OBJECTIVES: Patients in the intensive care unit (ICU) with hypoxaemic respiratory failure receive supplemental oxygen. In the Handling Oxygenation Targets in the ICU (HOT-ICU) trial such patients were allocated to a lower target of partial pressure of arterial oxygen (PaO₂) of 8 kPa versus a higher PaO₂ target of 12 kPa. The objectives of this study were to evaluate health care resource use, associated costs, labour market participation, and quality adjusted life years (QALYs) at one year after randomisation.

METHODS: A cost-utility analysis based on Danish registry data was performed with a time horizon of one year after randomisation in the HOT-ICU trial. The registry data provided information on health service use and associated costs. QALYs were calculated based on the EuroQol five-dimension five-level (EQ-5D-5L) questionnaire. The registries and QALYs combined give an insight into the long-term (one-year) economic aspects of different oxygenation targets during ICU stay.

RESULTS: In total, 2,282 Danish patients were included in the trial (1,143 allocated to the lower and 1,139 to the higher oxygenation target group). We obtained access to data from somatic hospitals for a subsample of 186 patients, making this study a pilot analysis of the one-year outcomes associated with HOT-ICU. The highest cost category was somatic hospital stays. We found no differences in one-year costs or QALY gains but found a tendency towards a decreased number of bed days in somatic hospitals and an increased number of people in employment with the lower oxygenation target when compared to the higher oxygenation target. However, these effect sizes were small.

CONCLUSIONS: In these preliminary analyses, targeting a lower vs. a higher oxygenation in ICU patients with acute hypoxaemic respiratory failure did not result in QALY differences or one-year cost differences. The results will likely be less uncertain when assessing outcomes in all included Danish patients.