Healthcare practitioners are often unable to say how one treatment e.g. a specific medication, compares with another, e.g. increasing exercise, on reducing the signs or symptoms of a given disease, unless there are trials directly comparing these two kinds of treatments. Most effective behavioural change interventions for chronic disease e.g. increasing exercise, smoking cessation and alcohol reduction have not been directly compared against biomedical interventions e.g. medications or surgery. Therefore, there are critical knowledge gaps around what are the best treatments for many chronic diseases. This project used a methodology that has been developed relatively recently in biostatistics i.e. network meta-analysis, to compare blood pressure medications with exercise interventions on blood pressure reduction in people with high blood pressure i.e. hypertension. Most of the identified randomised controlled trials (87%) included in our network meta-analysis focused on medications. The results indicated that antihypertensive medications were more effective than exercise but there was insufficient evidence to suggest that first-line medications significantly reduced BP to a greater extent than did the exercise interventions. In addition to conducting the analyses and disseminating the findings of this work we also provided training in network meta-analysis for health researchers by hosting training events and instructional publications for an international audience on using the method to compare behavioural and biomedical interventions. The project has helped advanced knowledge about comparative effectiveness analyses of key health interventions for hypertension and has contributed to creating a larger international community of researchers trained to use this powerful analytical tool.
