Researchers analyzing pharmacy and medical claims among nearly 94,000 individuals with ongoing statin prescriptions found that prescription filling behavior before and after the index fill had the highest predictive value for continued adherence.
Researchers analyzed pharmacy and medical claims among nearly 94,000 individuals with ongoing statin prescriptions to assess which of their 10 models was most successful at predicting adherence. Their findings, published in the Journal of Managed Care & Specialty Pharmacy, indicated that prescription filling behavior before and after the index fill had the highest predictive value for continued adherence.
Prior studies have found that models based on short-term prescription fill behaviors are most accurate at predicting adherence among patients initiating a new therapy, compared with other models constructed with medical claims data. However, little information existed on which models could best predict adherence among the much larger population of established medication users. As such, a group of researchers set out to evaluate the predictive value of 10 models for adherence among patients with ongoing statin prescriptions.
Using medical and claims data from a large commercial insurer, the study authors identified 93,777 individuals with an index statin fill in January 2012 who had filled at least 1 statin in the last 6 months of 2011. These data included socioeconomic and other demographic information, as well as claims for all procedures, visits, hospitalizations, and filled prescriptions. Adherence was determined by a cutoff of 80% proportion of days covered from the first prescription through the end of 2012 based on observed statin fills.
The investigators identified 6 groups of predictors that could drive adherence: pharmacy and demographic predictors or medical predictors chosen by the researchers, pre-index adherence, post-index filling, and high-dimensional propensity scores calculated with either pharmacy or medical claims data. These propensity scores were complex algorithms that identified the 400 prescription or medical codes most strongly associated with adherence, in an attempt to serve as a proxy for patient-level characteristics.
They then constructed 10 models using combinations of these predictors, and conducted analyses to assess the predictive power of each of the models across varying follow-up periods within the 12-month study period. For instance, the adherence outcomes linked to the post-index fill indicators were measured from April to December 2012 to “exclude temporal overlap between predictors and outcome.”
The researchers found that the models based on observed adherence, whether before or after the index fill, were most accurate at predicting which patients would maintain adherence. The strongest model incorporated pre-index pharmacy claims, demographic information, and medical claims in addition to post-index fills.
Interestingly, the complex propensity scores and the investigator-specified medical claims models performed the worst at predicting adherence. The authors wrote that these findings support the literature on predictors of adherence among new statin users and indicate that “adherence is not only a dynamic process, but also one with predictable patterns of fluctuation over the course of a patient’s lifetime on treatment.”
They recommended that efforts to improve adherence should be sure to provide ongoing support, instead of a one-time intervention. They also noted that the simplicity of the best predictive model could make it easier for clinicians to tackle the complex problem of medication adherence.
“In particular, the success of the 6 simple adherence-based pharmacy claims-derived predictors could form the basis of an intervention that could be easily implemented in a pharmacy or clinical outpatient setting,” they concluded.
Overcoming Employment Barriers for Lasting Social Impact: Freedom House 2.0 and Pathways to Work
April 16th 2024To help celebrate and recognize National Minority Health Month, we are bringing you a special month-long podcast series with our Strategic Alliance Partner, UPMC Health Plan. Welcome to our second episode, in which we learn all about Freedom House 2.0 and the Pathways to Work program.
Listen
Pegcetacoplan for PNH More Cost-Effective Than Anti-C5 Monoclonal Antibodies
April 18th 2024A cost-utility analysis conducted from the perspective of the Italian health system found that pegcetacoplan was more effective and less costly than 2 complement 5 (C5) inhibitors for the treatment of paroxysmal nocturnal hemoglobinuria (PNH).
Read More
Making Giant Strides in Maternity Health Through Baby Steps
April 9th 2024To help celebrate and recognize National Minority Health Month, we are kicking off a special month-long podcast series with our strategic alliance partner, UPMC Health Plan. Welcome to our first episode, which is all about the Baby Steps Maternity Program and its mission to support women throughout every step of their pregnancy journey.
Listen
Collecting SDOH Data Can Assess Risk of Medical Nonadherence, Improve HEI and Star Ratings
April 18th 2024At the Academy of Managed Care Pharmacy (AMCP) 2024 annual meeting, a panel of presenters explored changes coming to Medicare that incorporate social determinants of health (SDOH) data to improve patient and health system outcomes.
Read More
Many Patients With Psoriasis in Clinical Trials Experience Nocebo Effects, Study Finds
April 18th 2024Half of patients exposed to placebo in clinical trials experienced adverse events (AEs), which may be partially explainable by nocebo effects, according to a recent review and meta-analysis.
Read More