No reduction in medical errors with increased direct supervision of resident teams

Reference

Finn KM, Metlay JP, Chang Y, et al. Effect of increased inpatient attending physician supervision on medical errors, patient safety, and resident education. JAMA Intern Med 2018;178(7):952-959.

Synopsis

In this single center study from Boston, these investigators compared 2 different levels of attending supervision on patient safety and educational outcomes on an inpatient general medicine teaching service over a course of 9 months. In the control arm, attendings provided standard supervision, joining residents on morning work rounds for newly admitted patients only, followed by afternoon "card flip" rounds for established patients. In the intervention arm, attendings joined residents on bedside work rounds for both new and established patients, providing increased direct supervision. Attendings selected for the study were highly rated teachers with a career focus in resident education and varying levels of experience. Of the 22 attending physicians who participated in the study, one-third had fewer than 5 years of experience and one-third had more than 15 years of experience. Each attending spent one 2-week block each in the control arm and the intervention arm. The primary patient safety outcome was rate of medical errors, either preventable adverse events or near misses. The primary educational outcome was the result of a time-motion study of control and intervention teams. Although the trend favored the intervention group, there was no statistically significant difference detected in the rate of medical errors, either number of preventable adverse events or near misses, when comparing standard and increased supervision (overall medical errors: relative risk 0.85; 95% CI 0.64 - 1.09; P = .21). Additionally, the number of radiology studies and consultations ordered, length of stay, intensive care unit transfer, mortality, and discharge disposition were similar for patients in the 2 arms. As far as educational outcomes, mean duration of work rounds (202 minutes) and new admission presentations (105 minutes) were similar despite differing levels of supervision. However, interns were noted to speak for a shorter amount of time in the increased supervision arm (55 minutes vs 64 minutes; P = .008). Residents who were surveyed perceived that increased supervision led to less efficiency and autonomy while quality of care remained the same. Attendings, on the other hand, rated the quality of care higher in the increased supervision arm.