The Impact of Hospital Costing Methods on Cost-Effectiveness Analysis: A Case Study

BACKGROUND: Several methods exist to cost hospital contacts when estimating the cost effectiveness of a new intervention. However, the implications of choosing a particular approach remain unclear. We compare the use of the three main diagnosis-related group (DRG)-based national unit costs in England to determine whether choice of approach can impact on economic evaluation results. METHODS: A cost-utility model was developed to compare secondary fracture prevention models of care for hip fracture patients, using data from large primary and hospital care administrative datasets in England. A healthcare and personal social services payer perspective was adopted, and utilities were informed by a meta-regression. Hospital resource use was valued using three DRG-based unit costs, and regression-based costing models were developed using data from 13,906 patients to inform the model health states. RESULTS: Finished consultant episode (FCE)-level reference costs resulted in the highest costs on admission (£9075) and in the year of the fracture (£14,440). Relative to FCE-level costs, spell-level tariffs led to the lowest total hospital care costs per patient within 1 year of fracture (− £3691) compared with spell-level reference costs (− £2106). At a £20,000/quality-adjusted life-year threshold, using spell-level reference costs or spell-level tariffs, the introduction of a nurse-led fracture liaison service model of care was the cost-effective alternative. However, using FCE-level reference costs, usual care was the cost-effective option. CONCLUSIONS: Our results show that, conditional on the set of national unit costs adopted, the costs of hip fracture may vary considerably and different decisions may be reached regarding the introduction of new healthcare interventions. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (10.1007/s40273-018-0673-y) contains supplementary material, which is available to authorized users.