Published ahead of print on November 4, 2005, doi:10.1164/rccm.200504-545OC
© 2006 American Thoracic Society doi: 10.1164/rccm.200504-545OC
Association between Adrenal Insufficiency and Ventilator WeaningDepartment of Thoracic Medicine II, Chang Gung Memorial Hospital, Taipei, Taiwan Correspondence and requests for reprints should be addressed to Horng-Chyuan Lin, M.D., Department of Thoracic Medicine II, Chang Gung Memorial Hospital, 5 Fushing Street, Gueishan Shiang, Taoyuan, Taiwan. E-mail: lin53424{at}ms13.hinet.net
Rationale: Adrenal insufficiency is a common disorder in critically ill patients with mechanical ventilation and is usually associated with higher mortality and poor clinical outcome. Objectives: To determine whether stress dose corticosteroid supplementation can improve ventilator weaning and clinical outcome in patients with adrenal insufficiency. Methods: A prospective, randomized, placebo controlled, double-blinded study was conducted in the intensive care unit of a tertiary teaching hospital. A total of 93 mechanically ventilated patients were enrolled in the ventilator weaning trial. Adrenal function was assessed in all patients. Patients with adrenal insufficiency were randomized to the treatment group (50 mg intravenous hydrocortisone every 6 h) and the placebo group. Measurements and Main Results: The successful ventilator weaning percentage was significantly higher in the adequate adrenal reserve group (88.4%) and in the stress dose hydrocortisone treatment group (91.4%) than in the placebo group (68.6%). The weaning period was shorter in the hydrocortisone treatment group than in the placebo group. No significant adverse effects were observed in the corticosteroid treatment group. Conclusions: For patients with respiratory failure, early identification of adrenal insufficiency and appropriate supplementation with stress dose hydrocortisone increase the success of ventilator weaning and shortens the weaning period.
Key Words: adrenal insufficiency • hypothalamic-pituitary-adrenal axis • stress dose hydrocortisone • ventilator weaning The hypothalamic-pituitary-adrenal (HPA) axis plays an important role in the host stress response. During stress, stimulation of the HPA axis increases corticotropin release and enhances adrenal secretory activity (1, 2). Adrenal insufficiency (AI) is not uncommon in the intensive care unit (ICU), and may occur during severe sepsis (3–6). The diagnosis can be confirmed through laboratory assessment of adrenocortical function. However, diagnostic criteria vary (7–11), which can cause misdiagnosis. Application of corticosteroid treatment in patients with sepsis has remained debatable since the clinical introduction of cortisone in 1949. The safety and effectiveness of corticosteroid supplements have not been proved for all patients with sepsis (12, 13). Despite this, some studies (14–17) found that extremely ill patients with persistent shock requiring vasopressors and prolonged mechanical ventilation may benefit from physiologic corticosteroid supplement. These results implied that critically ill patients may have relative AI. Ventilator weaning (VW) is the transitional period between total ventilatory support and spontaneous breathing and is a crucial issue in ICUs (18). Rapid and successful VW can shorten the length of stay, decrease morbidity and mortality, and reduce costs (19). Although the reasons remain unclear, patients requiring reintubation after failed extubation have a poor prognosis and higher hospital mortality (20). Several pathophysiologic mechanisms explain why some patients fail a VW trial (21), but few clinically practical indices have been defined. In the past, adrenal function has usually been evaluated after VW failure. The relationship between AI and VW remains largely unclear for critical care clinicians. On the basis of the concept that early diagnosis and treatment of AI may influence patient outcomes, we hypothesized that corticosteroid repletion in patients with impaired adrenal reserve would improve VW. Part of the preliminary results of the study has been previously reported in the form of an abstract (22).
Patients Chang Gung Memorial Hospital is a 3,000-bed tertiary teaching facility. The ICU is a 26-bed unit that cares for patients with critical medical conditions. The study started on May 1, 2003, and ended on December 31, 2003. All patients enrolled in the study were under mechanical ventilation for more than 72 h via endotracheal tubes. The primary etiology of their respiratory failure was recovering, and vasopressor agents and sedative medication had been discontinued for at least 24 h before the study commenced. These patients were hemodynamically stable, neurologically intact (Glasgow coma score > 11), and waiting for VW. Patients received regular corticosteroid medication before or during the admission, and patients without adequate cough reflex were excluded. The study was performed with the approval of the hospital ethics committee. Informed consent from each enrolled patient or a close relative was obtained.
Study Protocol
All enrolled patients underwent a protocol-driven VW strategy. The respiratory therapists screened for the weaning criteria daily. The criteria for starting weaning were as follows:
Patients meeting the criteria entered a 2-h spontaneous breathing T-piece trial. The criteria to terminate spontaneous breathing trial were as follows:
If the patient went through the spontaneous breathing trial successfully, extubation was performed. Successful weaning was defined as the patient not requiring reintubation or additional respiratory support, such as noninvasive positive-pressure ventilation within 48 h of extubation. If the patient was not weaned successfully and the total ventilator period was more than 14 d, the patient was defined as weaning failure. Tracheostomy was recommended for long-term care. All complications during VW were recorded.
Statistical Analysis
Patient Characteristics During the study period, 472 ventilated patients were admitted in the ICU and 179 patients were evaluated for this study. Ninety-three patients who met the inclusion criteria were enrolled in the study (Figure 1). After initial evaluation of adrenal function, 23 patients were found to have normal adrenal function and 70 were diagnosed with AI. Patients with AI were randomly selected by computer to the treatment group (n = 35) and the placebo group (n = 35). Demographic data for the groups are shown in Table 1. No statistical differences among the three groups were found in terms of age, sex, APACHE III, RSI, ventilator days before randomization, or PaO2/FIO2. The mean morning cortisol level was significantly higher in the group with adequate adrenal reserve than in the AI groups. In the AI groups, there was no statistical difference between the corticosteroid treatment group and the placebo group in comparisons of morning cortisol levels. The main respiratory failure etiologies and the underlying diseases are shown in Table 2.
Successful Weaning In adequate adrenal reserve group, 20 patients were successfully weaned from ventilator but three patients failed after extubation. In corticosteroid treatment group, 32 patients were successfully weaned, one failed the T-piece trial, and two failed after extubation. In the placebo group, 24 patients were successfully weaned, two failed the T-piece trial, and nine failed after extubation (Table 3). For patients with adequate adrenal reserve and the patients with AI receiving corticosteroid treatment, the success rate for weaning was similar. For patients with AI receiving placebo treatment, the successful weaning rate was significantly lower than in the other two groups (p = 0.035). The results of ICU length of stay, hospital stay, duration of weaning, and hospital mortality in the three groups are presented in Table 3. The results of multivariate logistic regression analysis of the demographic characteristics and underlying diseases of the corticosteroid supplement group and the placebo group are presented in Table 4.
Complications No statistically significant complications, such as new-onset hyperglycemia, nosocomial infection, and gastrointestinal bleeding, were found to be increased in the corticosteroid treatment group.
In this study, stress dose corticosteroid supplementation before extubation of patients with AI led to a significantly higher success rate for VW and a shorter weaning duration than in the placebo group. This is the first study to use morning cortisol level and ACTH stimulation tests to determine adrenal reserve before extubation, thereby demonstrating an association between AI and extubation outcomes. The results suggest that evaluation of adrenal function before weaning is useful in routine ICU practice. Although cortisol is one of the most widely studied hormones, it is probably one of the least understood. The definition of AI remains controversial (3, 5, 23, 26). A cortisol level greater than 25 µg/dl was considered adequate adrenal reserve in many textbooks and recent studies (8, 10). ACTH stimulation is essential to exclude AI, especially when the cortisol level is below 25 µg/dl. The cut-off value of increment of cortisol after ACTH stimulation test remains debatable but an increment of less than 9 µg/dl (250 nmol/L) was considered insufficient in critically ill patients (5, 23). The morning cortisol level and ACTH stimulation test provide a reliable evaluation of adrenal function. In addition to playing a role in septic shock, physiologic corticosteroid replacement may be beneficial in patients with other critical illnesses, including trauma, burns, and medical and surgical conditions where AI is evident (27, 28). To date, there are no published data indicating that stress dose corticosteroids improve the success of VW in the ICU. Our study suggests that patients with adequate adrenal function have a similar success rate of VW to patients with AI receiving stress dose hydrocortisone. It can be claimed that ICUs should routinely test adrenal function before VW. In the present study, hospital mortality, ICU length of stay, and hospital stay did not statistically differ between the hydrocortisone treatment group and the placebo group, although the mean ICU length of stay was shorter in the hydrocortisone treatment group compared with the placebo group. Because such results could reflect the influence of multiple factors on hospital mortality, ICU length of stay, and hospital stay, corrected AI may not produce improvement. The limitation of our study is the small patient number in each group. However, the result inspired us to conduct a larger and multicenter study to prove the present finding. Another limitation of the study is that we cannot explain the result from physiology. A good animal model study is required to clarify the underlying mechanism. Although corticosteroid supplementation has benefited patients with septic shock in recent studies (4), the actual pathophysiology remains unclear. Keh and coworkers (29) reported that low-dose hydrocortisone supplementation increased arterial pressure and systemic vascular resistance, and decreased inotropic agent requirements in patients with septic shock. The increased respiratory work in a VW trial challenges the cardiopulmonary system and HPA axis, potentially leading to hemodynamic instability. Hypoglycemia caused by AI may also contribute to hemodynamic instability. Possibly, low-dose hydrocortisone therapy improves VW success rates by improving hemodynamic stability. A high incidence of AI (> 75%) in patients with severe sepsis was reported by Annane and coworkers (4). The high prevalence (75.3%) of AI in our study was surprising and implied a problem of ignorance. Presentation of AI is nonspecific and asymptomatic. Cooper and Stewart (23) suggested that the threshold for investigating AI should be low because of the limitations of physical examination, particularly in patients with septic shock. What test should be included in routine weaning profiles is still controversial. Manthous and coworkers (30) suggested that most currently used weaning parameters are better at identifying the cause of respiratory failure rather than predicting who will succeed with spontaneous breathing. Shikora and coworkers (31) noted that patients meeting accepted predictive work of breathing criteria frequently require reintubation within 48 h of extubation. Correcting AI may prove an ideal measure of VW.
Conclusions
The authors thank An-Chen Feng for assistance in statistical advice.
Originally Published in Press as DOI: 10.1164/rccm.200504-545OC on November 4, 2005 Conflict of Interest Statement: Neither author has a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Received in original form April 8, 2005; accepted in final form October 31, 2005
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
METHODS
2 test (with Yates' correction) was applied to successful VW results and hospital mortality. Multivariate logistic regression was used to analyze the demographic variables and underlying disease between the corticosteroid group and the placebo group. A p value of 0.05 or less was considered to be statistically significant. 
