Search

Background: Environmental exposures account for a growing proportion of global mortality. Large cohort studies are needed to characterize the independent impact of environmental exposures on mortality in low-income settings. Methods: We collected data on individual and environmental risk factors for a multiethnic cohort of 50,045 individuals in a low-income region in Iran. Environmental risk factors included: ambient fine particular matter air pollution; household fuel use and ventilation; proximity to traffic; distance to percutaneous coronary intervention (PCI) center; socioeconomic environment; population density; local land use; and nighttime light exposure. We developed a spatial survival model to estimate the independent associations between these environmental exposures and all-cause and cardiovascular mortality. Findings: Several environmental factors demonstrated associations with mortality after adjusting for individual risk factors. Ambient fine particulate matter air pollution predicted all-cause mortality (per μg/m3, HR 1.20, 95% CI 1.07, 1.36) and cardiovascular mortality (HR 1.17, 95% CI 0.98, 1.39). Biomass fuel use without chimney predicted all-cause mortality (reference = gas, HR 1.23, 95% CI 0.99, 1.53) and cardiovascular mortality (HR 1.36, 95% CI 0.99, 1.87). Kerosene fuel use without chimney predicted all-cause mortality (reference = gas, HR 1.09, 95% CI 0.97, 1.23) and cardiovascular mortality (HR 1.19, 95% CI 1.01, 1.41). Distance to PCI center predicted all-cause mortality (per 10km, HR 1.01, 95% CI 1.004, 1.022) and cardiovascular mortality (HR 1.02, 95% CI 1.004, 1.031). Additionally, proximity to traffic predicted all-cause mortality (HR 1.13, 95% CI 1.01, 1.27). In a separate validation cohort, the multivariable model effectively predicted both all-cause mortality (AUC 0.76) and cardiovascular mortality (AUC 0.81). Population attributable fractions demonstrated a high mortality burden attributable to environmental exposures. Interpretation: Several environmental factors predicted cardiovascular and all-cause mortality, independent of each other and of individual risk factors. Mortality attributable to environmental factors represents a critical opportunity for targeted policies and programs. [ABSTRACT FROM AUTHOR]

Machine learning algorithms are increasingly used in the clinical literature, claiming advantages over logistic regression. However, they are generally designed to maximize the area under the receiver operating characteristic curve. While area under the receiver operating characteristic curve and other measures of accuracy are commonly reported for evaluating binary prediction problems, these metrics can be misleading. We aim to give clinical and machine learning researchers a realistic medical example of the dangers of relying on a single measure of discriminatory performance to evaluate binary prediction questions. Prediction of medical complications after surgery is a frequent but challenging task because many post-surgery outcomes are rare. We predicted post-surgery mortality among patients in a clinical registry who received at least one aortic valve replacement. Estimation incorporated multiple evaluation metrics and algorithms typically regarded as performing well with rare outcomes, as well as an ensemble and a new extension of the lasso for multiple unordered treatments. Results demonstrated high accuracy for all algorithms with moderate measures of cross-validated area under the receiver operating characteristic curve. False positive rates were < 1%, however, true positive rates were < 7%, even when paired with a 100% positive predictive value, and graphical representations of calibration were poor. Similar results were seen in simulations, with the addition of high area under the receiver operating characteristic curve (> 90%) accompanying low true positive rates. Clinical studies should not primarily report only area under the receiver operating characteristic curve or accuracy. [ABSTRACT FROM AUTHOR]

Objective: To characterize the hospitalization and death rates among patients with inflammatory arthritis (IA) affected by coronavirus disease 2019 (COVID‐19) and to analyze the associations of comorbidities and immunomodulatory medications with infection outcomes. Methods: Data on clinical and demographic features, maintenance treatment, disease course, and outcomes in individuals with IA (rheumatoid arthritis and spondyloarthritis) with symptomatic COVID‐19 infection were prospectively assessed via web‐based questionnaire followed by individual phone calls and electronic medical record review. Baseline characteristics and medication use were summarized for hospitalized and ambulatory patients, and outcomes with the different medication classes were compared using multivariable logistic regression. Results: A total of 103 patients with IA were included in the study (80 with confirmed COVID‐19 and 23 with high suspicion of COVID‐19). Hospitalization was required in 26% of the participants, and 4% died. Patients who were hospitalized were significantly more likely to be older (P < 0.001) and have comorbid hypertension (P = 0.001) and chronic obstructive pulmonary disease (P = 0.02). IA patients taking oral glucocorticoids had an increased likelihood of being admitted for COVID‐19 (P < 0.001), while those receiving maintenance anticytokine biologic therapies did not. Conclusion: Among patients with underlying IA, COVID‐19 outcomes were worse in those receiving glucocorticoids but not in patients receiving maintenance anticytokine therapy. Further work is needed to understand whether immunomodulatory therapies affect COVID‐19 incidence. [ABSTRACT FROM AUTHOR]