Postoperative Pulmonary Complications in the ENIGMA II Trial: A Post Hoc Analysis.

BACKGROUND: Nitrous oxide promotes absorption atelectasis in poorly ventilated lung segments at high inspired concentrations. The Evaluation of Nitrous oxide In the Gas Mixture for Anesthesia (ENIGMA) trial found a higher incidence of postoperative pulmonary complications and wound sepsis with nitrous oxide anesthesia in major surgery compared to a fraction of inspired oxygen of 0.8 without nitrous oxide. The larger ENIGMA II trial randomized patients to nitrous oxide or air at a fraction of inspired oxygen of 0.3 but found no effect on wound infection or sepsis. However, postoperative pulmonary complications were not measured. In the current study, post hoc data were collected to determine whether atelectasis and pneumonia incidences were higher with nitrous oxide in patients who were recruited to the Australian cohort of ENIGMA II. METHODS: Digital health records of patients who participated in the trial at 10 Australian hospitals were examined blinded to trial treatment allocation. The primary endpoint was the incidence of atelectasis, defined as lung atelectasis or collapse reported on chest radiology. Pneumonia, as a secondary endpoint, required a diagnostic chest radiology report with fever, leukocytosis, or positive sputum culture. Comparison of the nitrous oxide and nitrous oxide-free groups was done according to intention to treat using chi-square tests. RESULTS: Data from 2,328 randomized patients were included in the final data set. The two treatment groups were similar in surgical type and duration, risk factors, and perioperative management recorded for ENIGMA II. There was a 19.3% lower incidence of atelectasis with nitrous oxide (171 of 1,169 [14.6%] vs. 210 of 1,159 [18.1%]; odds ratio, 0.77; 95% CI, 0.62 to 0.97; P = 0.023). There was no difference in pneumonia incidence (60 of 1,169 [5.1%] vs. 52 of 1159 [4.5%]; odds ratio, 1.15; 95% CI, 0.77 to 1.72; P = 0.467) or combined pulmonary complications (odds ratio, 0.84; 95% CI, 0.69 to 1.03; P = 0.093). CONCLUSIONS: In contrast to the earlier ENIGMA trial, nitrous oxide anesthesia in the ENIGMA II trial was associated with a lower incidence of lung atelectasis, but not pneumonia, after major surgery.

Stan and BART for Causal Inference: Estimating Heterogeneous Treatment Effects Using the Power of Stan and the Flexibility of Machine Learning.

A wide range of machine-learning-based approaches have been developed in the past decade, increasing our ability to accurately model nonlinear and nonadditive response surfaces. This has improved performance for inferential tasks such as estimating average treatment effects in situations where standard parametric models may not fit the data well. These methods have also shown promise for the related task of identifying heterogeneous treatment effects. However, the estimation of both overall and heterogeneous treatment effects can be hampered when data are structured within groups if we fail to correctly model the dependence between observations. Most machine learning methods do not readily accommodate such structure. This paper introduces a new algorithm, stan4bart, that combines the flexibility of Bayesian Additive Regression Trees (BART) for fitting nonlinear response surfaces with the computational and statistical efficiencies of using Stan for the parametric components of the model. We demonstrate how stan4bart can be used to estimate average, subgroup, and individual-level treatment effects with stronger performance than other flexible approaches that ignore the multilevel structure of the data as well as multilevel approaches that have strict parametric forms.