The “night stalker” effect: Are quality improvements with a dedicated night call rotation sustained?

The authors assessed whether the addition of a second-year diagnostic radiology resident assigned to cover the night shift at a major urban university hospital has a sustained effect on the number and clinical significance of “missed” radiologic findings. Radiographs interpreted overnight in the emergency department by radiology residents between January 1992 and December 1992 were reviewed daily by emergency radiology attending staff. A list of patients for whom there was a modification in the final radiologic interpretation was given to the emergency department physicians, who reviewed each case, scored the urgency of patient recall, and estimated the likelihood of patient morbidity attributable to the miss. The relative performance of after-hours residents was compared on the five nights per week with the dedicated night resident vs. the two nights per week without the dedicated night resident (control group). Of 22,295 after-hours examinations performed during the study period, 304 (1.36%) misses were recorded, nearly identical to the miss rate for the preceding 6 months. The percentage per examination interpreted (and number) of missed cases stratified by recall score for the control and dedicated night resident groups, respectively, were: (a) immediate, 0.62% (34) and 0.29% (49); (b) within 48 hours, 0.31% (17) and 0.32% (54); (c) no recall, 0.71% (24) and 0.29% (39); (d) finding already recognized by emergency department physicians, 0.44% (24) and 0.23% (39); total, 2.09% (114) and 1.13% (190). The difference in total discordance rates is statistically significant (P < 1 × 10⁻¹⁵). Our previously reported improvement in the quality of after-hours radiographic interpretation due to the addition of a dedicated night shift resident is sustained in a new group of residents. This confirms that the improvement is real and not a manifestation of the measurement methods.