OBJECTIVES: Methods such as network meta-analysis (NMA) are used frequently by the National Institute for Health and Care Excellence (NICE; United Kingdom), to indirectly compare technologies’ clinical effectiveness. NICE Decision Support Unit (DSU) Technical Support Documents (TSD) encourage the use of WinBUGS to conduct NMAs. We explored indirect comparison software trends within NICE Technology Appraisal (TA) submissions.

METHODS: We reviewed a random sample of 50% of NICE TAs submitted between March 2000 to May 2022 and recorded whether an indirect comparison was included, and which software was used. We excluded terminated TAs and TAs that did not assess investigational drugs. Main data sources included the company submission and Evidence Assessment Group critique, including clarification questions.

RESULTS: We included 299 TAs; 167 reported at least one indirect comparison. The most frequently reported analyses were NMA and indirect or mixed treatment comparison (MTC; 95/167), population-adjusted indirect comparison (PAIC; 33/167) and Bucher indirect treatment comparison (ITC; 29/167), and. Of those, 110 TAs reported what software was used, including 21 TAs that used more than one software. Of the TAs submitting an NMA/MTC, 92 reported the software used (BUGS: 672; R, either as standalone or as interface to another software: 27; JAGS: 10). TAs reporting PAIC methods were more likely to use R (12/17), followed by BUGS (4/17). Convergence issues were mentioned in approximately one-third of TAs, and 30% that used BUGS developed their own code. Use of R has increased within TAs since 2015 and use of JAGS has been reported from 2016 onwards.

CONCLUSIONS: Our analyses show that NICE TA submissions often include indirect comparisons, for which BUGS is commonly used. However, the increased use of R and JAGS in recent years demonstrates the diversified use and acceptance of statistical software beyond those recommended by the NICE DSU.