Use of steroids to reduce high fever in head injury—is it acceptable?

Abstract

Sir: I thank Dr Llompart–Pou and coauthors for their valuable and accurate comments on the use of steroids in head-injured patients. In principle I share their view. Steroids should not be used as a general therapy for head injured patients due to their potential side effects, as supported by the large clinical MRC CRASH study to which Llompart–Pou et al. refer. As they also point out, steroid treatment will induce deviation from normal physiology and therefore is against the very essence of the Lund therapy. This does not mean, however, that the use of steroids is prohibited under all circumstances. I will explain why steroid treatment was still accepted under specific conditions in the Lund therapy [1]. Mainly due to fewer lung complications, the number of patients with high fever has been reduced with the Lund therapy [1], but patients with extravascular blood, extensive posttraumatic cerebral inflammation or direct hypothalamic damage may still suffer from high fever. As the effect of paracetamol is small, active cooling and steroids are the only known ways to effectively reduce a high temperature. This gives us three choices: (1) accept the high fever, (2) use active cooling or (3) treat fever with a bolus dose of steroids. All three choices involve side effects, and we are on the horns of a dilemma, as we do not know the extent of these effects or their importance for the outcome. The current literature strongly supports the view that high fever is deleterious and lowering of a high fever will improve outcome for the head-injured patient [2]. Accepting this view, we have to choose between active cooling and steroids. By creating a difference between the set thermostat temperature and body temperature, active cooling reduces temperature in an unphysiological way, activating the endocrine, autonomic and motor systems. Patients exposed to such adaptations may be extremely stressed even though they are sedated. This situation must be evaluated against a condition where the temperature has been lowered in a more physiological way, by resetting the thermostat temperature level towards more normal values. This is the reason why a deleterious high fever is reduced by steroid therapy in the Lund concept instead of active cooling. After all, the adverse effects of one bolus dose of a steroid [1] must be small compared with the longer-term steroid treatment used in clinical studies [3], and most likely also smaller than those induced by long-term active cooling. The temperature-reducing effect of one bolus dose is marked and often persistent. The patient becomes more stable haemodynamically, and a raised ICP is often reduced. The induced hyperglycaemia can be treated by insulin and a transient adrenal suppression may be compensated by a small dose of hydrocortisone. The proportion of patients suffering from high fever (> 38.5 °C) when treated strictly according to the Lund Concept is relatively small [1], and therefore the number of candidates for treatment with the recommended single bolus dose of Solu-Medrol will be limited. References