Is diagnostic severity grading for head injuries possible?

The classification of head injured patients is more difficult than that for most other disease processes. The quantum of data to be embedded into each patient's grading code depends on the purpose to which that grading is used. If the information is merely required for broad epidemiological surveys, it may be confined to a double rubric which represents the most significant diagnostic component and an arbitrary index of associated severity. For this purpose, diagnostic severity grading is possible provided the task is delegated to experienced members of the neurosurgical team. If the grading is to be used in attempts to compare one patient group with another or for predictions of complications or outcome, a more detailed data-set is required. This may be accomplished with the use of multiple ICD diagnostic codes but assignations of severity to each diagnostic component requires very subjective judgement. Such an approach is unlikely to be successful and the only alternative is to define a data-set of "pure" information which includes all the relevant clinical, radiological, and operative findings without resorting to artificial data compression by using potentially misinterpretable deduced codes.