Factors associated with post-COVID-19 asymptomatic hypoxemia

Annier J. Fajardo-Quesada, School of Medical Sciences of Bayamo, Universidad de Ciencias Médicas de Granma, Granma, Cuba
Ilian E. Tarife-Romero, Medical Degree Program, Facultad de Ciencias Médicas Mayabeque, Mayabeque, Cuba
Luis M. Abreu-Pereira, Medical Degree Program, Facultad de Ciencias Médicas Mayabeque, Mayabeque, Cuba

Introduction: Asymptomatic hypoxemia or felonious hypoxia during and after severe acute respiratory syndrome coronavirus 2 infection posed clinical and organizational challenges: patients with dangerously low SpO2 without dyspnea may delay seeking care, which is why home oximetry and the development of predictive models have been promoted. Objective: To identify factors associated with hypoxemia and asymptomatic hypoxia in patients screened after coronavirus disease 2019 (COVID-19). Material and methods: A prospective analytical study was conducted in 125 patients screened after COVID-19 in Matanzas in 2022. Variables included age, sex, and vaccination schedule, time since hospital discharge due to COVID-19, SpO2, heart rate, respiratory rate, and blood pressure. Bivariate tests (Mann–Whitney, χ2), multivariate logistic regression models (β, standard errors [SE], p, odds ratio, 95% confidence interval), and Pearson correlation analyses were performed. Model stability was assessed (pseudo R2, SE, and signs of multicollinearity). Results: No significant associations were found between hypoxemia and sociodemographic or clinical variables; time since symptom onset showed a trend (p = 0.149). The multivariate model for hypoxemia did not identify significant predictors and had a low pseudo R2 (0.1095). For festering hypoxia, the model showed a very high pseudo R2 (0.999) but coefficients with large SE. The strongest correlation found was age with systolic blood pressure (r = 0.448). Conclusions: No robust predictors of hypoxemia or asymptomatic hypoxia were identified; the models present problems of explanatory power and instability that require interpretive caution, a larger sample size, diagnosis and correction of multicollinearity, the use of penalized techniques, and external validation before applying predictive rules in clinical practice.

Keywords: Coronavirus disease 2019. Severe acute respiratory syndrome coronavirus 2. Oximetry. Asymptomatic. Hypoxia.