PROGNOSIS

Criteria abstracted from The Users' Guides to the Medical Literature series in JAMA

Intensive Care Med 2007 33(9):1609-13.[abstract]

Review posted April 16, 2008

1. What is being studied?
   
   
   1. Study Objective:
      
      To describe the serum cortisol profile, evaluate the adrenal response in children with septic shock, and to determine the influence of adrenal response on outcome and mortality
      
      
   2. Study Design:
      
      This was a prospective observational study.
      
      
2. Are the results of the study valid?

Primary questions

1. Was there a representative and well defined sample of patients at a similar point in the course of the disease?
   
   Yes. The authors were interested in the adrenal response in children with septic shock. They defined adrenal insufficiency as, a baseline serum cortisol level < 690 nmol/L (25 g/dl) and/or a corticotrophin response < 250 nmol/L (9 g/dl) from the baseline. Septic shock was defined according to the 1992 criteria specified in the consensus statement of the American College of Chest Physicians and the Society of Critical Care Medicine and adapted for children.

   Patients included were children with septic shock between the ages of 1 month and 15 years admitted to one of two intensive care units in southern Brazil during the seven month interval, May through November 2003. Patients were excluded if they had heart disease or recent cardiac surgery, a requirement for vasopressor or inotropes of dopamine of < 5 ug/kg/min, a pre-existing condition associated with hypothalamic-pituitary dysfunction or any corticosteroids given during the 4 weeks before the admission.
   
   The severity of illness in patients with septic shock was defined by PRISM II scores in the first 24 hours after the diagnosis of septic shock.
   
   

2. Was follow-up sufficiently long and complete?
   
   Yes. There were no patients lost to follow up and patients were followed until the time of discharge from the ICU or until the time of death

Secondary questions:

3. Were objective and unbiased outcome criteria used?
   
   Yes. Outcome measures were defined at the beginning of the study as length of ICU stay, duration of catecholamine treatment and death during ICU admission all of which are objective data and can easily be measured.
   
   
4. Was there adjustment for important prognostic factors?
   
   Partly. There was no significant difference in age, sex, weight, requirement for mechanical ventilation or duration of catecholamine use when patients with baseline cortisol <690 nmol/L (25.0 ug/dL) were compared to those with baseline cortisol >690 nmol/L (25.0 ug/dL), or when patients with corticotrophin response >250 nmol/L (9.06 ug/dL)were compared to those with corticotrophin response <250 nmol/L (9.06 ug/dL). Patients with corticotrophin response <250 nmol/L (9.06 ug/dL) had greater occurrence of catecholamine resistant shock and higher PRISM II scores than patients with corticotrophin response >250 nmol/L (9.06 ug/dL). The authors controlled for gender and PRISM II scores in their regression analysis that evaluated the relationship between baseline cortisol, response to corticotrophin and mortality.

3. What are the results?
   
   
   1. How large is the likelihood of the outcome event in a specified period of time?
      
      Baseline cortisol of 690 nmol/L (25.0 ug/dL) or more was highly associated with increased mortality. Of those with baseline cortisol <690 nmol/L (25.0ug/dL), 0/9 (0%) died. Of those with baseline cortisol >690 nmol/L (25.0 ug/dL, 7/13 (53%) died. This represents an absolute risk increase (ARI) of 53%!

      Failure to increase cortisol more than 250 nmol/L (9.06 ug/dL) in response to corticotrophin stimulation was associated with death in 6 of 10 cases versus death in 1 out of 12 patients who demonstrated an appropriate response to corticotrophin stimulation. The relative risk is 7.2 (0.6/0.083). Therefore patients who demonstrated an inadequate response to corticotrophin stimulation had a 7 fold greater risk of death than patients who responded appropriately to corticotrophin stimulation.
      
      Although univariate analysis showed that baseline cortisol >690 nmol/L (25.0 ug/dL) was associated with death and that cortisol response to corticotrophin <250 nmol/L (9.06 ug/dL) was also associated with death, regression analysis showed that only the combination of high baseline cortisol (>690 nmol/L (25.0 ug/dL)) and low response to corticotrophin (<250 nmol/L 9.06 ug/dL) was associated with death when controlled for gender and PRISM II score. They do not provide the regression coefficients to estimate the independent risk of this combination.
      
      

   2. How precise are the estimates of likelihood?
      
      The RR of 7.2 provides only an estimate of the true risk. The confidence interval will allow for a determination of the precision of this estimate. The 95% CI is 1.03 to 50.3. The relative risk of death in pediatric patients with septic shock who fail to respond to corticotrophin stimulation is somewhere between 1.03 and 50.3. This is a wide confidence interval and therefore not very precise.

      Relative risk and confidence intervals for the regression analysis are not provided.
      
      

4. Will the results help me in caring for my patients?
   
   
   1. Were the study patients similar to my own patients?
      
      The study patients described seem very similar to patients that we see in our intensive care unit; therefore the results appear to be generalizable to my practice. However, we do not have an overall mortality rate of 32% from septic shock. The difficulty in addressing this condition in any PICU population is the lack of a universally agreed upon definition for adrenal insufficiency. Hatherhill et al [1] defined adrenal insufficiency as a serum cortisol < 200nmol/l (7.24 ug/dL) following corticotrophin (145ug/m2 to maximum of 250ug) stimulation. Pizarro et al[2], described adrenal insufficiency as a serum cortisol < 248nmol/l (8.98 ug/dL) in response to corticotrophin (250ug) stimulation while Menon et al [3] defined adrenal insufficiency as a serum cortisol of < 496nmol/l (17.97 ug/dL) in response to corticotrophin stimulation. It may be that a poor response to such a small dose of corticotrophin as used in the Casartelli study (0.5 g / 1.73 m2) in a critically ill child is the true definition since this study was the only one to show an association between AI and mortality. The finding in the Casartelli study that was most unusual was that all seven of the children who died had a baseline cortisol > 690 nmol/L (25.0 ug/dL).
      
      
   2. Will the results lead directly to selecting or avoiding therapy?
      
      An assessment of baseline cortisol levels and response to corticotrophin in septic patients could identify patients at high risk for death but further study is necessary to determine whether treatment (steroid replacement) would be benefit these patients.
      
      
   3. Are the results useful for reassuring or counseling patients?
      
      In this study inadequate response to corticotrophin was found to be positively associated with increased mortality. This information is useful in that once identified, patients who are at a higher risk for mortality could be treated more aggressively.

References

1. Hatherhill M, Tibby SM, Hilliard T, Turner C, Murdoch IA (1998) Adrenal insufficiency in septic shock. Arch Dis Child 80: 51 -55
   

2. Pizarro CF, Troster EJ, Damiani D, Carcillo JA (2005) Absolute and relative adrenal insufficiency in children with septic shock. Crit Care Med 33:855-859
   

3. Menon K, Clarson C (2002) Adrenal function in pediatric critical illness. Pediatr Crit Care 3:112-116

Last Updated: April 16, 2008

Fill out our Feedback Form | Editorial Notes | About Us