Medical interviewing is a skill with which physicians ascertain pertinent information that leads to proper diagnoses, and excellent interviewing skills may also strengthen the bond between physicians and patients. Properly using interviewing skills increases patients' satisfaction, which has been associated with improvements in patients' compliance and health status, as well as fewer malpractice suits.1
According to the biopsychosocial model,2 every patient has a story that communicates the complex interactions among the biological, psychological, and social components of his or her life.2 Unfortunately, physicians frequently emerge from a patient's interview without an understanding of the patient's complete story because current interviewing practices often eschew the human dimension in favor of a purely biological story.2 Patient-centered interviewing (PCI) strives to produce a more human interaction whereby patients are heard and understood in a novel way.3 Patients' needs and the physician—patient relationship are the main foci of this model, and it complements the physician-centered interviewing model taught in most medical schools.3–6
Previous studies have shown the short-term effectiveness of intensive block training in PCI at the postgraduate level. Residents gained more knowledge, confidence, and skills in gathering data, handling patients' emotions, and managing patients with psychosocial problems.1,3,6–8 These studies have also suggested a trend toward greater patients' satisfaction.6 However, there has been no study to show that PCI-trained residents effectively retain these skills, and whether they continue to use this technique throughout their residency.
The purpose of this study was to determine (1) whether early intensive PCI training in residency leads to long-term retention of these skills, and (2) how frequently residents incorporate PCI skills into their usual patient encounters two years after their intensive training.
This study had two components: (1) a prospective study of the retention of PCI skills two years after the intensive training and (2) a cross-sectional survey comparing the use of PCI skills between housestaff who received training and those who had not.
Fourteen medicine housestaff from Long Island Jewish Medical Center (LIJMC) received intensive training in PCI during their internship between September 1996 and April 1997. Control groups for the cross-sectional survey consisted of 14 PGY-1s from LIJMC prior to their PCI training (1998 interns) and 14 medicine residents from another institution whose level of training (PGY-3) matched that of the intervention group, but who had not been trained in PCI. The exposure to PCI of the control groups during medical school was not evaluated.
The curriculum used in this study was based on the work of Smith1,6,8 and Novack.10 Training in PCI during internship consisted of an intensive experiential one-month course, a detailed description of which is published elsewhere.9 In summary, the training consisted of 20 core sessions incorporating seminars, role plays, and supervised interactions with patients. These encounters were videotaped for immediate feedback and further discussion. The main foci of these sessions were efficient data gathering, emotion handling, patient education, and management of common psychosocial and psychiatric problems encountered by primary care physicians. During this rotation the housestaff had no other clinical responsibilities except their outpatient clinics.
Residents were assessed prior to, immediately after, and late after the training course through three directly observed interviews with random in-hospital patients at LIJMC. Each interview was scored by one of the three faculty members responsible for the training block using the Rhode Island Hospital Resident Interview Checklist (RIC), formerly called Brown Interview Checklist.10 Only four residents (44%) had all three evaluations scored by the same faculty member. The late evaluation was obtained two years after training to identify retention of acquired PCI skills.
The evaluation instrument, developed by Novack and colleagues, is divided into four parts: (1) flow of the initial interview (opening and exploration of problems), (2) interpersonal skills (facilitation and relationship skills), (3) clinical reasoning skills, and (4) key content areas (the first two items are shown in Table 1). For the purposes of this study, only the initial two sections of the RIC were evaluated because the study focused on data-gathering and physician—patient-relationship skills.
Neither the faculty nor the residents were blinded to the purpose of the evaluation. Statistical analyses were restricted to the scores of residents who completed all three evaluations.
For the cross-sectional portion of this study, we surveyed the intervention and the control groups for residents' actual use of PCI in December 1998. The instrument consisted of five clinical vignettes and five Likert-scale questions on the use of different interviewing skills based on three PCI skills: (1) flow of the initial interview (introduction of self and others and optimization of settings for interview), (2) exploration of problems (open-to-closed-ending questioning cone and segment summary), and (3) relationship skills (reflection of patient's emotion). Housestaff in the intervention and control groups were not aware of the specific purpose of this survey, and this instrument also contained 35 general questions about interviewing and physical examination skills not related to PCI.
We analyzed the RIC scores from prior to and late after training using Sign and Wilcoxon signed-rank tests, and we used the same method to compare RIC scores from immediately after with those from the late evaluation to check for any loss of acquired PCI skills. The results of the PCI-use survey were analyzed using Mann—Whitney U tests.
Of the 14 residents who received intensive training in PCI, ten were evaluated prior to the training, nine were evaluated immediately after training, and 11 residents completed the late evaluation in December 1998. There was no statistically significant difference relative to age, gender, or proportion of international medical graduates between the intervention and control groups.
Table 1 shows the major areas evaluated during the supervised interviews. Checklist scores on both the immediate-post and late evaluations showed that the residents had improved in all areas when compared with the scores from the evaluation prior to training. Baseline evaluation scores showed that the residents already had good skills in opening the interview (verbal greeting and self-introduction), with some basic facilitation skills (eye contact and open posture).
Table 2 shows that in the evaluation of PCI skills immediately after intensive training the residents had significantly improved their abilities to effectively optimize patient interview setting (p < .01), establish and maintain narrative thread (p = .01), use segment summary (p < .01), use transition statements (p < .01), and use some facilitation skills (minimal feedback p = .03, reflection p < .03, and respect p < .01).
Table 2 also shows significant improvements at the late evaluation compared with the evaluation prior to training in the abilities to effectively optimize patient interview setting (p = .01), establish and maintain narrative thread (p = .03), use the open-to-closed-ended questioning cone (p < .01), avoid more than one question at a time (p < .01), and use some facilitation skills (open posture p < .05, minimal feedback p = .02, and silence as a facilitation skill p = .03).
No significant loss of PCI skills was evident on the late evaluation when compared with the scores of the evaluation immediately after training, with the exception of respect in facilitation skills.
Fourteen PGY-3 residents from the intervention group returned the survey about their use of PCI skills. In the control groups, 13 of the 14 interns from LIJMC and 11 of the 14 PGY-3s from the other program returned the questionnaire. One survey from the latter control group was excluded because it contained the same answer for all 40 questions.
The intervention group scored better than did the two control groups on all five scales the survey evaluated. However, statistically significant differences were found for only the use of reflection of patients' emotions. The rates for use of skills related to exploration of problems (open-to-closed-ended questioning cone and segment summary) and introduction of self and others were very high in both control groups despite their lack of exposure to the intensive PCI training. Similar results were obtained when the intervention group was separately compared with each control group.
The delicate balance of providing outstanding medical care to our patients within an environment that has become more cost-conscious has become more difficult with the loss of time and attention that physicians can devote to their patients. As a result, frustration and dissatisfaction have accumulated in both physicians and patients.
This study demonstrated that medical housestaff were able to retain PCI skills over the course of their medical residency. They recognized and seemed better able to respond to patients' emotions after acquiring PCI training. There were improvements in all PCI skills at late evaluation (two years after training) compared with the baseline evaluation scores. We also saw no loss of skills in most areas when the late evaluation scores were compared with scores from the evaluation immediately after the training. The preservation of this technique may enable doctors throughout their careers to better understand their patients by improving data gathering and strengthening the patient—physician bond.
The lack of a control group with similar level of experience in interviewing is a limitation to this study. In theory, the residents could have acquired interview skills without exposure to the intensive block training. Other limitations of this study are intraobserver variability and the possible bias that the evaluators could have introduced because they were not blinded to the purpose of the evaluation. And, the survey instrument we used was not evaluated for test—retest reliability or validity through direct observation of housestaff behavior.
The cross-sectional portion of this study addressed the question of the frequency with which residents incorporate PCI skills into their real patient encounters. Our results suggest improvements in the use of all PCI skills among those who received PCI training. However, the only statistically significant improvement we found was in the residents' use of relationship skills, which may reflect a lack of emphasis on patient—physician relationship skills in the traditional physician-centered interviewing model that is frequently taught at medical schools. The lack of significance in other areas may have been a result of our small sample size. It may also reflect a need to expand the curriculum to include didactic sessions and role-plays on how to effectively use PCI skills in a time-conscious manner. Anecdotally, we have found time to be the most common concern expressed by the housestaff in regard to using the PCI technique.
Further studies are needed to determine whether successful postgraduate training of physicians in PCI will lead them to continue to use these skills during their professional lives. In addition, the effects of PCI skills on clinical outcomes and on patients' satisfaction, therapeutic alliance, and compliance also need to be explored further. Nevertheless, it is important that medical doctors learn how to identify the social and psychological aspects of their patients' lives. This skill may play an important role in strengthening the patient—physician relationship and improving patients' compliance and health status.6
1. Smith RC, Lyles JS, Mettler JA, et al. A strategy for improving patient satisfaction by the intensive training of residents in psychosocial medicine: a controlled, randomized study. Acad Med. 1995;70:729–32.
2. Engel GL. The clinical application of the biopsychosocial model. Am J Psychiatry. 1980;137:535–44.
3. Smith RC, Hoppe RB. The patient's story: integrating the patient- and physician-centered approaches to interviewing [see comments]. Ann Intern Med. 1991;115:470–7.
4. Levenstein JH, McCracken EC, McWhinney IR, Stewart MA, Brown JB. The patient-centered clinical method. A model for the doctor—patient interaction in family medicine. Fam Pract. 1986;3:24–40.
5. Levenstein JH, Brown JB, Weston WW, Stewart M, McCracken EC, McWhinney IR. Patient-centered clinical interviewing. In: Communicating with Medical Patients. London, U.K.: Sage Publications, 1989:107–20.
6. Smith RC, Lyles JS, Mettler J, et al. The effectiveness of intensive training for residents in interviewing. A randomized, controlled study [see comments]. Ann Intern Med. 1998;128:118–26.
7. Shore BE, Smith RC. Patient responses to a training interview. J Med Educ. 1984;59:901–3.
8. Smith RC, Osborn G, Hoppe RB, et al. Efficacy of a one-month training block in psychosocial medicine for residents: a controlled study. J Gen Intern Med. 1991;6:535–43.
9. Gordon JH, Walerstein SJ, Pollack S. The advanced clinical skills program in medical interviewing: a block curriculum for residents in medicine. Int J Psychiatry Med. 1996;26:411–29.
10. Novack DH, Dube C, Goldstein MG. Teaching medical interviewing. A basic course on interviewing and the physician—patient relationship. Arch Intern Med. 1992;152:1814–20.