OBJECTIVES

The National Centre for Pharmacoeconomics (NCPE) assesses the clinical effectiveness, cost-effectiveness and potential BI of drugs seeking reimbursement in Ireland. Cost-effectiveness is one component of the reimbursement decision process post-HTA, which may also include price negotiations, patient access schemes and the consideration of unmet clinical need in keeping with Irish Health (Pricing and Supply of Medical Goods) Act 2013. The importance of a reliable BI prediction is arguably greater for reimbursement decisions for drugs with an ICER above the cost-effectiveness threshold. This study aimed to determine the reliability of BI predictions for such drugs.

METHODS

To be eligible for inclusion, drugs were submitted to the NCPE for assessment between 2012-17, have a BI prediction available, have an ICER above the cost-effectiveness threshold and have expenditure data available for one-year post-reimbursement. The drug’s real-world expenditures at their agreed price were extracted from national reimbursement data. The expected expenditure derived from BI prediction was compared to the real-world expenditure for the first year after reimbursement.

RESULTS

42 drugs met the inclusion criteria. When compared to the real-world drug expenditure, the BI was underestimated for 71% of drugs. The mean BI underestimate per drug was 58.8% of the real-world expenditure. Four drugs had a manifold overestimate of BI, skewing the results for overestimates (mean 1027%, median 65.7%). BI was underestimated for 92.3% of drugs for cancer compared with 58.3% of drugs for orphan conditions and 64.7% in all other categories. 33% of drugs had an estimated BI within 40% of the real-world expenditure.

CONCLUSIONS

BI predictions for this group of drugs were frequently underestimated and most predictions were unreliable. This unreliability may be due to subsequent price negotiations, market conditions or unpredictable drug utilisation. However, it presents a significant challenge for decision makers.