P. Dequen

Presentational approaches used in the UK for reporting evidence synthesis using indirect and mixed treatment comparisons

Objectives To establish current guidance and practice in UK on presentation of indirect comparison and mixed treatment comparison analyses; to provide recommendations to improve indirect comparison/mixed treatment comparison reporting and to identify research priorities for improved presentation. Methods Existing institutional guidance for conducting indirect comparison/mixed treatment comparison alongside current practice in health technology assessment was reviewed. Reports published in UK by the Health Technology Assessment programme since 1997, which utilized indirect comparison/mixed treatment comparison methods, were reviewed with respect to the presentation of study data, statistical models and results. Recommendations for presentation were developed. Results Guidance exists that provide the details necessary to conduct a successful indirect comparison/mixed treatment comparison analysis but recommendations on presentation are limited. Of 205 health technology assessment reports that contained evidence synthesis for effectiveness, 19 used indirect comparison/mixed treatment comparison methods. These reports utilized numerous presentational formats from which the following key components were identified: network table/diagram for presenting data; model description to allow reproducibility; and tables, forest plots, matrix tables and summary forest plots for presenting a range of results. Recommendations were developed to ensure that reporting is explicit, transparent and reproducible. Approaches most understandable by non-technical decision makers, and areas where future research is required, are outlined. Conclusions There is no standard presentational style used in UK for reporting indirect comparison/mixed treatment comparison, and the use of graphical tools is limited. Standardization of reporting and innovation in graphical representation of indirect comparison/mixed treatment comparison results is required.

USE OF IMPLICIT AND EXPLICIT BAYESIAN METHODS IN HEALTH TECHNOLOGY ASSESSMENT

OBJECTIVES The aim of this study was to examine the use of implicit and explicit Bayesian methods in health technology assessments and to identify whether this has changed over time. METHODS A review of all health technology assessment (HTA) reports of secondary research published by the UK National Institute of Health Research (NIHR) between 1997 and 2011. Data were extracted on the use and implementation of Bayesian methods, whether defined as such by the original authors (i.e., explicit) or not (i.e., implicit). RESULTS A total of 155 of 375 (41 percent) NIHR HTA reports, identified as relevant to this review, contained a Bayesian analysis. Of these, 128 (83 percent) contained an implicit Bayesian analysis, 3 (2 percent) an explicit Bayesian analysis and 24 (15 percent) both implicit and explicit Bayesian analyses. Of the twenty-seven reports that explicitly used Bayes theorem, only six included prior information in the form of (informative) prior distributions. Over time, the percentage of HTA reports that used Bayesian (implicit and/or explicit) methods increased from 0 percent in 1997 to nearly 80 percent in 2011. CONCLUSIONS This review has shown that there has been an increase in the use of Bayesian methods in HTA, which is likely to be a result of the increase in freely available resources to implement the approach. Areas where Bayesian methods have the potential to advance healthcare evaluations in the future are considered in the discussion.

Searching for Indirect Evidence and Extending the Network of Studies for Network Meta-Analysis: Case Study in Venous Thromboembolic Events Prevention Following Elective Total Knee Replacement Surgery

OBJECTIVE To evaluate the effect of study identification methods and network size on the relative effectiveness and cost-effectiveness of recommended pharmacological venous thromboembolic events (VTEs) prophylaxis for adult patients undergoing elective total knee replacement surgery in the United Kingdom. METHODS A stepwise literature search specifically designed to identify indirect evidence was conducted to extend the original clinical review from the latest National Institute for Health and Care Excellence (NICE) VTE technology appraisal. Different network sizes or network orders, based on the successive searches, informed three network meta-analyses (NMAs), which were compared with a replicated base case. The resulting comparative estimates were inputted in an economic model to investigate the effect of network size on cost-effectiveness probabilities. RESULTS Searches increased the number of indirect comparisons between VTE interventions, progressively widening the relevant network of studies for NMA. Precision around mean relative treatment effects was increased as the network was extended from the base case to first-order NMA, but further extensions had limited effect. Cost-effectiveness analysis results were largely insensitive to variation in clinical inputs from the different NMA orders. CONCLUSIONS No standard methodology is currently recommended by NICE to identify the most relevant network of studies for NMA. Our study showed that optimizing the identification of studies for NMA can extend the evidence base for analysis and reduce the uncertainty in relative effectiveness estimates. Although in our example network extensions did not affect the acceptability of available treatments in VTE prevention based on cost-effectiveness results, it may in other applications.

Network Meta-Analysis of Biological Response Modifiers in Rheumatoid arthritis Including real World Evidence at Multiple time Points

free survival (PFS) and/or overall survival (OS) as reported outcomes and to create a RCT network accordingly. Fixed and random effects FP models of first/second order were applied on these data and tested for goodness of fit using deviance information criteria. Finally, the best fitting model was used to estimate the hazard function, median PFS, median OS and uncertainty of treatment effect. Results: Literature review found 8 RCTs and 5 RCTs which reported PFS and OS respectively, for 7 different mRCC treatments (interferon-alfa (IFN), bevacizumab+IFN, temisrolimus+bevacizumab, sunitinib, pazopanib, cediranib, placebo). The best fitting FP model was second order random effect model for both, PFS and OS NMA. Hazard functions varied significantly. Estimated median PFS was the longest with sunitinib (10.8 months; 95% credible interval (CI): 9.5-11.8), followed by pazopanib and temsirolimus+bevacizumab. Similarly, sunitinib was estimated with the longest median OS (28.8 months; 95% CI: 25.7-31.0) followed by pazopanib and bevacizumab+IFN. ConClusions: Synthesis of NMA evidence for 1LmRCC treatments identified sunitinib to be the treatment with favourable PFS and OS. When dealing with multiple sources, hazards proportionality assumption is violated, and proposed method should be the method of choice.