Numerous studies indicate that, although communication skills can be learned, they can also deteriorate as students progress through medical school, particularly in the clinical years as students learn medical problem solving.1,2,3,4,5 The good news is that this deterioration in communication skills can be prevented or reduced with more rigorous training. This was the surprise finding of Davis and Nicholaou,6 who compared the communication skills of first- and fourth-year medical students. They found that fourth-year students had superior facility in communication skills, which is attributable to a greater emphasis on the importance of communication and increased training in the curriculum. To be effective, communication training must provide bridges between theory, knowledge, practice, and exposure—with exposure providing students contact with patients through clinical observation and clinical consultation. Students acquire the most effective interviewing skills when they interact with patients during their clinical training,7 so exposure to a wide variety of clinical situations is essential. Prior training for such clinical encounters helps students develop working knowledge, understanding, and communication skills for dealing with challenging doctor—patient interactions.8 Students must fulfill three conditions to demonstrate appropriate communication skills.7 First, they need to know and understand a minimum of the corpus of knowledge and theory underlying communication exchanges in general and consultation processes. Second, they need to have a positive attitude towards using these skills in their interactions with patients. According to Bandura,9 this attitude is best developed through positive role modeling. Third, students need to be trained in a repertoire of specific communication skills and techniques and be placed in situations where these can be practiced successfully with patients.10
The purpose of this study was to examine students' exposures to and confidence in communication skills, the relationship between exposure and confidence, and the relationship between exposure and performance of patient—doctor communication skills among students graduating from an undergraduate medical program. By exposure we mean observing, assisting, or performing the skill. The four categories of communication skills we studied were interviewing, breaking bad news, crisis management, and counseling. We refer to the last three of these as “higher-order” skills, as they involve progressively more challenging and complex communication interactions.
Preclerkship Curriculum (Years One and Two). A problem-based learning (PBL) curriculum was begun at Dalhousie University Faculty of Medicine in 1992. The primary vehicle used to instruct students in communication skills is a module on interviewing skills in the first-year Patient—Doctor unit. Students are videotaped interviewing a standardized patient, and they practice their skills in small groups. They also receive lectures and written material. The students observe and practice basic history taking in clinical settings in their first and second years.
Clerkship Curriculum (Years Three and Four). At the time of the study, the clerkship comprised an 86-week continuum of experience, with significant flexibility and student choice. The students received some formal training in communication skills during their family medicine and psychiatry rotations. However, in other rotations, instruction occurs in clinical settings on an ad hoc basis, without a formal curriculum, as needs are identified.
The students in the sample comprised the first two classes (n = 172) to graduate from the new PBL curriculum at Dalhousie (classes of 1996 and 1997).
A locally-developed questionnaire was used to obtain students' self-assessments of exposure and confidence. It consisted of four sections that asked students to indicate their levels of exposure to a set of ten communication skills (see Table 1). They also were asked to rate their confidence with respect to each skill, using a 6-cm visual analog scale with the ends marked “low” and “high.” This is a useful and rarely used approach to assessing students' confidence in their skills. The rating scale for exposure consisted of the categories: never encountered, observed only, assisted senior staff member, performed once, performed two or more times. Students in the classes of 1996 and 1997 also participated in a two-hour objective structured clinical examination (OSCE) with simulated patients. However, three ten-minute communication stations were added to the 1997 OSCE, dealing with (1) requesting an organ donation from the husband of a woman declared “brain dead,” (2) counseling a middle-aged woman with depression, and (3) managing a 70-year-old woman brought to the emergency department by her daughter after a fall. All students were rated in each station by a trained physician—examiner, using a standard rating scale.
The students took the two-hour OSCE on the day following completion of their final clerkship rotations at the end of medical school. While awaiting their results in a large room, they completed a series of questionnaires, including the one used in this study. Students' identities were masked before coding to ensure confidentiality.
The data were analyzed using a statistical software package, and means, standard deviations, and frequencies were calculated. The five exposure categories were recombined into three: never encountered, observed or assisted, and performed one or more times. This was done retrospectively so that the number of students in each category would be high enough for statistical comparison. Two one-way ANOVAs were run across these three categories of exposure, one to compare the students based on their confidence levels and the other to compare them based on their OSCE performances.
The response rate for this study was 88% (148/172). Table 1 presents the results for level of exposure and confidence.
Nearly all students in both classes had taken a general adult history (99.3%) and a general pediatric history (97.3%). In fact, closer examination of the data showed that most students had performed these skills two or more times. The majority of the classes also had elicited, one or more times, a sexual history (96.7%), a history of drug or alcohol abuse (94.0%), and a history of sexual or physical abuse (59.6%). With respect to the higher-order communication skills, smaller proportions of the classes had performed these at all: breaking bad news to a patient or relative (50.7%), managing a patient seeking drugs (37.8%), managing a violent or hostile patient (37.8%), counseling for drug or alcohol abuse (29.1%), and counseling for victims of physical or sexual abuse (10.6%).
The students in the graduating classes of 1996 and 1997 rated their confidence in interviewing relatively high for general adult history (84.4%), general pediatric history (76.8%), eliciting sexual history (71.1%), eliciting history of drug or alcohol abuse (76.0%), and eliciting history of sexual or physical abuse (53.4%). In the areas of breaking bad news and crisis management, ratings were around or below 50% (see Table 1). Lower confidence ratings were given to the counseling areas (i.e., drug and alcohol abuse (43.0%), physical or sexual abuse (30.5%).
Since the complexity of the higher-order skills may have contributed to lower confidence, seven individual skills were examined (see Table 2). The students in the graduating classes of 1996 and 1997 were more confident as their levels of exposure increased for each communication skill.
Confidence levels were higher for each of the seven skills examined for the group that had observed or assisted than they were for the group that had never encountered the skill. More dramatic differences were observed between the group that had performed the skill one or more times than for the group that had simply observed or assisted.
An ANOVA on the total score across the three OSCE communication stations (1997 class) was conducted for each of the following exposure groups: low exposure across all ten communication skills, medium exposure, and high exposure. The scores across the three stations were combined in order to achieve a more adequate representation of the students' performances. Since skills are context-dependent, this combined score yielded a more valid and reliable outcome measure. In order to provide a more defensible measure for the variable, exposure was defined as total exposure across all ten skills. We felt that a total exposure score would better represent students' actual medical school experiences in doctor—patient communication. The results showed that OSCE performances increased from the low-exposure group (n = 9; mean = 59.8) to the medium-exposure group (n = 80; mean = 64.3) to the high-exposure group (n = 58; mean = 66.3). These differences were statistically significant (F = 3.1; p =.05).
In this study, graduating medical students had higher levels of exposure to standard communication skills than to higher-order communication skills, and their confidence levels were lower for the higher-order communication skills. One possible alternative explanation for the lower levels of confidence with respect to higher-order communication skills is that these skills are more demanding. Therefore, we decided to compare the confidence levels of students for each individual communication skill, as a function of type of exposure. The students who had performed each skill had much higher confidence levels than did those who had only observed or assisted. Also, the students who had observed or assisted with the skill had much higher confidence levels than did the students who had not encountered the skill at all. However, our findings suggests that observing or assisting is insufficient to develop confidence to an educationally significant degree; the more substantial gains were observed when students had performed the skill one or more times.
Although increased exposure increases confidence, a crucial question is whether increased exposure also leads to improved performance in applying these skills. The results of this study showed that this is indeed the case. The students who had had more overall exposure to the ten communication skills in this study performed at higher levels on the three OSCE stations emphasizing communication skills. Although not all ten communication skills were assessed in the three OSCE stations, these skills are composed of many common subskills, such as developing rapport, listening actively, explaining, and planning. Students with more exposure overall to the ten skills would have developed these subskills to a greater extent, and would most likely have better applied them in the OSCE.
It is important to note that students with less confidence in their abilities to exercise a skill may have avoided performing the skill in the clinical setting. Therefore, a causal relationship between exposure and confidence in this study should not be assumed. Although the results of the study confirmed our hypotheses, the exposure scale used did not measure actual level of exposure, i.e., number of times observed, assisted, or performed. Because the exposure scale simply measured students' recall of exposures, some bias may have been introduced. More important, the study surveyed only two medical school classes, and only one class's performance, so a broader survey is needed to confirm our findings.
The results of this study indicate that undergraduate students may not be getting sufficient opportunities to observe and practice complex communication skills in clinical or classroom settings, which results in low confidence levels. Factors affecting students' confidence do relate to clinical exposure, but they are also influenced by students' training in communication skills through structured programs that provide opportunities for learning and practice. The focus of this training appears to be on basic interviewing and interpersonal skills and not necessarily on higher-order skills. This was the case with the graduating classes in this study. All students had been given instruction in basic interviewing techniques. They had had opportunities to learn these techniques by observing videos and through role playing, by practicing their skills on each other and with simulated patients, and by receiving feedback on their skills from other students, course instructors, and simulated patients. For the higher-order skills, the students had been given exposure to breaking bad news through video programs and discussion as part of their training in palliative care; however, they had not had the opportunity to practice and receive feedback in these skills, as had been the case in their interviewing skills program. The students had been given no classroom training in crisis management or active counseling skills.
Both types of exposure may have to become more orchestrated for students during their undergraduate training. Providing effective training in higher-order communication skills as a core part of the undergraduate curriculum, where students have ongoing opportunities to observe, practice, and receive feedback in these skills, is a significant first step. This could occur in the clerkship years using the same techniques employed in learning basic interviewing skills. For this training, however, the use of role playing and standardized patients becomes particularly important. Once students have practiced these skills, they need to be provided with the opportunity to use them under supervision in a clinical setting. This practice will require some faculty development to ensure that physicians have the necessary skills to supervise effectively.
1. Kauss DR, Robbins AS, Abrass I, Bakaitis RF, Anderson LA. The long term effectiveness of interpersonal training in medical school. J Med Educ. 1980;55:595–601.
2. Alroy G, Ber R, Kramer D. An evaluation of the short term effects of an inter-personal skills course. Med Educ. 1984;18:85–9.
3. Kendrick T, Freeling P. A communications skills course for preclinical students: evaluation of general practice based teaching using group methods. Med Educ. 1993;27:211–7.
4. Craig JL. Retention of interviewing skills learned by first-year medical students: a longitudinal study. Med Educ. 1992;26:276–81.
5. Engler CM, Saltzman GA, Walker ML, Wolf FM. Medical student acquisition and retention of communication and interviewing skills. Med Educ. 1981;56:572–9.
6. Davis H, Nicholaou T. A comparison of the interviewing skills of first and final year medical students. Med Educ. 1992;26:441–7.
7. Evans BJ, Sweet B, Coman GJ. Behavioral assessment of the effectiveness of communication programs for medical students. Med Educ. 1993;27:344–50.
8. Dignan M, et al. Helping students respond to stressful interactions with cancer patients and their families: a pilot program. J Cancer Educ. 1989;4:179–83.
9. Bandura A. Self-efficacy mechanism in human agency. Am J Psychol. 1982;37:112–47.
10. Sanson-Fisher RW, Fairbairn S, Maguire P. Teaching skills in communication to medical students: a critical review of the methodology. Med Educ. 1981;15:33–7.
Research in Medical Education: Proceedings of the Thirty-ninth Annual Conference. October 30 - November 1, 2000. Chair: Beth Dawson. Editor: M. Brownell Anderson. Foreword by Beth Dawson, PhD.