Background:

Little information exists about how much work individual patients actually contribute to the workload of a physician. We conducted this study to determine the amount of time internal medicine interns spend on new patient admissions.

Methods:

We conducted a prospective time–motion study on general internal medicine wards at a single VAMC, with IRB approval. We consented internal medicine interns and patients admitted by those interns. Inclusion criterion for interns was being on a general medicine team; there were no exclusion criteria for interns. Inclusion criterion for the patients was being admitted to a consented intern. Exclusion criteria were being (1) unable to give consent, (2) admitted after 1 AM, and (3) admitted directly to the ICU. We trained observers to shadow interns during an on call period. We used specialized software on laptop computers to continuously record the tasks performed by interns. We also recorded work performed by the interns for consented patients. We calculated the total time spent on individual patients at 4, 6, and 8 h after admission. We also calculated the amount of time spent on different types of work (e.g., direct patient care) for each patient.

Results:

Twenty–five of 36 (69%) interns and 26 of 43 (60%) of patients agreed to participate. Mean age of interns was 28.6 (SD 2.4) and mean age of patients was 62.5 (SD 14.2); 98% of the patients were men. Interns spent a mean of 69 [pm] 31 min with each new admission in the first 4 h after admission, 89 [pm] 41 min after 6 h and 107 [pm] 87 min at 8 h. In the first 4 h, interns spent a mean of 32.0 [pm] 19.9 min (47%) in documentation tasks, 16.5 [pm] 13.3 min (25%) communicating with other healthcare professionals, and 15.90 [pm] 15.03 min (22%) at the bedside of each new patient. The remainder was spent in other activities such as teaching about the patient (2 min or 2%). Care for patients occurred episodically, with a mean of 36 tasks devoted to each patient in the first 4 h of admission. We examined correlations between the number of hours spent on individual patients and other workload parameters. We found that the amount of time that interns predicted that they would spend on the patient was negatively correlated with amount of time they actually spent (r =-0.43, P = 0.03). The number of months an intern had been in training was also negatively correlated with the amount of time actually spent (r =-0.38, P = 0.06). Team census, intern census and number of patients cross–covered were not significantly correlated with time spent on individual patients.

Conclusions:

We have demonstrated that it is possible to measure the amount of time interns spend on new admissions. These admissions take a significant amount of time. This work is fragmented, and predominantly composed of indirect patient care. Less experienced interns spend more time than more experienced interns, which reinforces the need for graduated responsibility of patient care.