Independent and Combined Effects of Age and COVID on Patient Outcomes

PURPOSE: To evaluate the independent effect and interactive effect of age and positive Covid-19 status on patient survival and number of days spent on ECMO. METHODS: Single center data was gathered for patient's currently receiving ECMO treatment. The main effect and interaction effect of patients’ age and COVID status was evaluated to investigate the impact of these factors on patient outcomes of discharged deceased and days on ECMO. A logistic regression model of 204 patients was used to evaluate the outcome of discharged deceased, and a Poisson regression model of 129 patients was used to evaluate the outcome of patient days spent on ECMO. RESULTS: In the logistic regression on discharged deceased adjusting for COVID status (n=204), age was associated with higher odds of death (OR=1.05 per year older, CI95 1.03-1.07). COVID was strongly associated with mortality (OR=4.81, CI95 2.46-9.43). Incorporating an interaction between age and COVID status, being discharged deceased was significantly associated with older age but COVID status and the interaction between COVID and age were not significant predictors of mortality. In the Poisson regression on days on ECMO (n=129), main effects of both age and COVID were noted. In the test of the interaction of age by COVID status, the interaction was also significant as were both main effects. CONCLUSION: Both age and positive COVID status were found to be independent risk factors for increased patient mortality, however only age was associated with increased patient mortality when interaction between age and COVID was incorporated. Conversely, older patients with and without COVID exhibited decreased days on ECMO, whereas COVID was associated with a significant increase in days on ECMO.