Non-adherence an important challenge

The results also show there are differences in persistence at the end of the first year of treatment comparing three commercially available PGA agents with an advantage for latanoprost (Xalatan, Pfizer) versus both bimatoprost (Lumigan, Allergan) and travoprost (Travatan, Alcon Laboratories), said Gail F. Schwartz, MD.

"Persistence with ocular hypotensive therapy has been investigated in a number of studies using a variety of methods," said Dr. Schwartz, a glaucoma specialist in private practice, and assistant professor, Wilmer Eye Institute, Johns Hopkins University School of Medicine, Baltimore.

Continued research targeting reasons for lack of adherence and best methods to detect and improve medication adherence remain a crucial component to successful therapy, according to Dr. Schwartz.

"As many studies have shown, it is difficult to assess which patients are non-adherent," she said. "Education, open-ended inquiry, and discussion with the patient are important for their potential to improve adherence."

Dr. Schwartz was a co-author of a retrospective study [Reardon G, Schwartz GF, Kotak S. Persistence on prostaglandin ocular hypotensive therapy: an assessment using medication possession and days covered on therapy. BMC Ophthalmol. 2010;10:5] that was based on a U.S. pharmacy claims database.

More than 42,000 patients were identified who had a prescription dispensed for one of the three PGA agents between Jan. 1 and Dec. 31, 2004.

Persistence study

After applying exclusion criteria, 1,464 patients treated with bimatoprost, 4,994 patients treated with latanoprost, and 1,415 patients treated with travoprost were retained for the analysis. Persistence was calculated based on two methods:

• Whether the last prescription filled provided a sufficient supply to achieve medication possession at the end of the first year of therapy ("medication possession").

• By determining the number of days during the first therapy year for which there was available medication supply ("days supply").

For both measures, three separate imputation methods for days supply were applied to address uncertainty regarding the number of days supply available in a given prescription fill of a study drug. Regression analysis techniques were applied to investigate differences in persistence among the three medications with adjustments for gender, age, and previous ocular hypertension diagnosis.

Depending on the imputation method used, the medication possession rates for the total patient population ranged from 28% to 48%, while the number of days covered during the first year of therapy ranged from 131 to 236. For both measures of persistence, there were statistically significant differences favoring latanoprost over each of the other PGA agents.

Across the multiple analyses and using latanoprost as the reference group, the odds of achieving medication possession at the end of the first year were 26% to 34% lower for bimatoprost and 34% to 36% lower for travoprost. Compared with the latanoprost group, the number of days covered during the first year was 21 to 29 fewer days lower in the bimatoprost group and 33 to 42 fewer days lower in the travoprost group.

Analysis methods

The methods for analyzing persistence were based on those used in a previous study of patients taking PGA agents by Wilensky et al. However, though the Wilensky study excluded patients who discontinued therapy for lack of efficacy or for side effects, Dr. Schwartz and colleagues adjusted for potential selection bias in excluding those patients.

"Pharmacy claims-based analyses are very useful in determining actual rates of persistence in terms of prescription filling behavior, although they are unable to identify reasons for lack of adherence or account for the influence of medication sampling," Dr. Schwartz said. "The methodology of these claims-based studies can affect the outcome strongly, and hence, it is imperative to apply inclusion criteria that do not inadvertently eliminate a sample of the population, such as early failures to refill, providing a representative snapshot of the real-world patients.

"A patient who discontinues the medication early is non-adherent whether it be for tolerability, efficacy, or one of many more personal reasons, and it is as important to account for such failures as it is to assess longer-term failure in patient persistence with therapy," she added.

Persistence also was analyzed using an "early failures" test to screen for patients unable to refill their medication within 90 days of the first fill. Consistent with the findings of the other measures of persistence, the early failure rate was lower for latanoprost users compared with bimatoprost and travoprost, 37% versus 47% and 49%, respectively. Failure to refill the index agent within the initial 90 days was a strong predictor of poor persistence.

FYI

Gail F. Schwartz, MD
E-mail: schwartzgf@aol.com

Dr. Schwartz is a consultant to Pfizer.