Citation

Caserta MT, Qiu X, Tesini B, Wang J, Murphy A, Corbett A, Falsey AR, Holden-Wiltse J, Walsh EE. Development of a Global Respiratory Severity Score (GRSS) for Respiratory Syncytial Virus Infection in Infants. J Infect Dis. 2017; 215 (5): 750-756. doi: https://doi.org/10.1093/infdis/jiw624

Variables used in this score

Please see table below for the list of 10 variables used in the calculation, along with their data type, \(\beta\) values and mean values used in the calculation (See other sections below for details).

Mean imputation of Missing Values

For subjects with missing values, we use mean imputation to substitute NA values. Here “mean” refers to the mean values obtained from our study and these values are recorded in column 5 of the above table.

Adjusted Respiratory Rate

The maximum respiratory rate needs to be adjusted by the following formula \(Adjusted Resp. Rate = Resp. Rate - (0.8 * Age)\)

Calculating the GRSS

\(GRSS = \sum_{j=1}^9 \beta_jX_j + c, c=6.438577.\)

Remarks

• Visit Age, SaO2, Respiratory Rate and 6 out of 7 symptoms are required and cannot be missing.
• Warning messages will be produced when there is more than one missing (NA) value. This is because mean imputation is a method that only works when missing data is minimal.