HOUSTON – Prehospital use of systemic corticosteroids does not prevent the development of acute lung injury in at-risk patients, according to data reported at the annual congress of the Society of Critical Care Medicine.
In the first study to specifically evaluate the prophylactic value of prehospital systemic corticosteroids in patients with at least one risk factor for acute lung injury (ALI), Dr. Lioudmila Karnatovskaia of the Mayo Clinic in Jacksonville, Fla., and colleagues found a statistically similar incidence of ALI among at-risk patients who were and were not taking systemic corticosteroids at the time of hospitalization.
The investigators also determined that prehospital use of systemic corticosteroids did not affect the need for mechanical ventilation or overall mortality – a finding that appears to contradict previous studies that have linked preventive steroids in at-risk patients with increased rates of ALI and acute respiratory distress syndrome, Dr. Karnatovskaia said.
The study was a planned exploratory subgroup analysis of the Lung Injury Prediction Score cohort of the U.S. Critical Illness and Injury Trials Group, which prospectively enrolled 5,584 patients who were admitted to 22 acute care hospitals and who had predisposing conditions for ALI, including sepsis, shock, pancreatitis, pneumonia, aspiration, high-risk trauma, and high-risk surgery. The primary outcome was the development of ALI, and secondary outcomes were need for invasive ventilation and ICU and hospital mortality, Dr. Karnatovskaia said, noting that the data were analyzed using univariate, logistic regression, and propensity score–based analyses.
For the propensity analysis, "the propensity score balanced all of the covariates. Of the 458 patients on systemic corticosteroids, 443 were matched up to 1:4 to those not on systemic corticosteroids, for a total of 1,332 matched patients," she said. "We calculated adjusted risk for acute lung injury, invasive ventilation, and in-hospital mortality from the propensity score–matched sample using a conditional logistic regression model."
Of the 5,584 patients, 458 were on systemic corticosteroids at the time of hospitalization and 5,126 were not. Among the systemic corticosteroid group, 34 (7.4%) developed ALI, compared with 343 (6.7%) of those not taking them, Dr. Karnatovskaia reported. In the systemic corticosteroid group, 104 patients (23%) required mechanical ventilation and 35 patients (8%) died, compared with 1,752 (34%) and 172 (3%) of those not taking systemic corticosteroids, she said.
On univariate analysis, systemic corticosteroid patients were more likely to be older, to be white, and to have diabetes, chronic obstructive pulmonary disease, malignancy, or previous chest radiation, Dr. Karnatovskaia said, noting that they were also more likely to have a lower body mass index and to be on a statin drug, inhaled steroid, inhaled beta-agonist, proton pump inhibitor, ACE inhibitor, angiotensin receptor blocker, or insulin and were less like likely to abuse alcohol or smoke tobacco.
After adjustment for significant covariates, systemic corticosteroid use was not independently associated with the development of ALI or the need for invasive ventilation, but did appear to be an independent predictor of ICU and hospital mortality, Dr. Karnatovskaia said. The latter association fell away, however, in the propensity score–based analysis. "Following propensity score–based analysis with matching, the association of prehospital systemic corticosteroids with mortality no longer remained significant," she said.
The findings are limited by the lack of data on the indication for systemic corticosteroid therapy, its duration, "and even whether it was continued throughout the hospital stay," as well as the fact that patients on prehospital systemic corticosteroids appeared to have worse functional status, which might have influenced their outcomes, according to Dr. Karnatovskaia. Although using the propensity score with matching addressed this as well as other hidden biases, "the potential for unmeasured effects remains," she said.
The study’s strengths include the large number of patients at risk for ALI enrolled from different centers and regions in the United States, as well as two hospitals in Turkey, and the use of comprehensive propensity score–based analysis with matching in addition to traditional logistic regression, Dr. Karnatovskaia said.
Ideally, the finding that prehospital use of systemic corticosteroids does not mitigate the development of ALI would be validated in a randomized controlled trial to best address any causal relationship, "but such a study would not be practical," Dr. Karnatovskaia said.
Dr. Karnatovskaia reported having no relevant financial disclosures.