OBJECTIVES: The use of PAIC has substantially increased in reimbursement submissions. Guidance on performing PAICs recommends adjusting for EMs only, in the case of anchored PAICs, and all EMs and prognostic factors for unanchored PAICs. The accurate identification of relevant factors is therefore essential. This research investigates what methods are currently being applied to identify EMs and prognostic factors.

METHODS: A targeted review of National Institute for Health and Care Excellence (NICE) appraisals in five cancer types (lung, breast, ovarian, colorectal, and endometrial) published from January 2019 to May 2023 was performed. Appraisals that featured population adjustment were identified and methods for prognostic factors and EM identification were extracted. These were grouped into three broad categories; literature informed, use of expert opinion or conduct of trial data analysis.

RESULTS: Of the 66 appraisals reviewed, 24 (36.4%) considered a PAIC: 13 in lung, 5 in breast, 3 in ovarian, 2 in colorectal and 1 in endometrial. The most common method applied was use of expert opinion (75.0%). Literature review and data analyses were used in 50.0% and 45.8% of appraisals, respectively. Multiple methods were often used (62.5%), a quarter (25.0%) combined all three and approximately one third (37.5%) used two methods. Of these, 55.6% paired expert opinion and literature information, and 44.4% paired expert opinion and data analysis.

CONCLUSIONS: In the NICE appraisals reviewed, there was clear variation in the approaches used to identify EMs and prognostic factors. Whilst multiple methods including expert opinion were used in most appraisals, there was no single preferred set of methods for variable selection with literature information and/or data analysis frequently used in combination with expert opinion. Overall, our findings suggest that further guidance on how to identify these adjustment variables would be beneficial to improve consistency in the methods used in NICE appraisals.