Drug development is a risky business. Manufacturers are faced with the dilemma of whether or not to invest at any stage in the development process. Even once marketing authorization has been attained, payers are becoming increasingly demanding of evidence to justify price premiums in the face of increasing budgetary pressures. Cost-effectiveness is a critical decision-making criterion for many payers, and restrictions to sub-populations is common. Early economic modelling at very early phases of the development pathway can inform optimal investment decision-making, including go/no-go decisions and clinical trial design, particularly in population selection. To test the hypothesis of changing payer requirements, we carried out a study on the trends in reimbursement submissions where payers approved but ultimately restricted the population compared to the marketing license or the company's target population.

A systematic literature review of all single technology appraisals (STAs) by the National Institute for Health and Clinical Excellence (NICE) was carried (01/01/2006- to 16/11/2018). We used a linear regression model to examine the relationship between frequency of optimizations and time.

In total, 357 STAs outcomes were identified, 55 percent were recommended and 26 percent were optimized. The proportion of optimized recommendations increased over time vs all other outcomes (p = 0.01), with more technologies being optimized over time (p < 0.01).

The results indicate an increasing trend by NICE towards maximization of value through approval of drugs in select groups of patients. From a manufacturer's perspective, prediction of such outcomes at an early stage is fundamental for investment purposes and to maximize financial returns. An early stage model provides a framework to examine these issues as well as identifying data gaps, where real world evidence can be planned to support the value argument for products, and to inform clinical trial design through value of information analysis.