Pmu94 Data Generated by the French Compassionate Use Program: How Can They Contribute to the Health Technology Assessment?

Abstract

We investigated how utility values (UVs) were selected by manufacturers and the evidence review groups (ERGs) and what were discussed on UVs during the appraisal process in NICE. Methods: We reviewed NICE Technology Appraisal (TA) guidance and committee papers available on the NICE website, including the types of instrument of UVs; source of utility data and utilization (i.e., mapping) in the submission by manufacturers and the ERGs; the final decision by NICE. Our review included individual TAs for pharmaceuticals published between April 2018 and March 2019. Results: After exclusion of 13 TAs (3 multiple TAs; 2 cost-comparison TA; 1 non-pharmaceutical TA; 7 terminated TAs) from 56 TAs (TA517-TA572), 43 TAs which contain 150 unique health state UVs were identified. Sixty-four UVs (43%) and 13 UVs (9%) were obtained by using EQ-5D (3L) and EQ-5D-5L in the manufacturers' submissions, respectively. Seventy-three UVs (49%) were mainly taken from pivotal clinical trials, followed by 62 (41%) from published studies including previous NICE TAs. Mapping was performed for 33 UVs (22%) in the manufacturers' submissions. The ERGs agreed to 74 UVs (49%) selected by manufacturers, while 67 UVs (45%) were criticized and 9 UVs (6%) were not clear from TAs and committee papers. Of the 67 UVs criticized by the ERGs, 47 UVs (70%) were finally replaced to the values selected by the ERGs, while 13 UVs (19%) remained in the final appraisals. Conclusions: EQ-5D, which was recommended in the NICE 2013 reference case, was a major measure of the health-related quality of life in the manufactures' submissions, but there were substantial differences in the selection of UVs between manufactures and the ERGs during the appraisal process. Nearly half of the UVs used in NICE TAs were criticized by the ERGs, and only around half of the original UVs submitted by manufacturers were adopted in the final guidance.