DESIGN: Prospective trial comparing 3-month blocks of intensive prompting, modest prompting, and no prompting on adverse event reporting by housestaff at morning report.

SETTING: Inpatient internal medicine service at a university-affiliated, Veterans Affairs Medical Center teaching hospital.

INTERVENTIONS: Intensive prompting (daily), modest prompting (once or twice weekly), and no prompting of medical residents to report hospital-associated adverse events.

MEASUREMENTS AND MAIN RESULTS: The number, type, and severity of hospital-acquired adverse events occurring on an internal medicine service were determined during the various periods of intervention on a per houseofficer basis. Residents were reminded to record events once or twice weekly, daily, or not at all. These data were compared with those identified by usual hospital surveillance. The addition of housestaff reporting to usual hospital surveillance increased the numbers of adverse events reported. There was little overlap in episodes reported by the two strategies. Increasing the level of prompting increased the number of reports per houseofficer. Housestaff prompting increased reporting at all levels of adverse event severity from mild to serious and detected a wide variety of types of adverse events, especially adverse drug reactions and procedure complications.

CONCLUSIONS: Our study demonstrates that physician self-reporting of adverse events adds to the usual hospital surveillance adverse event reporting, and finds that such reporting can be easily accomplished within the context of a daily teaching activity. The information provided about adverse events by housestaff at morning report is additive to that obtained by usual surveillance methods. The use of such a strategy provides information in a timely fashion.