An Analysis of NICE’s Optimised Decisions from 2015 to 2024

One component of this is NICE’s Technology Appraisal (TA) programme, wherein NICE assess the cost-effectiveness and clinical effectiveness of medical technologies and makes recommendations accordingly.

NICE states that 84% of its recommendations are positive; however, NICE categorises several decision outcomes as ‘positive,’ including recommended decisions, optimised decisions, and decisions allowing a medicine to be used in the Cancer Drug Fund (CDF). Further analysis of this statement reveals that 36% of NICE’s positive decisions since 2000 have been optimised, excluding CDF decisions.

Optimised decisions can have the effect of limiting patient access to medicines, as they often only recommend use of the medicine by a smaller population than for which the medicine is indicated in its marketing authorisation. However, the extent to which optimised decisions restrict patient access is unclear. Thus, the aim of this report is to better understand the impact of NICE’s optimised decisions on patient access to medicines from January 2023 to December 2024.

We relied on O’Neill and Devlin’s (2010) method for quantifying this impact, which involves calculating an M-score for each optimised decision:

M = (p/P) * 100%

where M is a measure of patient access, with 100% indicating full access and 0% indicating no access at all. P represents the population eligible for treatment as indicated in the medicine’s marketing authorisation, and p represents the subset of this population for which NICE recommends treatment.

Between January 2023 and December 2024, NICE made 135 positive decisions, of which 64 (47%) were labelled as optimised, excluding CDF decisions. This result is consistent with the long-term trend of optimised decisions increasing as a proportion of positive decisions. Of the 64 optimised decisions made by NICE, 56 were carried forward for our analysis, and of those, 30 (47%) included sufficient information to allow for the estimation of M scores.

Among those 30 TAs, the mean M score was 31%, meaning that, on average, each optimised decision by NICE provides access to 31% of potential candidates for treatment, which compares to average M scores of 45% and 39% for the periods of 2020-2022 and 2015-2019, respectively. Moreover, we found that 77% of TAs recommended treatment for under half of the eligible patient population, with 43% having M scores below 25%. Between 2020 and 2022, these figures were 56% and 25%, and between 2015 and 2019, they were 65% and 35%. Thus, these results represent a decrease in patient access during the period of study. Figure 1 presents a visual representation of these changes.

M-scores below 50% and 25% from previous OHE analyses

The average level of patient access was particularly low for the following disease areas: infections (average M score 8%); neurological conditions (average M score of 15%); and diabetes and other endocrinal, nutritional and metabolic conditions (average M score of 17%). Medicines for digestive tract conditions, skin conditions, and eye conditions, in contrast, had average M scores over 50%.

Overall, these findings highlight the fact that optimised decisions are increasing in frequency and becoming more restrictive. A detailed examination of the drivers underlying this trend is beyond the scope of this analysis. An initial assessment of whether NICE report if optimisation is at the request of the company submitting evidence, suggests that the factors underlying this trend are complex.

This consulting report, An Analysis of NICE’s Optimised Decisions from 2015 to 2024, was commissioned and funded by The Association of the British Pharmaceutical Industry (ABPI).