Comparative effectiveness of CO@h—mortality.

Introduction
The pandemic saw widespread use of home pulse oximeters to patients diagnosed with COVID-19 to support early detection of low oxygen saturation levels and appropriate care. Rapid implementation made conventional evaluation challenging, highlighting the need for rigorous non-randomised methods to support decision-making about future use of these technologies. We used routine data to explore the benefits of pulse oximetry in Greater Manchester, under the ‘COVID-19 oximetry at home’ (CO@h) programme.
Methods
We used data from the Greater Manchester Secure Data Environment and defined study parameters using a ‘target trial’ model to compare patients receiving pulse oximetry under the CO@h programme, with matched controls using various comparator groups. Primary outcomes were unplanned hospitalisation and all-cause mortality. This study is based on data from the Greater Manchester Care Record (GMCR), using anonymised, routinely collected data provided in a de-identified format for research. Informed written consent is needed for primary care patient data to be collected for service improvement and research, before data extraction to the GMCR. The study was approved under protocol GMCR RQ-048, on 12/05/2022. As indicated by the University of Manchester ethics decision tool, formal ethical approval was not required for this study.
Results
The adjusted odds ratios for an unplanned hospitalisation were higher among patients receiving pulse oximetry: OR 1.86 (95% CI 1.54–2.25) at 28 days, 1.5 (95% CI 1.3–1.74) at 90 days and 1.63 (95% CI 1.44–1.83) at 1 year. Overall odds of mortality were lower among patients receiving pulse oximetry: adjusted ORs of 0.5 (95% CI 0.25–0.98) at 28 days, 0.5 (95% CI 0.32–0.78) at 90 days and 0.58 (95% CI 0.44–0.76) at 1 year. The results were robust to different comparison groups.
Conclusion
Use of pulse oximetry at home under the CO@h programme, through the resulting prioritisation for appropriate care, was associated with a higher frequency of unplanned admissions and a reduction in the risk of mortality up to 1 year later. Therefore, it is likely effective for early detection of clinical deterioration and timely intervention among patients with COVID-19. Further research is needed to understand whether this is a cost-effective use of healthcare resources.