Objective.-To survey a broad sample of Japanese internists regarding the effects of postgraduate training on their ability to provide ambulatory care.
Design.-Analysis of internists trained at university or non-university hospitals (as designated by the Ministry of Health and Welfare) by written questionnaire.
Results.-74.0% of respondents (127 in total) had undergone ambulatory care training. 30.7% were trained in outpatient clinic sessions that were held more than once a week on average. 85.8% of respondents said they recognized the differences between ambulatory care and inpatietn care. Furthermore, the percentage of respondents who understood such differences during their postgraduate training periods was higher in the group that had had ambulatory care training than the group that had not. Many respondents also suggested the necessity of training in non-internal medicine specialties (e. g. dermatology, OB-GYN, otolaryngology, orthopedics, urology and ophthalmology) that were not popular postgraduate training programs.
Conclusion.-Postgraduate training had a large effect on the ability of internists to provide ambulatory care. Japanese postgraduate training is still not sufficient in this regard. Residency training programs should put more emphasis on ambulatory care.

Our previous study concerning evaluation methods of postgraduate clinical training showed a usefulness in improving clinical training, especially in the fields of clinical skills and knowledge. However, it also pointed out the unsolved problem of assessment of those behaviors in which no significant differences were shown throughout the residency period. In order to solve this problem, a feedback system has been introduced, consisting of self-evaluation during the residensy, and appropriate counseling regarding hortcomings.
In the present study, the effect of this feedback system on clinical training, particularly on physician behavior, has been analyzed. The study was performed by comparing evaluation scores, based on a five-point scale, from two groups: 35 residents who did not undergo the feedback system, and 84 residents who had. Clinical skills, knowledge, positiveness, behavior with patients and colleagues, and patients' records, were each evaluated after discussions by the teaching staff.
Our results showed that there was a significant improvement not only in clinical skills and knowledge, but also in physician behavior in the group using the feedback system. Thus, it appears that this feedback system is useful in improving postgraduate clinical training.