OBJECTIVES: The Medical Technologies Advisory Committee (MTAC) makes recommendations to the National Institute for Health and Care Excellence (NICE) on medical devices after conducting an evaluation of clinical and cost effectiveness evidence. Companies are required to submit relevant evidence including a cost model which demonstrates cost-saving compared with current care. This research reviews the cost models submitted for evaluation and was undertaken to inform the development of a model template [available for company submissions].

METHODS: Twenty-two models were analysed and categorised by type. Data were extracted by 1 reviewer, with a second reviewer checking a sample. Information was extracted for 17 categories. Information not available from the model was sourced from other documents considered by MTAC. Themes were then analysed based on type.

RESULTS: Models were built using either Excel^® (n=20, 91%) or TreeAge^® (n=2, 9%). Ten models (46%) were developed by a health economics consultancy. The developer was the company (n=2, 9%) or unknown (n=10, 46%) for the remainder. Cost-minimisation analyses were most common, 95% (n=21), with 5% (n=1) being a cost-utility analysis. Twelve models were structured as a decision tree (48%), 9 (36%) were cost calculators and 4 (16%) were Markov models. Complexity of the model structure adopted varied substantially. Thirteen models had a time horizon of 1 year or greater (51%) of which 2 (9%) inappropriately omitted discounting. Data from the clinical evidence submission were commonly used to populate the model but adverse event data were often excluded. Most models included sensitivity analysis (97%), commonly as univariate deterministic sensitivity analysis (n=18, 51%). Only 4 (18%) models included probabilistic sensitivity analysis.

CONCLUSIONS: This review increased knowledge of the nature of models submitted for evaluation by MTAC. This information informed the development of a model template.