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

Specific questions in this review are based on Jaeschke RZ, Meade MO, Guyatt GH, Keenan SP, Cook DJ. How to use diagnostic test articles in the intensive care unit: diagnosing weanability using f/Vt. Crit Care Med. 1997;25(9):1514-2. [abstract]

Article Reviewed:

Please visit the new Evidence Based Journal Club Reviews

Adrenal insufficiency during septic shock.

Marik PE, Zaloga GP.

Crit Care Med. 2003;31(1):141-5. [abstract]

Reviewed by Priya Prabhakaran MD, SRMCRI, India

Review posted November 14, 2003

I. What is being studied?

Study objective:

To determine the sensitivity and specificity of the routinely used low dose (1 mcg), and high dose (250 mcg) corticotropin stimulation tests using a random serum cortisol concentration of 25 mcg/dl as the gold standard.

Study design:

a) Prospective, non-randomized, non-blinded evaluation of three tests used to diagnose adrenal insufficiency (random serum cortisol concentration <25 mcg/dl, low dose corticotropin stimulation test and high dose corticotropin stimulation test).

The patients included:

59 consecutive adult patients with septic shock as defined by the SCCM/ACCP criteria who were admitted to a medical ICU between March 1999 and May 2000.

The patients excluded:

Patients with HIV infection, pre-existing adrenal insufficiency, and those who had received any steroids within the previous year were excluded.

Methods:

After volume loading with lactated Ringer's, all patients were on norepinephrine infusions at the initiation of the study. Dopamine was not used. Dobutamine was used in patients who had a CI< 3.5 L/min/m2. None of the patients received any drugs that are known to suppress adrenal function. A baseline serum cortisol was drawn on all patients. Subsequently, they underwent the low dose (LD) corticotropin stimulation test, followed 60 minutes later by the high dose (HD) test. Serum cortisol concentrations were drawn at 30 and 60 minutes after both tests. All patients then received 100 mg hydrocortisone IV every 8 hours.

Patients who could be weaned off of pressors within 24 hours of beginning steroid replacement were considered steroid responsive. Hydrocortisone was continued in steroid responsive patients as well as those who had a random cortisol concentration < 25 mcg/dl.

II. Are the results of the study valid?

Primary questions:

1. Was there an independent, blind comparison with a reference standard?

A serum cortisol level < 25 mcg/dl and the hemodynamic response to steroid replacement were the reference tests used. The hemodynamic response to steroid replacement was defined as the need for or cessation of norepinephrine to maintain a mean arterial pressure > 65 mm Hg within 24 hours of the first dose of hydrocortisone. These results were, in any case only available the following day, by which time steroid responsiveness or not was already ascertained. The authors are not entirely clear that hemodynamic responsiveness was determined blindly, i.e., without knowing the results of the cortisol levels. They state that cortisol levels were available "within" 24 hours of testing.

2. Did the patient sample include an appropriate spectrum of patients to whom the diagnostic test will be applied in clinical practice?

Yes. Since sepsis and hypotension are frequent issues in children admitted to the ICU, as well as very rigorous stressors of the hypothalamic-pituitary-adrenal axis, this patient sample is appropriate to evaluate adrenal insufficiency. Valuable information could perhaps be obtained by including patients with serious trauma and post-operative patients.

Secondary questions:

3. Did the results of the test being evaluated influence the decision to perform the reference standard?

No, all patients were started on steroid replacement after the baseline cortisol concentration was drawn, and the LD and HD tests performed. The hemodynamic response to steroid replacement was assessed in all patients regardless of the results of the random serum cortisol concentration, or the stimulation tests.

4. Were the methods for performing the test described in sufficient detail to permit replication?

Yes. A baseline serum cortisol was drawn on all patients. Subsequently, they underwent the low dose (LD, 1 mcg) corticotropin stimulation test, followed 60 minutes later by the high dose (HD, 249 mcg) test. Serum cortisol concentrations were drawn at 30 and 60 minutes after the LD and HD tests.

III. What are the results?

1. Are the test's sensitivity, specificity, and likelihood ratios presented (or are the data necessary for their calculation provided)?

When a random serum cortisol < 25 mcg/dl was used as the diagnostic test, 61% of patients were adrenal insufficient. This was opposed to 37% of patients who were steroid responsive in this trial. A random serum cortisol concentration < 25 mcg/dl had a sensitivity of 95% in diagnosing steroid responsiveness and specificity of 57%. A positive LD test had a sensitivity of 54% and specificity of 97% in diagnosing steroid responsiveness. A positive HD test had a sensitivity of 22% and specificity of 100% in diagnosing steroid responsiveness.

In predicting steroid responsiveness, receiver operating curve (ROC) analysis revealed that a serum cortisol level of 23.7 mcg/dl had the best sensitivity (86%) and specificity (66%). This level had a likelihood ratio of 2.6, a positive predictive value (PPV) of 0.62, and a negative predictive value (NPV) of 0.88. The LR of 2.6 means that a subject with a serum cortisol level < 23.7 is 2.6 times as likely as someone with a level ≥ 23.7 to have steroid responsiveness.

The diagnostic stress cortisol threshold of 25 mcg/dl that was used in this study closely approximates the value of 23.7 mcg/dl determined by the ROC analysis.

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

1. Will the reproducibility of the test result and its interpretation be satifactory in my setting?

While the test can be easily replicated in children, its interpretation in this population is a matter of debate and needs to be further evaluated. Although the authors do not tell us the methods used to measure serum cortisol it is likely to be easy to measure in children.

2. Are the results applicable to my patients?

The results are not directly applicable in children. It cannot be determined from this paper whether a random serum cortisol level of 25 mcg/dl is an appropriate reference or should be used as the cut-off value for determining steroid responsiveness for children.

3. Will the results change my management?

The results of this study reinforce the need for further evaluation of this issue in children, but will not directly influence my management. The LR of 2.6 indicates a small change from pre- to post- test probability.

4. Will patients be better off as a result of the test?

This issue has not been directly addressed in this study. One might speculate though, that early identification of a sub-group of patients who are most likely to respond to steroids (in this case, most patients with a random serum cortisol of < 25mcg/dl) might be beneficial. It is noteworthy that a few patients with levels > 25 mcg/dl also did respond to steroids, and that not all with a level less than this were steroid responsive. It is unclear what the benefit of steroid replacement would be in steroid-non-responsive patients with a level < 25 mcg/dl. Alternatively, if the main utility of steroids in this setting is hemodynamic stability, the response to steroid replacement alone can be used as a guide to determining adrenal insufficiency in this population of patients.

The issue of whether identifying adrenal insufficiency in patients with septic shock will actually improve outcome, i.e., decrease mortality, is still a matter of debate. The study by Annane et al (1) on 300 adult patients with septic shock revealed a decrease in mortality in the steroid treated non-responders to the short corticotropin stimulation test, but the 95% CI around the primary end point was wide, thus the true magnitude of the effect might actually be fairly small. The only pediatric study that has shown a reduction in mortality with steroid treatment in children with dengue shock syndrome is by Min et al. (2) The current recommendation is to consider high dose hydrocortisone in all children with septic shock who are ctecholamine resistant, and those with suspected or proven adrenal insufficiency. Further evaluation of a large cohort of children with septic shock would be needed before guidelines and limits for the diagnosis of adrenal insufficiency in this population can be established.

References:

1. Annane D, Sebille V, Charpentier C, Bollaert PE, et al. Effect of treatment with low doses of hydrocortisone and fludrocortisone on mortality in patients with septic shock. JAMA. 2002;288(7):862-71. [abstract, PedsCCM EBJC Review]
2. Min M, U T, Aye M, Shwe TN, Swe T. Hydrocortisone in the management of dengue shock syndrome. Southeast Asian J Trop Med Public Health. 1975 Dec;6(4):573-9. [abstract]

 

Comments

Back to the EB Journal Club Index