Overnight shift work: factors contributing to diagnostic discrepancies

Tarek N. Hanna, Thomas W Loehfelm, Faisal Khosa, Saurabh Rohatgi, Jamlik Omari Johnson

Research output: Contribution to journalArticle

6 Citations (Scopus)

Abstract

The aims of the study are to identify factors contributing to preliminary interpretive discrepancies on overnight radiology resident shifts and apply this data in the context of known literature to draw parallels to attending overnight shift work schedules. Residents in one university-based training program provided preliminary interpretations of 18,488 overnight (11 pm–8 am) studies at a level 1 trauma center between July 1, 2013 and December 31, 2014. As part of their normal workflow and feedback, attendings scored the reports as major discrepancy, minor discrepancy, agree, and agree—good job. We retrospectively obtained the preliminary interpretation scores for each study. Total relative value units (RVUs) per shift were calculated as an indicator of overnight workload. The dataset was supplemented with information on trainee level, number of consecutive nights on night float, hour, modality, and per-shift RVU. The data were analyzed with proportional logistic regression and Fisher’s exact test. There were 233 major discrepancies (1.26 %). Trainee level (senior vs. junior residents; 1.08 vs. 1.38 %; p < 0.05) and modality were significantly associated with performance. Increased workload affected more junior residents’ performance, with R3 residents performing significantly worse on busier nights. Hour of the night was not significantly associated with performance, but there was a trend toward best performance at 2 am, with subsequent decreased accuracy throughout the remaining shift hours. Improved performance occurred after the first six night float shifts, presumably as residents acclimated to a night schedule. As overnight shift work schedules increase in popularity for residents and attendings, focused attention to factors impacting interpretative accuracy is warranted.

Original languageEnglish (US)
Pages (from-to)41-47
Number of pages7
JournalEmergency Radiology
Volume23
Issue number1
DOIs
StatePublished - Feb 1 2016
Externally publishedYes

Fingerprint

Appointments and Schedules
Workload
Workflow
Trauma Centers
Radiology
Logistic Models
Education
Datasets

Keywords

  • Interpretation discrepancies
  • Medical errors
  • Night float
  • Radiology
  • Resident

ASJC Scopus subject areas

  • Emergency Medicine
  • Radiology Nuclear Medicine and imaging

Cite this

Overnight shift work : factors contributing to diagnostic discrepancies. / Hanna, Tarek N.; Loehfelm, Thomas W; Khosa, Faisal; Rohatgi, Saurabh; Johnson, Jamlik Omari.

In: Emergency Radiology, Vol. 23, No. 1, 01.02.2016, p. 41-47.

Research output: Contribution to journalArticle

Hanna, Tarek N. ; Loehfelm, Thomas W ; Khosa, Faisal ; Rohatgi, Saurabh ; Johnson, Jamlik Omari. / Overnight shift work : factors contributing to diagnostic discrepancies. In: Emergency Radiology. 2016 ; Vol. 23, No. 1. pp. 41-47.
@article{6a48e2b641e248ecb03a4e88440778df,
title = "Overnight shift work: factors contributing to diagnostic discrepancies",
abstract = "The aims of the study are to identify factors contributing to preliminary interpretive discrepancies on overnight radiology resident shifts and apply this data in the context of known literature to draw parallels to attending overnight shift work schedules. Residents in one university-based training program provided preliminary interpretations of 18,488 overnight (11 pm–8 am) studies at a level 1 trauma center between July 1, 2013 and December 31, 2014. As part of their normal workflow and feedback, attendings scored the reports as major discrepancy, minor discrepancy, agree, and agree—good job. We retrospectively obtained the preliminary interpretation scores for each study. Total relative value units (RVUs) per shift were calculated as an indicator of overnight workload. The dataset was supplemented with information on trainee level, number of consecutive nights on night float, hour, modality, and per-shift RVU. The data were analyzed with proportional logistic regression and Fisher’s exact test. There were 233 major discrepancies (1.26 {\%}). Trainee level (senior vs. junior residents; 1.08 vs. 1.38 {\%}; p < 0.05) and modality were significantly associated with performance. Increased workload affected more junior residents’ performance, with R3 residents performing significantly worse on busier nights. Hour of the night was not significantly associated with performance, but there was a trend toward best performance at 2 am, with subsequent decreased accuracy throughout the remaining shift hours. Improved performance occurred after the first six night float shifts, presumably as residents acclimated to a night schedule. As overnight shift work schedules increase in popularity for residents and attendings, focused attention to factors impacting interpretative accuracy is warranted.",
keywords = "Interpretation discrepancies, Medical errors, Night float, Radiology, Resident",
author = "Hanna, {Tarek N.} and Loehfelm, {Thomas W} and Faisal Khosa and Saurabh Rohatgi and Johnson, {Jamlik Omari}",
year = "2016",
month = "2",
day = "1",
doi = "10.1007/s10140-015-1355-0",
language = "English (US)",
volume = "23",
pages = "41--47",
journal = "Emergency Radiology",
issn = "1070-3004",
publisher = "Springer New York",
number = "1",

}

TY - JOUR

T1 - Overnight shift work

T2 - factors contributing to diagnostic discrepancies

AU - Hanna, Tarek N.

AU - Loehfelm, Thomas W

AU - Khosa, Faisal

AU - Rohatgi, Saurabh

AU - Johnson, Jamlik Omari

PY - 2016/2/1

Y1 - 2016/2/1

N2 - The aims of the study are to identify factors contributing to preliminary interpretive discrepancies on overnight radiology resident shifts and apply this data in the context of known literature to draw parallels to attending overnight shift work schedules. Residents in one university-based training program provided preliminary interpretations of 18,488 overnight (11 pm–8 am) studies at a level 1 trauma center between July 1, 2013 and December 31, 2014. As part of their normal workflow and feedback, attendings scored the reports as major discrepancy, minor discrepancy, agree, and agree—good job. We retrospectively obtained the preliminary interpretation scores for each study. Total relative value units (RVUs) per shift were calculated as an indicator of overnight workload. The dataset was supplemented with information on trainee level, number of consecutive nights on night float, hour, modality, and per-shift RVU. The data were analyzed with proportional logistic regression and Fisher’s exact test. There were 233 major discrepancies (1.26 %). Trainee level (senior vs. junior residents; 1.08 vs. 1.38 %; p < 0.05) and modality were significantly associated with performance. Increased workload affected more junior residents’ performance, with R3 residents performing significantly worse on busier nights. Hour of the night was not significantly associated with performance, but there was a trend toward best performance at 2 am, with subsequent decreased accuracy throughout the remaining shift hours. Improved performance occurred after the first six night float shifts, presumably as residents acclimated to a night schedule. As overnight shift work schedules increase in popularity for residents and attendings, focused attention to factors impacting interpretative accuracy is warranted.

AB - The aims of the study are to identify factors contributing to preliminary interpretive discrepancies on overnight radiology resident shifts and apply this data in the context of known literature to draw parallels to attending overnight shift work schedules. Residents in one university-based training program provided preliminary interpretations of 18,488 overnight (11 pm–8 am) studies at a level 1 trauma center between July 1, 2013 and December 31, 2014. As part of their normal workflow and feedback, attendings scored the reports as major discrepancy, minor discrepancy, agree, and agree—good job. We retrospectively obtained the preliminary interpretation scores for each study. Total relative value units (RVUs) per shift were calculated as an indicator of overnight workload. The dataset was supplemented with information on trainee level, number of consecutive nights on night float, hour, modality, and per-shift RVU. The data were analyzed with proportional logistic regression and Fisher’s exact test. There were 233 major discrepancies (1.26 %). Trainee level (senior vs. junior residents; 1.08 vs. 1.38 %; p < 0.05) and modality were significantly associated with performance. Increased workload affected more junior residents’ performance, with R3 residents performing significantly worse on busier nights. Hour of the night was not significantly associated with performance, but there was a trend toward best performance at 2 am, with subsequent decreased accuracy throughout the remaining shift hours. Improved performance occurred after the first six night float shifts, presumably as residents acclimated to a night schedule. As overnight shift work schedules increase in popularity for residents and attendings, focused attention to factors impacting interpretative accuracy is warranted.

KW - Interpretation discrepancies

KW - Medical errors

KW - Night float

KW - Radiology

KW - Resident

UR - http://www.scopus.com/inward/record.url?scp=84955710230&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84955710230&partnerID=8YFLogxK

U2 - 10.1007/s10140-015-1355-0

DO - 10.1007/s10140-015-1355-0

M3 - Article

C2 - 26475281

AN - SCOPUS:84955710230

VL - 23

SP - 41

EP - 47

JO - Emergency Radiology

JF - Emergency Radiology

SN - 1070-3004

IS - 1

ER -