | 0.89 | 2.9 | 0.03 | 2.6 | 0.23 |
| Proportion of patient encounters during which some time spent identifying/addressing SDH (overall mean)§ | 3.0 | 3.3 | 0.05 | 3.1 | 0.45 | 2.8 | 0.30 |
| Insurance status/out-of-pocket financial costs | 3.1 | 3.8 | 0.04 | 3.5 | 0.37 | 3.0 | 0.14 |
| Access affordable/nutritious food | 2.9 | 3.1 | 0.46 | 2.6 | 0.32 | 2.6 | 0.92 |
| Transportation to appointments | 3.2 | 3.5 | 0.23 | 3.4 | 0.24 | 3.5 | 0.68 |
| Housing conditions and stability | 3.2 | 3.3 | 0.68 | 3.2 | 1.00 | 2.5 | 0.03 |
| Eligibility for supplemental income and disability benefits | 2.4 | 3.1 | 0.03 | 2.9 | 0.14 | 2.6 | 0.30 |
Italics entries are significant p values
*Post intervention scores vs. pre-intervention scores (paired samples t test)
†6-month post intervention scores vs. pre-intervention scores (paired samples t test)
‡6-month post intervention scores vs. control scores (independent samples t test)
§Numbers correspond to a percentage range of patient encounters during the last three ambulatory blocks. 1 = 0%, 2 = 1–5%, 3 = 6–10%, 4 = 11–15%, 5 = 16–20%, 6 = > 20%
DISCUSSION
Our curriculum improved knowledge, confidence, and practice behavior related to SDH. The decrease in social work referrals may be due to increased self-sufficiency, a hypothesis supported by the lack of change in referrals by control residents during the same time period, and may indicate that similar SDH training could have meaningful benefits in clinics with fewer resources.
Study limitations include its single-institution setting, small sample size, risk for recall and desirability bias on survey responses, and lack of patient outcomes. Continued curricular reinforcement may result in a more sustained impact in individual SDH domains.
Our novel curriculum, using a trigger-based approach and focusing on applicability to patient care, addresses the need for training in SDH to help IM residents better care for their patients. Future work should focus on assessing additional measurable outcomes of resident behavior and clinical outcomes.
Acknowledgements
We would like to thank Carla Spagnoletti, MD, MS, for her invaluable feedback and assistance editing this manuscript. We would also like to thank Mike Elnicki, MD; Julia Fleming, MD; Asha Isable, MD; Jon Giftos, MD; Brian Dorsey, MD; and Maritza Casillas, LSW, for their feedback and contributions to this study.
Prior Presentations: This study was presented in part at the Society of General Internal Medicine Conference April 2017, Beyond Flexner Conference September 2016, and Society of General Internal Medicine Conference May 2016.
Compliance with ethical standards
Albert Einstein College of Medicine IRB approved this study.
Conflict of interest
The authors declare that they do not have a conflict of interest.
References
- 1.Association of American Medical Colleges. 2013. Why Teaching Hospitals Are Important to All Americans. Available at: https://www.aamc.org/advocacy/teachhosp/253374/teaching-hospitals.html. Accessed February 26, 2017.
- 2.Catalanotti JS, Popiel DK, Duwell MM, Price JH, Miles JC. Public Health Training in Internal Medicine Residency Programs. Am J Prev Med. 2014;47(5):S360–S367. doi: 10.1016/j.amepre.2014.07.024. [DOI] [PubMed] [Google Scholar]
- 3.American Medical Association Wire. What future doctors need to know about health determinants. November 20, 2014. Available at: https://wire.ama-assn.org/education/what-future-doctors-need-know-about-health-determinants. Accessed October 30, 2017.
- 4.World Health Organization Commission on Social Determinates of Health 2008. Closing the gap in alytygeneration: Health equity through action on the social determinants of health 2008). Available at: http://www.who.int/social_determinants/thecommission/finalreport/en/index.html. Accessed October 29, 2017.
- 5.Berkowitz SA, Hulberg AC, Hong C, Stowell BJ, Tirozzi KJ, Traore CY, Atlas SJ. Addressing basic resource needs to improve primary care quality: a community collaboration programme. BMJ Qual Saf. 2016;25(3):164–172. doi: 10.1136/bmjqs-2015-004521. [DOI] [PubMed] [Google Scholar]
- 6.Weiland ML, Beckman TJ, Cha SS, Beebe TJ, McDonald FS. Resident Physicians’ Knowledge of Underserved Patients: A Multi-Institutional Survey. Mayo Clin Proc. 2010;85(8):728–733. doi: 10.4065/mcp.2009.0703. [DOI] [PMC free article] [PubMed] [Google Scholar]