‘10 Small Steps’: A randomised controlled trial to improve health behaviours using computer-tailored advice in a general practice setting

Abstract

Objective: Consensus exists regarding the urgent need to address non-communicable diseases by developing effective strategies to modify health risk behaviours such as insufficient physical activity, smoking, alcohol use and diet. Sir Geoffrey Rose first proposed that prevention can be better achieved by targeting whole populations rather than high risk individuals; however, the area of non-communicable diseases remains shaped by the high risk approach. Research is needed to develop and test health behaviour intervention models that have wide reach and are easy to incorporate into existing practice. The ‘10 Small Steps’ study, a randomised controlled trial, is based on the notion that it might be possible to create an effective system in primary care to help large numbers of people to adopt and maintain healthier lifestyle. Aims: The overarching question addressed in this doctoral thesis is: ‘Can a low intensity computer-tailored intervention be used to motivate large numbers of individuals to adopt and maintain a healthier lifestyle and diet?’ The set of studies reported in the thesis aimed to: (i) develop a reliable and valid self-report summary measure to assess and provide personalised feedback on current health behaviours in the general practice setting; (ii) develop and test an intervention designed to produce short-term ( at 3 months after baseline) improvement in ten health behaviours; and (iii) test if health improvements can be maintained over a longer term (at 12 months after baseline) and if an additional 3-month contact improves maintenance of health behaviour change over and above a single one-off intervention. Methods: The ‘10 Small Steps’ study took place in Brisbane, Australia between October 2006 and December 2009. A pilot study involving two general practices and 113 patients sought to establish the reliability and general utility of the study questionnaire. The main intervention trial recruited and randomised 4678 adult participants from 21 general practitioners. Participants were randomised to four groups: contact at baseline only (‘single intervention’ and corresponding control group) and contact at baseline and 3 months (‘dual intervention’ and corresponding control group). At each contact the participants received a computer-tailored feedback and one page information sheet according to their allocation to intervention or control groups. Current recommended guidelines for ten key health behaviours were used to develop the ‘Prudence Score’, a summary health score that assessed an individual’s health behaviours. Data were collected at 3 months from half of the participants (the dual group) to test the short term effectiveness of the intervention measured as increase in the percentage of participants adhering to the suggested guidelines and as increase in the mean Prudence Score. At 12 months invitations to participate in the follow-up were sent to all participants with the aim of testing the long term maintenance of behaviour change. Results: At baseline the response was 56.5% (4678 of 8343 invited participants) and the study sample was primarily female (68.7 %), married or living as married (68.8 %), with a diploma or university degree (56.6 %) and an average age of 47 years. The mean Prudence Score was 5.80 (95% CI 5.75-5.85). Only 5.1% of the study population followed five important health behaviours. At 3 months participant response was 76.2% (1711 of 2309). The intervention group showed significant increase (F = 13.3, p<0.05) in the mean Prudence Score, with baseline to 3 months change in the intervention group of 0.37 (from 5.88 to 6.25) compared with improvement of 0.12 in the control group (from 5.84 to 5.96). More specifically, three individual health behaviours (salt, fish and butter intake) showed significant positive change (p<0.05). At 12 months participant response was 66.8% (3065 of 4676 allocated) and both intervention groups (dual and single) showed significant improvement in greater than half the individual health behaviours. The mean Prudence Score (net change of 0.47 in dual and 0.15 in single intervention group) compared to the control groups was observed. However, no additional positive change at 12 months was observed in the dual intervention group when compared to the single intervention group after adjusting for age, education, time and group. Conclusion: These findings demonstrate the potential for a low-intensity intervention to improve the adoption and maintenance of health behaviours in a primary care population and for general practice as a conduit for the primary prevention of non-communicable diseases. Although the individual behaviour changes resulting from the intervention were relatively small the Prudence Score, which is easy to calculate and interpret, appears to be a useful tool for promoting a population-wide primary prevention strategy. Further research is needed for translation of this model into practice, which possibly can provide a sustainable system in primary care to assist a large number segment of the population to improve lifestyle and can form a foundation of successful health behaviour change.