Abstract
BACKGROUND: Outpatient treatment studies of anorexia nervosa (AN) are notoriously hard to conduct given the ambivalence of the patient group and high drop-out rates. It is therefore not surprising that previous meta-analyses of pharmacological and psychological treatments for outpatient treatment of adult AN have proved to be inconclusive. Network meta-analysis (NMA) has the potential to overcome the limitations of pairwise meta-analysis, as this approach can compare multiple treatments using both direct comparisons of interventions within randomized controlled trials (RCTs) and indirect comparisons across trials based on a common comparator. To date there is no published example of this approach with eating disorders and the current study provides a protocol which will use NMA to advance knowledge about what outpatient therapy works best for which patients with AN by conducting both direct and indirect comparisons of different treatments and the moderating variables. METHODS: Searches of electronic data bases will be supplemented with manual searches for published, unpublished and ongoing RCTs in international registries, and clinical trials registries of regulatory agencies and pharmaceutical companies. Two reviewers will independently extract the data and where possible we will access individual data in order to examine moderators of treatment. Two primary outcomes will be selected: changes to body mass index and changes to global eating disorder psychopathology. The secondary outcome is the total number of patients who, at 12-month post-randomization, attained over the previous 28 day period: (i) BMI > 18.5, and (ii) global eating disorder psychopathology to within 1 SD of community norms. We will also provide a statistical evaluation of consistency, the agreement between direct and indirect evidence. RESULTS: Descriptive statistics across all eligible trials will be provided along with a network diagram, where the size of the nodes will reflect the amount of evidence accumulated for each treatment. We will use a contribution matrix that describes the percentage contribution of each direct meta-analysis to the entire body of evidence. DISCUSSION: Findings will make a major contribution to the literature by summarising individual data across rapidly accumulating outpatient trials of AN using state of the art NMA methodology. TRIAL REGISTRATION: PROSPERO registration number: CRD42017064429.