Purpose:
To document the incidence of neurogenic fever in the population with traumatic brain injury and identify surrounding characteristics.
Significance:
If predicting characteristics for neurogenic fever after brain injury can be identified, an earlier opportunity for the application of fever-prevention protocols may improve outcomes and decrease costs.
Design:
Fever in patients with traumatic brain injury can result in increased length of stays and mortality rates and decreased functional outcomes (Thompson et al, 2003; Mourad et al, 2003; Marik, 2000). In up to 37% of theses patients, a cause for fever cannot be found (Thompson et al, 2003; Meythaler and Stinson, 1994) and may relate to the degree of brain injury (Marik, 2000). The goal of the Society of Critical Care Medicine's practice parameters for evaluating new fever in critically ill adult patients is to determine if infection is present so additional testing can be avoided and therapeutic options can be made (O'Grady et al, 1998). These guidelines also state that in patients with certain neurologic pathology, temperature elevation is so predictable that diagnostic workups for infection should only take place if the patient is showing signs and symptoms of such an infection (O'Grady et al, 1998).
Methods:
The study population consisted of patients with traumatic brain injury admitted during a 6-month period and meeting specific inclusion/exclusion criteria. A retrospective chart review was done to identify subjects exhibiting a temperature 100.0[degrees]F or greater. Variables of interest were collected and analyzed using Statistical Product for Service Solution software, and a backward logistic regression was performed.
Results:
Of the 48 patients, 8 lacked documentation to support a complete diagnostic work up, 24 had infectious etiologies which were resolved with antiobiotic therapy and no longer exhibited hyperthermia. Three subjects exhibited infection states which did not respond to antiobiotic therapy and remained hyperthermic. Thirteen remained febrile with unknown etiology. A backward logistic regression was completed where the only significant predictor of neurogenic fever noted was temperature (p = 0.011). This group of subjects exhibited the highest means for temperature (103.7[degrees]F) and length of stay (25.8 days), and the lowest mean for Glasgow Coma Scale (5.2).
Conclusions:
Incidence of actual neurogenic fever could not be determined due to the lack of complete diagnostic workups for infectious or noninfectious etiologies. It was assumed that fevers no longer explained by other etiologies could be classified as neurogenic. Temperature was the only significant predictor of neurogenic fever.
Implications for Practice:
Neurogenic fever does not usually respond to standard treatment. By identifying incidence and common characteristics, earlier interventions may decrease the possibility of further neuronal damage to brain cells in already at risk traumatic brain injury patients. Potential to impact length of stay and costs could be quantified in further studies.