<p>Context: It is unclear whether the construct of a randomized controlled trial (RCT) itself could confer benefit or harm to trial participants beyond any effect of the experimental treatment under study (trial effect).</p> <p>Objective: To determine whether there is a trial effect appreciated by RCT participants (insiders) compared to similar patients who do not participate (outsiders). Although we are most interested in the pragmatic comparison of insiders to outsiders, we will also conduct the explanatory comparison of insiders to outsiders when the intervention is the same.</p> <p>Data Sources: We searched electronic health research databases, including CENTRAL (1960-2010), MEDLINE (1966-2010), EMBASE(1980-2010) and PsycINFO (1880- 2010).</p> <p>Study Selection: Eligible studies included those that reported the outcomes of insiders and a group of parallel or consecutive outsiders and reported the same health outcome at the same endpoint.</p> <p>Results: We included 147 articles out of the 42493 identified in our initial search. Five out of the 147 studies randomized patients to be insiders or outsiders, the remaining were observational designs. The heterogeneity of our overall result was reduced by grouping studies based on whether the intervention being investigated was effective and whether treatment inside and outside of the RCT was the same or different. There was no significant difference in outcomes between insiders and outsiders when the experimental intervention was ineffective (standard mean difference [95% confidence interval]: -0.03 [-0.1, 0.04]), or when it was effective and received by both insiders and outsiders (0.04 [-0.04,0.13]). If the experimental intervention was effective but was not administered to outsiders, they experienced worse health outcomes (-0.36 [-0.61, -0.12]).</p> <p>Conclusions: There is no evidence to support any benefit or harm associated with trial participation. There is some evidence that better outcomes are experienced by insiders who had access to effective treatments not offered or available to outsiders.</p> / Master of Health Sciences (MSc)
Identifer | oai:union.ndltd.org:mcmaster.ca/oai:macsphere.mcmaster.ca:11375/12435 |
Date | 10 1900 |
Creators | Fernandes, Natasha A. |
Contributors | Bryant, Dianne, Griffith, Lauren, Clinical Epidemiology/Clinical Epidemiology & Biostatistics |
Source Sets | McMaster University |
Detected Language | English |
Type | thesis |
Page generated in 0.0088 seconds