Do abnormal liver function tests predict inpatient imaging yield? An evaluation of clinical decision making

Diagnostic abdominal imaging is frequently performed in hospitalized patients to assess the cause of abnormal liver function tests (LFTs). We undertook this study to assess whether the extent and severity of LFTs abnormalities predicted the yield of inpatient imaging.

We retrospectively reviewed inpatients' abdominal imaging studies performed for abnormal LFTs during a 27 month period. Imaging results were matched to LFTs performed during a 5 day collection window surrounding the image request date. Five LFTs were categorized by severity of abnormality and were then collapsed into three classes based on pathophysiology.

Among 759 imaging studies completed for the indication of abnormal LFTs, 196 (26%) were positive (abnormal and explained the abnormal LFTs). Among the LFT classes, severity of laboratory test abnormality correlated with positive imaging examinations yield only for the transaminases: 18% for the normal-mildly abnormal transaminases compared with 31% for moderately-severely abnormal transaminases. The number of abnormal LFT classes per study correlated only slightly better: 21% of patients with none or one abnormal LFT class had a positive imaging study compared with 35% of patients with all three abnormal LFT classes.

The yield of inpatient abdominal imaging for abnormal LFTs correlates only weakly with both the severity and the extent of different abnormal LFTs. Further research is needed to define the optimal imaging strategies for evaluating inpatients with suspected hepatobiliary disease.