Increased Mortality Risk Among Critically Ill Non-COVID-19 Patients During the COVID-19 Pandemic in Ontario, Canada

The COVID-19 pandemic significantly disrupted healthcare systems globally, impacting access to care and potentially worsening patient outcomes. While the impact on COVID-19 patients is well-studied, the consequences for critically ill individuals without COVID-19 remain unclear due to conflicting data from smaller studies. In Ontario, Canada, the pandemic led to decreased emergency department visits and reallocation of ICU resources towards COVID-19 patients. Concerns arose regarding potential adverse outcomes for non-COVID-19 patients due to delayed presentations or altered care pathways. Furthermore, specific vulnerable populations, such as immigrants and those with chronic conditions like COPD, may have experienced disproportionately negative impacts on their healthcare access and outcomes. This study aimed to address these concerns by comparing the characteristics and outcomes of adult ICU patients without COVID-19 during the initial phase of the pandemic in Ontario to a comparable pre-pandemic period. The primary outcome of interest was all-cause in-hospital mortality, with secondary outcomes including ICU and hospital lengths of stay, discharge disposition, and use of invasive procedures.

Prior research on the impact of pandemics on non-COVID-19 patients yielded mixed results. Some small cohort studies reported no significant difference in mortality for critically ill patients without COVID-19 infection during the pandemic, possibly due to limitations in sample size or patient population representativeness. Other studies, including systematic reviews, suggested adverse outcomes for non-COVID-19 patients across various conditions but did not specifically focus on ICU populations. The limited existing evidence underscored the need for a large population-based study to evaluate the pandemic's effect on ICU patient outcomes in North America.

This population-based cohort study utilized linked health administrative databases from Ontario, Canada, encompassing over 14 million individuals. Data sources included the OHIP Claims Database, Discharge Abstract Database (DAD), Registered Persons Database, National Ambulatory Care Reporting System, Continuing Care Reporting System, Statistics Canada Census data, Ontario Drug Benefit Claims database, and the Ontario Mental Health Reporting System. Data on immigrant status was obtained from the Immigration, Refugees and Citizenship Canada (IRCC) Permanent Resident Database. The study included all adult patients (≥18 years old) admitted to an ICU in Ontario without a COVID-19 diagnosis between March 1 and June 30, 2020 (pandemic cohort) and compared them to a similar cohort from March 1 to June 30, 2019 (non-pandemic cohort). Patient characteristics, comorbidities (using data from up to two years prior), and outcomes such as in-hospital mortality, ICU/hospital length of stay, discharge disposition, and use of invasive procedures were analyzed. Statistical analyses included descriptive statistics, chi-square tests, one-way ANOVA, Kruskal-Wallis test, and logistic regression to account for potential confounders. The study obtained necessary data access approvals but did not require research ethics board approval due to its use of population-level data under section 45 of the Personal Health Information Protection Act of Ontario.

The study identified 32,486 patients in the pandemic cohort and 41,128 in the non-pandemic cohort. While the mean age and gender distribution were similar, the pandemic cohort showed a 21% reduction in ICU admissions compared to the non-pandemic cohort. There was a 33% relative decrease in the proportion of patients admitted from long-term care facilities during the pandemic. All-cause in-hospital mortality was significantly higher in the pandemic cohort (13.5%) compared to the non-pandemic cohort (12.5%), representing a 7.9% relative increase. This difference persisted after adjusting for various demographic and clinical factors in a multivariate logistic regression analysis (adjusted odds ratio 1.10; 95% CI, 1.05–1.56). Subgroup analysis revealed significantly higher mortality during the pandemic for patients admitted with COPD exacerbation (28.9% relative increase) and recent immigrant/refugee status (14% relative increase). ICU and hospital lengths of stay, as well as the overall proportion of patients receiving most invasive procedures, were largely similar between the two cohorts. However, there was a notable 50% relative decrease in the use of ECMO in the pandemic cohort, which may warrant further investigation.

The study's findings demonstrate a statistically significant increase in all-cause in-hospital mortality among critically ill non-COVID-19 patients admitted to ICUs in Ontario during the initial phase of the COVID-19 pandemic. This increase was not explained by differences in disease severity at admission, as indicated by similar Charlson Comorbidity Index and HOMR scores between the cohorts. The increased mortality likely stems from various institutional changes resulting from the pandemic response, including resource allocation challenges, infection control measures affecting usual care practices, staffing shortages, and potential delays in timely interventions. The observed disparities in mortality among specific subgroups, particularly immigrants and patients with COPD exacerbations, underscore the pandemic's disproportionate impact on vulnerable populations. The study's results align with findings from other studies indicating increased mortality in non-COVID-19 patients during the pandemic but contrasts with some smaller studies showing no significant difference, potentially due to differences in study design and sample size.

This large population-based study provides strong evidence of increased in-hospital mortality among critically ill patients without COVID-19 during the pandemic in Ontario. The findings highlight the indirect consequences of the pandemic on healthcare delivery and patient outcomes, underscoring the need to develop pandemic response strategies that effectively manage both COVID-19 and non-COVID-19 patient care needs while ensuring equitable access to high-quality care. Future research should focus on evaluating specific mechanisms contributing to increased mortality, investigating the impact of resource allocation policies, and exploring strategies to mitigate the disproportionate effects on vulnerable populations during pandemics.

As an observational study, this research is susceptible to inherent limitations, including potential selection bias and unmeasured confounders that could influence the observed association between the pandemic and mortality. The use of administrative data may limit the granularity of information available, potentially missing important clinical variables influencing outcomes. While the study controlled for several factors, it's possible that other unmeasured confounders may have affected the results. Furthermore, the study primarily focuses on in-hospital outcomes and doesn't account for out-of-hospital events that might contribute to increased overall mortality.