An epidemiological study suggests that non-adherence to therapy is a significant cause of re-admission for heart failure, however treatment of these patients is generally straightforward.

Heart failure (HF) is a significant cause of hospital admission, and many patients will be re-admitted – a factor that predicts poorer long term outcomes. It is not clear what proportion of re-admissions are due to non-adherence to the sometimes complex drug and dietary regimens used in treatment of HF, and the authors of this study intended to clarify these in a large adult patient population. The study used administrative data from a large US hospital network that includes academic and community hospitals. Data were collected as part of a quality improvement initiative, and for this analysis were used to determine factors associated with non-adherence and the association between non-adherence and length of stay (LOS) and in-hospital mortality.

Eligible patients were those re-admitted to hospital for HF where the cause of re-admission was documented. Those where re-admission was due to non-adherence (drug, diet, or both) were identified, and differences between these and adherent patients analysed. Study period was January 2005 to December 2007 inclusive.

In total, 95,127 potentially relevant patients were identified among 333 hospitals over the study period; of these, 54,322 patients from 236 hospitals were eligible for analysis. Non-adherence (dietary or medication) was documented as a reason for admission in 5,576 (10.3%) of the total. Non-adherent patients were more likely (P<0.0001 for all) to be male (60.4% of non-adherent vs. 48.4% of adherent), younger (mean age 64.2 vs. 73.6) and of minority ethnic origin (47.0% vs. 25.4%). They were also more likely to not have health insurance (11.7% vs. 3.3%).

In terms of baseline clinical characteristics, non-adherent patients were more likely (P<0.0001 for all) to have non-ischaemic cardiomyopathy (57.4% vs. 50.6%), atrial fibrillation (31.0% vs. 24.4%), and to have a history of frequent re-admissions (2 or more in the previous 6 months, 18.2% vs. 11.3%). They also had lower left-ventricular ejection fractions (34.9% vs. 39.6%) and higher brain natriuretic peptide levels (1,813 vs. 1,371 picograms/ml).

Patients who were non-adherent to diet were more likely to have diabetes and higher BMI, whereas the demographic characteristics (gender, age, and insured status) tended to identify those who were non-adherent to medication.

Despite having a number of apparently poor prognostic characteristics on admission, patients with non-adherence tended to have shorter lengths of hospital stay and fewer complications than adherent patients (mean length of stay 4.99 vs. 5.63; mortality 1.55% vs. 3.49%; both P<0.0001).

The authors conclude that non-adherence to therapy is a common cause of re-admission for HF. Although these patients may have an apparently higher risk profile, their length of stay and in-hospital mortality are lower than adherent patients, suggesting that resuming previous therapy is frequently sufficient to achieve disease control. In socio-demographic terms, non-adherent patients tend to be disadvantaged in comparison with those who adhere to treatment.