Background--Undertreatment with recommended pharmacotherapy is a common problem in heart failure and may influence prognosis. We studied initiation and persistence of evidence-based pharmacotherapy in 107 092 patients discharged after first hospitalization for heart failure in Denmark from 1995 to 2004.

Methods and Results--Prescriptions of dispensed medication and mortality were identified by an individual-level linkage of nationwide registers. Inclusion was irrespective of left ventricular function. Treatment with renin-angiotensin inhibitors (eg, angiotensin-converting enzyme inhibitors and angiotensin-2 receptor blockers), -blockers, spironolactone, and statins was initiated in 43%, 27%, 19%, and 19% of patients, respectively. Patients who did not initiate treatment within 90 days of discharge had a low probability of later treatment initiation. Treatment dosages were in general only 50% of target dosages and were not increased during long-term treatment. Short breaks in therapy were common, but most patients reinitiated treatment. Five years after initiation of treatment, 79% patients were still on renin-angiotensin inhibitors, 65% on -blockers, 56% on spironolactone, and 83% on statins. Notably, multiple drug treatment and increased severity of heart failure was associated with persistence of treatment. Nonpersistence with renin-angiotensin inhibitors, -blockers, and statins was associated with increased mortality with hazard ratios for death of 1.37 (95% CI, 1.31 to 1.42), 1.25 (95% CI, 1.19 to 1.32), 1.88 (95% CI, 1.67 to 2.12), respectively.

Conclusions--Persistence of treatment was high once medication was started, but treatment dosages were below recommended dosages. Increased severity of heart failure or increased number of concomitant medications did not worsen persistence, but nonpersistence identified a high-risk population of patients who required special attention. A focused effort on early treatment initiation, appropriate dosages, and persistence with the regimen is likely to provide long-term benefit.