Medical school is supposed to instill positive attitudes in students about the practice of medicine. However, several studies have reported an erosion of positive attitudes between entry and graduation,1–5 especially during the traditional third year, in which most students are immersed for the first time in hospital-based clinical medicine.6–8 In spite of growing speculation about what causes this phenomenon and how to prevent it, few if any evidence-based reports exist assessing the success of efforts to do so. On the basis of several innovations in clinical education introduced at Harvard Medical School (HMS), we present data on the impact of these new educational programs on students' attitudes. Specifically, we hypothesized (1) that students who experienced the new models would not demonstrate a loss of patient-centeredness compared with students in the traditional model, (2) that students in the new models would experience the impact of the hidden curriculum less strongly, and (3) that changes in students' patient-centered attitudes would be associated with their experience of the hidden curriculum.
In the traditional model at HMS, students rotate among a large number of Harvard-affiliated clinical sites to complete their seven core clerkships (medicine, surgery, obstetrics–gynecology, pediatrics, neurology, psychiatry, and radiology). Beginning in May 2008, however, all third-year students at HMS began to experience their principal clinical experience (PCE) at a single site. Several models of PCE have been developed.9,10 Each shares similar values and objectives, but each implements these in different ways. For instance, at most sites the students experience traditionally discipline-based rotations, but at Cambridge Hospital students learn by longitudinal contact with a carefully monitored patient cohort in order to receive exposure to all the core disciplines.9 What is common to each PCE site is that students have longitudinal contact with faculty and residents as well as longitudinal contact with patients. In addition, at all sites students go through their PCE as a single cohort, receive a good deal of mentoring, and are exposed to a range of specially designed curricular experiences such as longitudinal and interdisciplinary case conferences emphasizing the integration of the basic sciences.
The trend toward a degradation of attitudes has been demonstrated broadly. Students have been found to become less idealistic,6 less empathic,2 less patient-centered,11,12 less attuned to the needs of special populations,8,13,14 and less sensitive to ethical issues.7,15 This pattern has been found in both older5,7,16 and newer1,2,12,17 accounts, and internationally1,12,15,18,19 as well as in the United States.3,4,6,7 One of the primary explanations offered for this phenomenon is the hidden curriculum,20,21 in which students identify a strong disconnect between the values espoused by the faculty versus those that are modeled. Using an instrument developed to detect the presence of the hidden curriculum in clinical training, Haidet and his colleagues22,23 found that students in nine medical schools reported a wide range of experiences with faculty and housestaff in terms of support for patient-centered attitudes and behavior. The research reported below was, to our knowledge, the first to specifically investigate the relationship between students' experience of the hidden curriculum and changes in patient-centered attitudes.
Participants and research design
Before instituting the changes for all HMS students indicated above, pilot tests of the PCEs were conducted for small groups of students at each of three sites that were operational during the academic year 2005–2006: Brigham and Women's Hospital (BWH, n = 12 students), Beth Israel Deaconess Medical Center (BIDMC, n = 8 students), and Cambridge Integrated Clerkship (CIC, n = 12 students). Students in the PCE program were all volunteers who chose the site at which they would spend their entire year. The data in this report were collected as part of the evaluation of these pilots.
A control group of 17 HMS student volunteers was recruited to compare their experiences and attitudes with those of the 32 PCE students. Students in the control group rotated from site to site and discipline to discipline according to their own individually selected pattern in the traditional manner of students at HMS. Whereas the PCE students at Cambridge were at a different location and their experiences were structured very differently (i.e., they had no disciplinary blocks and had greater outpatient and longitudinal patient contact), it is important to note that PCE students at BWH and BIDMC and the control group students at those sites worked side by side. The PCE students at these two sites were not overtly identified in any way to the attendings or the housestaff, and there is little reason to believe that PCE and control students at these sites had differing exposure to patients, were given differing levels of responsibility, or were treated differently while on the clerkships themselves.
During orientation week for all clerkships, traditional and PCE based, students completed the Patient-Practitioner Orientation Scale (PPOS), a well-validated instrument that measures students' attitudes toward patient-centered care.11,12,18,24 The PPOS is an 18-item instrument in Likert format that assesses students' beliefs about sharing power and information with patients and the extent to which it is important to get to know patients as “whole people.” Students responded to this same instrument 12 months later at the completion of their clerkship year. Students each chose a unique four-digit personal identification number at the beginning of the year so that their anonymity could be maintained while still allowing for a repeated-measures design in which their pre- and postclerkship responses were matched and compared.
At the end of the year, students also completed the Community, Curriculum, and Culture (C3) measure,22,23 an instrument devised to measure students' experience of the hidden curriculum. The C3, which has been validated across a range of medical schools, asks medical students to report on their clinical experiences in three content areas relating to patient-centered care: Role Modeling of Faculty and House Staff; Own Patient Care Experiences; and Support Received for Patient-Centered Behaviors. Indices combining these items were created consistent with the validation work of the C3 study group. This research was approved by the HMS Committee on Human Studies.
The analysis was carried out in several steps. First, we sought to determine the comparability of the PCE groups and the control group at baseline on several different indicators. We calculated mean scores on two nationally standardized measures of ability and knowledge, MCAT and Step 1, and we compared the groups on preclerkship clinical skills based on students' mean scores on an end-of-second-year seven-station OSCE. Then, we compared students' stated future preferences in terms of time to be spent in practice, research, teaching, and administration. Mean scores on each of these measures were compared using one-way analysis of variance (ANOVA).
Next, we computed PPOS scores (higher scores indicate a stronger orientation toward patient-centeredness) and compared these scores for all four groups (three PCE sites plus control) at the beginning of the year using one-way ANOVA to determine whether baseline differences existed among them. We did the same at the end of the year to determine whether PPOS differences existed at that point in time. Next, for all four groups we also compared their mean pre- and post- scores on the PPOS using repeated-measures t tests.
Mean C3 scores for each of the three content areas were compared to determine whether differences existed among the four groups. Combining the PPOS and C3 in the same analysis, for each student we calculated the signed difference between pre- and post-PPOS scores (referred to as PPOS-delta, with negative scores indicating decreases in patient-centered attitudes) and determined the extent to which changes in patient-centered attitudes were associated with students' reported C3 experiences, using Pearson correlation coefficients.
None of the PCE groups differed significantly from one another on any of the baseline measures of knowledge, skills, or interests, nor did the PCE students differ from those in the control group. In addition, there were no statistically significant differences in the pairwise comparisons made between each of the clinical sites and the control group. Likewise, there were no preclerkship differences in attitudes as measured by the PPOS.
Comparing PPOS scores at the beginning and end of the clerkship year, we found that the mean PPOS scores for the control group went down significantly (P < .009), whereas the mean PPOS scores of the combined PCE groups remained flat (see Figure 1). The lack of attitude change was remarkably similar for all three PCE sites, with pre- and post- means almost identical for each (the pre- and post- means were 5.14 versus 5.13, 4.89 versus 4.83, and 5.14 versus 5.12 for sites 1, 2, and 3, respectively). At the end of the year, the mean PPOS score of the combined PCE groups was significantly higher than that of the control group (P < .007; see Figure 1).
The experiences of those PCE and control students who were side by side on the wards during the year were compared using the C3 instrument. For this analysis, we therefore excluded the CIC students, who were at a different site than the control students were and whose exposure to hospital staff and patients was organized and structured very differently. Comparing the combined means of the BWH and BIDMC PCE sites with those of the control group, the PCE means were significantly higher than those of the control group for both Own Patient Care Experiences (P < .001) and Support for Patient-Centered Behaviors (P < .02), with large effect sizes of 1.34 and 0.83, respectively (Table 1). In addition, on the summary measures for the attending and the housestaff role modeling questions, the PCE means were marginally higher than those of the control group (see Table 1), with moderate effect sizes (0.60 and 0.63, respectively).
Finally, correlations between C3 scores and the extent of student attitude change (using PPOS-delta) indicated that for the entire sample of students (across all sites and program types), two of the C3 content areas were significantly associated with changes in PPOS scores (PPOS-delta and Patient Care Experiences, r = 0.41, P < .01; and PPOS-delta and Support for Patient-Centered Behaviors, r = 0.44, P < .01).
The findings of this study demonstrate a degradation of attitudes for those students who went through their clinical rotations in the traditional manner. This is similar to changes reported by others.6–8 However, even though the three PCE groups followed different models of clinical education and implemented different curricula, the findings across these three groups showed no such negative change. These results constitute a noteworthy exception to a pattern that has disconcerted medical educators for years.
A second cross-site finding of note is the significant correlation between students' reports of their experiences on the wards (on the C3) and the extent to which their attitudes changed. The more positive the experiences that students reported, especially relating to their patient care experiences and perceived support for their own patient-centered behaviors, the more their attitudes toward patient-centeredness resisted negative change.
This result is made more intriguing by the C3 differences found between the PCE students at BWH and BIDMC compared with the traditional students who rotated through these same sites. We have no evidence, anecdotal or otherwise, to indicate that the attendings or housestaff were aware which students were doing traditional rotations and which were in the PCE, or that the responsibilities, exposures, and experiences of the two groups were any different. Nonetheless, the PCE students reported more positive experiences concerning reactions to and support for their own patient-centered behaviors. One possible explanation is that, somehow, students became identified to hospital staff as part of the PCE and came to be treated differently, whether overtly or subtly. However, the large size of the hospitals and the number and turnover of students, faculty, and housestaff make this explanation unlikely. Another plausible explanation is the “rose-colored glasses effect,” which suggests that students in the PCEs came to feel empowered and felt involved in a generally more supportive environment. As a result, despite having virtually similar experiences, they may have become less sensitive to negative aspects of life on the wards, or they may have interpreted such interactions in a more positive way. In essence, if the hidden curriculum tree fell in the clinical forest but these students did not hear it, perhaps it did not actually make a sound.
The unique contribution of each of the elements of the PCEs cannot be ascertained from our data; however, the similar pattern found across all the PCEs leads us to ask what common elements they shared that might account for this outcome. We have identified five aspects of the PCE experience that likely contributed to the immunization of these students against the erosion of their patient-centered orientations, each of which increased the intensity of student interaction with faculty, housestaff, patients, and peers:
* Longitudinality. Being in one site gives students a sense of continuity with faculty and housestaff, enables them to form meaningful relationships with patients over time, allows them to begin to identify with an institution, and enables feedback to be given and progress to be monitored over time.
* Mentoring. When students remain at one site, this allows for the explicit design of a longitudinal mentoring program (in fact, some sites assigned mentors and others used different mechanisms to ensure close and continuous faculty–student support and guidance). When students remain at one site and mentoring is built into the plan from the first day, levels of familiarity and comfort tend to grow and a sense of collegiality between mentor and mentee can develop over time.
* Planned interdisciplinary curriculum. Although more related to knowledge than attitudinal outcomes, the use of student cases and student-directed tutorials models a collaborative and supportive style of learning between faculty and students that may spill over to students' attitudes toward collaborative relations with patients.
* Opportunities for writing and/or self-reflection. Although the approaches used differed by site, each had protected time for students to come together guided by a faculty preceptor to discuss their experiences and to process feelings about their experiences in the hospital and their ambulatory settings. In addition, some sites had well-developed writing programs.
* Group support. After going through their preclinical experiences as a group (in tutorial groups, in societies, etc.), students typically experience their third year more individually because their unique patterns of rotation take them in various directions according to discipline, site, and time. PCE students go through the entire year together as a single cohort, offering them a built-in support group in which they can share experiences and, within a safe atmosphere, admit weaknesses and discuss reactions to the behaviors of patients, faculty, and housestaff.
The findings reported here are suggestive only, and we acknowledge that the conclusions are limited by several factors. First, the research was part of a pilot program at one medical school with a small number of participants in a single cohort. Second, as is common in field research and educational evaluation, it was impossible to assign students randomly to PCEs or to recruit control group students randomly from those doing traditional rotations. This nonrandom selection leaves open the possibility that any differences identified at the end of the project may have been attributable to preexisting differences among the groups. Aware of this possibility, we measured a broad array of variables that might have made the PCE and the traditional groups noncomparable at baseline, and we found no differences within and among the groups; yet, it is still possible that other unassessed variables may have differentiated the groups and biased the results. And, finally, because it was impossible to randomly assign students to groups, no cause-and-effect relationships can be inferred.
Nonetheless, in the context of so many reports, both anecdotal and empirical, that indicate the negative effects of students' first long-term and intense exposure to clinical medicine, the findings from this cohort of PCE students are encouraging. The fact that systematic and planned alterations in curricular and extracurricular experiences during the third year helped inoculate students' attitudes, colored their experience of the hidden curriculum, and were associated with resistance to negative attitude change provides an evidence base for the hope that these (and other) innovative changes to medical students' PCEs can yield positive, measurable outcomes. We hope that the approach we describe here may be useful to other schools as they seek ways to improve their students' clinical experiences.
1 Woloschuk W, Harasym PH, Temple W. Attitude change during medical school: A cohort study. Med Educ. 2004;38:522–534.
2 Hojat M, Mangione S, Nasca TJ, et al. An empirical study of decline in empathy in medical school. Med Educ. 2004;38:934–941.
3 Coulehan J, Williams P. Vanquishing virtue: The impact of medical education. Acad Med. 2001;76:598–605.
4 Pololi L, Price J. Validation and use of an instrument to measure the learning environment as perceived by medical students. Teach Learn Med. 2000;12:201–207.
5 Wolf TM, Balson PM, Faucett JM, Randall M. A retrospective study of attitude change during medical school. Med Educ. 1989;23:19–23.
6 Griffith CH, Wilson JF. The loss of student idealism in the 3rd-year clinical clerkships. Eval Health Prof. 2001;24:61–71.
7 Feudtner C, Christakis DA, Christakis NA. Do clinical clerks suffer ethical erosion? Students' perceptions of their ethical environment and personal development. Acad Med. 1994;69:670–679.
8 Davies BE, Nelson DB, Sahler OJ, McCurdy FA, Goldberg R, Greenberg LW. Do clerkship experiences affect medical students' attitudes toward chronically ill patients? Acad Med. 2001;76:815–820.
9 Ogur B, Hirsh D, Krupat E, Bor D. The Harvard Medical School–Cambridge integrated clerkship: An innovative model of clinical education. Acad Med. 2007;82:397–404.
10 Bell SK, Krupat E, Fazio SB, Roberts DH, Schwartzstein RM. Continuity as a central principle in a longitudinal third-year medical student clerkship. Acad Med. 2008;83:467–475.
11 Haidet P, Dains JE, Paterniti DA, Chang T, Tseng E, Rogers JC. Medical students' attitudes toward patient-centered care and standardized patients' perceptions of humanism: A link between attitudes and outcomes. Acad Med. 2001;76(10 suppl):S42–S44.
12 Tsimtsiou Z, Kerasidou O, Efstathiou N, Papaharitou S, Hatzimouratidis K, Hatzichristou D. Medical students' attitudes toward patient-centered care: A longitudinal survey. Med Educ. 2007;41:146–153.
13 Ubel PA, Jepson C, Silver-Isenstadt A. Don't ask, don't tell: A change in medical student attitudes after obstetrics/gynecology clerkships toward seeking consent for pelvic examinations on an anesthetized patient. Am J Obstet Gynecol. 2003;188:575–579.
14 Masson N, Lester H. The attitudes of medical students towards homeless people: Does medical school make a difference? Med Educ. 2003;37:869–872.
15 Rennie SC, Rudland JR. Differences in medical students' attitudes to academic misconduct and reported behavior across the years—A questionnaire study. J Med Ethics. 2003;29:97–102.
16 Marcotte DB, Held JP. A conceptual model for attitude assessment in all areas of medical education. J Med Educ. 1978;53:31031–31034.
17 Schwartz PL, Loten EG. Effect of year in school on medical students' perceptions evaluated with the Cognitive Behavior Survey, Attitudes Toward Social Issues in Medicine Survey, and Learning Environment Questionnaire. Teach Learn Med. 2004;16:3333–3344.
18 Ribeiro MM, Krupat E, Amaral C. Brazilian medical students' attitudes toward patient-centered care. Med Teach. 2007;29:1–5.
19 Shankar PR, Dubey AK, Subish P, Deshpane VY. Attitudes of first year medical students towards the doctor–patient relationship. J Nepal Med Assoc. 2006;45:196–203.
20 Hafferty FW. Beyond curriculum reform: Confronting medicine's hidden curriculum. Acad Med. 1998;73:403–407.
21 Browning DM, Meyer EC, Truog RD, Solomon RZ. Difficult conversations in health care: Cultivating relational learning to address the hidden curriculum. Acad Med. 2007;82:905–913.
22 Haidet P, Kelly PA, Bentley S, et al. Not the same everywhere. Patient-centered learning environments at nine medical schools. J Gen Intern Med. 2006;21:405–409.
23 Haidet P, Kelly PA, Chou C. Characterizing the patient-centeredness of hidden curricula in medical schools: Development and validation of a new measure. Acad Med. 2005;80:44–50.
24 Krupat E, Hiam CM, Fleming MZ, Freeman P. Patient-centeredness and its correlates among first year medical students. Intern J Psychiatry Med. 1999;29:347–356.