Hypoxia as a predictor of mortality among patients admitted with COVID-19 disease in three referral hospitals in Kenya, October 2020 to December 2021

IntroductionUsing peripheral oxygen saturation (SpO2) measurement is a cost-effective and reliable approach to assess the need for oxygen supplementation in patients admitted with coronavirus disease 2019 (COVID-19). Patients who have a SpO2 level of <95% are considered hypoxic and, per COVID-19 management guidelines, should receive oxygen supplementation. We sought to determine whether hypoxia at admission predicted adverse COVID-19 outcomes including mortality among patients admitted with COVID-19 disease in Kenya. MethodsThe study was a cross-sectional retrospective medical chart review of patients hospitalized with COVID-19 between October 1, 2020, and December 31, 2021 in three purposively selected health facilities in Kilifi, Nairobi and Kisumu. We explored the differences in proportions of categorical variables using the {chi}2 test and assessed predictors (selected a priori) of mortality among patients with hypoxia using Cox proportional hazards models. Using the Kaplan-Meier method, we also computed survival probabilities by hypoxia status for patients on room air or oxygen supplementation and produced survival graphs. ResultsOf the 1,124 COVID-19 patients, 94.8% had documented SpO2 measurements at admission, and 81.4% were found to have hypoxia, with 39.9% of hypoxic patients not exhibiting dyspnea. Hypoxic patients compared to those with normal oxygen saturation levels were significantly older (60+ years: 44.6 vs. 24.4%) and had a higher prevalence of dyspnea (60.1 vs. 36.9%), higher pulse rate (38.2 vs. 24.6%), and hypertension (40.4 vs. 25.8%), p<0.001. Oxygen supplementation was provided to only 68.6% of hypoxic patients. Mortality was notably higher in hypoxic patients versus those with normal SpO2 (38.0% vs. 13.6%, p<0.001), with hypoxia being a key predictor of death. Hypoxic, older patients ([&ge;]60 years), and those with dyspnea had a higher risk of death (adjusted hazard ratio: 1.9 [95% confidence interval (CI):1.2-2.8], 1.8 [95% CI 1.3-2.6] and 1.5 [95% CI 1.2-2.0], respectively). Regardless of dyspnea or oxygen supplementation, survival probabilities were worse for hypoxic patients (p<0.001). ConclusionsHypoxia was prevalent among hospitalized COVID-19 patients, even without respiratory distress symptoms. These findings underscore the importance early identification and management of hypoxia in COVID-19 patients, thereby guiding clinical care and improving outcomes, particularly for older or sicker patients. Author summaryHypoxia (low oxygen levels) is a common complication in severe COVID-19 and can be a key predictor of poor outcomes, including death. In low-resource settings like Kenya, the availability of advanced diagnostic tools is limited, and simple methods like peripheral oxygen saturation (SpO2) measurements are crucial for early identification and management of hypoxia. Despite the importance of oxygen supplementation for hypoxic patients, data on how hypoxia at admission correlates with COVID-19 outcomes in sub-Saharan Africa is limited. We conducted a cross-sectional retrospective study of 1,124 COVID-19 patients hospitalized between October 2020 and December 2021 in three Kenyan counties. We found that 81.4% of patients had hypoxia at admission, but 39.9% of these patients did not exhibit signs of respiratory distress (dyspnea). Hypoxia at admission was associated with higher mortality, especially among older patients ([&ge;]60 years) and those with comorbidities, such as hypertension. Survival probabilities for hypoxic patients were significantly lower, regardless of whether they showed signs of dyspnea or received oxygen supplementation. Hypoxia is highly prevalent among hospitalized COVID-19 patients, even in the absence of respiratory distress, highlighting the need for routine oxygen level screening using simple, cost-effective tools like oxymeters. Early detection and management of hypoxia, especially in older or high-risk patients, is essential to improving clinical outcomes in COVID-19. These findings provide evidence for refining COVID-19 management protocols in resource-limited settings, with an emphasis on early intervention to reduce mortality.