Article type
Year
Abstract
Objectives: To present methods for including observational studies in a network meta-analysis (NMA).
Description: Pairwise and network meta-analyses are often limited to synthesizing evidence from randomized controlled trials (RCTs). Observational evidence from non-randomized studies (NRSs) is often disregarded because researchers assume estimates of relative treatment effects obtained from NRSs are more likely to be biased. Although RCTs are considered the most reliable source of information on relative treatment effects, their strictly experimental setting and inclusion criteria may limit their ability to predict results in real-world clinical practice. Estimates of treatment effects from NRSs may complement evidence provided by RCTs, and potentially address some of their limitations. In this workshop we will discuss situations where NRSs can facilitate the decision-making process in the context of NMA. We will present statistical and graphical methods for assessing the agreement between the various types of evidence and we will introduce a range of alternative approaches that researchers can use to incorporate non-randomized evidence in a NMA of RCTs. We will illustrate all methods using a published network of 167 RCTs that compare 15 antipsychotics and placebo for schizophrenia, augmented by observational data on five interventions coming from a large cohort study.
Description: Pairwise and network meta-analyses are often limited to synthesizing evidence from randomized controlled trials (RCTs). Observational evidence from non-randomized studies (NRSs) is often disregarded because researchers assume estimates of relative treatment effects obtained from NRSs are more likely to be biased. Although RCTs are considered the most reliable source of information on relative treatment effects, their strictly experimental setting and inclusion criteria may limit their ability to predict results in real-world clinical practice. Estimates of treatment effects from NRSs may complement evidence provided by RCTs, and potentially address some of their limitations. In this workshop we will discuss situations where NRSs can facilitate the decision-making process in the context of NMA. We will present statistical and graphical methods for assessing the agreement between the various types of evidence and we will introduce a range of alternative approaches that researchers can use to incorporate non-randomized evidence in a NMA of RCTs. We will illustrate all methods using a published network of 167 RCTs that compare 15 antipsychotics and placebo for schizophrenia, augmented by observational data on five interventions coming from a large cohort study.