ASSESSING CROSS-COUNTRY AGREEMENT IN HTA RATINGS: THE CASE OF PORTUGAL AND FRANCE
Author(s)
Paula Costa, MSc, BE1, Ana Margarida Advinha, Professor, PhD2, Sofia Oliveira Martins, Professor, PhD3.
1PhD student, Faculdade de Farmácia da Universidade de Lisboa, Lisbon, Portugal, 2School of Health and Human Development, University of Evora, Évora, Portugal, 3Faculty of Pharmacy, University of Lisbon, Lisbon, Portugal.
1PhD student, Faculdade de Farmácia da Universidade de Lisboa, Lisbon, Portugal, 2School of Health and Human Development, University of Evora, Évora, Portugal, 3Faculty of Pharmacy, University of Lisbon, Lisbon, Portugal.
OBJECTIVES: To assess cross-country agreement, direction, and magnitude of discrepancies between Portuguese and French Added Therapeutic Value (ATV) ratings in drug reimbursement HTA decisions.
METHODS: A retrospective analysis included 75 medicines assessed by INFARMED and HAS for the same indication. The cohort comprised European Commission-authorized novel medicines (2014-2024) with FDA Breakthrough and/or EMA Accelerated Assessment status, used as proxies for high innovation or unmet need. Portuguese ATV ratings were harmonized into ordinal categories using a pre-specified mapping aligned with France’s ASMR (Amélioration du Service Médical Rendu) clinical benefit gradient. Agreement was assessed using percentage agreement, Cohen’s kappa (unweighted), and weighted kappa (linear and quadratic), and visualized via a heatmap. Discrepancies were further characterized by mean absolute differences and directional asymmetry. Subgroup analyses compared oncology (ATC L01) versus non-L01 medicines.
RESULTS: Exact agreement was 30.7%. Unweighted kappa indicated slight agreement (k=0.026; 95% CI: -0.10 to 0.16), while ordinal-weighted values were higher (linear k=0.14; quadratic K=0.248). Most discrepancies were minor, with 74.7% of medicines differing by no more than one category (mean absolute difference: 0.99). Portugal assigned more favourable ratings in 41.3% of cases versus 28.0% for France. Oncology medicines showed higher agreement than non-L01 products (35.0% vs. 25.7%), with directional asymmetry driven by more favourable Portuguese ratings.
CONCLUSIONS: Despite limited exact agreement, discrepancies were predominantly minor, suggesting broadly aligned clinical benefit assessments but differing grading intensity. Directional asymmetry and subgroup variation indicate non-random divergence. These findings suggest that Joint Clinical Assessment may not fully eliminate national heterogeneity in HTA value appraisal.
METHODS: A retrospective analysis included 75 medicines assessed by INFARMED and HAS for the same indication. The cohort comprised European Commission-authorized novel medicines (2014-2024) with FDA Breakthrough and/or EMA Accelerated Assessment status, used as proxies for high innovation or unmet need. Portuguese ATV ratings were harmonized into ordinal categories using a pre-specified mapping aligned with France’s ASMR (Amélioration du Service Médical Rendu) clinical benefit gradient. Agreement was assessed using percentage agreement, Cohen’s kappa (unweighted), and weighted kappa (linear and quadratic), and visualized via a heatmap. Discrepancies were further characterized by mean absolute differences and directional asymmetry. Subgroup analyses compared oncology (ATC L01) versus non-L01 medicines.
RESULTS: Exact agreement was 30.7%. Unweighted kappa indicated slight agreement (k=0.026; 95% CI: -0.10 to 0.16), while ordinal-weighted values were higher (linear k=0.14; quadratic K=0.248). Most discrepancies were minor, with 74.7% of medicines differing by no more than one category (mean absolute difference: 0.99). Portugal assigned more favourable ratings in 41.3% of cases versus 28.0% for France. Oncology medicines showed higher agreement than non-L01 products (35.0% vs. 25.7%), with directional asymmetry driven by more favourable Portuguese ratings.
CONCLUSIONS: Despite limited exact agreement, discrepancies were predominantly minor, suggesting broadly aligned clinical benefit assessments but differing grading intensity. Directional asymmetry and subgroup variation indicate non-random divergence. These findings suggest that Joint Clinical Assessment may not fully eliminate national heterogeneity in HTA value appraisal.
Conference/Value in Health Info
2026-11, ISPOR Europe 2026, Vienna, Austria
Value in Health, Volume 29, Issue 12S
Code
PT40
Topic
Health Policy & Regulatory, Health Technology Assessment
Topic Subcategory
Value Frameworks & Dossier Format