Free
Original Contribution  |   June 2018
Reducing Health Disparities: Understanding the Unintended Effects of Health Care Professional and Patient Characteristics on Treatment
Author Notes
  • From the Department of Communication at Berry College in Mount Berry, Georgia (Dr S. Nazione), and the Floyd Medical Center in Rome, Georgia (Dr A. Nazione). 
  • Financial Disclosures: None reported. 
  • Support: None reported. 
  •  *Address correspondence to Samantha Nazione, MA, PhD, Berry College, 2277 Martha Berry Hwy NW, Box 299, Mount Berry, GA 30149-9707. Email: snazione@berry.edu
     
Article Information
Endocrinology / Medical Education / Professional Issues / Diabetes
Original Contribution   |   June 2018
Reducing Health Disparities: Understanding the Unintended Effects of Health Care Professional and Patient Characteristics on Treatment
The Journal of the American Osteopathic Association, June 2018, Vol. 118, 376-383. doi:10.7556/jaoa.2018.081
The Journal of the American Osteopathic Association, June 2018, Vol. 118, 376-383. doi:10.7556/jaoa.2018.081
Abstract

Context: The responsibility-affect-helping model proposes that helping behavior is a function of perceived responsibility and affect.

Objective: To examine the effect of medical students’ degree (DO or MD) and gender on attitudes toward patients and how these factors could act as moderators in the responsibility-affect-helping model.

Methods: This 2×3 experimental study included third- and fourth-year osteopathic (ie, DO) and allopathic (ie, MD) medical students. Students were given a survey that included the medical record and photograph of a fictitious male patient with diabetes and a message from the patient regarding his diet nonadherence. The patients differed in race (black or white) and the cause of diet nonadherence (healthy foods don't taste good, no reason given, or inability to access healthy foods). Survey items measured students’ perception of the patient's responsibility for his nonadherence, level of anger, intention to help, level of sympathy, and ethnocentrism. Data were analyzed using a multivariate analysis of covariance with ethnocentrism as a covariate.

Results: Of 1520 potential students, 231 were included in the study. Mean (SD) responsibility scale scores showed that DO students viewed the patient who gave dislike of healthy food or no reason for their diet nonadherence as more responsible for his nonadherence than did MD students (4.69 [0.99] vs 3.93 [1.00] and 4.35 [0.88] vs 3.65 [1.01], respectively). Conversely, mean (SD) responsibility scores showed that DO students viewed patients who indicated lack of access to healthy food as his reason for diet nonadherence as less responsible for his nonadherence than did MD students (2.45 [0.94] vs 2.59 [1.08]) (F2,228=3.21, P<.05, η2=.03). Furthermore, female students perceived patients to be less responsible for their diet nonadherence than did male students (3.28 [1.22] vs 3.88 [1.22]) (F2,228=8.87, P<.01, η2=.04). Ethnocentrism was consistently a significant covariate for students’ perception of patient characteristics, predicted patient behaviors, perception of the patient's responsibility for his nonadherence, students’ level of anger, students’ intention to help, and students’ level of sympathy.

Conclusion: Survey results showed that DO students perceived patients who reported dislike of healthy food or no reason for diet nonadherence as more responsible for their health issue and patients who indicated lack of access to healthy food as less responsible for their nonadherence than did MD students. Additionally, female students perceived patients to be less responsible for their health issue than did male students. Results of the current study indicate that physician demographic factors could be taken into account as proxy variables when using the responsibility-affect-helping model in the health care field.

Health disparities, defined as “differences in health status or in the distribution of health determinants between different population groups,”1 are costly and well documented.2 Reduction of health disparities is an overarching goal of the Healthy People 2020 objectives.3 One pathway to meeting this goal is through the examination of prejudice within the patient-physician relationship. Prejudice held by physicians may affect their behavior and attitudes toward patients, diagnosis decisions, and self-efficacy levels,4 which may result in lower-quality health care. 
The responsibility-affect-helping model (RAHM) may assist in reducing health disparities. This model states that people perceived as responsible for their dilemma will solicit anger from observers, which will lead to avoidance behavior.5-8 Conversely, people perceived as not responsible for their dilemma will solicit pity and sympathy from observers, which will lead to helping behavior.5-8 Research has shown empirical support for this model. A meta-analysis of 64 studies found consistent support for a causal relationship between the level of personal control/responsibility of a person and solicitation of avoidance or helping behavior.9 Additionally, positive correlations between patient responsibility and a health care professional's willingness to help,10 prescribe medication,11 and perceive patients as treatable12 have been documented. 
Medical education could be a large factor in shaping a physician's opinions; namely, whether the physician is an osteopathic physician (ie, DO) or allopathic physician (ie, MD) could alter perceptions, given the difference in medical philosophies. Carey et al13 showed that DOs discuss social, familial, and emotional factors with patients more regularly than MDs, which led to our first research question: how would students’ degree influence their perception of patient characteristics, predicted patient behaviors, perception of a patient's responsibility for his nonadherence, and students’ level of anger, level of sympathy, and intention to help? Also, Hojat et al14 found that female medical students showed more empathy, a closely related construct to sympathy, toward patients than did male medical students, which led to our second research question: how would students’ gender influence their perception of patient characteristics, predicted patient behaviors, perception of a patient's responsibility for his own health, and students’ level of anger, level of sympathy, and intention to help? Additionally, ethnocentric views, a belief in cultural superiority,15 have been found to negatively correlate to health care professionals’ cultural competence.16 Hence, ethnocentrism was used as a covariate. The current study investigated the effect of third- and fourth-year medical students’ degree and gender on how they viewed patients while controlling for ethnocentrism. 
Methods
This 2×3 experimental study was approved by Michigan State University's institutional review board. 
Participants
Medical schools were identified for possible inclusion in this study if they had a diverse student population and their universities offered both DO and MD programs. Sixteen medical schools that fit the criteria were identified, and their respective offices of medical education or student affairs were contacted via email and invited to participate. Four medical schools agreed to participate. An email with a link to the Websurveyor survey was sent to each participating school's representative to send out to their third- and fourth-year students in July 2011. Surveys were returned by October 2011. 
Survey
A fictitious male patient with type 2 diabetes mellitus was presented first. The patient reported a chief complaint of foot numbness, detailed that he had not been controlling his blood sugar levels, and noted that he understood that he needed to improve his diet. His photograph, medical record (created by a medical student [A.N.] and approved by a physician), and a message to his physician regarding his reason for diet nonadherence were provided before the survey. After clicking “I agree” to an online consent form, students were randomly assigned 1 of the 6 versions of the patient's case followed by the survey. All versions of the case had the same medical record, but they differed in race (black or white) and reason for diet nonadherence. The patient either identified that healthy foods don't taste good; gave no reason for his diet nonadherence; or identified his inability to access healthy foods. 
After reading the case, students were asked to complete a series of questions to ensure that they understood the information provided in the medical record and patient message. Questions measuring students’ perceptions of the patient were adapted from studies by Schulman et al17 and van Ryn and Burke.18 Students were asked to rank 6 patient characteristics on 7-point scales ranging from 1 to 7 (ie, hostile [1] to friendly [7]; unintelligent [1] to intelligent [7]; lacking self-control [1] to self-controlled [7]; pleasant [1] to unpleasant [7]; irrational [1] to rational [7]; independent [1] to dependent [7]; ignorant [1] to knowledgeable [7]; and poor communicator [1] to good communicator [7]). Students were also asked to rank the likelihood of 7 predicted patient behaviors, such as continued nonadherence or suing for malpractice, on a 7-point Likert scale ranging from 1 to 7, in which 1 represented the lowest predicted likelihood of a patient exhibiting a specific behavior and 7 represented the highest. 
Then, students completed survey items regarding the study's dependent variables, which measured their perception of the patient's responsibility for his nonadherence, their level of anger, their intention to help, their level of sympathy, and their level of ethnocentrism. Each variable was ranked on a 7-point scale ranging from 1 to 7, in which 1 represented the strongest level of disagreement and 7 represented the strongest level of agreement. The 7-item perceived responsibility scale created for this study asked students to rank their level of agreement with statements regarding the patient's responsibility for their nonadherence, including “The reason for non-compliance rests solely with the patient” and “This patient's lack of compliance makes him irresponsible.” Sympathy was measured using 7 items adapted from a study by Davis.19 For example, students were asked to rank their level of agreement with statements such as “In this situation, I would make an effort to put myself in the shoes of the patient” and “I am touched by this patient's circumstances.” Anger was measured using 4 items adapted from work by Schmidt and Weiner.5 Items included “This situation would make me irritated with the patient” and “The patient's actions would upset me.” A 6-item intention-to-help scale was created using questions from past RAHM studies.5,20,21 Example questions for this scale included “I would exert extra effort to help this particular patient” and “I would spend extra time to help this particular patient.” A reduced 6-item version of Neuliep and McCroskey's15 generalized ethnocentrism scale was also included in the survey. Example items included “My culture should be the role model for other cultures” and “Lifestyles in other cultures are just as valid as those in my culture” (reverse coded). Student demographics were collected at the end of the survey. 
Statistical Analysis
Data were analyzed using SPSS software (IBM). Missing data (less than 2% of total data) were replaced with mean scores for the specific variable. Descriptive statistics, including mean, SD, and frequency, were used on all variables. Confirmatory factor analyses and Cronbach α scores were analyzed to measure validity and reliability, respectively. Message manipulations were checked through a 1-way analysis of variance using Tukey post-hoc and Pearson χ2 tests. The research questions were analyzed with multivariate analysis of covariance using Tukey post-hoc tests and ethnocentrism as a covariate. Post-hoc analyses for the second research question dropped ethnocentrism as a covariate and added it as a dependent variable before independent t tests were performed. η2 scores were calculated to report effect sizes. Statistical significance was defined as P<.05. 
Results
Of the 1520 students who received invitations to participate in this study, 231 returned surveys, providing a response rate of 15.2%. Of the 231 students, 122 (53%) were third-year and 109 (47.2%) were fourth-year. More than half were MD students (155 of 231 [67.4%]), 75 of 231 were DO students (32.6%), and 1 did not report his or her degree program (0.4%). Slightly more female students (124 of 231 [53.7%]) than male students (105 of 231 [45.5%]) were included in the study, with 2 students (0.9%) not indicating their gender. The mean (SD) age of the students was 25.37 (2.05) years. The majority of the students were white (161 of 231 [69.7%]), followed by Asian American (41 of 231 [17.7%]), students who indicated more than 1 race/ethnicity (11 of 231 [4.8%]), students who indicated “other” (10 of 231 [4.3%]), Hispanic/Latino (4 of 231 [1.7%]), and black (2 of 231 [0.9%]). Two of 231 students (0.9%) did not report their race/ethnicity. Of the 231 students, 39 (16.9%) received the medical record of a black patient whose reason for nonadherence was dislike of healthy food, 44 (19.0%) received the medical record of a white patient whose reason for nonadherence was dislike of healthy food, 36 (15.6%) received the medical record of a black patient who gave no reason for diet nonadherence, 36 (15.6%) received the medical record of a white patient who gave no reason for diet nonadherence, 40 (17.3%) received the medical record of a black patient whose reason for nonadherence was lack of access to healthy food, and 36 (15.6%) received the medical record of a white patient whose reason for nonadherence was lack of access to healthy food. 
Responses from the series of questions included to ensure that students understood the information provided in the medical record and patient message indicated that the sex, age range, health issue, race, income, and responsibility status of the fictitious patients were perceived as intended. The mean (SD) responsibility scale score of students who were assigned a patient whose reason for diet nonadherence was inability to access healthy foods was significantly lower than students whose assigned patient reported dislike of healthy foods or gave no reason for diet nonadherence (3.46 [1.06], 4.86 [0.84], and 5.27 [0.80], respectively) (F2,228=85.87, P<.001, η2=0.43) (large effect). Students who were assigned a patient whose reason for nonadherence was inability to access healthy foods had a lower mean (SD) score when asked to rank the patient's income range, than did students whose assigned patient reported dislike of healthy foods or no reason for their nonadherence (2.25 [0.75], 3.23 [1.02], and 3.15 [1.08], respectively) (F2,228=24.66, P<.001, η2=.13) (medium effect). Furthermore, all but 1 student correctly identified the patient as male, and all students identified the patient's correct race and health issue. Descriptive statistics and reliability for each factor are presented in Table 1, and confirmatory factor analysis results regarding validity are presented in Table 2. 
Table 1.
Descriptive Statistics and Reliability for Each Factor Included in a Survey Assessing How Medical Students View a Diabetic Patient Who Was Nonadherent to Diet
Factor Range Mean (SD) Score Cronbach α
Favorable perceptions of patient characteristics 3.17-7.00 5.34 (0.74) 0.76
Favorable perceptions of patient behavior 3.00-7.00 5.65 (0.77) 0.80
Patient perceptions of the patient's responsibility for his health issue 1.00-6.71 3.56 (1.26) 0.91
Student level of anger 1.00-6.00 2.58 (1.24) 0.89
Student intention to help 3.33-7.00 6.00 (0.77) 0.82
Student level of sympathy 2.57-7.00 5.76 (0.75) 0.82
Students’ ethnocentrism 1.00-5.17 2.41 (0.87) 0.70

a Variables were assessed on 7-point scales in which 1 indicates the lowest level of the construct and 7 indicates the highest level of the construct.

Table 1.
Descriptive Statistics and Reliability for Each Factor Included in a Survey Assessing How Medical Students View a Diabetic Patient Who Was Nonadherent to Diet
Factor Range Mean (SD) Score Cronbach α
Favorable perceptions of patient characteristics 3.17-7.00 5.34 (0.74) 0.76
Favorable perceptions of patient behavior 3.00-7.00 5.65 (0.77) 0.80
Patient perceptions of the patient's responsibility for his health issue 1.00-6.71 3.56 (1.26) 0.91
Student level of anger 1.00-6.00 2.58 (1.24) 0.89
Student intention to help 3.33-7.00 6.00 (0.77) 0.82
Student level of sympathy 2.57-7.00 5.76 (0.75) 0.82
Students’ ethnocentrism 1.00-5.17 2.41 (0.87) 0.70

a Variables were assessed on 7-point scales in which 1 indicates the lowest level of the construct and 7 indicates the highest level of the construct.

×
Table 2.
Variable Confirmatory Factor Analysis for Each Factor Included in a Survey Assessing How Medical Students View a Diabetic Patient Who Was Nonadherent to Diet
Factor χ2 P Value Confirmatory Fit Index Tucker-Lewis Index Root Mean Square Error of Approximation
Favorable perceptions of patient characteristics 14.75 .098 0.98 0.97 0.05
Favorable perceptions of patient behavior 27.25 .018 0.97 0.96 0.06
Perceptions of the patient's responsibility for his health issue 45.6 <.001 0.97 0.95 0.10
Student level of anger 4.41 .11 0.99 0.99 0.07
Student intention to help 19.34 .022 0.98 0.97 0.07
Student level of sympathy 17.7 .171 0.99 0.99 0.04
Students ethnocentrism 19.1 .024 0.96 0.93 0.07
Table 2.
Variable Confirmatory Factor Analysis for Each Factor Included in a Survey Assessing How Medical Students View a Diabetic Patient Who Was Nonadherent to Diet
Factor χ2 P Value Confirmatory Fit Index Tucker-Lewis Index Root Mean Square Error of Approximation
Favorable perceptions of patient characteristics 14.75 .098 0.98 0.97 0.05
Favorable perceptions of patient behavior 27.25 .018 0.97 0.96 0.06
Perceptions of the patient's responsibility for his health issue 45.6 <.001 0.97 0.95 0.10
Student level of anger 4.41 .11 0.99 0.99 0.07
Student intention to help 19.34 .022 0.98 0.97 0.07
Student level of sympathy 17.7 .171 0.99 0.99 0.04
Students ethnocentrism 19.1 .024 0.96 0.93 0.07
×
Mean (SD) responsibility scale scores showed that DO students perceived their assigned patient as more responsible for his diet nonadherence than did MD students (3.99 [1.30] vs 3.36 [1.19]) (F2,228=13.81, P<.001, η2=.05). Ethnocentrism was a significant covariate for students’ perception of patient characteristics (F2,228=15.63, P<.001, η2=.06) (small effect), predicted patient behaviors (F2, 228=12.51, P<.001, η2=.05), perception of the patient's responsibility for his nonadherence (F2,228=13.33, P<.001), level of anger (F2,228=13.42, P<.001, η2=.05) (small effect), intention to help (F2,228=41.43, P<.001, η2=.15) (medium effect), and level of sympathy (F2,228=21.33, P<.001, η2=.09) (small effect). 
The relationship between student degree and how students viewed their patient based on his reason for diet nonadherence is shown in the Figure. Mean (SD) responsibility scale scores showed that DO students viewed the patient who indicated dislike of healthy foods or no reason for nonadherence as more responsible for his nonadherence than did MD students (4.69 [0.99] vs 3.93 [1.00] and 4.35 [0.88] vs 3.65 [1.01], respectively) (P<.05). Conversely, DO students were found to view the patient who gave lack of access to healthy foods as his reason for diet nonadherence as less responsible for his nonadherence than did MD students (2.45 [0.94] vs 2.59 [1.08]) (F2,228=3.21, P<.05, η2=.03) (small effect). With reason for nonadherence, ethnocentrism was again a significant covariate for students’ perception of patient characteristics (F2,228=14.73; P<.001; η2=.06) (small effect), predicted patient behaviors (F2,228=12.42; P<.01; η2=.05) (small effect), perception of the patient's responsibility for his nonadherence (F2,228=15.99; P<.001; η2=.04) (small effect), level of anger (F2,228=13.03; P<.001; η2=.05) (small effect), intention to help (F2,228=40.37; P<.001; η2=.15) (medium effect), and level of sympathy (F2,228=21.12; P<.001; η2=.08) (small effect). 
Figure.
Mean perceived responsibility scale scores of osteopathic (ie, DO) and allopathic (ie, MD) medical students asked to rank a fictitious patient's responsibility for his own health issue on a scale of 1 to 7, in which 1 represents the lowest amount of responsibility and 7, the highest, after viewing his photograph, medical record, and reason for nonadherence to a recommended diet.
Figure.
Mean perceived responsibility scale scores of osteopathic (ie, DO) and allopathic (ie, MD) medical students asked to rank a fictitious patient's responsibility for his own health issue on a scale of 1 to 7, in which 1 represents the lowest amount of responsibility and 7, the highest, after viewing his photograph, medical record, and reason for nonadherence to a recommended diet.
The mean (SD) responsibility scale score for female students was lower than for male students (3.28 [1.22] vs 3.88 [1.22]) (F2,228=8.87, P<.01, η2=.04). With student gender, ethnocentrism was again a significant covariate for students’ perception of patient characteristics (F2,228=17.94; P<.001, η2=.07) (small effect), predicted patient behaviors (F2,228=12.50; P=.001, η2=.05) (small effect), perception of the patient's responsibility for his own nonadherence (F2,228=9.98; P<.01, η2=.04) (small effect), level of anger (F2,228=15.11; P<.001, η2=.06) (small effect), intention to help (F2,228=37.21; P<.001, η2=.14) (small effect), and level of sympathy (F2,228=18.66; P<.001, η2=.15) (medium effect). 
Mean (SD) sympathy and intention to help scale scores showed that female students expressed more sympathy than did male students (5.87 [0.74] vs 5.62 [0.74]) (t227=−2.53, P<.05, η2=.17) (medium effect) and greater intentions to help patients (6.14 [0.69] vs 5.89 [0.85]) (t227=−2.50, P<.05, η2=.03) (small effect). Female students had a lower mean (SD) ethnocentrism scale score than did male students (2.24 [0.82] vs 2.63 [0.89]) (t227=3.36, P<.01, η2=.22) (medium effect), as well as a lower mean (SD) responsibility scale score (3.28 [1.22] vs 3.88 [1.22]) (t227=-3.71, P<.001, η2=.24) (medium effect). 
Discussion
Prejudice is an important variable to understand in the health care setting because it can lead to poor health outcomes for patients.4 Poor communication between patients and physicians can also cause poor outcomes for patients.22 One model from which knowledge might be gleaned to work toward goals of reducing prejudice in health care and improving communication between patients and physicians is RAHM. 
Results from the current study show that physician factors, such as degree and gender, need to be taken into account when using RAHM. The finding that DO students viewed the patient who gave lack of access to healthy foods as his reason for nonadherence as less responsible for his health issue than MD students aligns with a previous study that found that DOs were more likely than MDs to discuss social issues, such as access or lifestyle factors, with patients.13 This finding may be indicative of greater awareness of social issues by DOs, and it shows how DOs may help alleviate health disparities by viewing patients as a unit of body, mind, and spirit.23 On the other hand, DO students’ perception of a patient having greater responsibility for his nonadherence when his reason for nonadherence was a dislike of healthy foods should be addressed as well. Although greater perceived responsibility led to greater anger and reduced intentions to help patients, DO students may have viewed this patient as more responsible because they are more likely to consider social issues and, therefore, might also expect patients to be more participative in their health care. 
Ethnocentrism as a covariate produced small to medium effect sizes for all dependent variables. It has been shown that female students are more empathic than male students,14 which may be linked to lower ethnocentrism in female students compared with male students. This factor will be an important covariate to include in future studies. In addition to the fact that student degree and gender were supported by small effect sizes, it is important to note that these variables likely serve as proxies in this study. Degree type is a proxy for receiving a medical education driven by the osteopathic philosophy. However, DO and MD students likely differ in other ways aside from medical philosophies, such as region of birth or socioeconomic status. Before advocating for osteopathic philosophy as a source of these results, future studies should seek to measure one's medical philosophy more directly to understand how it may influence RAHM. Similarly, student gender may serve as a proxy variable for the trait of greater sympathy. Still, an important lesson can be taken from this study's findings; physicians should ask patients to disclose barriers they face in seeking quality health care. Retrieving this social history, specifically from patients not responsible for their inability to enact healthy behaviors, may lead to increased intention to help via increased sympathy. This act may benefit those facing health disparities for such reasons as access issues. 
This study was not without limitations. Ethnocentrism was important to study outcomes. However, the ethnocentrism scale had somewhat low reliability (α=0.70). Furthermore, a confound of socioeconomic status was found in this study's manipulation, so results may be partially based on perceived income. However, perceived income was not found to be significantly correlated with students’ perception of the patient responsibility for his health issue, level of anger, intention to help, or level of sympathy. Furthermore, the effect size for the responsibility manipulation (0.43) was much larger than the effect size of socioeconomic status (0.13). Students included in this study were mostly white, and more MD than DO students participated. Future studies should include a more diverse study population and a larger DO sample to better understand the effect of a physician's race on the factors that may affect RAHM and to obtain more generalizable results. 
Conclusion
Results of the current study indicate that physician demographic factors could be taken into account as proxy variables when using RAHM in the health care field. Additionally, female students perceived patients to be less responsible for their health issue than did male students, regardless of their reason for nonadherence, which may indicate greater sympathy and intentions to help patients among women than men in a clinical setting. 
References
Health Impact Assessment (HIA) glossary of used terms. World Health Organization website. http://www.who.int/hia/about/glos/en/index1.html. Accessed January 20, 2017.
CDC Health Disparities and Inequalities Report—United States, 2013. Atlanta, GA: Centers for Disease Control and Prevention; 2013. http://www.cdc.gov/minorityhealth/CHDIReport.html. Accessed December 9, 2016.
United States Dept of Health and Human Services. Disparities. HealthyPeople website. https://www.healthypeople.gov/2020/about/foundation-health-measures/Disparities. Accessed December 9, 2016.
van Ryn M, Fu SS. Paved with good intentions: do public health and human service providers contribute to racial/ethnic inequalities in health? Am J Public Health. 2003;93(2):248-255. doi: 10.2105AJPH.93.2.248 [CrossRef] [PubMed]
Schmidt G, Weiner B. An attribution-affect-action theory of behavior: replications of judgments of help-giving. Pers Soc Psychol Bulletin. 1988;14(3):610-621. doi: 10.1177/0146167288143021 [CrossRef]
Weiner B. A cognitive (attribution)-emotion-action model of motivated behavior: an analysis of judgments of help-giving. J Pers Soc Psychol. 1980;39(2):186-200. doi: 10.1037/0022-3514.39.2.186 [CrossRef]
Weiner B. Human Motivation. Hillsdale, NJ: Lawrence Erlbaum Associates; 1989.
Weiner B, Perry RP, Magnusson J. An attributional analysis of reactions to stigmas. J Pers Soc Psychol. 1988;55(5):738-748. [CrossRef] [PubMed]
Rudolph U, Roesch S, Greitemeyer T, Weiner B. A meta-analytic review of help giving and aggression from an attributional perspective: contributions to a general theory of motivation. Cogn Emotion. 2004;18(6):814-848. doi: 10.1080/02699930341000248 [CrossRef]
Stanley B, Standen PJ. Carers’ attribution for challenging behavior. Br J Clin Psychol. 2000;39(2):157-168. [CrossRef] [PubMed]
Leggett J, Silvester J. Care staff attributions for violent incidents involving male and female patients: a field study. Br J Clin Psychol. 2003;42(4):393-406. doi: 10.1348/014466503322528937 [CrossRef] [PubMed]
Reid D, Millard R. The application of attribution theory to perceptions of dangerousness and treatability. In: Deu N, Roberts L, eds. Dangerous, Disordered and Doubly Deviant: Issues in Criminological and Legal Psychology. . Leicester, England: British Psychological Society; 1997:4-9.
Carey TS, Motyka TM, Garrett JM, Keller RB. Do osteopathic physicians differ in patient interaction from allopathic physicians? an empirically derived approach. J Am Osteopath Assoc. 2003;103(7):313-318. [PubMed]
Hojat M, Gonnella JS, Mangione S, et al. Empathy in medical students as related to academic performance, clinical competence and gender. Med Educ. 2002;36(6):522-527. [CrossRef] [PubMed]
Neuliep JW, McCroskey JC. The development of a U.S. and generalized ethnocentrism scale. Commun Res Rep. 1997;14(4):385-398. doi: 10.1080/08824099709388682 [CrossRef]
Capell J, Dean E, Veenstra G. The relationship between cultural competence and ethnocentrism of health care professionals. J Transcul Nurs. 2008;19(2):121-125. doi: 10.1177/1043659607312970 [CrossRef]
Schulman KA, Berlin JA, Harless W, et al The effect of race and sex on physicians’ recommendations for cardiac catheterization. N Engl J Med. 1999;340(8):616-626. doi: 10.1056/NEJM199902253400806 [CrossRef]
van Ryn M, Burke J. The effects of patient race and socio-economic status on physicians’ perceptions of patients. Soc Sci Med. 2000;50(6):813-828. [CrossRef] [PubMed]
Davis MH. Measuring individual differences in empathy: evidence for a multidimensional approach. J Pers Soc Psychol. 1983;44(1):70-87. doi: 10.1037/0022-3514.44.1.113 [CrossRef]
Menec VH, Perry RP. Reactions to stigmas among Canadian students: testing an attribution-affect-help judgment model. J Soc Psychol. 1998;138(4):443-453. [CrossRef]
Sharrock R, Day A, Qazi F, Brewin CR. Explanations by professional care staff, optimism and helping behaviour: an application of attribution theory. Psychol Med. 1990;20(4):849-855. [CrossRef] [PubMed]
Agency for Healthcare Research and Quality. 2015 National Healthcare Disparities Report and 5th Anniversary Update on The National Quality Strategy. Rockville, MD: United States Dept of Health and Human Services; 2016. http://www.ahrq.gov/sites/default/files/wysiwyg/research/findings/nhqrdr/nhqdr15/2015nhqdr.pdf. Accessed December 9, 2016.
Seffinger MA, King HH, Ward RC, Jones JMIII, Rogers FJ, Patterson MM. Osteopathic philosophy. In: Chila A, executive ed. Foundations of Osteopathic Medicine. 3rd ed. Baltimore, MD: Lippincott Williams & Wilkins; 2011:15-16.
Figure.
Mean perceived responsibility scale scores of osteopathic (ie, DO) and allopathic (ie, MD) medical students asked to rank a fictitious patient's responsibility for his own health issue on a scale of 1 to 7, in which 1 represents the lowest amount of responsibility and 7, the highest, after viewing his photograph, medical record, and reason for nonadherence to a recommended diet.
Figure.
Mean perceived responsibility scale scores of osteopathic (ie, DO) and allopathic (ie, MD) medical students asked to rank a fictitious patient's responsibility for his own health issue on a scale of 1 to 7, in which 1 represents the lowest amount of responsibility and 7, the highest, after viewing his photograph, medical record, and reason for nonadherence to a recommended diet.
Table 1.
Descriptive Statistics and Reliability for Each Factor Included in a Survey Assessing How Medical Students View a Diabetic Patient Who Was Nonadherent to Diet
Factor Range Mean (SD) Score Cronbach α
Favorable perceptions of patient characteristics 3.17-7.00 5.34 (0.74) 0.76
Favorable perceptions of patient behavior 3.00-7.00 5.65 (0.77) 0.80
Patient perceptions of the patient's responsibility for his health issue 1.00-6.71 3.56 (1.26) 0.91
Student level of anger 1.00-6.00 2.58 (1.24) 0.89
Student intention to help 3.33-7.00 6.00 (0.77) 0.82
Student level of sympathy 2.57-7.00 5.76 (0.75) 0.82
Students’ ethnocentrism 1.00-5.17 2.41 (0.87) 0.70

a Variables were assessed on 7-point scales in which 1 indicates the lowest level of the construct and 7 indicates the highest level of the construct.

Table 1.
Descriptive Statistics and Reliability for Each Factor Included in a Survey Assessing How Medical Students View a Diabetic Patient Who Was Nonadherent to Diet
Factor Range Mean (SD) Score Cronbach α
Favorable perceptions of patient characteristics 3.17-7.00 5.34 (0.74) 0.76
Favorable perceptions of patient behavior 3.00-7.00 5.65 (0.77) 0.80
Patient perceptions of the patient's responsibility for his health issue 1.00-6.71 3.56 (1.26) 0.91
Student level of anger 1.00-6.00 2.58 (1.24) 0.89
Student intention to help 3.33-7.00 6.00 (0.77) 0.82
Student level of sympathy 2.57-7.00 5.76 (0.75) 0.82
Students’ ethnocentrism 1.00-5.17 2.41 (0.87) 0.70

a Variables were assessed on 7-point scales in which 1 indicates the lowest level of the construct and 7 indicates the highest level of the construct.

×
Table 2.
Variable Confirmatory Factor Analysis for Each Factor Included in a Survey Assessing How Medical Students View a Diabetic Patient Who Was Nonadherent to Diet
Factor χ2 P Value Confirmatory Fit Index Tucker-Lewis Index Root Mean Square Error of Approximation
Favorable perceptions of patient characteristics 14.75 .098 0.98 0.97 0.05
Favorable perceptions of patient behavior 27.25 .018 0.97 0.96 0.06
Perceptions of the patient's responsibility for his health issue 45.6 <.001 0.97 0.95 0.10
Student level of anger 4.41 .11 0.99 0.99 0.07
Student intention to help 19.34 .022 0.98 0.97 0.07
Student level of sympathy 17.7 .171 0.99 0.99 0.04
Students ethnocentrism 19.1 .024 0.96 0.93 0.07
Table 2.
Variable Confirmatory Factor Analysis for Each Factor Included in a Survey Assessing How Medical Students View a Diabetic Patient Who Was Nonadherent to Diet
Factor χ2 P Value Confirmatory Fit Index Tucker-Lewis Index Root Mean Square Error of Approximation
Favorable perceptions of patient characteristics 14.75 .098 0.98 0.97 0.05
Favorable perceptions of patient behavior 27.25 .018 0.97 0.96 0.06
Perceptions of the patient's responsibility for his health issue 45.6 <.001 0.97 0.95 0.10
Student level of anger 4.41 .11 0.99 0.99 0.07
Student intention to help 19.34 .022 0.98 0.97 0.07
Student level of sympathy 17.7 .171 0.99 0.99 0.04
Students ethnocentrism 19.1 .024 0.96 0.93 0.07
×