Improving prediction of COVID-19 mortality using machine learning in the Spanish SEMI-COVID-19 registry

COVID-19 is responsible for high mortality, but robust machine learning-based predictors of mortality are lacking. To generate a model for predicting mortality in patients hospitalized with COVID-19 using Gradient Boosting Decision Trees (GBDT). The Spanish SEMI-COVID-19 registry includes 24,514 pseudo-anonymized cases of patients hospitalized with COVID-19 from 1 February 2020 to 5 December 2021. This registry was used as a GBDT machine learning model, employing the CatBoost and BorutaShap classifier to select the most relevant indicators and generate a mortality prediction model by risk level, ranging from 0 to 1. The model was validated by separating patients according to admission date, using the period 1 February to 31 December 2020 (first and second waves, pre-vaccination period) for training, and 1 January to 30 November 2021 (vaccination period) for the test group. An ensemble of ten models with different random seeds was constructed, separating 80% of the patients for training and 20% from the end of the training period for cross-validation. The area under the receiver operating characteristics curve (AUC) was used as a performance metric. Clinical and laboratory data from 23,983 patients were analyzed. CatBoost mortality prediction models achieved an AUC performance of 84.76 (standard deviation 0.45) for patients in the test group (potentially vaccinated patients not included in model training) using 16 features. The performance of the 16-parameter GBDT model for predicting COVID-19 hospital mortality, although requiring a relatively large number of predictors, shows a high predictive capacity.

Paid employment in subjects with and without chronic obstructive pulmonary disease in five Latin American cities: the PLATINO study

BACKGROUND

Chronic obstructive pulmonary disease (COPD) is a costly condition that frequently causes permanent work disabilities. Little information exists regarding the impact of COPD on work force participation and the indirect costs of the disease in developing countries.

OBJECTIVE

To examine the frequency of paid employment and factors influencing it in a Latin-American population-based study.

METHODS

Post-bronchodilator FEV(1)/FVC < 0.70 (forced expiratory volume in 1 s/forced vital capacity) was used to define COPD. Information regarding paid work was assessed by the question 'At any time in the past year, have you worked for payment?'

RESULTS

Interviews were conducted with 5571 subjects; 5314 (759 COPD and 4554 non-COPD) subjects underwent spirometry. Among the COPD subjects, 41.8% reported having paid work vs. 57.1% of non-COPD (P < 0.0001). The number of months with paid work was reduced in COPD patients (10.5 ± 0.17 vs. 10.9 ± 0.06, P < 0.05). The main factors associated with having paid work in COPD patients were male sex (OR 0.33, 95%CI 0.23-0.47), higher education level (OR 1.05, 95%CI 1.01-1.09) and younger age (OR 0.90, 95%CI 0.88-0.92). COPD was not a significant contributor to employment (OR 0.83, 95%CI 0.69-1.00, P = 0.054) in the entire population.

CONCLUSIONS

Although the proportion of persons with paid work is lower in COPD, having COPD appears not to have a significant impact on obtaining paid employment in the overall population of developing countries.

Pericardial tamponade in a patient with polymyalgia rheumatica

We report a patient who developed pericarditis and pericardial tamponade coinciding with polymyalgia rheumatica onset. Our patient did not show any clinical sign of vasculitis; temporal artery biopsies were negative for giant cell arteritis. Pericardial biopsy in our case shows inflammatory perivascular lymphocytary infiltrates thus we believe pericardial effusion has an inflammatory-immunologic origin. Cardiac manifestations are exceptional in polymyalgia rheumatica, though it should be considered in the differential diagnosis in patients with pericarditis over 50 years. The recognition of this uncommon manifestation is very important due to the good response to corticosteroid treatment.

Worldwide burden of COPD in high- and low-income countries. Part II. Burden of chronic obstructive lung disease in Latin America: the PLATINO study

SETTING

Five Latin American cities: São Paulo, Brazil; Mexico City, Mexico; Montevideo, Uruguay; Santiago, Chile; Caracas, Venezuela.

OBJECTIVE

To describe the burden of chronic obstructive pulmonary disease (COPD) in Latin America.

DESIGN

This is a multicentre study. Post-bronchodilator spirometry was used and the main outcome measure was FEV(1)/FVC < 0.7 (fixed ratio criterion). Global Obstructive Lung Disease (GOLD) stages were also analysed.

RESULTS

The combined population aged > or =40 years in the five countries included in the study was approximately 85.3 million. Of these, it was estimated that 12.2 million have airflow obstruction, which corresponds to our prevalence estimate of 14.3%. The proportion of subjects in Stages II-IV of the GOLD classification was 5.6%. Risk factors presenting the highest aetiological fractions for COPD were age, current smoking, indoor exposure to coal and exposure to dust in the workplace. Smoking, the modifiable factor with the strongest aetiological fraction for COPD, affects 29.2% of adults aged > or =40 years in these cities, corresponding to approximately 25 million smokers in this age group.

CONCLUSION

Prevention of smoking and exposure to pollutants, such as coal and dust, are the interventions most likely to succeed against COPD in Latin America. The information obtained by a collaborative study has been vast and encouraging for other similar studies.

Tuberculosis and airflow obstruction: evidence from the PLATINO study in Latin America

The aim of the present study was to evaluate the association between history of tuberculosis and airflow obstruction. A population-based, multicentre study was carried out and included 5,571 subjects aged > or =40 yrs living in one of five Latin American metropolitan areas: Sao Paulo (Brazil); Montevideo (Uruguay); Mexico City (Mexico); Santiago (Chile); and Caracas (Venezuela). Subjects performed pre- and post-bronchodilator spirometry and were asked whether they had ever been diagnosed with tuberculosis by a physician. The overall prevalence of airflow obstruction (forced expiratory volume in one second/forced vital capacity post-bronchodilator <0.7) was 30.7% among those with a history of tuberculosis, compared with 13.9% among those without a history. Males with a medical history of tuberculosis were 4.1 times more likely to present airflow obstruction than those without such a diagnosis. This remained unchanged after adjustment for confounding by age, sex, schooling, ethnicity, smoking, exposure to dust and smoke, respiratory morbidity in childhood and current morbidity. Among females, the unadjusted and adjusted odds ratios were 2.3 and 1.7, respectively. In conclusion, history of tuberculosis is associated with airflow obstruction in Latin American middle-aged and older adults.

Prevalence of oxygen desaturation and use of oxygen at home in adults at sea level and at moderate altitude

The aim of this study was to determine the prevalence of oxygen desaturation in adults aged>or=40 yrs as altitude above sea level increases. A population-based, cross-sectional study with a multistage cluster sampling of 1,063 subjects from metropolitan Mexico City (Mexico; 2,240 m above sea level), 1,357 from Caracas (Venezuela; 950 m) and 943 from Montevideo (Uruguay; 35 m). The mean of six measurements of arterial oxygen saturation (SP,O2) was estimated using a pulse oximeter. Mean SP,O2 decreased with altitude. No subject from Montevideo had a mean SP,O2<or=88%. Prevalence of subjects with SP,O2<or=88%, a common criterion used for oxygen prescription, was 1.0% in Caracas and 6.0% in Mexico City. Additional predicting factors for hypoxaemia in multiple regression models were age, body mass index and a low forced expiratory volume in one second as a percentage of the predicted value. In Mexico City and Caracas, <10% of the hypoxaemic subjects reported use of domiciliary oxygen, whereas at least half of the subjects reporting use of oxygen at home had a resting SP,O2>88%. In conclusion, the prevalence of hypoxaemia was closely related to altitude. Priorities for oxygen prescription must be defined in moderate altitudes because it is unfeasible to provide it to all subjects fulfilling the criteria commonly used.