The infrequency of liver function testing in patients using nonsteroidal anti-inflammatory drugs

To describe the monitoring of liver function for patients using nonsteroidal anti-inflammatory drugs (NSAIDs), we reviewed the patterns of liver function testing in a medium-sized health maintenance organization.

We examined the interval between start of therapy and first performance of a liver function test during courses of therapy of the NSAIDs diclofenac sodium, naproxen and naproxen sodium, and piroxicam. For comparison, we also studied courses of lovastatin as a "positive control," in which the anticipated frequency of liver function testing was high.

The frequency of liver function tests in patients using NSAIDs was generally low, although testing was more common in patients who used diclofenac. The probability of liver function testing was higher for patients treated in recent calendar years, for patients treated by rheumatologists, for patients who previously used NSAIDs, and for patients who had undergone a liver function test sometime in the 6 months preceding the onset of therapy.

Physicians ordered liver function tests less frequently than recommended, but the observed testing patterns appear rational in light of the very low reported frequency of serious hepatic disease in large, monitored populations of patients using NSAIDs.