The decision of whether or not to admit a patient with COVID-19 to the hospital and the extent to which they should be monitored while outside of the hospital require guidelines for primary care physicians. Therefore, researchers aimed to construct and evaluate easy-to-use risk ratings that do not rely on laboratory tests or imaging to predict hospitalization for COVID-19 outpatients. In a polymerase chain reaction study, patients 12 years and older who tested positive for SARS-CoV-2 were tallied. Patients presenting before March 2021 were analyzed using logistic regression to generate a risk score, which was then verified using a cohort between March and September 2021 and an Omicron cohort between December 2021 and January 2022. Hospitalization rates were 4.0% among early 5,843 derivation cohort members (those seen before 3/1/21), 4.2% among late validation cohort members (those seen after 3/1/21), and 1.2% among Omicron cohort members. A person’s age, the severity of dyspnea, and the presence of any comorbidities were the foundation of their risk score. Temperature, respiration rate, and/or oxygen saturation are some additional scores that can be provided. All were highly accurate overall (AUC 0.85–0.87), with around half of the patients being placed into the low-risk group with a hospitalization probability of less than 1%. Hospitalization rates were 0.22%, 1.3%, and 8.7% in the Omicron cohort. About 2 alternative risk scores that were created from the outside identified more low-risk patients but carried a higher total probability of hospitalization. Over half of COVID-19 outpatients can be classified as having a very low risk, with a 0.22% hospitalization risk in the Omicron cohort, using a simple risk assessment appropriate for outpatient and telemedicine settings.
Source: jabfm.org/content/early/2022/09/15/jabfm.2022.AP.220056
