Background The MRC framework describing the development-evaluation-implementation process for phase III complex intervention trials has been extended to include implementation research. We present i) the effectiveness evaluation of the PACE-UP pedometer-based walking intervention and ii) an exploration of implementation planning in routine practice using this extended framework.

Methods i) 1023 inactive 45–75 year old primary care patients were randomised to: usual care (338); postal pedometer intervention (339); nurse-supported pedometer intervention (346). Intervention groups received pedometers, 12-week walking programmes, and physical activity (PA) diaries. The nurse group were offered three PA consultations. The primary and main secondary outcomes were changes from baseline to 12-months in average daily step-counts and time in moderate-to-vigorous PA (MVPA) in ≥10 minute bouts, measured over 7 days by accelerometry. ii) the extended MRC framework was used for implementation planning: long-term follow-up and dissemination; implementation development; feasibility and piloting of implementation; and phase IV controlled studies.

Results i) 956 (93%) provided outcome data. Baseline average daily step-counts for the whole cohort were 7479 (s.d. 2671) steps/day and average time in MVPA ≥10 minute bouts was 94 (s.d. 102) minutes/week. Both intervention groups significantly increased step-counts and MVPA compared to controls at 12 months, with no significant differences between interventions. Additional steps/day: postal 641 (95% CI: 328, 954), nurse-support 682 (95% CI: 371, 994); additional MVPA in bouts (minutes/week) postal 33 (95% CI: 17, 49), nurse-support 35 (95% CI: 19, 51). ii) Long-term follow-up is funded and in progress; dissemination is underway, including adoption by the local CLAHRC and liaison with national bodies such as Public Health England; funding for intervention development including mobile apps and online resources has been applied for; feasibility testing and piloting of different patient recruitment methods from primary care is planned locally through the CLAHRC; a funding application for a phase IV controlled evaluation study is being developed.

Conclusion A primary care pedometer-based walking intervention in inactive 45–75 year olds was effective at increasing step-counts by approximately a tenth and time in MVPA in bouts by approximately a third. Nurse delivery had no greater effect on 12-month PA outcomes than simpler, resource-efficient postal delivery. A primary care postal pedometer intervention could help address the public health physical inactivity challenge. The MRC extended framework for implementation is helpful for structuring the complex implementation process required to translate research findings into practice.