Background
: Chest P-A interpretation is very important because chest P-A is one of the important diagnostic tools in primary care. Primary care physicians often make decision of patient management based on clinical impression and radiologic findings. This study assessed the accuracy of family physicians in interpretationg radiographs taken in the seven family practice offices.

Methods : This study compared the interpretation of family physicians with those of one radiologist on 183 office radiographs using a uniform protocol. Nine family physicians participated in the study along with one radiologist. A total of 183 chest P-Ad were evaluated.

Results : The results were as follows 1) the family physicians' discordancy rate with the radiologist was 19.6% 2) 80% of the discordant films were clinically significant misses of those were read as "normal" by one radiologist 4) only 5% of chest P-As which were read as "normal" by family physicians were read as "abnormal" by radiologist 5) The sensitivity of family physician's interpretation compared with radiologist was 90%, the specificity was 86%, and the positive predictability was 71%.

Conclusion : This result suggest that it is not necessary to consult radiologist for interpretation of Chest PAs which were read as "normal" by family physician, because 95% of Chest PAs which were read as "normal" by family physicians were read as "normal" by radiologist. Radiologically abnormal films may need to be consulted with clinical information. Physicians should take more active role in extending cervical screening corverage.

Background
: Recent studies reported that doctors of different specialty use different diagnostic and therapeutic methods to the same clinical conditions.
Though this difference has significant influence on the quality and cost of medical care, study about this problem was never been done in Korea.

Methods : The diagnostic methods of third year residents in family practice(N=6) and internal medicine(N=6) were compared with respect to "functional gastrointestinal disorder(FGID)", common ambulatory patient problem. One "standardized patient" was presented.

Results : Analysis of the recorded interviews showed that family practice residents took more medical history(p<0.05), ordered more laboratory investigations(p<0.05). There was no difference in the selection of physical examination items. There was difference in the diagnosis reached by these two groups. Six family practice residents diagnosed FGID. In an analysis of the "commonness" of questions asked by internal medicine residents and family practice residents, it was found that internal medicine residents used more common questions.

Conclusion : There were some significant differences in diagnostic and therapeutic strategies between internal medicine and family medicine residents. These findings have implications for the future training of primary care physicians.