Therapy
Effect of treatment with low doses of hydrocortisone and fludrocortisone on mortality in patients with septic shock
This review may be edited
Summary
Posted By: Julian Castro and Gord Doig
E-Mail: castro_julian@hotmail.com gdoig@med.usyd.edu.au
Posted Date: 2 Sept 2002
Title: Effect of treatment with low doses of hydrocortisone and fludrocortisone on mortality in patients with septic shock
Authors: Annane D et al
Reference: JAMA 2002 August; 288(7): 862-871.
Link: Click here for a direct link to the paper. A password may be required for access to fulltext.
Abstract: CONTEXT: Septic shock may be associated with relative adrenal insufficiency. Thus, a replacement therapy of low doses of corticosteroids has been proposed to treat septic shock.
OBJECTIVE: To assess whether low doses of corticosteroids improve 28-day survival in patients with septic shock and relative adrenal insufficiency.
DESIGN AND SETTING: Placebo-controlled, randomized, double-blind, parallel-group trial performed in 19 intensive care units in France from October 9, 1995, to February 23, 1999.
PATIENTS: Three hundred adult patients who fulfilled usual criteria for septic shock were enrolled after undergoing a short corticotropin test.
INTERVENTION: Patients were randomly assigned to receive either hydrocortisone (50-mg intravenous bolus every 6 hours) and fludrocortisone (50- micro g tablet once daily) (n = 151) or matching placebos (n = 149) for 7 days.
MAIN OUTCOME MEASURE: Twenty-eight-day survival distribution in patients with relative adrenal insufficiency (nonresponders to the corticotropin test).
RESULTS: One patient from the corticosteroid group was excluded from analyses because of consent withdrawal. There were 229 nonresponders to the corticotropin test (placebo, 115; corticosteroids, 114) and 70 responders to the corticotropin test (placebo, 34; corticosteroids, 36). In nonresponders, there were 73 deaths (63%) in the placebo group and 60 deaths (53%) in the corticosteroid group (hazard ratio, 0.67; 95% confidence interval, 0.47-0.95; P =.02). Vasopressor therapy was withdrawn within 28 days in 46 patients (40%) in the placebo group and in 65 patients (57%) in the corticosteroid group (hazard ratio, 1.91; 95% confidence interval, 1.29-2.84; P =.001). There was no significant difference between groups in responders. Adverse events rates were similar in the 2 groups.
CONCLUSION: In our trial, a 7-day treatment with low doses of hydrocortisone and fludrocortisone significantly reduced the risk of death in patients with septic shock and relative adrenal insufficiency without increasing adverse events.
 
Are the Results Valid?
1. Was the assignment of patients to treatments randomized? ( Was allocation concealment maintained?)
Yes. Randomization was centrally performed, concealed, and stratified by center in blocks of 4.
2. Were all patients who entered the trial properly accounted for and attributed at its conclusion?
2a. Was followup complete?
Yes. Outcomes are avaiable on all patients except one. One patient withdrew consent to follow-up.
2b. Were patients analyzed in the groups to which they were randomized?
Yes. All patients except the one who withdrew consent were included in the intention-to-treat analysis. A more conservative approach would have been to assume a poor outcome for the patient who withdrew consent and include them in the analysis, however since only one patient was involved, their omission would be highly unlikely to have biased results.
3. Were patients, health workers, and study personnel blind to treatment?
Yes. Blinding was achieved through the use of placebo injectate and pills.
4. Were the groups similar at the start of the trial?
Yes. Review of table 1 and 2 suggests that there were no imbalances at baseline.
5. Aside from the experimental intervention, were the groups treated equally?
At table of concomitant interventions is not presented.
What are the Results?
1. How large was the treatment effect?
Intention to treat analysis
Primary outcome:
28-day survival distribution (time-to-death)
In the intention to treat analysis, the authors report an increase in median time to death from 13 days in the placebo group to 19.5 days in the steroid group. (P-value for direct comparison unavailable, the authors report p=0.03 from a multivariate model controlling for baseline cortisol, cortisol response, McCabe classification, LOD score, arterial lactate and PaO2/FiO2 ratio. Since none of these factors appear to be in imbalance at baseline, interpretation of results of the multivariate model would be strengthened if the authors provided a direct, univariate comparision of survival time.)
Secondary outcomes: Mortality rate at day 28, ICU discharge, hospital discharge and 1 year follow-up
In the intention to treat analysis, the authors report no significant differences between mortality rates. At day 28, 61% of placebo patients had died compared to 55% of steroid patients. At ICU discharge, mortality was 68% placebo vs. 60% treated. At hospital discharge, morality was 69% vs. 63%. At one year, mortality was 75% vs. 68%.
Sub-group analyses:
Primary outcome:
28-day survival distribution (time-to-death)
Nonresponders: The median time to death was 12 days in the placebo group vs. 24 days in the steroid treated group. (P-value for direct comparison unavailable, the authors report p=0.02 from a multivariate model controlling for baseline cortisol, cortisol response, McCabe classification, LOD score, arterial lactate and PaO2/FiO2 ratio. Since none of these factors appear to be in imbalance at baseline, interpretation of results of the multivariate model would be strengthened if the authors provided a direct, univariate comparision of survival time.)
Responders The median time to death was 14 days in the placebo group and 16.5 days in the steroid treated group (no significant differences).
Secondary outcomes: Mortality rates
Nonresponders
  • 28-day mortality: 73/115(63%) placebo vs. 60/114(53%) steroid treated. P-value from chi-square = 0.126 [Calculated by Gord]. The authors report p=0.04 from a multivariate model controlling for baseline cortisol, cortisol response, McCabe classification, LOD score, arterial lactate and PaO2/FiO2 ratio. Since none of these factors appear to be in imbalance at baseline, interpretation of results of the multivariate model are strengthened if they are consistent with the univariate chi-square.)
  • ICU mortality: 81/115(70%) placebo vs. 66/114(58%) steroid treated. P-value from chi-square = 0.066 [Calculated by Gord]. The authors report p=0.02 from a multivariate model controlling for baseline cortisol, cortisol response, McCabe classification, LOD score, arterial lactate and PaO2/FiO2 ratio. Since none of these factors appear to be in imbalance at baseline, interpretation of results of the multivariate model are strengthened if they are consistent with the univariate chi-square.)
  • Hospital mortality: 83/115(72%) placebo vs. 70/114(61%) steroid treated. P-value from chi-square = 0.112 [Calculated by Gord]. The authors report p=0.04 from a multivariate model controlling for baseline cortisol, cortisol response, McCabe classification, LOD score, arterial lactate and PaO2/FiO2 ratio. Since none of these factors appear to be in imbalance at baseline, interpretation of results of the multivariate model are strengthened if they are consistent with the univariate chi-square.)
  • 1-Year mortality: 88/115(77%) placebo vs. 77/114(68%) steroid treated. P-value from chi-square = 0.17 [Calculated by Gord]. The authors report p=0.07 from a multivariate model controlling for baseline cortisol, cortisol response, McCabe classification, LOD score, arterial lactate and PaO2/FiO2 ratio. Since none of these factors appear to be in imbalance at baseline, interpretation of results of the multivariate model are strengthened if they are consistent with the univariate chi-square.)
    Responders There were no significant differences in mortality between placebo and steroid treated groups however, it does appear that the mortality rate in the steroid treated group was consistently 10% (absolute) higher.
  • 2. How precise was the estimate of the treatment effect?
    Please refer to paper.
    Will the Results Help Me In Caring For My Patients?
    1. Can the results be applied to my patient care?
    The survival benefit, as measured by increased median time to death, is only apparent in the group of nonresponders to the synactin test after controlling for many baseline factors that were not in imbalance at time of randomization. If this benefit is real, then we must only treat nonresponders. Given that patients were treated VERY early in their course of septic shock, if we are to obtain benefit, we must be able to identify and treat nonresponder within 8 hours of the onset of shock. In our hospital, the synactin test would usually be available the next day and thus, without major changes to our hospital, we would be unable to initiate treatment early enough. That is, if you believe the documented treatment effects are real.
    2. Were all clinically important outcomes considered?
    The authors did a very good job of following patients for 1 year survival. Considering that the only major benefit detected was in duration of survival (increased median time to death), it becomes extremely important to document quality of life during this extra survival time. If the patient does not report a clinically important improvement in quality of life during this extra time, it is possible that this extra time is harmful.
    3. Are the likely treatment benefits worth the potential harms and costs?
    There was no reported difference in adverse event rates. Given that there the authors do not report that significantly more patients survived, and the only documented treatment effect was an increased time to mortality (longer length of stay), it is impossible to calculate 'costs per life saved' however, we are fairly certain that increasing length of stay increases overall costs. In summary, this study provides an excellent basis for a properly powered study to investigate the effect of steroids in patients with septic shock.

    What other people had to say about:
    Effect of treatment with low doses of hydrocortisone and fludrocortisone on mortality in patients with septic shock

    Add Your Comment
    [ Back ]

    Any questions or comments please contact Gordon S. Doig
    Implemented and designed by John Soer and Gordon Doig
    Page last modified on Friday August 10, 2001
    www.EvidenceBased.net
      THIS IS A SPACE MESSAGE. THIS IS STILL THE BEST METHOD TO SPACE A TABLE FOR NETSCAPE AND INTERNET EXPLORER. IF YOU SEE THIS MESSAGE CONTACT WEBADMIN@EvidenceBased.net