Measuring Asthma Controller Adherence

Jun 1, 2012

Is It Appropriate to Measure Asthma Controller Adherence Using Pharmacy Claims Data? Published in Value in Health.


Background: Although possession-based adherence measures are typically used to compare medication adherence using pharmacy claims, these measures are difficult to verify for inhaled asthma medications because the claim-reported days’ supply may not reflect the actual duration of use.

Objectives: To examine the impact of pharmacy claim reported days’ supply on medication adherence for asthma medications between retail and mail order pharmacies.  Specifically comparing the medication adherence between patients using Flucitasone/Salmeterol disk with inhalation device (combination inhaler, CI) to those using leukotriene inhibitors (LT) which are more likely to accurately reflect days’ supply in the claim.

Methods: Commercially insured U.S. patients aged 12-63 and medically diagnosed for asthma were followed for one year after initiation of an asthma controller using a retrospective, claims-based design. Adherence was defined using Medication Possession Ratio (MPR). CI and LT subsets included all patients using one of these as the index medications. The multivariate-adjusted relationship between adherence and channel was evaluated using a generalized linear model.

Results: 6,014 patients were included in the overall study cohort, with 3,410 in retail and 2,640 in mail-order.  Of those using either a CI or LT (n=4,106), 54% (2,222) were in  the CI subgroup and 46% (1,884) were in the LT subgroup. The adjusted MPR for asthma controllers in the retail pharmacy cohort was 39.7% (95% CI 37.08-42.52%) compared to 62.4% (95% CI 58.19-66.97%) in the mail-order cohort. In comparison, the adjusted MPR for retail and mail-order pharmacy cohorts were 38.3% (95% CI 34.58 - 42.49%) and 57.9% (95% CI 52.13-64.20%) in the CI subgroup and 49.0% (95% CI 43.83-54.79%) and 69.3% (95% CI 52.13-77.56%) in the LT subgroup, respectively.

Conclusions: In a large, nationally representative cohort, adherence to asthma controllers and differences in adherence across dispensing channels were similar for the subgroups with more accurate days’ supply information (i.e., CI and LT). Thus the findings suggest that inaccuracies on the days’ supply had no impact when comparing the effect of MPR between retail and home delivery.  It may be reasonable to use possession-based asthma adherence measures derived from pharmacy claims, despite potential errors in capturing days’ supply.