Download files
Abstract
Modification of the host response to traumatic injury of the central nervous system by induced cooling of core temperature may have beneficial effects through reducing cerebral metabolic rate and tissue oedema, and modifying the inflammatory response. In the last century research has primarily focused on evaluating the effect of reducing temperature into the hypothermic range (below 36°C) either immediately following injury or after the development of raised intracranial pressure, with an assumption that avoiding pyrexia is effective standard care. This thesis uses a mixed methods approach to evaluate the feasibility of addressing a fundamental gap in knowledge: does avoiding pyrexia reduce disability and death after traumatic brain injury? The main research findings are:
1. In descriptive studies of temperature patterns and management in Australia and New Zealand:
a. Pyrexia, defined as temperature > 37°C, is common.
b. Clinicians commonly use paracetamol and physical cooling interventions in a step-wise strategy to reduce temperature modestly.
c. Induced hypothermia and non-steroidal anti-inflammatory drugs are rarely used.
2. In a large international retrospective cohort study that investigated the association between peak temperature in the 1st 24 hours after intensive care admission and risk-adjusted hospital outcome:
a. There was no change in the risk of death between 37°C and 39°C.
b. There was an increase in the risk of death above 39°C.
3. In a phase 2b clinical trial investigating the efficacy of paracetamol in reducing core temperature:
a. Regular intravenous paracetamol is at best a modest anti-pyretic compared to placebo (effect size < 0.5°C)
and is ineffective at maintaining temperature below 37°C.
4. In a systematic review of modest cooling (temperature target of 36-37.5°C), clinical trials evaluating strict control of pyrexia for traumatic brain injury are needed.
In the context of the thesis research findings, a protocol for a funded feasibility study evaluating pyrexia control compared to standard care, with a mechanistic sub study, is presented in the concluding chapter.