ABSTRACT
The underlying principles and basic methodological steps needed for conducting and reporting systematic reviews of pharmacological or nonpharmacological intervention trials are similar. The basic steps for conducting a systematic review appropriately have been well established and reviewed recently [1]. Regardless of the nature of health intervention under review, the general sequence of broadly de ned steps/activities needed for the conduct of systematic reviews includes the following 12 steps:
1. Identi cation of gaps in knowledge 2. Formulation and re nement of the review question 3. Development of the review protocol by establishing the objectives,
study eligibility criteria (design of primary study, study population,
CONTENTS
13.1 Methodology of Systematic Reviews of Nonpharmacological Intervention Trials ..................................................................................... 169
13.2 Reporting Issues of Systematic Reviews of Nonpharmacological Interventions ............................................................................................... 172 13.2.1 Why Is Reporting Important? ...................................................... 172 13.2.2 Reporting Guidelines .................................................................... 173 13.2.3 Unique Issues in Reporting Nonpharmacological
Systematic Reviews ........................................................................ 177 13.3 Conclusion .................................................................................................. 180 References ............................................................................................................. 181
experimental intervention, control intervention, and outcome measures), literature search (search strategy, bibliographic sources, period of time to be covered), methods/tools to assess risk of bias of individual studies, and analytic approaches to describe and synthesize data
4. Assembling identi ed bibliographic records 5. Selection of the identi ed bibliographic reports of primary studies
(using broad and strict screen) 6. Extraction of relevant data using a pre-speci ed piloted form 7. Assessment of risk of bias of the included individual studies 8. Analysis and synthesis of data from results of primary study (quali-
tative and/or quantitative meta-analytic summary) 9. Presentation of the review results (study selection process diagram,
tabulate study and population characteristics, study results regarding outcome measures, and risk of bias for each relevant for the review outcome)
10. Interpretation of the review results (risk of bias of individual studies, the strength and applicability of evidence)
11. Dissemination of the review results (e.g., publication, presentation) 12. Updating the review (since evidence continuously accumulates over
time, there is a need to maintain the review results up-to-date)
Ideally, a systematic review of nonpharmacological intervention trials should additionally consider unique methodological aspects inherent to such trials. The methodological challenges in conducting nonpharmacological intervention trials have been widely acknowledged [2-9]. Unlike pharmacological interventions, the mechanisms of action of many nonpharmacological interventions are not clear and they are usually administered as single or discrete applications (e.g., acupuncture, massage, manipulation, surgery, diagnostic test, education/counseling) [10]. Since these interventions represent a combination of treatment-speci c (i.e., intervention as a procedure) and contextual effects (e.g., intervention delity, maturity of intervention, operator’s skill set, care provider’s age and length of experience, pre-and post-intervention care, study subject’s expectation), they are termed as “complex interventions” [8,11]. In this situation, it is often hard to tease out the speci c effects of an intervention from those of context.
