Criteria abstracted from The Users' Guide to Medical Literature, from the Health Information Research Unit and Clinical Epidemiology and Biostatistics, McMaster University

Highlighted lines and questions below provide links to the pertinent description of criteria in The EBM User's Guide, now available at the Canadian Centres for Health Evidence

Article Reviewed:

Please visit the new Evidence Based Journal Club Reviews

Randomized, placebo-controlled, double blinded trial of dexamethasone in African children with sepsis

Slusher T, Gbadero D, Howard C, Lewison L, Giroir B, Toro L, Levin D, Holt E, McCracken GH Jr.

Pediatr Infect Dis J 1996;15:579-583 [abstract]

Reviewed by Mignon McCulloch, MD and Max Klein, MD, Department of Paediatrics and Child Health,
University of Cape Town and Red Cross War Memorial Children's Hospital

Review posted August 18, 1998

---

I. What is being studied?:

The study objective:

To determine the effect of moderate dose dexamethasone administered before antibiotics on the outcome of African children with sepsis.

The Setting:

Two hospitals in Africa: a periurban hospital in Nigeria and a rural hospital in Kenya.

The study design:

Randomized, double blinded, placebo-controlled trial.

The patients included:

All patients thought to be moderately or severely ill secondary to an underlying infectious etiology if they met criteria for either sepsis syndrome or shock with evidence of infection.

Sepsis syndrome was defined by modifying the Bone criteria to include all of the following: 1.) clinical evidence of infection; 2.) fever; 3.) tachypnea (respiratory rate > 1 SD for age); 4.) tachycardia (heart rate > 1 SD for age); 5.) decreased end organ perfusion including one or more of (a.) hypoxemia, (b.) urine output < 1 ml/kg/h, (c.) capillary refill > 2 seconds, (d.) decreased peripheral pulses, (e.) altered mental status.

Shock was defined as: 1.) systemic hypotension (defined as 1 SD below systolic and/or diastolic for age; OR 2.) three or more of the following: (a.) tachycardia, (b.) sudden drop in systolic or diastolic blood pressure by 10 mm Hg or more, (c.) decreased peripheral pulses (d.) capillary refill > 2 seconds, (e.) signs of central nervous system hypoperfusion, (f.) urinary output < 1 ml/kg/h, (g.) altered respiratory function, defined as tachypnea for age, increased work of breathing or cyanosis.

The patients excluded:

Children whose parents refused to consent, children receiving parenteral antibiotics within 24 hours before admission to the study, children who were diagnosed with cerebral malaria.

The interventions compared:

Dexamethasone sodium phosphate 20 mg/kg/dose (maximum 4 mg) or placebo (0.45% sodium chloride solution) administered 5-10 minutes prior to the first dose of antibiotics, then every 8 hours for 2 days.

The outcomes evaluated:

The outcomes evaluated were survival to discharge, hemodynamic stability at 48 hours, fever status at 48-72 hours, median length of hospital stay, normality at discharge, normality at post-discharge follow-up.

II. Are the results of the study valid?

Primary questions:

1. Was the assignment of patients to treatments randomized?

Yes. Randomization was done in blocks of 8 patients.

2. Were all patients who entered the trial properly accounted for and attributed at its conclusion?

Was followup complete?

Yes, for some outcome variables. 72 patients were enrolled and the outcomes were evaluated for all 72 patients for survival, afebrile at 48-72 hours, median hospitalization time and functional status at hospital discharge. One patient was not followed up to evaluate hemodynamic stability at 48 hours in the placebo group. Long-term follow-up after hospital discharge was incomplete with 10 patients missing in the treatment and 7 patients in the control group.

Were patients analyzed in the groups to which they were randomized?

Yes. No patients crossed over.

Secondary questions:

3. Were patients, health workers, and study personnel "blind" to treatment?

Although the authors state that the trial was double-blinded, they do not state the methods used to do so.

4. Were the groups similar at the start of the trial?

Maybe. Although no statistically significant differences were found, as compared with the steroid group, controls were considerably younger (median age 15 vs 30 months), comprised more males, had a shorter duration of symptoms before onset of treatment (3 vs 4 days), had more positive blood cultures (42% vs 35%) and had fewer positive CSF gram stains (27% vs 43%). Also of concern is the fact that groups were not compared in regard to nutritional or HIV status which one would expect to be relevant issues in the populations under study.

5. Aside from the experimental intervention, were the groups treated equally?

Maybe. All patients received Ceftriaxone with an initial dose of 100 mg/kg/day (maximum 2 g) intravenously divided every 12 hours. This dose was reduced to 50 mg/kg/day if meningitis was excluded. Antibiotic therapy was altered if clinically indicated.

All patients had blood samples drawn for blood culture and most had blood drawn for blood count and malaria smear. Other investigations were performed as clinically indicated and available (including examination of the cerebrospinal fluid). One hospital had laboratory facilities and support that limited investigations. This may have led to differences between the two study groups in how aggressive investigators were in evaluating patients for the etiology of infection. Staffing variability may have led to differences in the intensity of recording outcomes.

III. What were the results?

1. How large was the treatment effect?

There was a trend towards increased mortality in the dexamethasone group. Mortality was 17% (6/36) in the dexamethasone group versus 11% (4/36) in the placebo group (Relative Risk 1.50; 95% Confidence Interval 0.46, 4.87). There was a trend towards increased risk of hemodynamic instability at 48 hours in the dexamethasone group. The risk of hemodynamic instability was 67% (24/36) in the dexamethasone group versus 51% (18/35) in the placebo group (Relative Risk 1.30; 95% Confidence Interval 0.87, 1.93). There was a trend towards decreased risk of fever at 48-72 hours in the dexamethasone group. The risk of fever was 39% (14/36) in the dexamethasone group versus 51% (10/36) in the placebo group (Relative Risk 0.70; 95% Confidence Interval 0.42, 1.16).

2. How precise was the estimate of the treatment effect?

As reported above, the 95% confidence interval for the effect of steroids on mortality is broad, ranging from a reduction of mortality of 54% to an increase in mortality of 480%. This wide confidence interval is related to the low (11%) mortality in the placebo group. Similarly, the confidence interval for the steroid effect on fever is broad (58% reduction - 16% increase).

IV. Will the results help me in caring for my patients?

1. Can the results be applied to my patient care?

Uncertain. The study was done in Africa and 40% of the children were malnourished. However, the pathogens were similar to organisms seen in developed countries.

2. Were all clinically important outcomes considered?

Six outcome criteria were used (see above) but precisely how they were applied to each of the diagnostic groups is not specified. Criteria of normality at discharge and follow-up were not defined for any of the diagnostic groups: pneumonia, meningitis, septic arthritis and osteomyelitis. In addition, the follow-up period is not specified.

3. Are the likely treatment benefits worth the potential harms and costs?

No, there is no evidence to support the use of steroids in pediatric patients with sepsis. This study supports the findings of a meta-analysis critically ill adults which found no benefit from the use of corticosteroids in sepsis or septic shock, and suggest that their use may be harmful (6).

Comments from the Reviewers:

Given the limitations of this study the failure to demonstrate an effect cannot be taken to mean that there were no effects. Members of the same group have shown that dexamethasone, at a dose similar to that used in this study, is actually beneficial in children with bacterial meningitis when given early in the course of the disease and about 20 minutes before the administration of antibiotic (2). Thus, the failure to exclude patients with meningitis or, at least, to use them as positive controls is surprising. A subgroup analysis is not reported but the study lacks sufficient statistical power to do so. According to the way the data were analyzed, any benefit from steroids in meningitis would disguise adverse effects in other groups.

The literature leaves no doubt that steroids are not a panacea in the treatment of sepsis per se. But one does need to bear in mind the fact that steroids are actually beneficial in bacterial meningitis (1-5). So there is at least one defined septic condition which does benefit from steroids, and there may possibly be others. This study will be counterproductive if it discourages the search for other specific infections which may be steroid responsive.

The authors state that "Informed consent was obtained from the parents or guardians of the infants and children who served as subjects of this investigation." However, it is not stated whether consent was written. But this is a legal technicality and of less concern than that consent should be informed. It is probable that naïve care-givers from poor and deprived communities would have difficulty grasping the implications of a double-blind placebo-controlled study or the risk of steroids. An idea of the degree to which consent is informed can be gained from the number of persons refusing enrollment, but the number of refusals is not specified.

This article is worth studying for the insight it gives into the conditions of medical practice in most of the world and the huge obstacles to medical research in developing countries. It is a cruel irony that research should be so difficult amongst the poor and illiterate - the most numerous victims of sepsis and potentially the greatest beneficiaries of simple, cheap and effective interventions.

The authors deserve credit for attempting a study under difficult conditions and for presenting their findings in a forthright manner without attempting to disguise its flaws.

References

1. Lebel MH, Freij BJ, Syrogiannopoulos GA, et al. Dexamethasone therapy for bacterial meningitis: results of two double-blind, placebo-controlled trials. N Engl J Med. 1988;319:964-971.
2. Odio CM, Faingezicht I, Paris M, et al. The beneficial effects of early dexamethasone administration in infants and children with bacterial meningitis. N Engl J Med 1991;324:1525-31
3. Paris MM, Hickey SM, Uscher MI, et al. Effects of Dexamethasone on therapy of experimental penicillin- and cephalosporin-resistant pneumococcal meningitis. Antimicrob Agents Chemother. 1994;38:1320-1324.
4. Wald ER, Kaplan SL, Mason EO, et al. Dexamethasone therapy for children with bacterial meningitis. Pediatrics. 1995;95:21-28.
5. Commentary. The Role of Steroids in the Management of Children with Bacterial Meningitis. Pediatrics 1995;95: 29-31.
6. Cronin L, Cook DJ, Carlet J, et al. Corticosteroid treatment for sepsis: a critical appraisal and meta-analysis of the literature. Crit Care Med 1995;23:1430-9.

Comments

Back to the EB Journal Club Index