As workloads continue to steadily increase, radiologists have been expressing concern about the tradeoff of quantity reporting versus quality and accuracy. Radiologists are under continuous pressure to increase productivity by increasing workload volume. Studies have repeatedly shown that the accuracy of radiologic interpretation and the detection of pathologies decrease as the volume of interpreted studies increases and the viewing time per study decreases. A pilot study conducted by radiologists at Monmouth Medical Center in Long Branch, NJ, published online May 20 in the Journal of the American College of Radiology makes a sobering contribution to the literature.
Evgeniya Sokolovskaya, D.O, M.D. and colleagues conducted a study to determine if increasing radiologist reading speed results in more misses and interpretation errors. Based on the interpretations of five experienced radiologists, they found a significant positive correlation between faster reading speed and the number of major misses and interpretation errors.
A board certified radiologist specializing in body imaging with more than 25 years of experience selected a sample of 53 noncontrast or contrast-enhanced abdomen-pelvis CT scans of variable complexity. These were categorized as one of four levels. Level 1 included noncontrast CTs with normal or noncomplex findings that did not have any prior comparison exams. Level 2 included contrast-enhanced CTs with no complex findings and no prior comparison exams. Level 3 and Level 4 exams had priors available for comparison, included both contrast-enhanced and non-contrast CTs with moderate to significant complex findings. The researchers defined complex findings as post-operative states, known malignancies, free air, intra-abdominal abscess, recurrent neoplasm, and metastatic disease.
Abdomen and pelvis studies were selected because they are among the most common examinations performed. Additionally, interpretation errors can have a significant impact on patient care. Five radiologists representing different specialties, but all of whom were experienced in reading abdomen-pelvis CT studies participated. Baseline normal reading times were established for each radiologist; the mean regular reading time was 10 minutes 9 seconds for abdominopelvic CT scans
The radiologists interpreted the studies in two stages, one enabling interpretation at normal speed and the other in half the normal amount of time. They worked in the same environment as in everyday practice under identical viewing conditions on their diagnostic workstations, using the same speech recognition dictation system. Report findings were reviewed by an experienced board certified radiologist. Mistakes and misses were categorized as major and minor.
Two of the five radiologists made more mistakes than their peers, both with respect to rapid and normal reading. The total number of major misses for the five radiologists reporting at fast speed was 16 compared to six at normal interpretation speed. Errors were made at levels 1, 3 and 4. The authors stated that the fact that five major misses at rapid reporting speed involved interpretation of the simplest cases suggests that not only subtle but obvious pathology can be easily missed when viewing time decreases.
The authors pointed out a variety of situations in which a pressured radiologist might compensate, such as setting aside a complex case to be read when not so fatigued. “The findings of our study suggest that the speed of radiology interpretations may ave a strong effect on the quality of patient care,” they wrote. They recommend that a larger, multicenter study be conducted as a follow-up to the pilot study they undertook.
Sokolovskaya E, Shinde T, Ruchman RB et al. The Effect of Faster Reporting Speed for Imaging Studies on the Number of Misses and Interpretation Errors: A Pilot Study. J Am Coll Radiol. Published online May 20, 2015.Back To Top
The tradeoff dilemma: Rapid reporting or risk of interpretation errors . Appl Radiol.
Cynthia E. Keen is a freelance writer based in Sanibel Island, Florida.