The effects of three different implementation strategies for heart failure guidelines on the management of heart failure in New Zealand primary care: a cluster randomised trial

Abstract

This thesis aimed to evaluate the effectiveness of three implementation strategies on change over time in the primary care management of heart failure (HF) in New Zealand. Emergence of the Internet as a tool for continuing medical education (CME) prompted the study, which compared an Internet-based CME course, small–group education sessions and a passive mail-out of national guidelines. These were “one-off” educational interventions as is common in CME. A single-blind stratified cluster randomised controlled trial (cRCT) design was used. Sixty nine practices were randomised to one of the three implementation strategies. The two active arms participated in almost identical education sessions based on four recommendations of the 2001 New Zealand Heart Foundation HF guideline. These recommendations were selected as primary study outcomes because their performance was suboptimal. Two of the recommendations – use of echocardiography and use of high dose angiotensin converting enzyme inhibitors (ACEi) – were not new in HF management. Introduction of β-blockers to the treatment schedule illustrated a dramatic change in accepted treatment. The indication for reintroducing spironolactone for HF had changed. Patients were identified using a HF scoring system. Twenty four practices completed the study and 359 patients participated. Retrospective data were collected from the patient cohort for up to five years. Most echocardiograms were performed before the educational intervention. In each group, ACEi prescription decreased over time, β-blocker use increased and spironolactone use remained static. None of the three implementation strategies had a statistically significant effect on outcomes. Two patient variables that negatively predicted echocardiography referral and prescribing were older age and female sex. The increase in β-blocker prescribing was mainly attributed to initiations that occurred in secondary care before the educational intervention. The intervention arms did not affect the dose prescribed. Older age was negatively associated with dose. None of the implementation strategies was superior in promoting change in primary care. Changes that occurred either were negative or could be attributed to factors operating beyond the control of the study, not to education. Ongoing work is required to develop interventions that will reduce barriers to changing HF management in primary care.