OBJECTIVES : The Cancer Drugs Fund (CDF) in England is an example of a managed access agreement (MAA) in which reimbursement of health technologies is conditional on the collection of additional data to reduce uncertainty. Entry into the CDF can be recommended by the National Institute for Health and Care Excellence (NICE) for oncology products that demonstrate ‘plausible’ cost-effectiveness. This review examined the extent to which probabilistic sensitivity analysis (PSA) results have been used to assess ‘plausible’ cost-effectiveness for oncology products entering the CDF.

METHODS : Published NICE guidance for oncology products appraised from April 2016 (launch of the new CDF process) were identified and those with a CDF recommendation were included in the review. The following PSA results were extracted for analyses corresponding to the final, decision-making incremental cost-effectiveness ratio (ICER): the probability of the intervention being the most cost-effective at the relevant thresholds, and the average probabilistic ICER.

RESULTS : Thirty-two appraisals with a CDF recommendation were identified. In 14 appraisals the decision-making ICERs were confidential, precluding the extraction of relevant PSA results. In 13/18 (72%) of remaining appraisals, PSA results corresponding to the decision-making ICER were not reported. Even in appraisals that went directly to Final Appraisal Determination (n=3), only one appraisal reported relevant PSA results. The probability of the intervention being the most cost-effective at the relevant thresholds for each appraisal ranged from 39% to 89% across the four appraisals that reported this result.

CONCLUSIONS : PSA results corresponding to the final, decision-making ICER have either been unavailable to inform the majority of past NICE recommendations for entry into the CDF or have been poorly reported in public documents. Based on the information available for this review, PSA results have typically not influenced the assessment of whether the cost-effectiveness estimates that led to the CDF recommendation were ‘plausible’.