Machine learning-based scoring system to predict in-hospital outcomes in patients hospitalized with COVID-19

Assessing the Utility of Prediction Scores PAINT, ISARIC4C, CHIS, and COVID-GRAM at Admission and Seven Days after Symptom Onset for COVID-19 Mortality

The COVID-19 pandemic underscores the need for accurate prognostic tools to predict patient outcomes. This study evaluates the effectiveness of four prominent COVID-19 prediction scores-PAINT, ISARIC4C, CHIS, and COVID-GRAM-at two critical time points: at admission and seven days post-symptom onset, to assess their utility in predicting mortality among hospitalized patients. Conducted at the Clinical Emergency Hospital Pius Brînzeu in Timișoara, this retrospective analysis included adult patients hospitalized with confirmed SARS-CoV-2 infection. Eligible patients had complete data for the scores at both time points. Statistical analysis involved ROC curves and logistic regression to assess the scores' predictive accuracy for mortality. The study included 215 patients, split into 139 survivors and 76 non-survivors. At admission, the PAINT, ISARIC4C, CHIS, and COVID-GRAM scores significantly differentiated between the survival outcomes (p < 0.0001). The best cutoff values at admission were 6.26 for PAINT, 7.95 for ISARIC4C, 5.58 for CHIS, and 0.63 for COVID-GRAM, corresponding to sensitivities of 85.47%, 80.56%, 88.89%, and 83.33% and specificities of 77.34%, 82.12%, 75.01%, and 78.45%, respectively. By day seven, the cutoff values increased, indicating deteriorating conditions in patients who eventually succumbed to the virus. The hazard ratios at admission for exceeding these cutoffs were significant: PAINT (HR = 3.45), ISARIC4C (HR = 2.89), CHIS (HR = 4.02), and COVID-GRAM (HR = 3.15), highlighting the scores' abilities to predict severe outcomes. One week post symptom onset, these scores' predictive values and corresponding hazard ratios increased, further validating their prognostic significance over time. The evaluated COVID-19 prediction scores robustly predict mortality at admission and become more predictive by the seventh day of symptom onset. These findings support the use of these scores in clinical settings to facilitate early identification and intervention for high-risk patients, potentially improving patient outcomes during the ongoing global health crisis.

Predictors of mortality in hospitalised patients with COVID-19: a 1-year case-control study

BACKGROUND

A paucity of predictive models assessing risk factors for COVID-19 mortality that extend beyond age and gender in Latino population is evident in the current academic literature.

OBJECTIVES

To determine the associated factors with mortality, in addition to age and sex during the first year of the pandemic.

DESIGN

A case-control study with retrospective revision of clinical and paraclinical variables by systematic revision of clinical records was conducted. Multiple imputations by chained equation were implemented to account for missing variables. Classification and regression trees (CART) were estimated to evaluate the interaction of associated factors on admission and their role in predicting mortality during hospitalisation. No intervention was performed.

SETTING

High-complexity centre above 2640 m above sea level (masl) in Colombia.

PARTICIPANTS

A population sample of 564 patients admitted to the hospital with confirmed COVID-19 by PCR. Deceased patients (n=282) and a control group (n=282), matched by age, sex and month of admission, were included.

MAIN OUTCOME MEASURE

Mortality during hospitalisation.

MAIN RESULTS

After the imputation of datasets, CART analysis estimated 11 clinical profiles based on respiratory distress, haemoglobin, lactate dehydrogenase, partial pressure of oxygen to inspired partial pressure of oxygen ratio, chronic kidney disease, ferritin, creatinine and leucocytes on admission. The accuracy model for prediction was 80.4% (95% CI 71.8% to 87.3%), with an area under the curve of 78.8% (95% CI 69.63% to 87.93%).

CONCLUSIONS

This study discloses new interactions between clinical and paraclinical features beyond age and sex influencing mortality in COVID-19 patients. Furthermore, the predictive model could offer new clues for the personalised management of this condition in clinical settings.

Sick leave one year after COVID-19 infection: a nationwide cohort study during the first wave in Sweden

This study aimed to investigate the patterns of sick leave, as well as factors associated with sick leave due to COVID-19 during one year after the COVID-19 diagnosis, and sex-related aspects on sick leave. This nationwide study involved 11,902 individuals who received sickness benefits for COVID-19 during the first wave of the pandemic. Data from three Swedish registries were analyzed for sick leave that commenced between March 1 and August 31, 2020, with a follow-up period of 12 months. Sick leave due to COVID-19 was counted as the number of days with sickness benefits and required to include at least one registered COVID-19 diagnosis. The median duration of sick leave was 35 days, and 347 (2.9%) individuals continued their sick leave during the entire follow-up period. Furthermore, 1 year later, the cumulative incidence of sick leave was slightly higher in males (3.5%) compared to females (2.7%). Older age, being single with no children, diagnosed with the virus, medium income level, history of sick leave, and need for inpatient care were significantly associated with a higher duration of sick leave due to COVID-19, both in the total population and when stratified by sex. These results indicated that three out of 100 (3%) patients were still on sick leave 1 year after their COVID-19 diagnosis. Aspects regarding the importance of sick leave duration differed between males and females and comprised sociodemographic characteristics and need for inpatient care. The results indicated the complexity of sick leave due to COVID-19.