Traumatic brain injury occurs as a result of direct impact to the brain, such as after a road traffic accident or fall from a height. It is a major cause of death and disability, experienced by about 5.5 million people worldwide every year. Damage occurs in two stages: firstly, at the time of impact; and secondly, in the hours and weeks following injury. Management of a traumatic brain injury aims to reduce the impact of this secondary damage. There is some evidence to suggest that people with a normal body temperature after injury may have a better outcome than those with a higher temperature.
This review assessed whether medicines or physical cooling treatments that do not use medicine, given to reduce body temperature to between 35 °C and 37.5 °C, affect outcomes in adults or children who are in hospital after a traumatic brain injury.
We searched for randomised controlled trials (RCTs) up to 28 November 2019. RCTs assign participants to treatment group by chance, and provide the most reliable type of evidence.
We found only one small study with 41 people who had recently been admitted to the intensive care unit after a traumatic head injury. This study assessed the effects of paracetamol, given intravenously (through a needle or tube inserted into a vein) for 72 hours, compared to an intravenous salt solution that was disguised to look like the paracetamol solution.
We found no full reports of studies that gave people with traumatic brain injury other medicines or physical cooling treatments, such as blankets, fans, or ice to cool the surface of the skin, or cooled intravenous fluids. We found a short report about one very small study of a physical cooling treatment. The report did not give enough information for us to include its data, but we may include it in future if more information becomes available.
We also found four ongoing studies, which may provide more evidence for the review once they have completed. Some studies include participants who have other types of brain injury. We will only be able to include these studies in the review if they report the findings for people with traumatic brain injury separately from people with other conditions.
We were not sure whether paracetamol affected the number of people who died within 28 days of injury. This study did not report information on the outcomes of interest to us, which included: whether people had a poor outcome (defined as death or dependency); additional serious bleeding inside the brain; bleeding of the head outside the brain; increased pressure within the skull (intracranial pressure); pneumonia or other serious infections.
Certainty of the evidence
Although the study that provided the data generally appeared to use good methods, we judged it to have a high risk of bias because the study authors did not record in advance that they planned to report on the number of people who died, even though they had shared their study plans before they started the study. Having only one study, with very few participants, means that we cannot be certain whether the results would be the same in larger studies with more people.
We judged the evidence for death to be very low certainty. This means we are not very confident about the key results in this review. The true effect of treatments may be very different from the key results reported here.
We are uncertain of the effects of medicines and other physical cooling treatments to reduce body temperature to 35 °C to 37.5 °C, when given to people in hospital after a traumatic head injury. More studies are needed to assess this question. In future updates, we will assess the side effects of treatments, such as nausea and vomiting, or discomfort, and we hope that ongoing and future studies will also look at these important outcomes.
One small study contributed very low-certainty evidence for mortality to this review. The uncertainty is largely driven by limited research into reduction of body temperature to 35 °C to 37.5 °C in people with TBI. Further research that evaluates pharmacological or physical interventions, or both, may increase certainty in this field. We propose that future updates of the review, and ongoing and future research in this field, incorporate outcomes that are important to the people receiving the interventions, including side effects of any pharmacological agent (e.g. nausea or vomiting), and discomfort caused by physical therapies.
Traumatic brain injury (TBI) is a major cause of death and disability, with an estimated 5.5 million people experiencing severe TBI worldwide every year. Observational clinical studies of people with TBI suggest an association between raised body temperature and unfavourable outcome, although this relationship is inconsistent. Additionally, preclinical models suggest that reducing temperature to 35 °C to 37.5 °C improves biochemical and histopathological outcomes compared to reducing temperature to a lower threshold of 33 °C to 35 °C. It is unknown whether reducing body temperature to 35 °C to 37.5 °C in people admitted to hospital with TBI is beneficial, has no effect, or causes harm. This is an update of a review last published in 2014.
To assess the effects of pharmacological interventions or physical interventions given with the intention of reducing body temperature to 35 °C to 37.5 °C in adults and children admitted to hospital after TBI.
We searched the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, Embase, Web of Science, and PubMed on 28 November 2019. We searched clinical trials registers, grey literature and references lists of reviews, and we carried out forward citation searches of included studies.
We included randomised controlled trials (RCTs) with participants of any age admitted to hospital following TBI. We included interventions that aimed to reduce body temperature to 35 °C to 37.5 °C: these included pharmacological interventions (such as paracetamol, or non-steroidal anti-inflammatory drugs), or physical interventions (such as surface cooling devices, bedside fans, or cooled intravenous fluids). Eligible comparators were placebo or usual care.
Two review authors independently assessed studies for inclusion, extracted data, and assessed risks of bias. We assessed the certainty of the evidence with GRADE.
We included one RCT with 41 participants. This study recruited adult participants admitted to two intensive care units in Australia, and evaluated a pharmacological intervention. Researchers gave participants 1 g paracetamol or a placebo intravenously at four-hourly intervals for 72 hours.
We could not be certain whether intravenous paracetamol influenced mortality at 28 days (risk ratio 2.86, 95% confidence interval 0.32 to 25.24). We judged the evidence for this outcome to be very low certainty, meaning we have very little confidence in this effect estimate, and the true result may be substantially different to this effect. We downgraded the certainty for imprecision (because the evidence was from a single study with very few participants), and study limitations (because we noted a high risk of selective reporting bias). This study was otherwise at low risk of bias.
The included study did not report the primary outcome for this review, which was the number of people with a poor outcome at the end of follow-up (defined as death or dependency, as measured on a scale such as the Glasgow Outcome Score), or any of our secondary outcomes, which included the number of people with further intracranial haemorrhage, extracranial haemorrhage, abnormal intracranial pressure, or pneumonia or other serious infections.
The only other completed trial that we found was of a physical intervention that compared advanced fever control (using a surface cooling device) versus conventional fever control in 12 participants. The trial was published as an abstract only, with insufficient details to allow inclusion, so we have added this to the 'studies awaiting classification' section, pending further information from the study authors or publication of the full study report.
We identified four ongoing studies that will contribute evidence to future updates of the review if they measure relevant outcomes and, in studies with a mixed population, report data separately for participants with TBI.