OBJECTIVES:

Over time, more and more treatment options are becoming available for many conditions and diseases. This makes resource allocation decisions increasingly complex. In particular, the effectiveness and cost-effectiveness of drugs might vary depending on the sequence in which they are used. To date, there is no standardised guidance on handling treatment sequences in health economic evaluations. This systematic review aims to investigate current practice and related issues regarding modelling treatment sequences in health economic evaluations.

METHODS: NICE Technology Appraisals (TAs) published between March 2000 and November 2019 were screened. TAs without adequate documentation were excluded. Company Submissions, Evidence Review Group reports and Final Appraisal Documents were reviewed to answer the following research questions: (1) How often is treatment sequencing considered an issue in TAs? (2) Why have sequential treatment models (not) been used? (3) How have treatment sequences been handled or modelled?

RESULTS: Of 460 identified TAs, 381 were included and 86% (322/381) involved discussion around treatment sequences, with this becoming more common over time. The issue was most prominent in oncology and autoimmune diseases (94.5% and 93.8 % TAs, respectively). Whilst treatment sequencing was regularly identified as an issue in TAs, sequences were explicitly modelled much less often. Sequential models were more often adopted when previous models in the same disease area incorporated sequencing. By contrast, data availability and procedural issues may result in omission of treatment sequence models.

CONCLUSIONS: Treatment sequence related discussions were prevalent in NICE TAs. However, the number of TAs including an explicit sequential treatment model remains low. Barriers to modelling treatment sequences include data availability and procedural issues. Economic evaluations that do not model treatment sequences when they are relevant may produce misleading results and sub-optimal resource allocation. Further research is required to investigate methods for comparing the cost-effectiveness of alternative treatment sequences.