Evaluation of Clinical Predictors for Major Outcomes in Patients Hospitalized With COVID-19: The Potential Role of the Rothman Index

ABSTRACT

Introduction The Rothman Index (RI, PeraHealth, Inc. Charlotte, NC, USA) is a predictive model intended to provide continuous monitoring of a patient's clinical status. There is limited data to support its use in the risk stratification of patients infected with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). We hypothesized that low admission RI scores would correlate with higher rates of adverse outcomes in patients hospitalized for coronavirus disease 2019 (COVID-19). Methods Medical records of adult patients admitted to a single 1,200-bed tertiary academic center were retrospectively reviewed for demographic data, baseline characteristics, RI scores, admission to intensive care unit (ICU), need for mechanical ventilation, and inpatient mortality. Statistical analyses were performed using STATA statistical software, version 17 (Stata Corp LLC, College Station, TX, USA). Continuous variables were analyzed using the Mann-Whitney test, and categorical variables were analyzed using Fisher’s exact test. Both univariate and multivariate analyses were performed. A p-value <0.05 was considered statistically significant. Results Median admission RI score for the entire cohort was 63.0 (IQR 45.0 - 77.1). The cohort was divided by admission RI into a low-risk group (RI ≥70;n=70) and a high-risk group (RI <70;n=107). Compared to patients with low-risk RI, patients with high-risk RI had higher mortality (95.2%, 95% CI 85.8 - 105 vs 4.8%, 95% CI -5 - 14.2, p < 0.01), were more likely to require ICU admission (90.2%, 95% CI 81.9 - 98.5 vs 9.8%, 95% CI 1.5 - 18.1, p < 0.01) and mechanical ventilation (89.7%, 95% CI 78.3 - 101 vs 10.3%, 95% CI -1 - 21.7, p < 0.01), and had a longer median hospital length of stay (12 days, 95% CI 9 - 14 vs 5 days, 95% CI 4 - 7, p < 0.01). Conclusions High-risk RI was associated with increased admission to the ICU, mechanical ventilation, and mortality. These results suggest that it may be used as a tool to aid provider judgment in the setting of COVID-19.