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

Impact of immunomodulating therapy on morbidity in patients with severe sepsis.

Pittet D, Harbarth S, Suter PM, et al.

Am J Respir Crit Care Med 1999;160:852-7. [abstract] [full-text for subscribers]

Reviewed by Natalie Cvijanovich, MD, Primary Children's Medical Center, Salt Lake City, Utah

Review posted October 7, 2000

---

I. What is being studied?:

The study objective:

To study the impact of therapy with tumor necrosis factor neutralizing receptor fusion protein (p55-IgG) on 28-day mortality and the incidence of end-organ failures in patients with severe sepsis or early septic shock. The hypothesis was that decreased 28-day mortality is associated with a prolonged duration of intensive care to support failing organs, and that unless treatment has a favorable impact on morbidity as well as mortality, resource use may increase in the treatment group.

The study design:

Patients with severe sepsis or early septic shock were randomized in Phase II double-blind, multicenter, placebo-controlled trial to receive single dose (0.008, 0.042, or 0.083 mg/kg) p55-IgG or placebo. Interim analysis showed no evidence of benefit among patients who received 0.008 mg/kg of p55-IgG, so this arm of the study dosage was discontinued.

The patients included:

Two hundred forty seven hospitalized adults ( 18 yo) who fulfilled criteria for sepsis syndrome without refractory shock. Criteria for sepsis syndrome included: 1) objective signs of acute infection, 2) at least 3 signs of a systemic inflammatory response, and 3) evidence of inadequate organ function, as defined by Abraham, et al. (1).

The patients excluded:

Patients in the original study prospectively stratified into the late (refractory) septic shock group were excluded. Looking at the original study by Abraham et al.(1), those patients with acute rejection, underlying diseases with < 3 month survival, > 120 kg, or chronic vegetative state were excluded, as well as pregnant patients, neutropenic patients (unless due to sepsis), those with HIV, cardiogenic shock, dialysis-dependent, burns, uncontrolled hemorrhage, or DNR. Patients who had signs of inadequate organ function for greater than 12 hours from study drug eligibility were also excluded from the original study.

The interventions compared:

Initially three dosages (0.008 mg/kg, 0.042 mg/kg, and 0.083 mg/kg), and then ultimately one dosage (0.083 mg/kg) of p55-IgG were compared with placebo and with each other. The planned interim analysis in the original study showed no evidence of benefit in the arm receiving the 0.008 mg/kg dose and this group was therefore discontinued. In the current study, only the placebo group and the group receiving 0.083 mg/kg of p55-IgG were included because there was a trend toward a statistically significant reduction in 28-day mortality in the latter group compared to placebo.

The outcomes evaluated:

28-day all cause mortality, ICU length of stay (LOS) after study entry, use and duration of intubation, number and duration of organ failures, number of organ failure-free days, ventilator and ARDS-free days were evaluated.

II. Are the results of the study valid?

Primary questions:

1. Was the assignment of patients to treatments randomized?

Yes. In the original study, patients were prospectively stratified into severe sepsis or late shock and each group was then separately randomized to receive 1 of 3 doses of p55-IgG or placebo. The method of randomization is described neither in this study nor in the original study.

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

Was followup complete?

Yes. All the patients were accounted for.

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

Yes. There was no cross-over between groups.

Secondary questions:

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

Yes.

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

Yes. Age, admission diagnosis, and documented infection on presentation were similar between groups. Overall, there was a trend toward lower predicted mortality in the placebo group than in the treatment group. However, the placebo group had a lower predicted mortality than the treatment group if you look at low vs. high mortality (0-50% vs. > 50% by APACHE II), Chi-square, p = 0.02. Also, although there was no statistically significant difference between groups in overall organ failure at baseline, among individual organ systems there was. For example, the baseline percentage of patient with ARDS in the placebo group was 24%, compared with only 15% in the treatment group, and 21% of patients in the placebo group had renal failure at baseline, compared with 8% of the treatment group. Especially in a study that measures ARDS-free days as an outcome measure, such a difference in baseline rates of ARDS between groups may affect the interpretation of the results.

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

There was no specific treatment protocol for patients. The authors simply state that "patients received standard supportive medical/surgical therapy (fluids, vasopressors, antibiotics, hemodynamic monitoring, and surgery, if needed)". Since this was a multicenter, international trial, it is difficult to know whether there were significant differences in therapeutic strategies among institutions.

III. What were the results?

1. How large was the treatment effect?

Among the group treated with p55-IgG (0.083 mg/kg), there was a 36% relative risk reduction in mortality compared with placebo (23% vs. 36%), but this was not statistically significant (p = 0.07; 95% CI -3%, 76%). There was no difference in hospital LOS. ICU LOS was statistically significantly shorter among the treatment group (9.8 vs. 12.4 days; 95% CI 0.2, 5.0).

The relative risk of developing new organ failure was 0.57 in the treatment group, but this was not significant (95% CI 0.29, 1.1). However, the overall incidence-density of organ failure (defined as the number of days when patients met the criteria for the outcome of interest per 1,000 patient-days) was significantly lower in the treatment group, (RR 0.65; 95% CI 0.60 to 0.71).

The number of days with ARDS was significantly higher among the controls (n = 290 days, incidence-density = 182 ARDS days/1,000 pt. days) than among the treatment group (n = 142 days, incidence-density = 71 ARDS days/1,000 pt. days). The authors then calculate the ratio of the incidence-density as a relative risk (RR = 0.39, 95% CI 0.32 to 0.48).

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

Although a reduction in mortality of 36% was not statistically significant, it would certainly be clinically important. However, the 95% confidence interval is quite wide, so that the relative risk reduction could be as high as 74% for the treatment group, or, conversely, the treatment group could have a 3% relative risk increase. It would have been helpful to stratify the groups prospectively according to the predicted mortality upon presentation. Had the authors done this, they could have assessed the relative risk reduction for each subgroup according to mortality.

For the relative risk of developing new organ failure, the 95% confidence interval is 0.29 to 1.10, indicating that there is no significant decrease in risk. In calculating the overall incidence-density of organ failure (i.e., incidence of organ failure per 1000 patient-days), the confidence interval for the risk reduction is quite tight (0.60 to 0.71) in favor of treatment with p55-IgG.

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

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

Since this study was done in adults, it may not be applicable in children.

2. Were all clinically important outcomes considered?

Mostly. One outcome that was not addressed was nosocomial infection, which is a significant cause of morbidity and mortality.

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

Maybe. It is not clear that any of the dosages of p55-IgG provided any significant benefit. The low dose initially used in the original study was discarded at the planned interim analysis, and the intermediate dose was discarded in the current study. The differences in baseline organ dysfunction make interpretation of any benefit shown, e.g., reduction in incidence of ARDS, difficult to attribute to the treatment. The complications and costs of the treatment cannot be derived from the data provided (as mentioned by the authors in their limitations), other than the potential reduction in costs if the treatment truly results in a shorter length of stay and lower morbidity.

Discussion

This is a subgroup analysis from a larger dose-finding study, and its results, although interesting, cannot lead to any definite conclusions since the study was not designed to answer the questions asked in this subgroup. In light of the fact that one of the limitations of the study (as described by the authors) is that a subsequent phase III trial did not show a significant reduction in mortality in the subgroup described, it seems unlikely that this p55-IgG will become a standard therapy in sepsis. However, the question of whether a treatment that reduces mortality results in higher resource utilization because of increased morbidity or longer LOS is an interesting one.

References

1. Abraham E, Glauser MP, Butler T, et al., p55 Tumor Necrosis Factor Receptor Fusion Protein in the Treatment of Patients with Severe Sepsis and Septic Shock: A Randomized Controlled Multicenter Trial Ro 45-2081 Study Group. JAMA. 1997;277(19):1531-8. [abstract]

Comments

Back to the EB Journal Club Index