Criteria abstracted from The Users' Guides to the Medical Literature series in JAMA
Induced Hypothermia in Severe Bacterial Meningitis: A Randomized Clinical Trial
Mourvillier_B, Tubach F, van de Beek D, Garot D, Pichon N, Georges H, Lefevre LM, Bollaert PE, Boulain T, Luis D, Cariou A, Girardie P, Chelha R, Megarbane B, Delahaye A, Chalumeau-Lemoine L, Legriel S, Beuret P, Brivet F, Bruel C, Camou F, Chatellier D
JAMA 2013 310(20):2174-2183.[abstract]
Review posted March 29, 2014
- What is being studied?:
98 patients who were at least 18 years of age were diagnosed with either suspected or proven bacterial meningitis.
Meningitis as defined by:
1. Cerebrospinal fluid (CSF) white blood cell count greater than 100/µL and glucose CSF/blood ratio of less than one-third, a CSF protein concentration of greater than 2.2 g/L
2. Microorganisms observed in CSF Gram stain.
3. A positive soluble antigen test or polymerase chain reaction for Streptococcus pneumoniae or Neisseria meningitidis,
4. Positive CSF cultures.
All study patients had a GCS of 8 or lower for less than 12 hours and received appropriate antibiotic therapy.- The study objective:
To test the hypothesis, induced hypothermia improves functional outcome at three months compared to normothermia, in comatose patients with bacterial meningitis.
-
The study design:
This is an open-label, multicenter, sequential randomized controlled clinical trial in 49 adult intensive care units in France, February 2009–November 2011.
- The patients included:
- The patients excluded:
Exclusion criteria included pregnancy, a positive cryptococcus test, brain abscess, purpura fulminans, and therapeutic hypothermia for cardiac arrest. They were also excluded if the physician responsible decided to limit life support, or if they were not insured. Finally, patients already enrolled in another interventional study were excluded.
- The interventions compared:
Intentional moderate hypothermia (32-34◦C) versus maintenance of normothermia (37◦C) for a period of 48 hours. The institutions’ standard practice determined the modality of induced hypothermia and included cold saline, ice packs or cooling blankets. The patients were passively rewarmed after 48 hours.
-
The outcomes evaluated:
A Glasgow outcome score (GOS) (1) was obtained at three month post randomization follow up, by a structured telephone interview. GOS of 1, 2, 3 or 4 which meant death, vegetative state, severe or moderate disability respectively was defined as poor outcome while a GOS of 1, i.e., no disability was defined as a good outcome.
Secondary outcomes included overall mortality at three months, hearing impairment at 3 months, muscle strength at intensive care discharge and 3 months. Complications within first 7 days, then weekly for 28 days and a 6 month GOS. Three investigators evaluated the causes of death.
The DSMB monitoring board reviewed severe adverse events and mortality rates after enrollment of 60 new patients.
- The study objective:
- Are the results of the study valid?
Primary questions:
- Was the assignment of patients to treatments randomized?
Yes. They were randomized via a centralized website, into balanced blocks of variable and undisclosed size and stratified on the hypothermia technique (intravascular cooling vs other cooling techniques)
- Were all patients who entered the trial properly accounted for and attributed at its conclusion?
Yes, one hundred and thirty patients were assessed for eligibility, and 32 were excluded.
Was follow-up complete?
Yes, no one was lost to follow-up. The intention to treat principle was used for the two patients with incomplete data.
Were patients analyzed in groups to which they were randomized?
Yes. There was no cross-over of patients between groups.Secondary questions:
- Were patients, health workers, and study personnel "blind" to treatment?
No, the patients, the treating physicians were not blinded. However, the physicians obtaining the 3 and six-month GOS, as well as, the investigators reviewing the charts for the cause of death were “blind” to treatment.
- Were the groups similar at the start of the trial?
Yes, they were similar in mean age, sex of the patient, timing of the first dose of antibiotics and had a mean GCS score of 7.
All patients received mechanical ventilation and were severely ill with high median Simplified Acute Physiology Score II (SAPS II) scores (57 in the intervention group and 53 in the control group). There was a higher percentage of septic shock among the intervention group (37%) as compared to the control (20%) that was not statistically significant (P=0.14).
In both groups, Streptococcus pneumoniae was the predominant organism.- Aside from the experimental intervention, were the groups treated equally?
Yes, patients were treated according to published guidelines and expert opinions. For example, all patients received appropriate antimicrobial therapy, which included ceftriaxone 100mg/kg/day or cefotaxime 200-300mg/kg/d. All the study patients were maintained at a mean arterial pressure of greater than 70 mmHg, normocapnia, and normal ranges for electrolytes. Hypothermia was stopped prematurely in 7 patients due to various reasons.
- What were the results?
- How large was the treatment effect?
The DSMB stopped the trial prematurely after 98 patients were enrolled due to increased mortality in the intervention arm at three months. 51% of the hypothermia group died compared to 31% of the control group, the relative risk was 1.99 with a 95% Confidence Interval (CI) 1.05 to 3.77; P= 0.04. 46% of the patients who received intravascular cooling had increased mortality at three months versus 40% of those cooled by other methods (P=0.36).
The trial investigators expected an absolute risk reduction of 15%, based on studies in the cardiac arrest trial and dexamethasone in adults [2]. At three months, an unfavorable outcome occurred in forty-two of forty nine patients in the hypothermia group versus thirty-six of forty nine patients in the control group. The relative risk was 1.17, with a 95% CI 0.95to1.13; P=0.13.
The relative risk reduction (RRR) for having an unfavorable outcome was -0.17.There was no difference in the number of patients who had a good outcome.
Post-hoc analysis of the observed data showed the probability of the study concluding in favor of futility was 0.977.
The median interquartile range(IQR) of length of ICU and hospital stay were longer in the hypothermia group as compared to control group, 15 days (9-25) vs seven days (6-15) and 33 days (21-42) vs 20 (14-30) days respectively.
Both groups had similar occurrences of aspiration pneumonia, hemorrhage and arrhythmia. - How precise was the estimate of the treatment effect?
The study was designed to have 276 patients, but was curtailed to 98 patients so the estimate of the treatment effect will likely be inaccurate. The relative risk for an adverse outcome in the treatment arm was 1.17 with a 95% CI 0.95 to 1.43, suggesting precise results due to the narrow range of values in the CI. However, the 95% CI crosses the number 1, suggesting that the relative risk could be in favor of the treatment arm at the lower end of the CI. The RR of death for the treatment arm was 1.9, 95% CI 1.29 to 2.81, was both precise and significant. The RR of hearing impairment was 3.38, 95% CI 1.08 to 10.57. The wide range of the CI suggests imprecise results. The RR of having a systemic cause of death in the treatment arm was statistically significant but imprecise RR 4.23, 95% CI 1.87to 9.58.
- How large was the treatment effect?
- Will the results help me in caring for my patients?
- Can the results be applied to my patient care?
No, the median age of patients in the study was 54. The patients in our practice are less than 19 years of age.
- Were all clinically important outcomes considered?
Yes, Investigators evaluated death, function, hearing impairment.
- Are the likely treatment benefits worth the potential harms and costs?
No, the study did not present any benefits, in fact; it caused harm.
- Can the results be applied to my patient care?
References
- J. Wilson, L. Pettigrew and G. Teasdale, "Structured Interviews for the Glasgow Outcome Scale and the Extended Glasgow Outcome Scle:Guidelines for their Use," Neurotrauma, vol. 15, no. 8, pp. 573-585, 1997.
- de Gans J, van de Beek D; European Dexamethasone in Adulthood Bacterial Meningitis Study Investigators. Dexamethasone in adults with bacterial meningitis. N Engl J Med. 2002;347(20):1549-1556.