RELATED CONTENT
SUBMIT A REQUEST FOR PROPOSAL
STATISTICS AND PSYCHOMETRICS
HTA SERVICES:
MIXED TREATMENT COMPARISONS
ECONOMIC EVALUATIONS
UTILITY ASSESSMENT
MARKET ACCESS SERVICES:
VALUE DEMONSTRATION
     EMAIL /     PRINT /     SHARE

Both for national reimbursement and local market access (i.e. local payers, budget holders and formulary committees) it is important to provide evidence that demonstrates the added value of medical technology. Systematic reviews are considered the standard practice to inform evidence-based decision making regarding the efficacy and safety of a medical technology. As part of a systematic review, quantitative results of several similar studies can be combined by means of meta-analysis to summarize the available evidence into a pooled estimate of the outcome of interest.

Mapi Values has performed many systematic literature reviews and meta-analyses for market access purposes.

Advances in Meta Analysis: Techniques for Including Multiple Study Designs, Multiple Endpoints, and Multiple Treatments

Usually, only results from randomized controlled trials (RCTs) with a parallel design are combined in meta-analyses. Cross-over trials are often ignored, although these designs may contribute evidence in a fifth of systematic reviews. Furthermore, single arm trials, while not providing information on relative effects, can provide valuable evidence on absolute effects. Especially for rare diseases, RCTs often account for the minority of information, and it can be of interest to draw together both randomized and single arm studies. In order to consider the totality of the evidence base, results from studies with different designs can be analyzed simultaneously by explicitly acknowledging the differences in designs.

Despite an interest in multiple endpoints, meta-analyses are often performed by endpoint. Analyzing multiple outcomes simultaneously has several advantages - for example by documenting relationships between endpoints which may be of clinical or scientific interest.

Traditional meta-analysis can also be extended by including multiple different pair-wise comparisons across a range of different interventions. These so called network meta-analyses or Mixed Treatment Comparisons allow for indirect comparisons in the absence of head-to-head data and may also strengthen the evidence base.

Traditionally, meta-analysis is performed with a frequentist approach, resulting in a pooled estimate along with a confidence interval. A drawback of this approach is the non-intuitive interpretation from a decision-making perspective, especially when there is no statistically significant finding. By contrast, a Bayesian approach delivers a joint posterior distribution of the parameters of interest with a natural interpretation that can be directly fed into a decision framework to support clinical or reimbursement decision making.