Sample size
The frequency rate of meningitis among our traumatic pneumocephalus patients was 21.5% (p 1). What we expect to be an intended (or at least acceptable) effect of the prophylactic antibiotics is 5 %(p 2) frequency of meningitis in the patients with antibiotics. To detect this difference with a sensitivity of 80% and an error probability of 5%, at least 62 patients per randomization group will be required using the following formula:
n = 7.84 * [p 1(1 - p 1) + p 2 (1 - p 2)]/[p
1 - p
2]2
Considering only comparing the placebo group with each one of the antibiotics group, our sample size should be at least 186 cases. To account for the possibility of loss to follow-up, our estimated sample size is 200 cases.
Exclusion criteria
Those patients who have received antibiotic therapy for other reasons within the prior 2 weeks, are on corticosteroids or allergic to the specified medications, individuals with penetrating traumatic brain injury, open skull fractures or operated for any causes, those who are discharged from hospital with personal consent, all cases with life threatening lesions including severe brain, abdominal or vascular injuries and death due to other causes will be excluded from this study.
Study design
The patients should be observed in the hospital, until occurrence of meningitis or at least 7 days after trauma. In the case of concomitant Cerebrospinal Fluid (CSF) leakage and need for close observation in the hospital, they will be hospitalized until recovery from leakage. After discharge the patients will be followed up until one month after trauma.
The following data is registered: age, sex, time interval between trauma and admission, time interval between admission and antibiotic therapy, the cause of trauma, GCS upon admission, intracranial air volume (at the time of admission and three days later), intracranial air location, presence of CSF rhinorrhea, CSF otorrhea, presence and location of the radiological signs of skull base fractures, presence and volume of the intracranial hemorrhage(at the time of admission and three days later), presence of meningitis, CSF findings, treatment and complications in the case of meningitis and one month follow-up note.
Intracranial air and hemorrhage volume are calculated in the CT scan using the formula
ABC/2, where
A is the greatest mass diameter by CT,
B is the diameter 90° to
A, and
C is the approximate number of CT slices with mass multiplied by the slice thickness. In the case of air in the subarachnoid space, this formula is not easily applicable and an approximate volume is calculated and determined whether the volume is less than 10 cc or not [
7].
The patients are divided into three groups; with intravenous antibiotics (IV), oral antibiotics (O) and placebo (P), according to the randomization list. In IV group, Ceftriaxone 2 grams BID plus oral placebo and in O group, Azithromycin 500 mg in the first day followed by 250 mg daily plus intravenous placebo for the rest will be continued for 7 days. Antibiotics should be started in less than 24 hours after trauma. Since antibiotics may mask the clinical presentation of the meningitis, these patients (IV and O) will remain hospitalized two more days for close clinical observation. In the case of meningitis, appropriate antibiotics are continued until patients' recovery. Corticosteroids should not be prescribed for the patients for any reason.
The physicians caring for the patients are not blinded to the regimen that is used.
Intracranial hemorrhages included subarachnoid hemorrhage, epidural and subdural hematoma and cerebral contusion or hematoma.
The diagnosis of bacterial meningitis is based upon CSF findings (increase in leukocyte count, decrease in CSF glucose [<60% of the level simultaneously measured in blood] increase in the CSF protein > 45 mg/dL and positive CSF smear or culture) in patients with compatible clinical findings (fever, headache, nausea and vomiting, change in the level of consciousness and/or meningismus not explainable with other causes). Probable meningitis is defined when positive CSF smear or culture is not available. In rare cases with relative contraindications for lumbar puncture (accompanying intracranial hemorrhage and clinically suspected raised intracranial pressure), the possible diagnosis of meningitis is based upon clinical findings and ruling out other causes.
Statistical analysis
Intention to treat analysis will be used.
Categorical variables are compared using a chi-square test, and the Student t-test is used to compare continuous variables between groups. Logistic regression with adjustment for other possible risk factors for meningitis is used to estimate the odds ratios and 95% confidence intervals.