Abstract
Context:
Physician empathy influences rapport with patients and improves outcomes, but it is not well understood as an outcome of osteopathic medical education.
Objective:
To determine how clerkship preceptors at Touro University College of Osteopathic Medicine-CA (TUCOM) in Vallejo define empathy and how they compare observed empathetic behavior of TUCOM students with that of other medical students.
Methods:
Cross-sectional data were obtained from a survey of TUCOM clinical preceptors comparing TUCOM students with other medical students on 10 behaviors. Results were analyzed with a 2-tailed z test of proportional difference at the 95% confidence level.
Results:
Of 650 preceptors contacted, 177 responded and were included in the final analysis (27%). Survey item reliability was high (Cronbach α=0.96). A majority of preceptors (59% to 71%) considered TUCOM students “similar” to other medical students for each behavior. A majority of preceptors (107 [60%]) shared a definition of empathy with one another and with the National Board of Osteopathic Medical Examiners and the American Association of Colleges of Osteopathic Medicine. Approximately 39% of preceptors rated TUCOM students “better” or “advanced” for “displays of empathy” compared with approximately 30% of preceptors who rated students as better or advanced across all 10 behaviors. Preceptors who shared a definition of empathy rated TUCOM students as better or advanced at a significantly higher rate for “displays of empathy” (z=1.982, P<.05) compared with preceptors who did not share a definition (n=70). Osteopathic preceptors (n=67) rated TUCOM students significantly higher on “displays of empathy” (z=2.82, P<.05) and “clear and effective communication to patients, families and co-workers” (z=2.83, P<.01) than did allopathic preceptors. No significant differences were found based on number of years as a preceptor or on the combination of types of students the preceptor taught.
Conclusion:
Most clinical preceptors shared a definition of empathy, and they were able to observe and rate displays of it in clerkship students. Preceptors rated TUCOM students more favorably than other students they taught for displays of empathy.
Empathy is a critical physician competency thought to be an essential condition for getting to know and then satisfying patient needs, and it is known to increase patient engagement in care and compliance with therapy while leading to more accurate diagnoses.
1-3 Empathy is multidimensional, with well-established cognitive, affective, and behavioral processes.
4-7 The need for both quantitative measures and qualitative research is well established, as no measure can capture each of these dimensions at the same time.
5,7,8-14
The majority of published empathy studies have been of allopathic medical students. The handful of studies of osteopathic students that have been conducted have shown inconsistent results.
15-22 Osteopathic medical students may develop a greater degree of empathy given that one of the distinctive skills they learn in medical school is osteopathic manipulative treatment (OMT). This treatment approach is a hands-on manipulative modality that requires them to become skilled in obtaining permission to perform it, thus becoming skilled at gaining patient trust. However, direct observations or evaluations of empathetic behavior in osteopathic vs allopathic medical students have not been studied, to our knowledge. Furthermore, outcome data typically have been obtained through self-report measures taken by the physician, resident, or medical student. The perception of osteopathic medical student empathy as rated by facilitators, trainers, and patients, parallel to similar studies of allopathic medical students, is missing.
23-25
Empathy is assessed as 1 of the standardized competencies within the Humanistic Domain of the Comprehensive Osteopathic Medical Licensing Examination-USA Level 2-Performance Evaluation (COMLEX-USA Level 2-PE), a licensure examination administered by the National Board of Osteopathic Medical Examiners (NBOME) usually taken at the end of the third year of osteopathic medical school. The student score report from the examination defines empathy as “the ability to demonstrate and communicate (verbally/nonverbally) understanding, concern, and interest in the patient's medical problem and life situation.”
26 It is defined by the American Association of Colleges of Osteopathic Medicine (AACOM) as a feature of the Humanistic Behavior component of the core competency of Professionalism: “Exhibit elements of altruism and empathy by listening to patients and respecting their views.”
27 It is described within the core competency of Interpersonal and Communication Skills as a prerequisite of competent patient interaction.
27 Medical schools intending to develop their students toward competence in these areas need to understand how empathy is observed and interpreted, as well as developmental milestones that can be assessed in the curriculum.
We were interested in whether an osteopathic curriculum leads to a distinctive level or kind of empathy demonstrated in patient interactions. Therefore, we measured the affective and behavioral dimensions of empathy by analyzing faculty perceptions of medical student behavior, consistent with other recent approaches that select methods to investigate particular aspects of interest.
23,24 The aim of the present study was to obtain outcome data on the following aspects of empathy: (1) how clinical preceptors define empathy, (2) how they perceive displays of empathy, and (3) how they compare those displays among groups of third- and fourth-year medical students they have taught from different osteopathic and allopathic medical schools.
Touro University College of Osteopathic Medicine-CA (TUCOM) in Vallejo is a Jewish-sponsored osteopathic medical school. It has an annual enrollment of approximately 500 students and had graduated approximately 1700 osteopathic physicians by fall 2015, when this study was conducted. A roster of approximately 650 clinical adjunct faculty members located mostly in California but representing 37 states and the District of Columbia was selected from a database maintained by TUCOM's Clinical Education Department. The inclusion criterion was being a credentialed physician whose teaching appointment was current and based in year-3 and year-4 clerkship training. There were no exclusion criteria. A 28-item survey was distributed electronically in late October 2015 using the Qualtrics survey platform. A reminder was sent 3 days after the invitation, and the survey was closed in late November 2015.
The survey was not validated. It included 10 questions consisting of 28 items. Seven questions were single-item multiple-choice demographic questions regarding the setting of clerkship training and the frequency and years’ experience training students. A 10-item Likert-type question, from which outcome data were derived, asked respondents to compare TUCOM students with “students [they] have taught from other medical schools” (they did not identify the other medical schools) on 10 behaviors involving aspects of student-physician and student-patient professionalism and interpersonal and communication skills. Among the behaviors were 4 that were chosen because they align with definitions of empathy according to the NBOME
26 and AACOM.
27 The scale choices were “deficient,” “below,” “similar,” “better,” and “advanced,” in order of increasingly favorable comparison of TUCOM student behavior. A 9-item Likert-type scale question asked whether or not each of 9 given behaviors met respondents’ definition of physician empathy, and a final open item for feedback was included. The survey format conformed to the definition of a probability-based internet survey of specifically named individuals.
28
Responses were exported from Qualtrics into Microsoft Access, where they were tabulated and then imported for analysis into Microsoft Excel (Microsoft Corporation) and STATA / MP 13 (StataCorp LLC). Integrity of results was evaluated by calculating response rate and margin of sampling error. Response distribution was assessed for bias by calculating median, mode, and skew. Results were compared among demographic cohorts based on respondent background (osteopathic physician [ie, DO] vs allopathic physician [ie, MD]), experience teaching (veteran vs novice, other DO students, MD students, students per year), and selecting the same vs a different list of behaviors as definitive of empathy. Statistical significance of differences among cohorts was determined using a 2-sample z test of proportions. The minimum level of significance was set at 95% (P<.05) for all statistical tests. Assessment of these data was ruled exempt within the broader results of the survey of preceptor perceptions by the Institutional Review Board of Touro University California.
Of 650 preceptors contacted, 186 responded for a response rate of 29%. The survey had high item reliability (Cronbach α=0.96) and was sufficient to merit further statistical analysis. The margin of sampling error at the 95% confidence level was approximately 7.0%. Responses from 9 participants who did not answer any items comparing TUCOM students with others taught were excluded, so all further analysis was performed on a sample of 177 responses.
The majority of respondents (110 [62%]) were allopathic physicians. Most respondents (151 [85%]) had been precepting for between 1 and 8 years, most (156 [89%]) had taught an average of 1 to 8 TUCOM students per year, and most taught 0 to 4 other osteopathic (141 [80%]) or allopathic (144 [82%]) students per year. A majority (98 [55%]) was from non–primary care specialty practices, and some (13 [7%]) specialized in neuromusculoskeletal medicine/osteopathic manipulative medicine (
Table 1).
TUCOM students were rated as similar to other medical students by 105 (59%) to 126 (71%) preceptors across the 10 behaviors relating to competencies of professionalism and interpersonal and communication skills (
Table 2). Thus, on a 5-choice Likert-type scale from deficient to advanced, the preceptor evaluations of TUCOM students conformed to a strong central tendency of being rated as similar. For any given behavior, 0 to 4 respondents (2%) provided no answer.
Preceptors rated TUCOM student demonstrations of empathy in patient interactions as better or advanced compared with students from other medical schools at combined rates ranging from 42 (24%) to 69 (39%). The highest frequency (36 [39%]) of favorable ratings was for “displays of empathy” (
Table 2), with no respondents offering a rating of deficient and only 2 (1%) rating TUCOM students as below. On the other hand, preceptors rated TUCOM student behaviors as below or deficient compared with students from other medical schools at combined rates ranging from 1 (1%) to 16 (9%). The highest frequencies of unfavorable ratings were for case presentations, attendance and participation in didactics, and critical reasoning skills, none of which involve demonstrating empathy in patient interactions.
Respondents were asked to select which of 9 given behaviors they believe are attributes of empathy. A majority of respondents (107 [60%]) selected the same 4 behaviors (
Table 3), establishing the face validity of the survey. Construct validity is established by the correspondence of these 4 behavior statements to 2 external reference definitions of empathy by the NBOME
26 and AACOM,
27 as previously noted.
Compared with the cohort that did not share a definition of empathy, the cohort that did share one rated TUCOM students as better or advanced significantly more often compared with students from other medical schools for “displays of empathy” (z=1.98, P<.05) and for nonjudgmental medical history and physical examination interaction with patients (z=2.19, P<.05). Likewise, compared with MD preceptors, DO preceptors rated TUCOM students as better or advanced significantly more often compared with students from other medical schools for “displays of empathy” (z=2.82, P<.01) and for effective communication with patients and families (z=2.83, P<.01). No statistically significant differences were found between the ratings of novice preceptors (≤4 years of teaching) vs veteran preceptors (>4 years of teaching), or the ratings of preceptors who taught DO students from TUCOM and other colleges of osteopathic medicine vs those who only taught allopathic students and DO students from only TUCOM. Students at TUCOM were not rated as below or deficient compared with students from other medical schools significantly more often by any cohort of preceptors.
The survey results indicate that preceptors in our sample modally rate TUCOM students to be similarly empathetic compared with other medical students they have taught. When TUCOM students were rated as different, they were rated as better or advanced much more frequently than below or deficient, and the areas in which they were perceived by some preceptors to be below or deficient did not involve direct patient contact (
Table 2). This phenomenon is more pronounced among DO preceptors than MD preceptors, as well as among the cohort that shared a definition of empathy compared with the cohort that did not.
Compared with MD preceptors, DO preceptors rated students from TUCOM slightly more favorably. It would not be surprising if professional bias explains why the DO preceptors perceived DO students to be better or advanced in empathy compared with other medical students. However, Carey et al
29 and Licciardone
30 found that DOs are perceived to be more empathetic by patients than are MDs. It may be that the DO preceptors in our study noticed the same thing in students that patients are documented to have noticed in their DOs. Although our study design does not permit generalization to all DO students, it may nonetheless constitute preliminary evidence that DO students are more empathetic than MD students. Based on our findings, it is logical to hypothesize that DO training results in a different level or kind of empathy.
Limitations of the study include potential nonresponse bias. Approximately 40% of the surveys issued via e-mail were not opened, but it is not possible to determine how many of those were received and willfully declined by the intended recipient vs those that were delivered to dormant accounts, blocked by firewalls, or sequestered in spam folders. While we did not detect statistically significant differences in the ratings provided by novice preceptors compared with veterans, most had been teaching fewer than 5 years and seeing fewer than 5 TUCOM students per year. The relatively low average teaching experience may therefore be a source of bias in our results, and the survey design does not allow us to estimate an effect size.
Recall bias is also a possibility because our items asked respondents to evaluate and compare students after they were no longer available for observation. Moreover, the survey design may have biased the results because the questions were designed to extend our existing preceptor feedback on curriculum, and the survey was not validated or subjected to additional expert review. Perhaps most importantly, our approach of asking third-person preceptors to rate observed interactions excludes the opinions of the students and patients regarding whether the interactions were characterized by feelings of empathy; thus, our findings may incorrectly estimate empathy in both proportion and quality.
It is notable that a majority of respondents (60%,
Table 3) shared a definition of empathy both with one another and with the NBOME and AACOM, especially given the inconsistent definitions of empathy in other studies, as noted by Quince et al.
7 At the same time, the disjoin within our sample between the preceptor cohort that shares a definition of empathy with the NBOME and AACOM and those who do not implies that targeted faculty development could improve reliability of future assessments of medical student empathy. The results of this study indicate some distinction in how TUCOM students display empathy, but they do not address the cause of this distinction. Future research of DO compared with MD student empathy is warranted, as are studies that combine qualitative assessments of empathy, self-reporting, and patient perspectives to better understand how to develop physician empathy in undergraduate osteopathic medical education.
Clinical preceptors were able to evaluate displays of empathy in clerkship students and provide feedback to training programs about this important element of the patient-physician relationship. Future studies should examine how patients perceive empathy in their encounters with physicians and how empathy relates to clinical outcomes.
We thank Jay H. Shubrook, DO, for advice regarding how to investigate aspects of empathy in the clinical clerkship period, and Jaesin Sa, PhD, for advice regarding appropriate statistics to use in the comparison of survey responses to Likert-type questions.