Daytime, evening, and overnight: the 24-h radiology cycle and impact on interpretative accuracy

• Published in: Emergency radiology Vol. 30; no. 5; pp. 607 - 612
• Main Authors: Zhou, Shannon, Hanna, Tarek, Ma, Tianwen, Johnson, Timothy D., Lamoureux, Christine, Weber, Scott, Johnson, Jamlik-Omari, Steenburg, Scott D., Dunkle, Jeffrey W., Chong, Suzanne T.
• Format: Journal Article
• Language: English
• Published: Cham Springer International Publishing 01-10-2023; Springer Nature B.V
• Subjects: Abdomen; Daytime; Emergency medical care; Emergency Medicine; Evening; Imaging; Magnetic resonance imaging; Medicine; Medicine & Public Health; Onsite; Original Article; Radiology; Time of use; Ultrasonic imaging; Teleradiology; Shiftwork; Fellowship; Emergency; Medical error
• ISSN: 1438-1435; 1070-3004; 1438-1435
• DOI: 10.1007/s10140-023-02161-6

Purpose
To assess the influence of time of day when a study is interpreted on discrepancy rates for common and advanced studies performed in the acute community setting.
Methods
This retrospective study used the databank of a U.S. teleradiology company to retrieve studies between 2012 and 2016 with a preliminary report followed by a final report by the on-site client hospital. Neuroradiology, abdominal radiology, and musculoskeletal radiology studies were included. Teleradiologists were fellowship trained in one of these subspecialty areas. Daytime, evening, and overnight times were defined. Associations between major and minor discrepancies, time of day, and whether the study was common or advanced were tested with significance set at
p
= .05.
Results
A total of 5,883,980 studies were analyzed. There were 8444 major discrepancies (0.14%) and 17,208 minor discrepancies (0.29%). For common studies, daytime (0.13%) and evening (0.13%) had lower major discrepancy rates compared to overnight (0.14%) (daytime to overnight, RR = 0.57, 95%CI: 0.45, 0.72,
p
< 0.01 and evening to overnight, RR = 0.57, 95%CI: 0.49,0.67,
p
< 0.01). Minor discrepancy rates for common studies were decreased for evening (0.29%) compared to overnight (0.30%) (RR = 0.89, 95%CI: 0.80,0.99,
p
= 0.029). For advanced studies, daytime (.15%) had lower major discrepancy rates compared to evening (0.20%) and overnight (.23%) (daytime to evening, RR = 0.77, 95%CI: 0.61, 0.97,
p
= 0.028 and daytime to overnight, RR = 0.66, 95%CI: 0.50, 0.87,
p
≤ 0.01).
Conclusion
Significantly higher major discrepancy rates for studies interpreted overnight suggest the need for radiologists to exercise greater caution when interpreting studies overnight and may require practice management strategies to help optimize overnight work conditions. The lower major discrepancy rates on advanced studies interpreted during the daytime suggest the need for reserving advanced studies for interpretation during the day when possible.