• Suzanne C Freeman, Neil W Scott, Rachael Powell, Marie Johnston, Alex J Sutton, and Nicola J Cooper.
• NIHR Complex Reviews Support Unit, Biostatistics Research Group, Department of Health Sciences, University of Leicester, Leicester, UK; Medical Statistics Team, Division of Applied Health Sciences, University of Aberdeen, Aberdeen, UK. Electronic address: suzanne.freeman@leicester.ac.uk.
• J Clin Epidemiol. 2018 Jun 1; 98: 105-116.

ObjectivesTo apply component network meta-analysis (CNMA) models to an existing Cochrane review of psychological preparation interventions for adults undergoing surgery and to extend the models to account for covariates to identify the most effective components for improving postoperative outcomes.Study Design And SettingInterventions consisted of between one and four components of psychological preparation: procedural information (P), sensory information (S), behavioral instruction (B), cognitive interventions (C), relaxation (R), and emotion-focused techniques (E). We used CNMA models to assess the effect of each component for three outcomes: length of stay, pain, and negative affect.ResultsWe found evidence that the most effective component for reducing length of stay depends on the type of surgery and that R may improve pain. There was insufficient evidence that individual components contributed to the overall reduction in negative affect, but P and S emerged as the most likely beneficial components. Overall, we were unable to identify any one component as the most effective across all three outcomes.ConclusionThe CNMA method allowed us to address questions about the effects of specific components that could not be answered using standard Cochrane methodology.Copyright © 2018 The Authors. Published by Elsevier Inc. All rights reserved.

28 keywords...

hide…