RSV research tool could help predict hospitalization in infants
Click Here to Manage Email Alerts
Researchers are developing a tool that can predict severe respiratory syncytial virus in infants.
“Unfortunately, I don’t believe there is a well-validated, uniform accepted research tool for measuring disease severity in infants with primary RSV infection,” Mary T. Caserta, MD, professor of pediatrics at the University of Rochester Medical Center, told Healio.
Caserta and colleagues originally developed the Global Respiratory Severity Score (GRSS) as a research tool to identify factors associated with severe RSV disease in healthy infants in a separate study.
“We designed the GRSS as a research tool in order to measure the severity of RSV disease over the entire course of the illness,” Caserta said. “We used hospitalization as an outcome as it is a good measure of disease severity, but our primary goal was not focused on clinical care.”
In their most recent analysis, the researchers wanted to validate the GRSS in a new cohort of 184 infants — 98 hospitalized and 86 nonhospitalized — and determine whether it correlated with clinical outcomes like hospitalization and length of stay (LOS). The GRSS was calculated by applying the original training data formula to the new data. The authors compared the mean GRSS between hospitalized and nonhospitalized infants using the Welch two sample t-test and correlated it with hospitalization and LOS.
The hospitalized infants had a significantly (t = 9.334) higher GRSS (4.20 + 2.10) than nonhospitalized infants (1.76 + 1.41). Using a GRSS of 3.5 or less as the classification criterion, the authors correctly predicted the hospitalization status of 71.2% of infants (n = 131).
The authors noted that they are currently retraining the GRSS model, using both datasets, and they believe the new tool (iGRSS) will lead to an improved predictive power of hospitalization.
“The validation of the GRSS has led to the creation of the improved GRSS, or iGRSS,” Caserta said. “The new score contains less clinical variables and performs well in differentiating infants who are hospitalized from those who remain outpatients. The rate for accurately predicting hospitalization with the iGRSS is 85% with an [area under the curve] of 0.92. We believe these validation data are excellent and support the utility of the iGRSS in measuring disease severity.”
Collapse
Caserta MT, et al. Abstract 133. Presented at: IDWeek; Oct. 21-25, 2020 (virtual meeting).
Read more about
Click Here to Manage Email Alerts