Treatment Interruptions and Non-Adherence with Imatinib and Associated Healthcare Costs

Abstract

Identify treatment interruptions and non-adherence with imatinib; examine the clinical and patient characteristics related to treatment interruptions and non-adherence; and estimate the association between treatment interruptions and non-adherence with imatinib and healthcare costs for US managed care patients with chronic myeloid leukaemia (CML). This retrospective analysis utilised electronic healthcare claims data from a US managed care provider. Adult patients with CML (as determined by International Classification of Diseases, ninth revision, Clinical Modification [ICD-9-CM] diagnosis code) were identified who began treatment with imatinib from 1 June 2001 through 31 March 2004. Treatment interruptions (i.e. failure to refill imatinib within 30 days from the run-out date of the prior prescription) were identified during the 12-month follow-up period. Medication possession ratio (MPR), calculated as total days' supply of imatinib divided by 365, was also examined. Healthcare costs (i.e. paid amounts for all prescription medications and medical services received, including health plan and patient liability) were examined in three ways: (i) total healthcare costs; (ii) total healthcare costs exclusive of imatinib costs; and (iii) total medical costs. All costs were converted to US dollars (2004 values) using the medical component of the Consumer Price Index. MPR was modelled using ordinary least squares regression. Presence of treatment interruptions was modelled using logistic regression. The association between MPR and healthcare costs was estimated using a generalised linear model specified with a gamma error distribution and a log link. All models included adjustment for age, gender, number of concomitant medications, starting dose of imatinib and cancer complexity. A total of 267 patients were identified. Average age was approximately 50 years, and 43% were women. Mean MPR was 77.7%, with 31% of patients having a treatment interruption. However, all of these patients resumed imatinib within the study period. In this population, MPR decreased as the number of concomitant medications increased (p = 0.002), and was lower among women (p = 0.003), patients with high cancer complexity (p = 0.003) and patients with a higher starting dose of imatinib (p = 0.04). Women were approximately twice as likely as men to have a treatment interruption (p = 0.009), as were patients with a high cancer complexity (p = 0.03). After adjusting for the aforementioned covariates, MPR was found to be inversely associated with healthcare costs excluding imatinib (p < 0.001) and medical costs (p < 0.001). A 10% point difference in MPR was associated with a 14% difference in healthcare costs excluding imatinib and a 15% difference in medical costs. For example, patients with an MPR of 75% incur an additional 4072 US dollars in medical costs annually compared with patients with an MPR of 85%. Treatment interruptions and non-adherence with imatinib, both of which could lead to undesired clinical and economic outcomes, appear to be prevalent. Physicians and pharmacists should educate patients and closely monitor adherence to therapy, as improving adherence and limiting treatment interruptions may not only optimise clinical outcomes but also reduce the economic burden of CML.