Effects of the revised HCFA evaluation and management guidelines on inpatient teaching

Abstract

Objective: In 1996, the Health Care Financing Administration (HCFA) introduced new evaluation and management (E&M) guidelines mandating more intensive supervision and documentation by attending physicians. We assessed the effects of the guidelines on inpatient teaching.

Design: Pretest-posttest, nonequivalent control group design.

Setting: A university hospital and an affiliated county hospital where the guidelines were implemented and an affiliated VA medical center where they were not.

Participants: Sixty-one full-time faculty who had attended on the general medical wards for at least 1 month for 2 of 3 consecutive years prior to July 1996 and for at least 1 month during the 18 following months.

Measurements and main results: We evaluated standardized, confidential evaluations of attending physicians that are routinely completed by residents and students after clinical rotations at all three sites. Comparing 863 evaluations completed before July 1, 1996 and 497 completed after that date, there were no significant differences at any of the hospitals on any items assessed. There were also no differences between the university and county hospitals as compared with the VA. Eighty-seven percent of 39 university and county attending physicians returned a survey about their perceptions of inpatient teaching activities before and after July 1, 1996. They reported highly significant increases in time devoted to attending responsibilities but diminished time spent on teaching activities.

Conclusions: Physicians reported a dramatic increase in overall time spent attending but a decrease in time spent teaching following implementation of the revised E&M guidelines. Yet, evaluations of their teaching effectiveness did not change.

FIGURE 1

Box plots of the mean ratings given to university and county hospital attending physicians for each item on the teaching assessment form for the periods before and after July 1, 1996 (7/1/96). Differences are shown for pooled values within each attending on each scale. Each box covers 50% of the observations, with the median in the center and the quartiles at the ends. The outside lines are a crude approximation of a 95% confidence interval if the data were normally distributed. Outliers are shown as individual points.

FIGURE 2

Box plots of differences in the mean ratings for each item on the teaching assessment form for the periods before and after July 1, 1996 (7/1/96).

FIGURE 3

Histogram displaying selected responses to the questionnaire completed by attending physicians. Black bars represent responses related to activities prior to July 1, 1996, and shaded bars reflect activities after that date. P values were calculated from the permutation rank sum test using all response categories.