RESUMO
Multiple CT investigations in critical ill neurosurgical patients are useful for monitoring the course of the illness and for the early detection of complications. CT's however, are expensive and require transportation of the patient, which is often inconvenient and, in some cases, dangerous. The decision to perform CT scanning should be based on the quantitative knowledge of potential benefits and harms (as well as costs) of the procedure. In a prospective trial, in which 59 such decisions were considered, we found it to be absolutely necessary to order a CT-investigation whenever neurological deterioration occurs. Even in patients not showing changes of neurological symptoms, about 30% of CT findings gave reason for therapeutic intervention. Thus, in critical ill neurosurgical patients, especially in those under sedative medication and artificial ventilation, neurological findings alone are insufficient as sole criteria for the decision to order a CT scan. To optimize this decision more sensitive indicators of deterioration are needed.