Generalisability of the Rapid Screen of Concussion: A Dual-Centre Approach

Abstract

This study investigated the generalisability of the Rapid Screen of Concussion (RSC; a series of brief computerised tests) and the Digit Symbol Substitution Test (DSS; Wechsler, 1981) for discriminating between concussion (mTBI) and orthopaedic patient groups in two different centres, the Royal Brisbane Hospital, Brisbane (N = 177) and St. Vincent's Hospital, Sydney (N = 211). Group differences and interactions were assessed between test centre (Sydney or Brisbane) and injury type (mTBI or orthopaedic control) on a composite variable of the RSC and DSS subtests. Age, education and number of standard drinks of alcohol consumed were covaried for these analyses. While the mTBI patients obtained lower scores than orthopaedic patients, there were no significant differences between Brisbane and Sydney participants in performance on the RSC and DSS. There was no significant interaction between injury type and test centre. Similarities of correct classification rates of the RSC and DSS for Brisbane and Sydney were examined using a double cross-validation technique. When cross-validating from the Brisbane to Sydney sample it was revealed that sensitivity was 73% in Brisbane and 71% in Sydney, while specificity for the two samples was 76% and 67%, respectively. When cross-validating from the Sydney sample to the Brisbane sample sensitivity was 64% and 63%, respectively. Specificity was 81% in Sydney and 80% in Brisbane. These results indicated that correct classification rates were sufficiently similar between the cities and that the RSC and DSS could be generalised. When samples were combined, sensitivity was found to be 66% while specificity was 80%. The RSC and DSS are recommended as adjunct procedures to medical diagnoses of mTBI.