Medicine Compliance and Survival in Heart Failure

Abstract & Commentary

By Michael H. Crawford, MD, Professor of Medicine, and Chief of Clinical Cardiology, at the University of California, San Francisco. Dr. Crawford is on the speaker's bureau for Pfizer.

Source: Gislason GH, et al. Persistent Use of Evidence-Based Pharmacotherapy in Heart Failure Is Associated With Improved Outcomes. Circulation. 2007;116:737-744.

The current polypharmacy management approach to chronic heart failure due to systolic dysfunction is of concern because it may lead to noncompliance. However, little is known about compliance with heart failure medications and its effect on outcome. Thus, this analysis of the initiation and persistence of evidence based therapy in all 107,072 Danish patients who survived their fist hospitalization for heart failure over a 9-year period is of interest. Via national pharmacy records, initiation of renin-angiotensin-system blockers (RASB), beta blockers and spironolactone within 90 days of discharge and statins with 180 days was recorded. Patients were stratified into 4 heart-failure severity groups by their use of diuretics. Subsequent prescription refills were used to define the persistence of therapy. Nonpersistence was defined as a > 90 day break in therapy. Only 123 (0.1%) of patients were lost to follow-up. Results: Initiation of RASB therapy was noted in an average of 43%, but was 50% during the last year of the study; 27% and 43% for beta blockers; 19 and 25% for spironolactone; and 12 and 27% for statins. Drug doses were often considerably below targets, and with the exception of carvedilol, were rarely up-titrated. Short breaks in therapy were common. Setting those who initiated treatment at 100% at the end of 5 years 70, 65, 56 and 83% were still on RASB, beta blockers, spironolactone and statins. One year mortality decreased from 39% to 35% (p < 0.001) over 8 years. Higher disease severity (diuretic use) correlated with therapy persistence. Nonpersistence of RASB, beta blockers, spironolactone and statins was associated with increased mortality (HR 1.37, 1.25, and 1.88, respectively). The authors concluded that once medications were started, persistence of therapy was high, but often with lower than recommended doses. Nonpersistence was associated with higher mortality.


This study is particularly strong because it includes all Danes with a heart failure diagnosis, avoiding selection biases. It has near 100% follow-up over 10 years. Accurate pharmacy data are available and small co-payments were all that was needed to fill prescriptions, avoiding non-compliance due to cost. That patients who don't take their drugs, don't do well is no surprise, but the fact that multiple drug prescriptions did not equate to noncompliance was. Apparently when drug polytherapy is started early, persistence with therapy is enhanced. Preferably, heart failure drugs should be started in the hospital, even in small doses, to enhance compliance. What is also surprising is that the doses of all drugs but carvedilol were rarely up-titrated after discharge. Perhaps this improved compliance because adverse effects were avoided, but perhaps at the expense of less efficiency. Interestingly, in the CHARM study (candesartan in heart failure) compliance with study drug or placebo was associated with better outcomes. Perhaps compliance is a marker for patients that take better care of themselves in general so drug doses are less critical. The message is if you can get your patients to be compliant with therapy, even at below recommended doses, they will do better. Heroic attempts to increase drug doses to reach recommended levels may not be worth the effort.

There are limitations to this type of study and to this study in particular. This study used administrative databases, so there is a paucity of clinical data. For example the diagnosis of heart failure was not verified and the existence of contraindications to various drugs is not known. Also, drug initiation increased over the course of the study for unknown reasons. We know Denmark had special heart failure clinics, but not how many, when and who used them. Other studies have shown increased compliance with therapy and better outcomes when patients are enrolled in such clinics. The main message of this study is to start multiple drug therapy early, preferably in the hospital, and don't forget to try to up-titrate to doses known to be effective if you can. Don't reduce compliance by forcing up-titration in the face of adverse effects, because compliance with a regimen for heart failure, even if it is based upon a placebo, will improve survival.