Cost and consequences of noncompliance to oral bisphosphonate treatment

Abstract

BACKGROUND: Despite the favorable efficacy, safety, and cost-effectiveness profile of bisphosphonate (BIS) treatment for osteoporosis (OP), patient compliance remains suboptimal. A longer follow-up period could help to better characterize patient behavior as well as the predictors of noncompliance because of the extended durations of osteoporosis and time to a fracture. OBJECTIVE: To determine health care outcomes associated with compliance and noncompliance to BIS therapy in women diagnosed with OP. METHODS: This retrospective claims study focused on women with OP, who were aged 55 years and older and using oral BIS treatment. Patients were identified within the HealthCore Integrated Research Environment (HIRE) between January 1, 2007, through June 30, 2010. Patients were required to have ≥ 12 months of pre-index eligibility and ≥ 24 months of post-index health plan eligibility. Post-index eligibility was split into 2 periods: (1) the compliance time period (the first 12-month post-index period, in which compliance was determined) and (2) the cost and consequences time period (13- to 24-month post-index period during which time health care resource utilization, cost, and outcomes were assessed). Noncompliance was defined as medical possession ratio (MPR) < 70%. Descriptive statistics described outcome variables for the study population. A logistic regression model determined variables predictive of compliance. Further, a generalized linear model was used to examine associations between all-cause or OP-related medical/total costs and to estimate health care utilization. RESULTS: Of patients overall (N = 27,905), 59% were noncompliant, and 62% discontinued medication. Among noncompliant patients, 6.7% switched BIS therapy (after 64 days average); 97% discontinued (87 days average); and 21% restarted medication (218 days average). Of noncompliant patients, 14% had > 1 inpatient visits; 16% had > 1 emergency room visits; 94% had >1 outpatient visits; and 95% had > 1 office visits. Logistic regression results indicated that under aged 65 years (P = 0.012) predicted noncompliance. Relative to the compliant group, noncompliant patients had higher fracture rates at post-index second year, 3.3% vs. 2.4%, and combined second and third years, 6.0% vs. 4.8%, respectively. Compared with noncompliant patients, compliant patients had 9% (P = 0.007) lower OP-related costs, 3% lower all-cause costs during the second post-index year, and 11% (P = 0.016) lower OP-related costs. Mean 13- to 24-month post-index period all-cause costs were $7,237 for noncompliant patients versus $6,758 for compliant patients (P = 0.008). CONCLUSIONS: These results indicate high noncompliance rates in this population of older females. OP medication compliance was associated with lower fracture rates, OP- and all-cause costs, and health care utilization. These findings highlight the financial implications and treatment outcomes of BIS medication noncompliance within a female osteoporotic population.