The Revised Trauma Score (RTS) [Champion, Sacco, Copes et al, 1989], widely used to control for injury severity case mix in injury research, cannot be calculated for intubated patients [Rutledge, Lentz, Fakhry, 1996]. This leads to missing RTS values for the most severely injured. When analyzing injury research data with heterogeneous baseline risks, the presence of missing physiological data leads to either the exclusion of all observations with missing data or the elimination of physiological variables from data analyses. The first option leads to the loss of precious data observations and therefore statistical power and most likely results in bias results since data are not missing at random [Moore, Lavoie, Le Sage et al, 2005]. The second option may result in residual confounding if groups differ with respect to physiological injury severity. To our knowledge, no study has has yet proposed a valid solution to the problem of missing physiological data for injury research case mix control [Gabbe, Cameron, Finch, 2003]. The present study aimed to investigate the validity of Multiple Imputation (MI) for imputing missing RTS in Motor Vehicle Crash (MVC) patients. (Author/publisher)
Samenvatting