THE USE AND ACCEPTANCE OF NOVEL STATISTICAL ANALYSES TO SUPPORT TECHNOLOGY SUBMISSIONS TO HTA AUTHORITIES
Author(s)
Batson S, Mitchell SA, King D
Abacus International, Bicester, UK
Presentation Documents
OBJECTIVES Indirect comparisons are increasingly accepted to model the clinical- and cost-effectiveness of treatments. The purpose of this study was to (i) assess the literature reporting on the use of novel statistical methods [simulated treatment comparison (STC), and matching-adjusted indirect comparison (MAIC)]; and (ii) assess technology appraisals (TAs) submitted to the National Institute of Health and Carel Excellence (NICE) to determine whether these techniques have been accepted by reimbursement authorities. METHODS Embase, Medline and the Cochrane Library were interrogated to identify publications reporting on the use of MAIC or STC. NICE TAs published from 2011-2014 which reported MAIC or STC analyses were identified and the critique by the appraisal group was summarised. RESULTS Six publications reported on the use of MAIC in six indications. Results from these analyses concluded that MAIC offered several advantages over conventional meta-analysis methods that rely on aggregate data Findings from the review of NICE TAs indicated that these novel statistical techniques have not been widely used in manufacturers' submissions to date. Of the most recent 60 NICE TAs, analyses employing MAIC methodology have been presented in two submissions and a STC analysis in a single HTA, all in the oncology setting. In all cases the review group identified limitations with the statistical methodology presented, although their use as exploratory analyses supporting results from conventional meta-analyses was highlighted in one submission. In particular the use of non-randomised data from single treatment arms was highlighted as a potential weakness of STC. CONCLUSIONS In spite of the increasing published evidence base reporting on MAIC in a range of indications, both MAIC and STC have not been widely used in manufacturer's submission to NICE. Assessment bodies critiquing the technology submissions remain to be convinced of the appropriateness of these novel techniques for the robust assessment of relative efficacy.
Conference/Value in Health Info
2014-11, ISPOR Europe 2014, Amsterdam, The Netherlands
Value in Health, Vol. 17, No. 7 (November 2014)
Code
PRM190
Topic
Methodological & Statistical Research
Topic Subcategory
Confounding, Selection Bias Correction, Causal Inference
Disease
Multiple Diseases