|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2003 American Society for Clinical Oncology
The Impact of a Physician Awareness Group and the First Year of Training on Hematology-Oncology Fellows
From the Department of Hematology and Medical Oncology, The Cleveland Clinic Foundation, Cleveland, OH; the Harvard Program in Refugee Trauma, Massachusetts General Hospital; the Massachusetts General Hospital Cancer Center, Harvard Medical School, Boston; Clark University; and the Department of Family Medicine, University of Massachusetts, Worcester, MA. Address reprint requests to Mikkael A. Sekeres, MD, MS, Department of Hematology and Medical Oncology, The Cleveland Clinic Foundation, Desk R35, 9500 Euclid Ave, Cleveland, OH 44195; e-mail: sekerem{at}ccf.org.
Purpose: To assess the impact of a Balint-like physician awareness group on hematology-oncology fellows’ attitudes and measure changes in attitudes during the first fellowship year. Patients and Methods: We used a modified crossover design in which one half of a fellowship class at a time was exposed to the group intervention over a 2-year period (2000 to 2002). Two 14-fellow classes were followed for 1 year each and were given three "attitudes" questionnaires, at the beginning, middle, and end of the academic year. Results: Forty Balint group sessions were held during the 2-year study period; 82 questionnaires of the 84 administered (98%) were recovered. Instrument content and criterion validity were demonstrated, as was topic domain reliability. Overall, mean attitude scores increased following the group intervention, from 3.6 (95% CI, 3.5 to 3.7) to 3.7 (95% CI, 3.6 to 3.8; P = .09). Within domains, scores increased in a "fellow’s views of him/herself as a physician," from 3.8 (95% CI, 3.6 to 3.9) to 4.1 (95% CI, 3.9 to 4.2; P = .008) and "comfort dealing with emotional patient/clinical situations," from 3.5 (95% CI, 3.3 to 3.7) to 3.7 (95% CI, 3.6 to 3.9; P = .11). Changes in responses to individual questions included: an increase in fellows’ comfort with the technical aspects of being an oncologist (P < .03); an increase in fellows’ comfort with discussing the stress of home at work (P < .023); and an increase among fellows in feeling pressed for time to discuss psychosocial issues with patients (P = .035). Conclusion: A physician awareness group was feasible and enhanced fellows’ development as physicians. Further research is needed to determine how to incorporate such groups into oncology fellowships.
MEDICAL TRAINING at the intern, resident, and fellow levels can present substantial challenges, including sleep deprivation, confrontation with patients’ emotional reactions to serious illness, and isolation from family and friends, all in the setting of inexperience. Depression, substance abuse, marital discord, and suicide are among the byproducts of this stress.1–6 The training of future oncologists is even more stressful because of the substantial exposure to death and dying.7,8 Oncologists express dissatisfaction with their training and their ability to communicate bad news to patients and to support terminally ill patients.9–13 House officers, who are novices in identifying feelings engendered by patients, also recognize that they deal poorly or inadequately with terminally ill patients.14,15 Yet cancer patients depend on their doctors to communicate information clearly, accurately, and empathically; moreover, successful conveyance of such information is associated with patient satisfaction.10,16–18 Seriously ill patients (n = 340) in a recent survey 19 identified items rated as important in their relationships to health care professionals. In particular, patients wanted a physician with whom they could discuss personal fears and one who would be comfortable discussing death and dying.19 Physicians’ personal characteristics, past experiences, values, and biases can have important effects on communication with patients; awareness of these characteristics enhance communication.20 Training programs have acknowledged the impact of these factors on patient-doctor interactions and on physician stress by developing efforts that include individual and group meetings; weekly support meetings; psychiatrically oriented, case-centered discussions of patients; and one-time retreats with formal process-oriented group discussions.5,10,13,21 One such effort, common in primary care residency programs, is based on the work of Michael and Enid Balint.22–25 In these physician-led or physician- and psychologist-led Balint groups, members (house officers) bring cases to group meetings, during which the discussion focuses on patient-doctor interactions. Balint groups are present in almost half (48%) of family practice residencies surveyed.22 Balint-like groups or physician awareness groups, offshoots of classical Balint groups, are case-focused only a percentage of the time and may include as objectives providing support for participants; learning management strategies for various illnesses, and resolution of professional role conflict.22,26 Within hematology-oncology fellowships, there is a clear need for psychosocial programs to address the stresses fellows face in treating acutely ill, often terminal patients, as they learn the technical aspects of cancer care. These programs should also provide the psychosocial skills needed for the optimal care of cancer patients,27,28 given that the quality of care for a cancer patient depends on the quality of an oncology fellow’s development of professional skills. We instituted a Balint-like physician awareness group in two successive first-year hematology-oncology fellowship classes. Our main objectives were: (1) to assess the impact the group intervention had on the fellows’ perceptions of how they related to patients and colleagues; (2) to evaluate the validity and reliability of an "attitudes" questionnaire measuring the fellows’ perceptions; and (3) to describe changes in fellows’ attitudes during the course of the first fellowship year.
Subjects As part of the Dana-Farber/Partners hematology-oncology fellowship, 14 first-year hematology-oncology fellows spend 6 months at the Dana-Farber Cancer Institute/Brigham and Women’s Hospital (DFCI/BWH) and 6 months at the Massachusetts General Hospital (MGH), with the majority of time caring for hematology or oncology patients in both outpatient and inpatient settings. Within a class of 14 first-year fellows, seven fellows start at the DFCI/BWH site, and seven start at the MGH site. On January 1 of each year, fellows switch locations; those who started at DFCI/BWH go to MGH, and those who started at MGH switch to DFCI/BWH. This study covers two classes of first-year hematology-oncology fellows, from July 2000 through June 2002, for a total of 28 fellows.
Intervention A typical group meeting started with one of the group leaders asking an open-ended question of the participants, such as: "So how have the past couple of weeks been for you?" The ensuing discussion was initiated by a first-year fellow, based on events that had occurred in the 2 weeks preceding a meeting. One or two major topics were covered in each session. As opposed to a traditional Balint group, which gives physicians the opportunity to explore the doctor-patient interaction through their own clinical experiences and makes the doctor-patient interaction the topic focus, the Balint-like group initiated in our fellowship allowed discussions of other areas considered important to first-year fellows.29 These included: developing identity as a physician, family-of-origin issues (for both patient and physician), psychosocial issues (for both patient and physician), difficult patient interactions, difficult administrative and collegial issues, general stresses of fellowship, and balancing personal and professional roles. The group leaders consisted of a psychiatrist familiar with cancer patients (D.B.G.) and an upper-year hematology-oncology fellow (M.A.S.), neither of whom was in an evaluative role for the first-year fellows. The upper-year fellow attended a formal training workshop on becoming a Balint group leader. We believed this combination of leaders was important, because the upper-year fellow was able to address issues related directly to the fellowship structure and clinical issues faced by first-year fellows, essentially assuming an "older sibling" role; the psychiatrist, however, assumed a senior mentoring role, ensuring that the group process focused on constructive interactions and that expertise in patient-focused psychosocial issues would be provided. At least two fellows had to be present for a group session to take place. All material discussed in the group was considered strictly confidential, so that participants felt free to present difficult, sensitive, and often personal issues.
Evaluation Questionnaires were administered and collected anonymously, which compromised the ability to perform paired comparisons but allowed participants to respond openly and honestly. At the conclusion of the group intervention, each fellow was given a separate questionnaire to evaluate the effectiveness of the intervention.
Instruments
Group Evaluation This 21-item survey was adapted from one developed at the University of Massachusetts to evaluate its family practice residency Balint group. The University of Massachusetts survey was administered to 160 physicians who participated in the Balint group and who graduated from the residency program between 1972 and 1996. It was developed (by E.I.K. and D.H.L.) based on a critical review of research on other Balint groups and on the intended teaching goals of the leaders of the University of Massachusetts Balint group. Respondents again used a five-point Likert scale, anchored by "strongly disagree" and "strongly agree," to answer each question.
Statistics This study used a modified crossover design: Although subjects completed the attitudes questionnaire at three time points, repeated measurements on each subject were not identifiable. Therefore, the analysis used two-group methods. The overall null hypothesis under investigation was that there was no effect of the group on attitudes questionnaire scores versus the two-sided alternative hypothesis of either a rise or decrease in attitudes scores. In general, results were considered to be statistically significant if the P value of the test was less than .05. However, for the exploratory analyses, larger P values were considered to be indicative of trends. A Bonferroni correction was applied to adjust for multiple comparisons and did not change the significance of the change in attitudes measured by the domains. The primary hypothesis was that the group intervention would have an impact on overall mean and topic-domain attitude scores. Data for the time point just before the start of the group and just after the group were abstracted for each cohort. The normal approximation to the Wilcoxon rank sum test was used to assess whether the group had an effect on attitude scores for each topic domain, where the two categories consisted of pregroup scores and postgroup scores. To control for cohort effects, for the study span of two fellowship years, and for the different times during the year that fellows received the survey, a stratified Wilcoxon rank sum test was performed, using exact computational algorithms,34 using the site-year cohort as a stratification factor. The stratified Wilcoxon rank sum test combines the results of the four individual rank-sum tests within each of the cohorts. As fellowship training itself could have an impact on attitudes scores, an exact, stratified Wilcoxon rank sum test was performed on the training-period-only data to confirm whether or not the fellowship itself (in the absence of the group intervention), had an impact on attitudes. To compare changes during the course of the fellowship year, a one-way analysis of variance was performed on individual item scores, irrespective of start site (MGH or DFCI/BWH) or fellowship year. The group evaluation survey was analyzed descriptively, looking at the overall mean score and the 21 item scores. The overall mean score between two ad hoc subject groups was compared. Scores for those who attended six or more group sessions, and for those who attended fewer than six (the median number of sessions attended), were compared with normal theory t tests.
Ten group sessions were held every six months, for a total of 40 sessions during the 2-year study period. Topics discussed were consistent across different fellow groups and mirrored those topics cited in publications of Balint groups held in other settings.20,22 They included (1) Patient-focused issues, such as caring for extremely ill or dying patients; managing the suicidal patient; difficult patient or family interactions; unequal care for patients of different social or ethnic backgrounds; cultural issues patients bring to patient-health care professional interactions; and advocating for patients. For example, one discussion focused on a fellow’s role in mediating among a dying patient, a family who felt ill-prepared for the patient’s imminent death, and an attending who the fellow felt had distanced himself from the patient and family. (2) Fellow-focused issues, including the role of the fellow in relation to the patient, his or her family, and the oncology attending or nurse (professional role conflict); emotional responses to patients, including "letting go" of dying patients or saying goodbye to patients at the end of fellowship, and setting boundaries; cultural issues caregivers bring to a patient-caregiver interaction; the stress of fellowship and of being an oncologist, particularly the dialectic among fixing a disease, studying a disease, and palliating a patient’s symptoms; personal/family illness; balancing personal and professional responsibilities; dealing with the limits of medicine; and the idea of a "calling" to be an oncologist. A recurring topic involved which patients fellows would continue to follow in a continuity clinic as they entered the second year of fellowship and how fellows select those patients based on personal relationships or professional goals and inform other patients that they will be "leaving" at the end of the year. (3) Professional issues, such as the development of the physician over time (both personally and professionally, including future professional directions); balancing psychosocial and oncology education; interactions with other fellows or staff; the politics of a hospital, university, and government; and medical mistakes and their impact. As the academic year closed, many discussions turned to the balance of research, clinical, and personal time and how these issues related to securing salaries and/or grant support. Each topic listed in the Results section was discussed by at least two different fellow groups.
Each academic year, the same attitudes questionnaires were administered to all 14 fellows (for a total of 28 fellows during the 2-year period) three times each: at the beginning, middle, and end of the academic year. We retrieved a total of 82 completed questionnaires of the 84 administered (98%), and divided the questionnaire into four topic domains (Fig 1
Next, we assessed changes in topic domain scores pre- and post-Balint group intervention. Both mean and median summary scores are presented in Table 1
We then controlled for the effect of the timing of the group intervention (first half of the year v second half of the year) and the academic year of the intervention (first-year fellows in academic year 2000–2001 v first-year fellows in academic year 2001–2002) and did not detect a significant change from unstratified values (Table 1  ). We also found that scores did not change during these time periods in the group of fellows who did not receive the group intervention (the control arm), indicating that changes pre- and post-group seem to have resulted from the group itself, and not simply from time spent in training or from the timing of the intervention.
Responses to individual questionnaire items during the course of the academic year were examined irrespective of the group intervention. Scores on six items changed significantly over the course of the year, more than would be expected by chance alone (Table 2
A separate group evaluation questionnaire was given to fellows to learn their perceptions of the quality of the intervention. As expected, mean scores for questionnaire responses corresponded to the number of group sessions attended: Fellows who attended six or more sessions (n = 15) had higher scores, indicating a more positive experience, while those who attended fewer sessions (five or fewer, n = 11) had lower scores, indicating more dissatisfaction with the group experience (3.1 [95% CI, 2.6 to 3.5] v 2.4 [95% CI, 1.9 to 2.9]; P = .033). Overall, fellows attended a median of six (interquartile range, five to seven) sessions. Two respondents did not indicate the number of sessions they attended.
We divided responses to individual questions on the group evaluation questionnaire into high, medium, and low levels of agreement (Table 3
We initiated a Balint-like physician awareness group for first-year hematology-oncology fellows and assessed the fellows’ responses to questions on a survey of attitudes over the course of their first year of fellowship. We were able to demonstrate that this instrument had good content and criterion validity and topic domain reliability35 and that it was an acceptable assessment tool that could evaluate changes in repeated measures during a 1-year period. As a result of the group, there was a trend toward improved mean scores on the attitudes questionnaire as a whole and a significant improvement in a fellow’s view of himself or herself as a physician. A suggestion of improved scores also occurred in the topic domain assessing comfort in dealing with emotional patient and clinical situations. A majority of group sessions, all of which were initiated by fellows, focused on difficult patient interactions and led to discussions of underlying biases or unexplored individual experiences or beliefs on the part of the patient or physician; issues of emotional attachment (or detachment) and boundaries; and on the transition in responsibility from resident to fellow—precisely the items addressed in topic domains that showed a change or a suggestion of a change. Balint groups and Balint-like groups provide physicians with a greater understanding of the issues underlying difficult interactions,23,24 and thus with more insight into future interactions, making them less difficult. Nease et al36 noted that the essential outcome of a Balint group was meant to be a reconceptualization of the physician-patient relationship so that the physician could try new behaviors. To the extent that the discussion enables the physician to meet patients’ needs better, physician stress is reduced. As responses to these domains indicate, these issues warrant further exploration in a larger sample of fellows. Refinement of items within each domain to better reflect actual issues discussed in group session may improve the ability of these domains to detect change. The "stress in the work environment" domain and "discomfort with psychosocial issues" domains, on the other hand, did not appear to change as a result of the intervention. Although stress was a frequent topic in group sessions, discussions focused on causes of stress and management of stress, but not elimination of stress (which was a phenomenon of the first year of a rigorous training program, and thus somewhat immutable). Therefore, the actual level of stress, which this domain measured, was not diminished by the intervention. For both domains, sensitivity to change could be enhanced in the future by incorporating items based on issues raised in our Balint group experience. Scores for other individual questions changed during the course of the fellowship year with patterns we expected. Fellows felt freer to discuss the stresses of work and home while at work as the first fellowship year progressed. This may have been related to the existence of the group (and thus, a perceived openness in the work place to psychosocial issues), but more likely is a phenomenon of spending an intense year with colleagues who become a support group, and of having "safe zones" (the room in which the group was held and the Fellows’ Work Room) where such topics could be discussed openly. The improved attitudes we measured may have resulted from improved self-awareness, or were nonspecific effects of better social support. Balint-like groups, as opposed to classical Balint groups, have both objectives, and either would be acceptable to effect the desired outcome. During the course of the fellowship year, fellows expressed increased comfort with some of the interpersonal aspects of being an oncology fellow. Proficiency in these areas would be expected as fellows gained more experience with, for example, providing informed consent to patients for chemotherapy regimens. Interestingly, fellows were less likely to feel as if they worked in a supportive environment. It is difficult to know whether this was a mechanism to project the stresses of the fellowship year on the work environment itself or a marker of progressive independence; fellows felt less supported as they took more responsibility in managing patients and did not require constant attending intervention. Responses to some individual items did not change during the course of the fellowship year. For example, fellows were just as likely to "find it hard to talk with a patient about death and dying" and to "feel comfortable addressing code status" at the beginning of the academic year as at the end of the year, despite their more intensive exposure to a sicker patient population than during their residencies. The static nature of these responses may reflect the discomfort of physicians as a whole (including oncologists) in confronting these issues9–15 and indicates a need for workshops on breaking bad news to patients and on enhancing communication and patient relationship skills. This study has some limitations. First, sample size is small, and important effects of the intervention or of the first oncology fellowship year may have been missed. In addition, the results we report are conservative; we used two-group comparison analyses instead of paired comparisons because we could not link pre- and postgroup questionnaires for individual fellows, to guarantee anonymity. Still, the changes we showed were internally consistent even with a small cohort of fellows, and give us some insight into the fellowship experience. We also were confined to conducting the intervention at one fellowship program. Fellows in this program may not have been representative of Hematology-Oncology fellows across the United States. While this limits the generalizability of the results, it allowed us to maintain the quality of the intervention and eliminate effects of, for example, variability of group leaders or confounding effects of different fellowship experiences on measurable changes.
Last, the attitudes questionnaire we used has not been validated previously, and individual topic domains varied in their reliability, particularly the "Fellow’s views of him/herself as a physician" domain, which had the lowest Cronbach’s In conclusion, hematology-oncology fellows’ attitudes change over the course of the first fellowship year. Positive attitudes and development as caring physicians can be enhanced through the institution of a physician awareness group. The impact and effectiveness of the group intervention can be measured, and successful groups should improve the ability of physicians to communicate with their patients, and thus patient satisfaction. Future studies should assess whether improved quality of care results from the group intervention.
The authors indicated no potential conflicts of interest.
The authors acknowledge Robert J. Mayer, MD, and Lynda G. Montgomery, MD, for critical review of the manuscript, and the fellows who participated in the group.
This work was supported in part by a grant from the Kenneth B. Schwartz Foundation.
1. Butterfield PS: The stress of residency: A review of the literature. Arch Intern Med 148:1428–1435, 1988 2. Kash KM, Holland JC, Breitbart W, et al: Stress and burnout in oncology. Oncology (Huntingt) 14:1621–1637, 2000 3. Lurie N, Rank B, Parenti C, et al: How do house officers spend their nights? A time study of internal medicine house staff on call. N Engl J Med 320:1673–1677, 1989[Abstract]
4. Pitts FN, Jr., Schuller AB, Rich CL, et al: Suicide among US women physicians, 1967–1972. Am J Psychiatry 136:694–696, 1979
5. Stern TA, Prager LM, Cremens MC: Autognosis rounds for medical house staff. Psychosomatics 34:1–7, 1993 6. Herzog DB, Wyshak G, Stern TA, et al: Patient generated dysphoria in psychiatric residents. Int J Psychiatry Med 16:395–400, 1986[Medline] 7. Ramirez AJ, Graham J, Richards MA, et al: Burnout and psychiatric disorder among cancer clinicians. Br J Cancer 71:1263–1269, 1995[Medline] 8. Herzog DB, Wyshak G, Stern TA: Patient-generated dysphoria in house officers. J Med Educ 59:869–874, 1984[Medline]
9. Baile WF, Lenzi R, Parker PA, et al: Oncologists’ attitudes toward and practices in giving bad news: An exploratory study. J Clin Oncol 20:2189–2196, 2002 10. Baile WF, Kudelka AP, Beale EA, et al: Communication skills training in oncology: Description and preliminary outcomes of workshops on breaking bad news and managing patient reactions to illness. Cancer 86:887–897, 1999[CrossRef][Medline] 11. Holland JC, Almanza J: Giving bad news: Is there a kinder, gentler way? Cancer 86:738–740, 1999[CrossRef][Medline]
12. Ptacek JT, Eberhardt TL: Breaking bad news: A review of the literature. JAMA 276:496–502, 1996
13. O’Shea EM, Penson RT, Stern TA, et al: A staff dialogue on do not resuscitate orders: Psychosocial issues faced by patients, their families, and caregivers. Oncologist 4:256–262, 1999 14. Stern T, Jellinek M: Training issues in the intensive care unit. New Horizons 6:398–402, 1998 15. Herman TA: Terminally ill patients: Assessment of physician attitudes within teaching institution. N Y State J Med 80:200–207, 1980[Medline]
16. Ptacek JT, Ptacek JJ: Patients’ perceptions of receiving bad news about cancer. J Clin Oncol 19:4160–4164, 2001
17. Liang W, Burnett CB, Rowland JH, et al: Communication between physicians and older women with localized breast cancer: Implications for treatment and patient satisfaction. J Clin Oncol 20:1008–1016, 2002 18. Wiggers JH, Donovan KO, Redman S, et al: Cancer patient satisfaction with care. Cancer 66:610–616, 1990[CrossRef][Medline]
19. Steinhauser KE, Christakis NA, Clipp EC, et al: Factors considered important at the end of life by patients, family, physicians, and other care providers. JAMA 284:2476–2482, 2000
20. Novack DH, Suchman AL, Clark W, et al: Calibrating the physician: Personal awareness and effective patient care. Working group on Promoting Physician Personal Awareness, American Academy on Physician and Patient JAMA 278:502–509, 1997
21. Baile WF, Buckman R, Lenzi R, et al: SPIKES: A six-step protocol for delivering bad news—Application to the patient with cancer. Oncologist 5:302–311, 2000 22. Johnson AH, Brock CD, Hamadeh G, et al: The current status of Balint groups in US family practice residencies: A 10-year follow-up study, 1990–2000. Fam Med 33:672–677, 2001[Medline] 23. Scheingold L: Balint work in England: Lessons for American family medicine. J Fam Pract 26:315–320, 1988[Medline] 24. Margo K, Margo G: Tailoring the Balint group seminar for first year family medicine residents. Annals of Behavioral Science and Medical Education 1:38–42, 1994 25. Balint M: The doctor, his patient, and the illness. London, Pitman Medical, 1957 26. Brock CD, Stock RD: A survey of Balint group activities in US family practice residency programs. Fam Med 22:33–37, 1990[Medline] 27. Wise TN: Training oncology fellows in psychological aspects of their specialty. Cancer 39:2584–2587, 1977[CrossRef][Medline] 28. Artiss KL, Levine AS: Doctor-patient relation in severe illness: A seminar for oncology fellows. N Engl J Med 288:1210–1214, 1973[Medline] 29. Dornfest F: Balint Training: A "how to" manual in development. Sayre, PA, The American Balint Society, 1992, pp 3–5 30. Williamson P, Beitman BD, Katon W: Beliefs that foster physician avoidance of psychosocial aspects of health care. J Fam Pract 13:999–1003, 1981[Medline] 31. Ashworth CD, Williamson P, Montano D: A scale to measure physician beliefs about psychosocial aspects of patient care. Soc Sci Med 19:1235–1238, 1984[CrossRef][Medline] 32. Rost K, Gordon GH: The teacher simulation exercise: Changes in physician teaching emphasis and strategy. The SGIM Task Force on the Medical Interview. J Gen Intern Med 4:121–126, 1989[Medline] 33. Gordon GH, Rost K: Evaluating a faculty development course on medical interviewing, in Lipkin M, Putnam SM, Lazare (eds): The Medical Interview: Clinical Care, Education and Research. Frontiers of Primary Care. New York, NY, Springer-Verlag, 436–447, 1995 34. Mehta C, Patel N: StatXact4 for Windows: Statistical software for exact nonparametric inference. User Manual, Version 4.0.1. Cambridge, MA, CYTEL Software Corp, 2000 35. Streiner D, Norman G: Health Measurement Scales. Oxford, UK, Oxford University Press, 1989 36. Nease DE, Jr, Margo G, Johnson AH, et al: Role of Balint groups in caring for patients with unexplained symptoms. J Am Board Fam Pract 12:182–183, 1999 Submitted December 2, 2002; accepted July 11, 2003.
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2003 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
. Scores on a criterion question, "The statements on this questionnaire reflect the stresses I face as a doctor," also were assessed. 
