Association between the commercial characteristics of psychotropic drugs and their off-label use

Abstract

Background: Off-label prescribing, or the use of a medicine for non-Food and Drug Administration (FDA)-approved indications, is especially common for psychotropic therapies and often lacks scientific support. We quantified the association between 4 commercial characteristics of prescription medicines-product age, therapeutic class age, drug volume, and promotional expenditures-and off-label use of antidepressants, antipsychotics, and mood stabilizers from 1998 through 2009.

Methods: We linked data from the IMS Health National Disease and Therapeutic Index, a nationally representative audit of office-based physicians, with data from FDA@gov and the drug compendium DrugDex, to derive information regarding off-label use. Our primary outcome was the rate (per 1000 uses) with which a drug was prescribed for non-FDA-approved indications during a given calendar year. We used mixed-effects regression models with random intercepts for each drug, adding measures of commercial characteristics as fixed effects within this model.

Results: From 1998 through 2009, the average proportion of all uses that occurred off-label was 23.3% for antidepressants, 60.7% for antipsychotics, and 54.2% for mood stabilizers. There was a positive association between the annual rate of off-label use and drug volume [incidence rate ratio (IRR), 1.41; 95% confidence intervals (CI), 1.32-1.50], although the strength of this association was not uniform across the therapeutic classes examined. There was also a small but statistically significant association between product age (IRR, 0.96; 95% CI, 0.94-0.98) and class age (IRR, 1.03; 95% CI, 1.01-1.06) and the rate of off-label use, also varying across therapeutic class. There was a statistically significant inverse association between promotional expenditures and off-label use (IRR, 0.94; 95% CI, 0.93-0.96) when controlling for our other commercial characteristics. These associations were similar when examining scientifically unsupported rather than all off-label use.

Conclusions: Our findings suggest that drug prescription volume, rather than product age or therapeutic class age, should be scrutinized further to identify settings where the public health impact of unsupported off-label prescribing may be particularly important.