Potential Pathological, Clinical, and Symptomatic Findings of COVID-19 to Predict Mortality in Positive PCR Individuals Using Data Mining

Introduction:COVID-19 has placed immense burdens on healthcare systems and medical staff. To avoid spread, the statistician’s role and the use of appropriate predictive models -prediction of survivors versus non-survivors- is highly relevant. This study aimed to apply a model which avoids overfittin...

Full description

Saved in:
Bibliographic Details
Main Authors: Nasrin Talkhi, Nooshin Akbari sharak, Zahra Pasdar, Maryam Salari, Seyed Masoud Sadati, Mohammad Taghi Shakeri
Format: Article
Language:English
Published: Mashhad University of Medical Sciences 2023-01-01
Series:Patient Safety and Quality Improvement Journal
Subjects:
Online Access:https://psj.mums.ac.ir/article_22058_9896fceae375f67ab557abee509bcab5.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849245723405582336
author Nasrin Talkhi
Nooshin Akbari sharak
Zahra Pasdar
Maryam Salari
Seyed Masoud Sadati
Mohammad Taghi Shakeri
author_facet Nasrin Talkhi
Nooshin Akbari sharak
Zahra Pasdar
Maryam Salari
Seyed Masoud Sadati
Mohammad Taghi Shakeri
author_sort Nasrin Talkhi
collection DOAJ
description Introduction:COVID-19 has placed immense burdens on healthcare systems and medical staff. To avoid spread, the statistician’s role and the use of appropriate predictive models -prediction of survivors versus non-survivors- is highly relevant. This study aimed to apply a model which avoids overfitting and selection bias towards selecting predictors to predict COVID-19 mortality. Materials and Methods: The Conditional Inference Tree (CIT) model was used. Data from 59,564 hospitalized individuals with positive polymerase chain reaction (PCR) test results were collected from February 20, 2020, to September 12, 2021, in the Khorasan Razavi province, Iran. Results: The sensitivity and specificity of the model were 88.7% and 88.1%, respectively, the accuracy was 88.2%, and the area under the curve (AUC) was 73.0% on test data. Therefore, the model had considerable accuracy in prediction. The potential predictors involved in predicting survivors versus non-survivors were intubation, age, PO2 level, decreased consciousness level, presence of distress, anorexia, drug use, and kidney diseases. Conclusion:According to the findings, the CIT model showed high accuracy by avoiding overfitting and selection bias toward selecting predictors. Thus, the results of this study and the efforts of healthcare systems to stop the spread of this pandemic prove helpful.
format Article
id doaj-art-1b04acaa467d4118997cc7cc918af99a
institution Kabale University
issn 2345-4482
2345-4490
language English
publishDate 2023-01-01
publisher Mashhad University of Medical Sciences
record_format Article
series Patient Safety and Quality Improvement Journal
spelling doaj-art-1b04acaa467d4118997cc7cc918af99a2025-08-20T03:58:43ZengMashhad University of Medical SciencesPatient Safety and Quality Improvement Journal2345-44822345-44902023-01-01111132110.22038/psj.2023.70741.139022058Potential Pathological, Clinical, and Symptomatic Findings of COVID-19 to Predict Mortality in Positive PCR Individuals Using Data MiningNasrin Talkhi0Nooshin Akbari sharak1Zahra Pasdar2Maryam Salari3Seyed Masoud Sadati4Mohammad Taghi Shakeri5Department of Biostatistics, School of Health, Mashhad University of Medical Sciences, Mashhad, Iran.Department of Biostatistics, School of Health, Mashhad University of Medical Sciences, Mashhad, Iran.Institute of Applied Health Sciences, School of Medicine, Medical Sciences and Nutrition, University of Aberdeen.Assistant Professor in Biostatistics, Expert Management and Information Technology, Mashhad University of Medical Sciences, Mashhad, Iran.Center of Statistics and Information Technology Management, Imam Reza Hospital, Mashhad University of Medical Sciences, Mashhad, Iran.Social Determinants of Health Research Center, Mashhad University of Medical Sciences, Mashhad, Iran.Introduction:COVID-19 has placed immense burdens on healthcare systems and medical staff. To avoid spread, the statistician’s role and the use of appropriate predictive models -prediction of survivors versus non-survivors- is highly relevant. This study aimed to apply a model which avoids overfitting and selection bias towards selecting predictors to predict COVID-19 mortality. Materials and Methods: The Conditional Inference Tree (CIT) model was used. Data from 59,564 hospitalized individuals with positive polymerase chain reaction (PCR) test results were collected from February 20, 2020, to September 12, 2021, in the Khorasan Razavi province, Iran. Results: The sensitivity and specificity of the model were 88.7% and 88.1%, respectively, the accuracy was 88.2%, and the area under the curve (AUC) was 73.0% on test data. Therefore, the model had considerable accuracy in prediction. The potential predictors involved in predicting survivors versus non-survivors were intubation, age, PO2 level, decreased consciousness level, presence of distress, anorexia, drug use, and kidney diseases. Conclusion:According to the findings, the CIT model showed high accuracy by avoiding overfitting and selection bias toward selecting predictors. Thus, the results of this study and the efforts of healthcare systems to stop the spread of this pandemic prove helpful.https://psj.mums.ac.ir/article_22058_9896fceae375f67ab557abee509bcab5.pdfconditional inference treecovid-19data miningdecision treesmachine learning
spellingShingle Nasrin Talkhi
Nooshin Akbari sharak
Zahra Pasdar
Maryam Salari
Seyed Masoud Sadati
Mohammad Taghi Shakeri
Potential Pathological, Clinical, and Symptomatic Findings of COVID-19 to Predict Mortality in Positive PCR Individuals Using Data Mining
Patient Safety and Quality Improvement Journal
conditional inference tree
covid-19
data mining
decision trees
machine learning
title Potential Pathological, Clinical, and Symptomatic Findings of COVID-19 to Predict Mortality in Positive PCR Individuals Using Data Mining
title_full Potential Pathological, Clinical, and Symptomatic Findings of COVID-19 to Predict Mortality in Positive PCR Individuals Using Data Mining
title_fullStr Potential Pathological, Clinical, and Symptomatic Findings of COVID-19 to Predict Mortality in Positive PCR Individuals Using Data Mining
title_full_unstemmed Potential Pathological, Clinical, and Symptomatic Findings of COVID-19 to Predict Mortality in Positive PCR Individuals Using Data Mining
title_short Potential Pathological, Clinical, and Symptomatic Findings of COVID-19 to Predict Mortality in Positive PCR Individuals Using Data Mining
title_sort potential pathological clinical and symptomatic findings of covid 19 to predict mortality in positive pcr individuals using data mining
topic conditional inference tree
covid-19
data mining
decision trees
machine learning
url https://psj.mums.ac.ir/article_22058_9896fceae375f67ab557abee509bcab5.pdf
work_keys_str_mv AT nasrintalkhi potentialpathologicalclinicalandsymptomaticfindingsofcovid19topredictmortalityinpositivepcrindividualsusingdatamining
AT nooshinakbarisharak potentialpathologicalclinicalandsymptomaticfindingsofcovid19topredictmortalityinpositivepcrindividualsusingdatamining
AT zahrapasdar potentialpathologicalclinicalandsymptomaticfindingsofcovid19topredictmortalityinpositivepcrindividualsusingdatamining
AT maryamsalari potentialpathologicalclinicalandsymptomaticfindingsofcovid19topredictmortalityinpositivepcrindividualsusingdatamining
AT seyedmasoudsadati potentialpathologicalclinicalandsymptomaticfindingsofcovid19topredictmortalityinpositivepcrindividualsusingdatamining
AT mohammadtaghishakeri potentialpathologicalclinicalandsymptomaticfindingsofcovid19topredictmortalityinpositivepcrindividualsusingdatamining