OBJECTIVES : To conduct a systematic literature review (SLR) of published model structures assessing the cost-effectiveness of treatments for chronic rhinosinusitis with nasal polyps (CRSwNP).

METHODS : EMBASE, MEDLINE, and Cochrane Database of Systematic Reviews were searched through April 2021. Relevant conference proceedings and HTA guidance was identified from January 2015 to April 2021. The SLR followed PRISMA guidelines.

RESULTS : 8 unique cost-effectiveness models were identified, and all had surgery as a comparator. Model structures included a Markov model (1), a decision-analytic model (1), or a combination of both (6). The 1 Markov model compared 2 surgical approaches: endoscopic sinus surgery (ESS) and endoscopic polypectomy in clinic (EPIC). Health states included post-operative adverse events, symptom relief, and death. The 1 decision-analytic model compared polypectomy to intranasal budesonide. Of the 6 mixed-model structures, 1 compared ESS to EPIC. In the post-surgery space, there were 3 health states: healthy with no revision, symptomatic requiring revision surgery, and death. 4 models, all from the same author group, started with a decision tree for initial treatment and then used a Markov process to assess progression through post-treatment health states. 1 compared ESS to dupilumab and the rest compared ESS to other medical treatment. The final model compared ESS to exhalation delivery system with fluticasone. It was described as a Markov decision tree model but did not define health states within the available publication. All models were used to estimate incremental cost per QALY. Other than the 2 models comparing ESS to EPIC, no models utilized direct or indirect comparative clinical effectiveness data.

CONCLUSIONS : All cost-effectiveness models assessing treatments for patients with CRSwNP included surgery as a comparator. Common Markov health states included recovery after surgery, disease recurrence, and revision surgery. Overall, use of comparative efficacy data and robust methodology were lacking in these models.

FUNDING : GSK [214518]