Literature searches for systematic reviews
For each of the clinical questions, the research assistant:
- identified the relevant population(s), intervention(s) (or exposure), comparator(s) and outcome(s)
- developed a search strategy, in consultation with a clinical expert from the relevant working group
- ran literature searches in multiple databases (Cochrane Database of Systematic Reviews, MEDLINE, EMBASE, Cochrane Central Register of Controlled Trials, Database of Abstracts of Reviews of Effects)
- applied the inclusion and exclusion criteria.
For each study identified, the research assistant assessed the level of evidence using NHMRC levels of evidence1 and appraised the methodological quality using criteria developed a priori according to study design. Each study was allocated a risk of bias rating using a template developed by the Monash Centre for Health Research and Implementation for systematic reviews and randomised control trials2 and the Newcastle-Ottawa Scale for cohort studies.3
Search findings were summarised in a report to the working group, which identified the reference, year, country, type of study (study design and evidence level), intervention and comparator, population, study description, outcome measures, results and comments on study quality or conclusions. If one or more systematic reviews were identified for a clinical question, the research assistant flagged individual studies identified by the Australian Asthma Handbook literature searches that had been included in those systematic reviews.
The research assistant updated the searches before each working group met to formulate recommendations based on the systematic review.
Limited (non-systematic) structured searches
For each of the high-priority questions identified, the nominated researcher (medical writer or working group member) developed a search strategy and ran searches in PubMed. The following standard limits were applied: English language, Humans, publication date 1990–present, limited to systematic reviews, meta-analyses or randomised controlled trials. For aetiology clinical questions, publication type was extended to include cohort studies.
Search findings were summarised in a report to the working group, which identified the reference, year, country, type of study (study design), intervention and comparator, population, study description, outcome measures, results and comments on study quality or conclusions.
For most clinical questions, the literature search was updated once or twice during handbook development. If a high-quality systematic review (e.g. a Cochrane review) was published during guideline development, the working group considered whether it altered interpretation of the body of evidence on which a recommendation was based.
- National Health and Medical Research Council (NHMRC). NHMRC additional levels of evidence and grades for recommendations for developers of guidelines. NHMRC, Canberra, 2009. Available from: http://www.nhmrc.gov.au/guidelines/resources-guideline-developers
- Monash Centre for Health Research and Implementation (MCHRI) - Monash University and Monash Health. Evidence Synthesis Program templates for critical appraisal and risk of bias (adapted from Critical Appraisal Templates, Centre for Clinical Effectiveness, Southern Health, Melbourne, 2010). MCHRI - Monash University and Monash Health, Melbourne, 2013.
- Wells GA, Shea B, O’Connell D, et al. The Newcastle-Ottawa Scale (NOS) for assessing the quality of non-randomised studies in meta-analyses. Ottawa Health Research Institute, Ottawa, 1999. Available from: http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp