Assessment of Training in Health Disparities in US Internal Medicine Residency Programs

Key Points Question Are internal medicine residents being trained to address health disparities, and what are their perceptions of such training? Findings This survey study of 227 program directors and 22 723 internal medicine residents in training found that residents’ perceptions of training in health disparities and its quality were not associated with the presence of a curriculum. Meaning These findings suggest that adding health disparities curricula in their current forms may not adequately train residents to address gaps in care.


Introduction
In 2002, the Institute of Medicine, now the National Academy of Medicine, produced their seminal report "Unequal Treatment: Confronting Racial and Ethnic Disparities in Healthcare," 1 which summarized the stark differences in health and health care outcomes between patients in minority groups vs those not in minority groups.Despite significant attention to that report, the 2018 "National Healthcare Disparities Report" 2 found that while some disparities in health care based on race, ethnicity, socioeconomic status, disability, and in populations with special health needs were getting smaller, disparities persisted for poor and uninsured populations in all priority areas.The 2003 report 1 prompted calls for increased education for all health care practitioners (including physicians) to begin to address these disparities. 3,4This charge is further reflected in the Health Quality Pathways of the Clinical Learning Environment Review (CLER) 5 and required Accreditation Council for Graduate Medical Education (ACGME) competencies.[8] While these recommendations have existed for at least a decade, 9 a 2014 review of health care disparities training in residency programs in the US 10 demonstrated few published reports of graduate medical education programs that ensure that residents are competent to address disparities.The first national report of findings for the CLER program, 11 highlighted health disparities as an area of challenge and opportunity for graduate medical education. 12The major findings were that the primary source of learning involved clinical experiences with patients who were at risk, mainly occurring in primary care settings and was generic, not tailored to the specific populations served by their medical center or hospital.In the second report based on residency site visits, 32.8% of residents and fellows reported their health disparities training was specific to their at-risk populations. 13evious Association of Program Directors in Internal Medicine (APDIM) surveys have addressed some issues related to health disparities, including cultural competency training and quality improvement projects.In the 2012 survey, 16.6% of internal medicine programs reported the presence of a health disparities curriculum.There are studies describing educational curriculum within internal medicine residencies for special populations, [14][15][16][17] and a recent survey of internal medicine program directors highlighted disparities related to patients with limited English proficiency in their curricula. 182010 multi-institutional study of internal medicine residents' health disparities education 19 found that only 14% of residents surveyed felt confident in their knowledge of underserved populations.However, a 2011 national survey 20 showed that residents' attitudes toward topics regarding medically underserved populations and health disparities were generally favorable.The

Methods
The survey study was reviewed and deemed exempt by the Mayo Clinic institutional review board.
Program director surveys were granted a waiver of informed consent because data were

Program Director Survey
The APDIM Survey Committee develops and distributes an annual survey to internal medicine program directors to address important issues in graduate medical education.The survey process has been previously described. 21,22

Resident Physician Survey
Detailed information concerning the internal medicine In-Training Examination (IM-ITE) resident survey has been reported previously. 26Briefly, the IM-ITE is a standardized examination developed by a Committee of the American College of Physicians that serves as a self-assessment of medical knowledge for internal medicine residents.Nearly 100% of internal medicine residency programs accredited by the ACGME participate in the IM-ITE, which is administered every August through September.On completion of the IM-ITE, residents are asked to complete a voluntary survey, which is submitted with their online examination.The survey questions are designed to gain an understanding of residents' training environment, and residents are asked for their consent to allow responses to be used in research.The 2015 survey included 3 questions related to a residents' perception of their training in health disparities (eAppendix 2 in the Supplement).

Statistical Analysis
Univariate summaries of program director and resident responses were reported as count (with percentage) or mean (with SD), as appropriate.

Results
The

Resident Survey Results
The  1).There was an association between a resident's perceived receipt of training and their estimated proportion of patients who would be considered at risk for health disparities (Table 2).

Comparison of Program Director and Resident Responses
While less than 40% of program directors reported a curriculum for training in health disparities, most residents reported training in the care of patients who are at risk for health disparities (ie, those who are underserved, uninsured, unemployed, or experiencing homelessness).We observed no association between the program director-reported presence of a curriculum and the resident report  3).

Discussion
This survey study is the first to our knowledge to report on the breadth of health disparities training   31 this may provide an opportunity for meaningful learning in the context of graduate medical education.Point of care teaching around health disparities and social determinants of health can be a powerful mechanism for residents to internalize these complex topics through the eyes of their own patients. 32A comment on the program director survey summarized this practice: "…we take care of the patients in our community with health care disparities.We live it every day!"However, reliance on learning in the context of caring for patients who are underserved does not ensure residents will receive the training and acquire the skills they will need.Accordingly, a 2019 study found that increased care of at-risk populations did not translate into increased relevant knowledge among resident physicians. 33ditionally, while some data suggest a positive influence of care of underserved populations on the attitudes of residents, there was no improvement in patient care with additional training and an inverse association was found between self-reported skills and knowledge. 34Our finding that residents' rating of the quality of their training was not associated with the presence of a curriculum in health disparities in their program also raises a concern that perceptions may overestimate the acquisition of needed skills.
Strengths of this study include a large, representative, and comprehensive sample of all internal medicine residents and program directors in the US with high survey response rates.Furthermore, the ability to link responses of residents with their programs through coincident surveys in the same calendar year provided a powerful tool for comparison of resident perceptions and curriculum.

Limitations
The study has limitations.A major limitation was that residents were not asked directly if they were exposed to a curriculum in health disparities but rather if they received training in the care of patients who would be at risk, which raises the concern that we cannot distinguish between their recognition of a formal and informal curriculum.Additionally, while program directors were asked explicitly whether they had a curriculum, we cannot know with certainty that they were aware of all training, but given their reporting responsibilities, we feel it is a reasonable assumption.Furthermore, because the survey items were embedded in larger program director survey, we were limited in the ability to ask them to define more specifically the components of their health disparities curricula.

Conclusions
This survey study found that the existence of health disparities curricula among internal medicine residency programs were below goal levels established by national accrediting bodies and that existing curricula were not associated with resident's perception of training in these domains.
Residents who cared for more patients who are underserved reported higher disparities training despite a lack of formal curricula, highlighting the opportunity to teach around health disparities at the point of care, but also the need for standardized curriculum and capable faculty.Additionally, program directors reported a lack of practice-level data to examine health disparities, which emphasizes the opportunity for institutional collaboration with residency programs for quality improvement initiatives aimed at the reduction of health disparities.
Future research is needed to develop and assess the most relevant domains for health disparities curricula in graduate medical education, including for point-of-care encounters and quality improvement.There are opportunities to explore partnerships among residencies, institutional clinical practices, and communities for productive collaborations around disparities-related quality improvement projects to address gaps in health care that are specific to the populations they serve.
goals of this study were to describe what internal medicine residency training programs provide as curriculum and/or educational experiences on health disparities, to determine residents' perception of training in health disparities, and to determine the association of program curriculum with resident perception of training.
Data from the program director and resident surveys were linked by training program via a unique examination identification number for each internal medicine examinee prior to being deidentified to compare program director responses with residents' perceptions of training in health disparities.The inclusion of questions on the APDIM and IM-ITE surveys related to training in health disparities provided a unique opportunity to compare the program director-reported curriculum within their internal medicine training programs with the perceptions and experiences of the residents in training.
of training or their rated quality of their training in the merged data set of 225 program director responses matched with 11 583 resident responses (Table

Downloaded From: https://jamanetwork.com/ on 10/08/2023
Assessment of Training in Health Disparities in US Internal Medicine Residency Programs To test the association of resident report of receiving training on health disparities with the proportion of patients considered at-risk whom they cared for, , and the threshold for statistical significance was set at P < .01 to account for multiple comparisons.Statistical analyses were conducted using SAS statistical software version 9.4 (SAS Institute).Final analysis of the merged data set was completed in 2018.
a generalized linear mixed model with a binary response distribution was used to estimate odds ratios.To test the association of PD report of presence of a health disparities curriculum with resident rating of the quality of their training in care for underserved, a generalized linear mixed model with a Gaussian response distribution was used to estimate the mean difference between groups.Both models included random effects to account for nesting of residents within programs.P values wereJAMA Network Open | Medical EducationAssessment of Training in Health Disparities in US Internal Medicine Residency Programs JAMA Network Open.2020;3(8):e2012757.doi:10.1001/jamanetworkopen.2020.12757(Reprinted) August 10, 2020 3/9 Downloaded From: https://jamanetwork.com/ on 10/08/2023 2-sided For analyses requiring identifiably linked PD and resident survey responses, there were 11 583 resident responses available from these 225 responding programs.There were no differences in response rates based on geographical region or program type.

Table 1 .
internal medicine training programs in the US.While approximately 70% of program directors reported a CLER visit, suggesting awareness of the requirement for health disparities training, only 40% of program directors reported having a curriculum.The curriculum was often limited to a few hours and lecture-based, which may not effectively engage learners.There was limited assessment of the impact of the curriculum on outcomes for the learner or the patients they care for.Only 30% of Program Director and Resident Rating of Training in Health Disparities

Table 2 .
Association of Resident Perception of Health Disparities Training With Proportion of Underserved Patients in Their Practice

Table 3 .
Association of Program Director-Reported Health Disparities Curriculum and Resident's Perception of Quality of Disparities-Related Training Assessment of Training in Health Disparities in US Internal Medicine Residency Programsinternal medicine programs without a current curriculum planned to address this need within the next year.While this was an improvement over the 2012 survey, in which only 16.6% of programs reported a curriculum, our findings raise concerns that additional efforts are needed for internal medicine graduate medical education to increase health disparities curricula.Challenges to developing curriculum were identified by program directors, most significantly competing curricular priorities.There was the additional challenge of determining what a healthTo our knowledge there are no prior studies that have reported results from a national survey of internal medicine resident perceptions on training in health disparities.Overall, 70% of residents reported training.Residents who cared for a larger proportion of underserved patients perceived that they received health disparities training at a higher rate.Since an essential component of graduate medical education is direct patient care and academic medical centers provide care for a disproportionate proportion of underserved populations in the US, JAMA Network Open.2020;3(8):e2012757.doi:10.1001/jamanetworkopen.2020.12757(Reprinted) August 10, 2020 5/9 Downloaded From: https://jamanetwork.com/ on 10/08/2023