tetano
Editor, Senior Moderator
Mayo Clin Proc Innov Qual Outcomes
. 2021 May 12.
doi: 10.1016/j.mayocpiqo.2021.05.001. Online ahead of print.
Personalized Prediction of Hospital Mortality in COVID-19 positive patients
Daniel Rozenbaum[SUP] 1 [/SUP], Jacob Shreve[SUP] 1 [/SUP], Nathan Radakovich[SUP] 2 [/SUP], Abhijit Douggal[SUP] 3 [/SUP], Lara Jehi[SUP] 4 [/SUP], Aziz Nazha[SUP] 1 2 5 [/SUP]
Affiliations
Abstract
Objective: To develop predictive models for in-hospital mortality and length of stay (LOS) for COVID-19 positive patients.
Patients and methods: We performed a multicenter retrospective cohort study of hospitalized COVID-19 positive patients. A total of 764 patients admitted to 14 different hospitals within the Cleveland Clinic from 03/09/2020 to 05/20/2020 who had reverse transcriptase-polymerase chain reaction proven coronavirus infection were included. We used LightGBM, a machine learning algorithm, to predict in-hospital mortality at different time points (after 7 days, 14 days, and 30 days of hospitalization) and in-hospital LOS.
Results: Among 764 patients, 116 (15%) either died (n = 87) or were transitioned to hospice care (n = 29) during their hospitalization. The median LOS was 5 days (range 1 - 44 days) for patients admitted to the regular nursing floor and 10 days (range 1-38 days) for patients admitted to the intensive care unit (ICU). Patients who died during hospitalization were older, initially admitted to the ICU, more likely to be white and to have worse organ dysfunction compared to patients who survived their hospitalization. Using the 10 most important variables only, the final model's area under the Receiver Operating Characteristics curve was 0.86 for 7-day, 0.88 for 14-day, and 0.85 for 30-day mortality in the validation cohort.
Conclusions: We developed a decision tool that can provide explainable and patient-specific prediction of in-hospital mortality and LOS for COVID-19 positive patients. The model can aid healthcare systems in bed allocation and distribution of vital resources.
Keywords: COVID-19; COVID-19, Coronavirus disease 2019; ICU, Intensive care unit; LOS, Length of stay; Personalized prediction; ROC AUC, Area under the Receiver Operating Characteristics curve; mortality.
.
. 2021 May 12.
doi: 10.1016/j.mayocpiqo.2021.05.001. Online ahead of print.
Personalized Prediction of Hospital Mortality in COVID-19 positive patients
Daniel Rozenbaum[SUP] 1 [/SUP], Jacob Shreve[SUP] 1 [/SUP], Nathan Radakovich[SUP] 2 [/SUP], Abhijit Douggal[SUP] 3 [/SUP], Lara Jehi[SUP] 4 [/SUP], Aziz Nazha[SUP] 1 2 5 [/SUP]
Affiliations
- PMID: 34002167
- PMCID: PMC8114764
- DOI: 10.1016/j.mayocpiqo.2021.05.001
Abstract
Objective: To develop predictive models for in-hospital mortality and length of stay (LOS) for COVID-19 positive patients.
Patients and methods: We performed a multicenter retrospective cohort study of hospitalized COVID-19 positive patients. A total of 764 patients admitted to 14 different hospitals within the Cleveland Clinic from 03/09/2020 to 05/20/2020 who had reverse transcriptase-polymerase chain reaction proven coronavirus infection were included. We used LightGBM, a machine learning algorithm, to predict in-hospital mortality at different time points (after 7 days, 14 days, and 30 days of hospitalization) and in-hospital LOS.
Results: Among 764 patients, 116 (15%) either died (n = 87) or were transitioned to hospice care (n = 29) during their hospitalization. The median LOS was 5 days (range 1 - 44 days) for patients admitted to the regular nursing floor and 10 days (range 1-38 days) for patients admitted to the intensive care unit (ICU). Patients who died during hospitalization were older, initially admitted to the ICU, more likely to be white and to have worse organ dysfunction compared to patients who survived their hospitalization. Using the 10 most important variables only, the final model's area under the Receiver Operating Characteristics curve was 0.86 for 7-day, 0.88 for 14-day, and 0.85 for 30-day mortality in the validation cohort.
Conclusions: We developed a decision tool that can provide explainable and patient-specific prediction of in-hospital mortality and LOS for COVID-19 positive patients. The model can aid healthcare systems in bed allocation and distribution of vital resources.
Keywords: COVID-19; COVID-19, Coronavirus disease 2019; ICU, Intensive care unit; LOS, Length of stay; Personalized prediction; ROC AUC, Area under the Receiver Operating Characteristics curve; mortality.
.