Turner, Jane L. MD; Farquhar, Lynda PhD
The 2002 Institute of Medicine report, Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care,1 clearly established that health disparities are pervasive throughout the U.S. health care system. A large body of evidence supports the contention that socioeconomic position is a major predictor of health outcomes, avoidable morbidity, and mortality.2–5 People of racial and ethnic minorities tend to receive a lower quality of health care in the U.S. even when access-related factors, such as patients’ insurance and income, are controlled.6 This is the case for emergent conditions7 as well as preventable and treatable conditions such as cardiovascular disease, diabetes, asthma, and cancer.8
Patient factors, provider factors, and health system factors all contribute to health disparities.9 During the curriculum redesign at the College of Human Medicine (CHM) at Michigan State University, we focused on provider factors because we, as medical educators, can address those factors. Such factors include lack of knowledge about health care disparities,10 denial that such disparities exist,11 need for stronger communication skills,12 and cultural competence.13,14 To address the problem of needing a workforce prepared to care for those populations that have historically received a lower level of care, CHM committed itself to curriculum revision with a goal of producing a physician graduate who has the knowledge, skills, and attitudes to care for those insured by the Medicaid program and others who suffer from disparate health care. We labeled our efforts the “Contract for Social Commitment.” The impetus to launch the project came from internal and external sources. Faculty and administrators of the college were becoming increasingly concerned about the disparities in health and health care reported in the literature, including the IOM report. In particular, the new assistant dean for governmental affairs at CHM had previous experience as bureau chief within the Medicaid program, and the new assistant dean of the preclinical curriculum (J.T.) brought to the job a long-standing interest in health care equity. In addition, clinical faculty experienced difficulty finding specialists to see their patients on Medicaid. Students demonstrated their interest in addressing inequities by offering noon-hour electives, organizing an annual health care rally, and serving in a free clinic. Individuals in the Michigan Department of Community Health expressed interest in collaborating with medical educators and researchers to improve services to individuals and populations served by Michigan Medicaid.
CHM at MSU is a community-based school with approximately 100 in each class. Students spend the first two years in East Lansing, Michigan, for preclinical coursework and then at assigned clinical campuses in six sites across the state where they complete their clerkships in community hospitals and clinics. The Michigan Department of Community Health funded the Contract for Social Commitment, helping to support the salaries of faculty and administrators so they could dedicate a portion of their time to project activities. Faculty salaries were the most expensive part of the project. The first cohort of students to experience all aspects of the curriculum graduated in 2006.
We added new curricular material, modified existing material, and expanded the discussions of values and ethics already present across all four years of the curriculum. We specified the overall learning objectives for students as follows:
* to identify and understand the relationship between health conditions and poverty;
* to know the conditions that illustrate disparities in health status and health care;
* to recognize barriers to access to health care;
* to be capable of working effectively with immigrants and refugees in the community;
* to be capable of communicating successfully with individuals of all social classes, educational levels, and cultural backgrounds;
* to have the skills to work well with interpreters;
* to define primary, secondary, and tertiary prevention and how each relates to socioeconomic factors; and
* to work with other professionals (e.g., social workers, dieticians) to promote the welfare of people on Medicaid and others who are economically disadvantaged.
Once we, the project directors (J.T., L.F.), identified the student learning objectives, we took an inventory, using two sources (written objectives from course syllabi and a survey of course directors), of the content in existing courses. We completed this inventory during the summer and fall of 2002.
We mapped student learning objectives on a grid across courses and years and evaluated the treatment of each objective in terms of teaching strategies, assessment of student performance, and faculty development. We worked with course directors to identify topics needing more coverage and to suggest where in the curriculum they could best cover these topics. We helped to link suggested activities explicitly to learning objectives and charged the course directors to identify ways to assess student performance. For example, the surgery clerkship director proposed assigning students to participate in hospital discharge planning meetings. We asked, “What do you hope they will learn, and how will you know they learned it?” Next, we linked the activity to the objective of working with other professionals, and, to assess student performance, the clerkship director asked students to write about the patient’s discharge plan and community resources to be used after discharge from the hospital.
Course directors worked together in committees to tie assignments in individual courses with those in other courses for horizontal and vertical integration. For example, the director of the Integrative Clinical Correlation course timed a presentation including a low-income patient with diabetes to coincide with students’ interviews of patient volunteers with chronic health conditions in another course. The instructors designed the content in both courses to illustrate the impact of socioeconomic factors on access to care. Vertical integration came into play as communication skills were taught and assessed: students learned to ask about potential barriers to care in the basic interview in the fall of year one; they applied these skills again in the advanced interview in year two; and they were evaluated on the skills at the beginning of the medicine clerkship in year three. The committees mapped each learning objective across the curriculum to ensure that the preclinical activities prepared students for clinical assignments, which, in turn, prepared them for residency. We also linked these student learning objectives to the core competencies as identified by the Accreditation Council for Graduate Medical Education.
We developed new curriculum components beginning in 2002 and implemented them into the curriculum for the matriculating class of that year. We rolled out new units as the class moved through the curriculum, adding content either as stand-alone curriculum units or as modifications of existing lectures, activities, or small-group discussion topics. Descriptions can be found at (http://www.echt.chm.msu.edu) by clicking on Contract for Social Commitment.
We accomplished faculty development through course orientation sessions and additions to the discussion guide. Faculty members meet as a group with the course director for an orientation session before the onset of preclinical courses taught in small groups, and they earn teaching credit for attending these mandatory sessions, which are typically one to two hours long. The course director reviews the expectations for faculty and students and highlights key points in the content that are related to the Contract for Social Commitment. In addition, faculty members receive a preceptors’ guide which highlights new and significant themes to be addressed.
In the preclinical curriculum, we modified 25% of the courses; that is, courses that accounted for 20 of the 82 required credit hours of years one and two. The Clinical Skills Program (i.e., the introduction to clinical medicine) changed the most. We added student learning objectives related to understanding access and barriers to health care, to working with interpreters, and to communicating effectively with individuals of varied backgrounds to the communications modules of the program. We also wrote, with the help of a folklorist, new case scenarios with more social and economic content for encounters with simulated patients. For example, at least one simulated patient described difficulties with transportation to referred specialists; others asked about the costs of tests and revealed that they did not have health insurance. A trapper from Michigan’s Upper Peninsula presented a whole new set of issues regarding access to care and health beliefs. We fleshed out the social histories of all the simulated patients, so they could respond to student questions. We added items related to access and barriers to care to checklists for evaluating interviews, and we encouraged faculty preceptors to ask about social and economic issues when debriefing videotaped interviews.
We also addressed the relationship among health conditions and poverty, access to care, and communication with individuals from varied backgrounds in the Longitudinal Patient Centered Experience (LPCE), a required program in which we assign students to an individual with a chronic health condition. Students visit their LPCE patient in the patient’s home eight times in a 14-month period.15 We actively recruit patient volunteers who have either public insurance (Medicaid) or low incomes. Student assignments to small groups for discussion ensured that each group had at least one student whose patient volunteer had low income or Medicaid health insurance.
We modified year-three clerkships in pediatrics, family medicine, internal medicine, obstetrics–gynecology, and surgery; we reviewed, but did not alter, psychiatry. In year four, we changed both senior surgery and internal medicine, altering or replacing existing assignments to address student learning objectives. We offer some representative examples of curriculum changes below:
* The family medicine clerkship expanded with an online lecture about health care disparities, including instruction on working with interpreters when patients are not proficient in English. The Web format of the unit allowed students in all six community campuses to receive the same instruction. We added information about resources available to patients covered by public insurance to the already existing unit on smoking cessation.
* Early in the year-three internal medicine rotation, students interviewed and examined a standardized patient while a faculty member observed. We modified the rating instrument for the exercise to include items about health insurance and access to care, and faculty physicians addressed issues of access in the debriefing that followed each encounter.
* In the obstetrics–gynecology clerkship, students focused on the community resources available to address problems presented by a patient insured by the Medicaid program. Students completed a writing assignment in which they identified Web sites, agencies, facilities, and/or groups which could help the patient. They also visited one of the identified facilities.
* In the pediatrics clerkship, an assignment that focused on potential barriers to care for children with chronic health conditions and public insurance replaced a writing assignment focused on one complete patient history and physical. Students also learned about the roles of other professionals (e.g., pharmacists, social workers, pediatric psychologists, therapists) in the care of patients. During their inpatient rotation, students identified a pediatric patient with a chronic condition, interviewed the patient and/or a family member, reviewed the medications, and met with the hospital social worker or discharge planner. They wrote a paper about the child’s condition and psychological, social, and economic issues related to treatment.
* During the year-three surgery clerkship, students interviewed a patient with a low income to learn about barriers to adherence to treatment recommendations and how insurance status affects the patient. They also met with the hospital social worker/discharge planner to learn about plans for care after discharge and resources in the community. The assignments included both writing a paper about the patient’s circumstances and creating a list of problems related to the patient’s adherence to medical recommendations.
* In senior surgery, students identified a patient either with a low income or who had public insurance, and they focused on patient access to medications. The students investigated programs that provide medication to individuals with low socioeconomic status (SES), and they learned to use the Medicaid formulary.
* During the senior internal medicine clerkship, students revisited issues of access to care in depth while completing a Web-based module about managing diabetic ketoacidosis. They learned about the resources available to help patients with low SES pay for their care. This clerkship reinforced the lessons on the Medicaid program and use of formularies.
We took an inventory of project objectives periodically throughout the implementation of the revised curriculum. Table 1 shows which core clerkships addressed the student learning objectives at the beginning of the project, in 2002, and in 2005 after the curriculum revision when the first cohort entered the clerkships of year three.
We established a robust system of evaluation to provide feedback to course directors to guide implementation, and to assess the impact of the curriculum changes on student attitudes and skills. MSU’s institutional review board granted exempt approval. We gathered the following:
* student performance data in courses modified by the project,
* student evaluation of courses and clerkships modified by the project,
* student attitudes and professional goals,
* student self-assessment of skills,
* data regarding selected items on the Association of American Medical Colleges (AAMC) Graduation Questionnaire, and
* residency program directors’ views of students’ progress and abilities.
Student performance data
We chose two key performance indicators to look at skills most likely to be affected by curriculum changes: interviews with standardized patients in preclinical years, and the clinical skills component of U.S. Medical Licensing Exam (USMLE) Step 2. Students who experienced the revised curriculum performed slightly better than students in the previous class on key skills related to the project and on the interview overall, but the differences were not statistically significant. For example, on the item “ability to elicit patient’s personal social context,” 79 (78%) students in our cohort were rated as good or excellent by a faculty evaluator compared with 73 (72%) students in the previous class. Performance did not decline, though new material was added.
We compared student performance of this cohort on the USMLE Step 2 Clinical Skills Exam with that of the previous class. The pass rate for this cohort was 98%, compared with 95% for the previous year. In addition, the CHM student scores on the Communication and Interpersonal Skills component rose from 98% for the 2005 class to 100% in the 2006 class.
Student feedback about courses and clerkships
CHM at MSU routinely gathers student evaluation data for all courses and clerkships using Web-based questionnaires distributed at the end of courses and semesters. We used these data to examine students’ satisfaction with the courses overall and with specific components of the revised curriculum.
We surveyed students at the end of year two to learn whether they had encountered patients with low SES. In Clinical Skills, 59 (59.3%) of the students interviewed a simulated patient who presented as a patient with a low SES. The rest of the class was exposed to patients with low SES through observing interviews of their peers. Similarly, in the LPCE, 27 (27%) of the students reported that they interacted directly with a patient with a low SES, and the remainder reported that they observed and talked about a patient with a low SES in group discussion.
Student satisfaction with courses remained the same or improved slightly (not statistically significant) compared with the previous year. For example, the mean response on the item “Videotaping my simulated patient interview was helpful in skill acquisition” was 4.5 (on a Likert scale of 1 to 5, with 5 being “strongly agree”) in 2001 (n = 59) and 4.7 in 2002 (n = 101). (The change in number of students responding was attributable to a change in college policy to require students to complete evaluation questionnaires.)
For the clerkships, we surveyed students regarding whether and where they learned the material related to the Contract for Social Commitment. The information in Table 2 shows the number and percentage of students who reported learning these skills through their clerkships.
In addition to the numeric data on clerkships, we also asked students to tell us an anecdote or patient story about a low-income patient. Below are a few examples of students’ comments.
* I had a low-income patient who could not afford a Rx he would have benefited from. We had to treat him with less effective meds. Learned a lot.
* I had a low-income male patient with a stab wound needing immediate care. I was able to get Medicaid agency to have Rx paid for. Good experience.
* Patient was poorly insured, came in for chest pain. She was homeless, multiple issues with children in foster care. She was worked up for a Dx of anxiety, but also had Hep C and advanced disease. Her other issues came to light as relationship developed. Patient was discharged without a social work consult because of this late disclosure of these issues.
* Most encounters with low-income patients concern their desire for me to truly understand them.
* Patient came in with abdominal pain. Surgeons refused to operate due to “unknown etiology.” I wondered if there was an income/payment connection about this decision.
* I had experience with real low-income patients growing up in a rural area. Not all the B.S. you guys give us.
Most of the anecdotes indicated that the students had learned something meaningful from their encounters with patients with low SES. A few comments, like the last one above, suggested that our efforts were perceived as heavy handed, driven by political correctness or as unnecessary busy work by some students.
Student attitudes and professional goals
An 18-item questionnaire specifically developed for this project looked at students’ attitudes toward patients with low incomes and at students’ professional goals. This survey was administered to students at matriculation and at the end of years two, three, and four. Examples of items include “Medical schools should provide students with exposure to role models who work with the poor” and “Students should know about health access problems of the poor.” Mean responses were all above 4 (on a 5-point scale, with 5 being “strongly agree”) at all points in time, indicating strong agreement with these principles. This persisted across the four-year program. Students’ ratings of professional goals did not change significantly from matriculation to the end of year four. The goal “opportunity to serve others” was the second-highest-rated goal after “intellectually stimulating work is important to me” at the end of year four (4.54 and 4.58, respectively).
Student self-assessment of skills
Students completed questionnaires designed for the project at the end of years two, three, and four, on which they assessed their skills related to the learning objectives of the project. At the end of year two, all means were between 2.5 and 4.1 (on a 5-point scale, with 5 being “strongly agree”). Changes between years two, three, and four were analyzed using ANOVA. The means on 17 of the 18 items were higher by year four. Seven of the 18 items were rated significantly higher at the end of year four. Each of the items that changed significantly is listed below: “I have the skills to …
* work with patients who don’t speak English” (p ≤ .000).
* work with recent immigrants and refugees” (p ≤ .001).
“I know …
* what constitutes appropriate interpreter services” (p ≤ .000).
* what public health system services can assist low-income patients” (p ≤ .005).
* the roles of other health professionals” (p ≤ .000).
* what health conditions are prevalent among the poor” (p ≤ .01).
* what community-based services are available for low-income patients” (p ≤ .000).
AAMC Graduation Questionnaire
The Graduation Questionnaire of the AAMC includes four items that relate directly to our curriculum project. Students in medical schools across the country annually complete these questionnaires, which provide a means to compare our students’ attitudes both with those of CHM students in previous years and with those of all medical students across the country. Results on these items are in Table 3. CHM students were similar to students in “All Schools” in 2005, but the students who experienced our revised curriculum (graduating class of 2006) were more likely to agree with the statements than the previous class at CHM and than their classmates across the country.
CHM outcomes survey of residency program directors
CHM surveys residency program directors about our graduates at the end of the first postgraduate year (PGY 1). We obtain student permission to contact program directors. Of the 100 former students in the cohort, 14 did not consent and 4 could not be located; 82 residency directors were contacted in 2007, and 72 (88%) responded about graduates in our cohort. To look at CHM graduates’ performance on the objectives related to our project, we added three new items to the questionnaire. The items were
* “Resident is culturally competent,”
* “Resident works well with low-income/Medicaid patients,” and
* “Resident uses, locates, or knows about resources for these patients.”
We asked program directors to rate the CHM residents against the rest of the program’s residents (those from other medical schools) in that postgraduate year. Sixty-one of the program directors (85%) rated the CHM graduate in their program as above or substantially above average compared with other PGY 1 residents in the program with respect to cultural competence, 59 (82%) rated the CHM graduate above or substantially above average compared with other PGY 1 residents in the program regarding his or her ability to “[work] well with low-income/Medicaid patients,” and 57 (79%) rated the CHM graduate above or substantially above average compared with other PGY 1 residents in the program regarding use of resources. Program directors had the option of responding “not observed,” and seven to nine (9.7%–12.5%) of the program directors chose this option on items.
Our evaluation tells us that we did indeed successfully revise the curriculum to help increase medical graduates’ awareness of and ability to mitigate inequities in health care as shown by both faculty (Table 1) and student (Table 2) reports. We monitored student performance and feedback on courses closely because we were concerned that the addition of new content would interfere with students’ acquisition of skills and that student satisfaction with courses would suffer if they perceived excessive political correctness in the curriculum. Student performance on key skills and student satisfaction with revised courses did not deteriorate. Students reported improvement in the skills specifically taught in the revised curriculum. Self-report of efficacy by students has its limitations and can be misleading; however, feedback from residency program directors further supports the contention that our students acquired skills needed to care for patients with low SES.
Because of our concern that our efforts might backfire and students might become saturated with our efforts to reinforce their idealism and altruism, the results of our own annual surveys and the AAMC Graduation Questionnaire heartened us. Students’ agreement with the statement, “The opportunity to serve others is important to me,” remained high between matriculation and the end of year three. On the AAMC graduation questionnaire, the CHM 2006 cohort had higher levels of agreement with each of the relevant statements than did the previous CHM class and higher levels of agreement than “all schools.”
Can we be certain that the revised curriculum was responsible for students’ acquisition of skills and the maintenance of their altruism? We would need more historical data or a concurrent control group to make that claim unequivocally, both of which were beyond the scope of this curriculum-development project. However, all data points align to support the contention that the CHM curriculum prepares students to care for patients and populations with low SES as well as those with public insurance.
Factors contributing to success
What factors facilitated successful enhancement of the curriculum? Central leadership in the dean’s office (J.T. was assistant dean for preclinical curriculum at the time of the project) allowed us to see the curriculum as a whole and to coordinate across courses and across years to minimize redundancies and fill gaps. Recognizing that the curriculum was already full, we added no new material without eliminating existing material; we replaced or “tweaked” existing material to address project objectives. We attended to fundamentals of curriculum design and implementation. We used grids to track objectives across courses and years and to tie them to teaching strategies, performance assessment, and faculty development. We took care to ensure that we exposed students to fundamentals before we challenged them with higher-order skills.
Challenges to implementation of curriculum revisions
One major challenge came from multiple demands on the time of course directors. Faculty ownership of the curriculum strengthened the project, but when other duties pulled faculty members away from the Contract for Social Commitment, the development and implementation of new material stalled. External funding from the Michigan Department of Community Health was extremely helpful in overcoming this challenge. Salary support for course directors helped to gain and keep their attention when competing demands for their time drew them away. A polite reminder to the department chairperson helped to redirect attention to the curriculum. External funding also allowed course directors to purchase educational aids such as videos, books, and online products, which enriched the courses and saved faculty the time and effort needed to develop educational aids de novo.
Another challenge was that course directors tended to think in terms of activities rather than learning objectives. Many course directors are physicians who have extensive education and experience as clinicians but only on-the-job training as educators. We linked activities explicitly to learning objectives and assisted in the design of methods to assess student performance.
Will the faculty of CHM sustain the revisions in the curriculum now that the project is officially over and external funding has ended? Incorporation of changes in written course objectives and assessments ensures some sustainability; the interests of individuals do not drive revisions. Encouragingly, we have added some new units since the initial revisions have been in place. For instance, a patient education exercise in year one now includes an encounter with a simulated patient not proficient in English, and the student must work through an interpreter. Also, in a new year-three observed structured clinical exam, one of the communication stations involves an interpreter. The alignment of the goals of the project with the existing mission and culture of CHM was key to the success of the project and ensures that many of the revisions will continue or even improve.
We have much to do as a society to eliminate disparities in health and health care. The curriculum project at CHM contributes to that endeavor by preparing students with the knowledge, skills, and attitudes needed to provide high-quality care to individuals and populations who have Medicaid insurance and low SES.
The authors wish to thank Nicholas Kuzera and Patricia Mullan, PhD, for statistical analysis, Dr. Mullan for evaluation assistance, and Brian Mavis, PhD, for statistical consultation.