Methods
Inclusion Criteria
Study Design We considered study designs in the following order: overviews of systematic reviews and systematic reviews, randomised controlled trials, cluster-randomised controlled trials, quasi-randomised controlled trials, controlled before-and-after studies, and interrupted time-series analyses.
Setting. Primary health care/community.
Population. Adults (≥18 years) with low levels of PA or increased risk of, or diagnosis of, disease for which physical activity may be a protective factor, e.g. cardiovascular disease, and type 2 diabetes.
Intervention. Organised follow-up over a period of 10–14 weeks, individually or in groups, given within a local organisation or by a single health professional (excluding general practitioners), starting with individual goal-setting and aiming to support increased PA.
Comparisons. 1) Advice (with or without written information) about PA from health professionals without organised follow-up outside the office, 2) usual care, and 3) no treatment.
Outcomes. The primary outcome was PA behaviour, e.g. frequency, duration, intensity, achievement of pre-set goals, or indications of PA, e.g. energy expenditure, aerobic capacity.
Language. There were no language restrictions for the literature search. Publications in other languages than English or Scandinavian have been translated if judged relevant.
Literature Searches
We searched MEDLINE, EMBASE, Cochrane Database of Systematic Reviews, Cochrane Central Register, DARE and HTA (via Centre for Reviews and Dissemination), Cinahl, PsychINFO, Sociological Abstracts, and Social Science citation Index up to October 2011. In addition, we searched manually in a) reference lists of relevant systematic reviews identified in the electronic search, b) reference lists of included publications, c) the following journals that most commonly publish papers that could potentially match our inclusion criteria (publication dates were January 2009 – February 2012): American Journal of Preventive Medicine, BMC Family Practice, BMC Public Health, European Journal of Public Health, Preventive Medicine, Scandinavian Journal of Primary Health Care, and Scandinavian Journal of Public Health, and d) reference lists of relevant protocols identified in c). The search was updated by a search in MEDLINE in June 2012. The search strategy for the 2012 search is available in Additional file 1 http://www.biomedcentral.com/1471-2296/15/120/additional.
Two authors (ED and GEV, ED and VU or ED and RCB) independently screened titles and abstracts and assessed retrieved full texts against a set of pre-determined inclusion criteria. Discrepancies were solved by consensus or by a third person. There was no need to contact authors to provide additional data.
Data Extraction and Analysis
The first author (ED) extracted study characteristics (study design, population, intervention(s), comparison(s), and outcome(s)) and study results (descriptive discrete or continuous data, effect measures and effect estimates) using a pre-designed data extraction form. One other author (GEV, VU or RCB) verified the extracted data against the full text articles. Risk of bias at study level was assessed according to the Cochrane Handbook and independently by two authors. If consensus was not reached a third person was consulted. The GRADE approach was used to rate our confidence in the effect estimate for each outcome in each comparison. The interventions were categorized according to main content, and in each category effect measures and effect estimates were described for each comparison. When several outcome measures were reported, we chose the measure(s) that best reflected total physical activity.