FACTORS ASSOCIATED WITH THE OUTCOMES OF RAPID REVIEWS SUBMITTED TO THE NCPE
Author(s)
Varley A1, Carey N2, O'Donnell H1, Tilson L1, Barry M1
1National Centre for Pharmacoeconomics (NCPE), Dublin, Ireland, 2National Centre for Pharmacoeconomics (NCPE), Athlone, WH, Ireland
Presentation Documents
OBJECTIVES: The National Centre for Pharmacoeconomics (NCPE) conducts health technology assessments (HTAs) for all drugs seeking reimbursement in Ireland. The first step in the evaluation process is a Rapid Review (RR). Following completion of the RR, the NCPE indicates whether a full HTA is required, or if a recommendation can be made regarding reimbursement at this stage. Our objective is to examine the factors associated with RR outcomes. METHODS: All RRs commissioned between January 2015 and December 2018 were included. RR outcomes were categorised into a binary measure. Recommendations were designated ‘positive’ if the review group recommended reimbursement without further assessment or recommended price negotiation at this stage. Recommendations were designated ‘negative’ if a full HTA was recommended; if a HTA was not recommended until additional data is submitted or if reimbursement was not recommended. Multivariate logistic regression techniques were used to regress the RR outcome on factors hypothesised to be correlated, including therapeutic area, budget impact, reimbursement scheme, type of marketing authorisation, orphan status and other factors. RESULTS: A total of 213 RR submissions were analysed. Submissions for oncological indications (-27.4 percentage points [pps] p<0.01), orphan drugs (-32.6pps p<0.01) and ‘first-in-class’ therapies (-39.3pps p<0.01) were less frequently associated with a ‘positive’ RR outcome. Gross budget impact (BI) was found to be significantly associated with RR outcome, with every €1 million increase in 5-year gross drug BI reducing the probability of a ‘positive’ outcome by 0.3pps (p<0.05). CONCLUSIONS: RR submissions for oncology drugs, ‘first-in-class’ therapies, orphan conditions and those with large gross BI’s are less frequently associated with ‘positive’ recommendations. However, our analysis is subject to a number of limitations including omitted variable bias.
Conference/Value in Health Info
2019-11, ISPOR Europe 2019, Copenhagen, Denmark
Code
PNS251
Topic
Health Technology Assessment
Topic Subcategory
Decision & Deliberative Processes, Systems & Structure, Value Frameworks & Dossier Format
Disease
No Specific Disease