Heterogeneity in antihypertensive treatment discontinuation between drugs belonging to the same class

Discontinuation of antihypertensive treatment is known to be different for different classes of antihypertensive drugs. No information is available on whether this phenomenon differs for drugs belonging to the same class. This is clinically relevant because treatment discontinuation is mainly responsible for poor blood pressure control in the antihypertensive population.

We studied a large (n=131,472) cohort of patients aged 40-80 years who lived in Lombardy (Italy) and received their first antihypertensive drug prescription during 2005. Discontinuation was defined by the absence of any antihypertensive drug prescription during the 90-day period following the end of the latest prescription. Class-related and drug-related discontinuation rates were standardized according to the demographic and therapeutic structure of the entire cohort and expressed as number of patients who experienced discontinuation every 100 person-months.

Standardized rates of discontinuation ranged from 6.2 to 24.4 events every 100 person-months for patients who started monotherapy with an angiotensin receptor antagonist and a diuretic, respectively. However, there was a significant heterogeneity between treatment discontinuation rates within each class and the heterogeneity differed between classes. The highest discontinuation rate was 13.9-fold for channel blockers, but only 1.7-fold for angiotensin receptor antagonists. Within this class, losartan showed a discontinuation rate significantly greater than that of the other angiotensin receptor antagonists whose discontinuation rate was similar. A significant heterogeneity also characterized initial treatment with fixed-dose combinations of different angiotensin-converting enzyme inhibitors or angiotensin receptor antagonists with a diuretic.

Comparison of treatment discontinuation between antihypertensive drug classes masks the fact that this phenomenon is heterogeneous within any given class. This is relevant to calculations of the cost-benefit of treatment, which, thus, should be drug-based rather than class-based.