OBJECTIVES : To understand how prevalent fractional polynomial (FP) methodology is in National Institute for Health and Care Excellence (NICE) single technology appraisals (STAs) and how their use is received by the Evidence Review Group (ERG) and the appraisal committee.

METHODS : The NICE website was searched for the keywords “fractional polynomial”. All documents were filtered to remove those not relating to STAs and then grouped by their unique STA number. Each STA was then screened and information on the purpose of using FP methodology and how it was perceived by the ERG and appraisal committee were extracted for each STA. All data extraction was performed by two independent reviewers and all decisions were reached by consensus with an additional third reviewer.

RESULTS : A total of 19 NICE STAs were identified that used FP methodology. FPs were used to calculate either or both overall survival and progression-free survival in 17 STAs with the remainder of the STAs using the FP method to calculate post-recurrence survival. Where available, 12 of 16 STAs explored both first-order and second-order FPs with 3 STAs exploring second-order FPs only and 1 first-order only. The use of FP methodology in NICE STAs were generally very well received by the ERGs/committees when the proportional hazards assumption was not met. In some cases where the manufacturer did not use FP methods in the original evidence submission, it was suggested by the ERG. Criticisms of the application of the FP methods typically occurred when the approach had not been applied appropriately or when not enough exploration into various powers of the FPs was conducted.

CONCLUSIONS : The use of fractional polynomials in NICE STA submissions where the proportional hazards assumption is violated are well received by ERGs/committees. The use of FPs requires care in implementation and appropriate exploration into various parametrisations.