Testing for clinical inertia in medication treatment of bipolar disorder.

BACKGROUND: Clinical inertia has been defined as lack of change in medication treatment at visits where a medication adjustment appears to be indicated. This paper seeks to identify the extent of clinical inertia in medication treatment of bipolar disorder. A second goal is to identify patient characteristics that predict this treatment pattern.
METHOD: Data describe 23,406 visits made by 1815 patients treated for bipolar disorder during the STEP-BD practical clinical trial. Visits were classified in terms of whether a medication adjustment appears to be indicated, and also whether or not one occurred. Multivariable regression analyses were conducted to find which patient characteristics were predictive of whether adjustment occurred.
RESULTS: 36% of visits showed at least 1 indication for adjustment. The most common indications were non-response to medication, side effects, and start of a new illness episode. Among visits with an indication for adjustment, no adjustment occurred 19% of the time, which may be suggestive of clinical inertia. In multivariable models, presence of any indication for medication adjustment was a predictor of receiving one (OR=1.125, 95% CI =1.015, 1.246), although not as strong as clinical status measures. LIMITATIONS: The associations observed are not necessarily causal, given the study design. The data also lack information about physician-patient communication.
CONCLUSIONS: Many patients remained on the same medication regimen despite indications of side effects or non-response to treatment. Although lack of adjustment does not necessarily reflect clinical inertia in all cases, the reasons for this treatment pattern merit further examination.