DO HTA AGENCIES STICK TO THEIR ICER THRESHOLDS? CROSS-JURISDICTIONAL EVIDENCE FROM AUSTRALIA, CANADA, AND ENGLAND (2020-2025)
Author(s)
Ahmad Hecham Alani, PharmD1, Mackenzie Mills, PhD1, Panos Kanavos, BSc, MSc, PhD2.
1HTA-Hive (Hive Health Optimum Ltd.), London, United Kingdom, 2London School of Economics and Political Science, London, United Kingdom.
1HTA-Hive (Hive Health Optimum Ltd.), London, United Kingdom, 2London School of Economics and Political Science, London, United Kingdom.
OBJECTIVES: ICER thresholds are designed to provide a clear signal of value-for-money for the adoption of new medicines. This study examines the extent to which new medicines exceed ICER thresholds and how modelling uncertainty shapes HTA decision-making.
METHODS: A retrospective analysis was conducted on cost-utility analysis (CUA)-based submissions for new medicines appraised by NICE (England), CDA-AMC (Canada), and PBAC (Australia) between 2020-2025. Submissions with a reported ICER were included. Two attributes were assessed: (i) modelling uncertainty, measured by number of economic uncertainties identified by the agency and dichotomised at the sample mean (~9 uncertainties) into lower- and higher-uncertainty submissions; and (ii) cost-effectiveness relative to each agency’s threshold range (within/partial/above). Final recommendations (positive/negative) were cross-tabulated across the uncertainty × cost-effectiveness distribution and stratified by agency. Cross-agency differences were assessed using χ2 tests or Fisher’s exact tests where appropriate (α=0.05).
RESULTS: Of 376 CUA submissions with reported ICERs, all 25 partial-threshold ICERs received positive recommendations irrespective of modelling uncertainty. Positive recommendation rates were similarly high among submissions within conventional thresholds, including those with higher modelling uncertainty (NICE 96.2% [51/53], CDA-AMC 100% [1/1], PBAC 87.5% [7/8]; Fisher's exact p=0.38). Differences emerged among submissions exceeding conventional thresholds — for lower-uncertainty submissions, positive recommendation rates were 61.5% [8/13] for NICE, 85.6% [95/111] for CDA-AMC, and 43.8% [7/16] for PBAC (Fisher's exact p<0.001). For higher-uncertainty submissions, rates declined to 37.5% [6/16], 72.2% [26/36], and 22.2% [8/36], respectively (χ²=18.6, p<0.001). Interpretation of CDA-AMC results should account for recommendations preceding price negotiations, unlike NICE or PBAC binding decisions.
CONCLUSIONS: Agencies differ not only in their assessment of cost-effectiveness, but also in the extent to which they tolerate modelling uncertainty. This is most evident in submissions where ICER exceeds traditional thresholds. Greater understanding of these differences may improve interpretation of HTA decisions and predictability across jurisdictions.
METHODS: A retrospective analysis was conducted on cost-utility analysis (CUA)-based submissions for new medicines appraised by NICE (England), CDA-AMC (Canada), and PBAC (Australia) between 2020-2025. Submissions with a reported ICER were included. Two attributes were assessed: (i) modelling uncertainty, measured by number of economic uncertainties identified by the agency and dichotomised at the sample mean (~9 uncertainties) into lower- and higher-uncertainty submissions; and (ii) cost-effectiveness relative to each agency’s threshold range (within/partial/above). Final recommendations (positive/negative) were cross-tabulated across the uncertainty × cost-effectiveness distribution and stratified by agency. Cross-agency differences were assessed using χ2 tests or Fisher’s exact tests where appropriate (α=0.05).
RESULTS: Of 376 CUA submissions with reported ICERs, all 25 partial-threshold ICERs received positive recommendations irrespective of modelling uncertainty. Positive recommendation rates were similarly high among submissions within conventional thresholds, including those with higher modelling uncertainty (NICE 96.2% [51/53], CDA-AMC 100% [1/1], PBAC 87.5% [7/8]; Fisher's exact p=0.38). Differences emerged among submissions exceeding conventional thresholds — for lower-uncertainty submissions, positive recommendation rates were 61.5% [8/13] for NICE, 85.6% [95/111] for CDA-AMC, and 43.8% [7/16] for PBAC (Fisher's exact p<0.001). For higher-uncertainty submissions, rates declined to 37.5% [6/16], 72.2% [26/36], and 22.2% [8/36], respectively (χ²=18.6, p<0.001). Interpretation of CDA-AMC results should account for recommendations preceding price negotiations, unlike NICE or PBAC binding decisions.
CONCLUSIONS: Agencies differ not only in their assessment of cost-effectiveness, but also in the extent to which they tolerate modelling uncertainty. This is most evident in submissions where ICER exceeds traditional thresholds. Greater understanding of these differences may improve interpretation of HTA decisions and predictability across jurisdictions.
Conference/Value in Health Info
2026-11, ISPOR Europe 2026, Vienna, Austria
Value in Health, Volume 29, Issue 12S
Code
HPR62
Topic
Economic Evaluation, Health Policy & Regulatory, Health Technology Assessment
Topic Subcategory
Reimbursement & Access Policy
Disease
No Additional Disease & Conditions/Specialized Treatment Areas