OBJECTIVES

This research compares the methodological quality and transferability assessments of six checklists using ten economic evaluations of genetic-guided pharmacotherapy for venous thromboembolism and considers implications for systematic reviews and decision making.

METHODS

From a systematic search, we found ten eligible model-based economic evaluations. We applied six checklists (Philips, CHEC-Extended, NICE, EURONHEED, Caro, Welte) to each study to assess methodological quality, and transferability to the UK. The items on each checklist were categorized by three domains (model structure, data identification & analysis, internal & external consistency) and considered whether response options required reporting and/or subjective assessment. Each checklist was applied independently by two reviewers; final ratings were agreed via consensus and discussion with a third reviewer, with all decisions and reasoning documented. The Spearman Rank test compared ordering of study quality between checklists.

RESULTS

For quality, Philips had the widest coverage (≥1 item in most domain-response option pairs, whereas EURONHEED the narrowest (24/26 items reporting structure and data only). Caro gave the widest score range (8-63% “Yes”), followed by Philips (14-57%), NICE (18-55%), CHEC-Extended (55-85%) and EURONHEED (30-57%). However, all checklists identified similar studies with the highest or the lowest scores. Rank correlations were highest between Philips and CHEC-Extended (0.90,p<0.001), and lowest between EURONHEED-NICE (0.45,p>0.05).

For transferability, only EURONHEED had items focusing on reporting only (10/16 items). NICE and Caro identified the same one study as transferable to the UK, which also scored the highest on EURONHEED, whereas Welte did not consider any study transferable.

CONCLUSIONS

Content (domains, response options) of a checklist may affect the conclusions on methodological quality and transferability, which could affect decision on the range of interventions for reimbursement or clinical implementation. Choice of checklists should be justified given decision context and resources available, including a consideration on the use of two checklists for sensitivity analysis.