Predicting mortality in SARS-COV-2 (COVID-19) positive patients in the inpatient setting using a Novel Deep Neural Network
Maleeha Naseem, Hajra Arshad, Syeda Amrah Hashmi, Furqan Irfan, Lead Scientist, Fahad Shabbir Ahmed
doi:10.1101/2020.12.13.20247254
Background The second wave of COVID-19 pandemic is anticipated to be worse than the initial one and will strain the healthcare systems even more during the winter months. Our aim was to develop a machine learning-based model to predict mortality using the deep learning Neo-V framework. We hypothesized this novel machine learning approach could be applied to COVID-19 patients to predict mortality successfully with high accuracy.
Methods The current Deep-Neo-V model is built on our previously statistically rigorous machine learning framework [Fahad-Liaqat-Ahmad Intensive Machine (FLAIM) framework] that evaluated statistically significant risk factors, generated new combined variables and then supply these risk factors to deep neural network to predict mortality in RT-PCR positive COVID-19 patients in the inpatient setting. We analyzed adult patients (≥18 years) admitted to the Aga Khan University Hospital, Pakistan with a working diagnosis of COVID-19 infection (n=1228). We excluded patients that were negative on COVID-19 on RT-PCR, had incomplete or missing health records. The first phase selection of risk factor was done using Cox-regression univariate and multivariate analyses. In the second phase, we generated new variables and tested those statistically significant for mortality and in the third and final phase we applied deep neural networks and other traditional machine learning models like Decision Tree Model, k-nearest neighbor models and others.
Results A total of 1228 cases were diagnosed as COVID-19 infection, we excluded 14 patients after the exclusion criteria and (n=)1214 patients were analyzed. We observed that several clinical and laboratory-based variables were statistically significant for both univariate and multivariate analyses while others were not. With most significant being septic shock (hazard All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
References
Ahmad, Ali, Ul, Khattak, Hameed et al., A hybrid machine learning framework to predict mortality in paralytic ileus patients using electronic health records (EHRs), Journal of Ambient Intelligence and Humanized Computing
Ahmed, Ali, Joseph, Ikram, Mustafa et al., A statistically rigorous deep neural network approach to predict mortality in trauma patients admitted to the intensive care unit, J Trauma Acute Care Surg
Ahmed, Raza-Ul-Mustafa, Ali, Imad-Ud-Deen, Hameed et al., Machine Learning Can Predict Deaths in Patients with Diverticulitis During their Hospital Stay, medRxiv
Altschul, Unda, Benton, De, Ramos et al., A novel severity score to predict inpatient mortality in COVID-19 patients, Sci Rep
Becher, Frerichs, Mortality in COVID-19 is not merely a question of resource availability, The Lancet Respiratory medicine
Cacciapaglia, Cot, Sannino, Second wave COVID-19 pandemics in Europe: a temporal playbook, Sci Rep
Cdc) Cfdcap, CDC COVID Data Tracker: Maps, charts, and data provided by the CDC
Chawla, Bowyer, Hall, Kegelmeyer, SMOTE: Synthetic Minority Over-sampling Technique, J Artif Int Res
Cheng, Luo, Wang, Zhang, Wang et al., Kidney disease is associated with in-hospital death of patients with COVID-19, Kidney Int
Dong, Du, Gardner, An interactive web-based dashboard to track COVID-19 in real time, Lancet Infect Dis
Fda) Fda, and ensure the safety, effectiveness and quality of COVID-19 vaccines
Hendren, Drazner, Bozkurt, Cooper, Description and Proposed Management of the Acute COVID-19 Cardiovascular Syndrome, Circulation
Henry, De Oliveira, Benoit, Plebani, Lippi, Hematologic, biochemical and immune biomarker abnormalities associated with severe illness and mortality in coronavirus disease 2019 (COVID-19): a meta-analysis, Clin Chem Lab Med
Hsiang, Allen, Annan-Phan, Bell, Bolliger et al., The effect of large-scale anti-contagion policies on the COVID-19 pandemic, Nature
Hultström, Persson, Eriksson, Lipcsey, Frithiof et al., Blood type A associates with critical COVID-19 and death in a Swedish cohort, Critical Care
Hunter, Trying to "Protect the NHS" in the United Kingdom, New England Journal of Medicine
Keni, Alexander, Nayak, Mudgal, Nandakumar, COVID-19: Emergence, Spread, Possible Treatments, and Global Burden, Front Public Health
Leung, Risk factors for predicting mortality in elderly patients with COVID-19: A review of clinical data in China, Mech Ageing Dev
Li, Ma, Acute respiratory failure in COVID-19: is it "typical" ARDS?, Crit Care
Motwani, Dey, Berman, Germano, Achenbach et al., Machine learning for prediction of all-cause mortality in patients with suspected coronary artery disease: a 5-year multicentre prospective registry analysis, Eur Heart J
Naseem, Akhund, Arshad, Ibrahim, Exploring the Potential of Artificial Intelligence and Machine Learning to Combat COVID-19 and Existing Opportunities for LMIC: A Scoping Review, J Prim Care Community Health
Ozturk, Turgutalp, Arici, Odabas, Altiparmak et al., Mortality analysis of COVID-19 infection in chronic kidney disease, haemodialysis and renal transplant patients compared with patients without kidney disease: a nationwide analysis from Turkey, Nephrol Dial Transplant
Parikh, Manz, Chivers, Regli, Braun et al., Machine Learning Approaches to Predict 6-Month Mortality Among Patients With Cancer, JAMA Netw Open
Pranata, Soeroto, Huang, Lim, Santoso et al., Effect of chronic obstructive pulmonary disease and smoking on the outcome of COVID-19, Int J Tuberc Lung Dis
Rahim, Amin, Noor, Bahadur, Gul et al., Mortality of Patients With Severe COVID-19 in the Intensive Care Unit: An Observational Study From a Major COVID-19 Receiving Hospital, Cureus
Shi, Wang, Wang, Duan, Yang, Dyspnea rather than fever is a risk factor for predicting mortality in patients with COVID-19, J Infect
Sterne, Murthy, Diaz, Slutsky, Villar et al., Association Between Administration of Systemic Corticosteroids and Mortality Among Critically Ill Patients With COVID-19: A Meta-analysis, Jama
Wang, Duan, Han, Liu, Zhou et al., High incidence and mortality of pneumothorax in critically Ill patients with COVID-19, Heart Lung
Xu, Li, Beware of the second wave of COVID-19, Lancet
Yataco, Simpson, Coronavirus Disease 2019 Sepsis: A Nudge Toward Antibiotic Stewardship, Chest
Zhao, Chen, Hou, Graham, Li et al., Prediction model and risk scores of ICU admission and mortality in COVID-19, PLoS One
Zhu, Ge, Jiang, Zhang, Li et al., Deep-learning artificial intelligence analysis of clinical variables predicts mortality in COVID-19 patients, J Am Coll Emerg Physicians Open
DOI record:
{
"DOI": "10.1101/2020.12.13.20247254",
"URL": "http://dx.doi.org/10.1101/2020.12.13.20247254",
"abstract": "<jats:title>ABSTRACT</jats:title><jats:sec><jats:title>Background</jats:title><jats:p>The second wave of COVID-19 pandemic is anticipated to be worse than the initial one and will strain the healthcare systems even more during the winter months. Our aim was to develop a machine learning-based model to predict mortality using the deep learning Neo-V framework. We hypothesized this novel machine learning approach could be applied to COVID-19 patients to predict mortality successfully with high accuracy.</jats:p></jats:sec><jats:sec><jats:title>Methods</jats:title><jats:p>The current Deep-Neo-V model is built on our previously statistically rigorous machine learning framework [Fahad-Liaqat-Ahmad Intensive Machine (FLAIM) framework] that evaluated statistically significant risk factors, generated new combined variables and then supply these risk factors to deep neural network to predict mortality in RT-PCR positive COVID-19 patients in the inpatient setting. We analyzed adult patients (≥18 years) admitted to the Aga Khan University Hospital, Pakistan with a working diagnosis of COVID-19 infection (n=1228). We excluded patients that were negative on COVID-19 on RT-PCR, had incomplete or missing health records. The first phase selection of risk factor was done using Cox-regression univariate and multivariate analyses. In the second phase, we generated new variables and tested those statistically significant for mortality and in the third and final phase we applied deep neural networks and other traditional machine learning models like Decision Tree Model, k-nearest neighbor models and others.</jats:p></jats:sec><jats:sec><jats:title>Results</jats:title><jats:p>A total of 1228 cases were diagnosed as COVID-19 infection, we excluded 14 patients after the exclusion criteria and (n=)1214 patients were analyzed. We observed that several clinical and laboratory-based variables were statistically significant for both univariate and multivariate analyses while others were not. With most significant being septic shock (hazard ratio [HR], 4.30; 95% confidence interval [CI], 2.91-6.37), supportive treatment (HR, 3.51; 95% CI, 2.01-6.14), abnormal international normalized ratio (INR) (HR, 3.24; 95% CI, 2.28-4.63), admission to the intensive care unit (ICU) (HR, 3.24; 95% CI, 2.22-4.74), treatment with invasive ventilation (HR, 3.21; 95% CI, 2.15-4.79) and laboratory lymphocytic derangement (HR, 2.79; 95% CI, 1.6-4.86). Machine learning results showed our DNN (Neo-V) model outperformed all conventional machine learning models with test set accuracy of 99.53%, sensitivity of 89.87%, and specificity of 95.63%; positive predictive value, 50.00%; negative predictive value, 91.05%; and area under the curve of the receiver-operator curve of 88.5.</jats:p></jats:sec><jats:sec><jats:title>Conclusion</jats:title><jats:p>Our novel Deep-Neo-V model outperformed all other machine learning models. The model is easy to implement, user friendly and with high accuracy.</jats:p></jats:sec>",
"accepted": {
"date-parts": [
[
2021,
8,
29
]
]
},
"author": [
{
"affiliation": [],
"family": "Naseem",
"given": "Maleeha",
"sequence": "first"
},
{
"affiliation": [],
"family": "Arshad",
"given": "Hajra",
"sequence": "additional"
},
{
"affiliation": [],
"family": "Hashmi",
"given": "Syeda Amrah",
"sequence": "additional"
},
{
"affiliation": [],
"family": "Irfan",
"given": "Furqan",
"sequence": "additional"
},
{
"ORCID": "http://orcid.org/0000-0002-5857-5478",
"affiliation": [],
"authenticated-orcid": false,
"family": "Ahmed",
"given": "Fahad Shabbir",
"sequence": "additional"
}
],
"container-title": [],
"content-domain": {
"crossmark-restriction": false,
"domain": []
},
"created": {
"date-parts": [
[
2020,
12,
17
]
],
"date-time": "2020-12-17T05:47:20Z",
"timestamp": 1608184040000
},
"deposited": {
"date-parts": [
[
2021,
12,
15
]
],
"date-time": "2021-12-15T15:44:27Z",
"timestamp": 1639583067000
},
"group-title": "Infectious Diseases (except HIV/AIDS)",
"indexed": {
"date-parts": [
[
2024,
3,
1
]
],
"date-time": "2024-03-01T23:36:24Z",
"timestamp": 1709336184025
},
"institution": [
{
"name": "medRxiv"
}
],
"is-referenced-by-count": 0,
"issued": {
"date-parts": [
[
2020,
12,
14
]
]
},
"link": [
{
"URL": "https://syndication.highwire.org/content/doi/10.1101/2020.12.13.20247254",
"content-type": "unspecified",
"content-version": "vor",
"intended-application": "similarity-checking"
}
],
"member": "246",
"original-title": [],
"posted": {
"date-parts": [
[
2020,
12,
14
]
]
},
"prefix": "10.1101",
"published": {
"date-parts": [
[
2020,
12,
14
]
]
},
"publisher": "Cold Spring Harbor Laboratory",
"reference": [
{
"DOI": "10.1038/s41586-020-2404-8",
"article-title": "The effect of large-scale anti-contagion policies on the COVID-19 pandemic",
"doi-asserted-by": "crossref",
"first-page": "262",
"issue": "7820",
"journal-title": "Nature",
"key": "2021083112451597000_2020.12.13.20247254v3.1",
"volume": "584",
"year": "2020"
},
{
"DOI": "10.1016/S1473-3099(20)30120-1",
"doi-asserted-by": "publisher",
"key": "2021083112451597000_2020.12.13.20247254v3.2"
},
{
"DOI": "10.1056/NEJMp2032508",
"doi-asserted-by": "crossref",
"key": "2021083112451597000_2020.12.13.20247254v3.3",
"unstructured": "Hunter DJ . Trying to “Protect the NHS” in the United Kingdom. New England Journal of Medicine. 2020."
},
{
"DOI": "10.1038/s41598-020-72611-5",
"article-title": "Second wave COVID-19 pandemics in Europe: a temporal playbook",
"doi-asserted-by": "crossref",
"first-page": "15514",
"issue": "1",
"journal-title": "Sci Rep",
"key": "2021083112451597000_2020.12.13.20247254v3.4",
"volume": "10",
"year": "2020"
},
{
"DOI": "10.1097/TA.0000000000002888",
"article-title": "A statistically rigorous deep neural network approach to predict mortality in trauma patients admitted to the intensive care unit",
"doi-asserted-by": "crossref",
"first-page": "736",
"issue": "4",
"journal-title": "J Trauma Acute Care Surg",
"key": "2021083112451597000_2020.12.13.20247254v3.5",
"volume": "89",
"year": "2020"
},
{
"DOI": "10.1007/s12652-020-02509-7",
"doi-asserted-by": "crossref",
"key": "2021083112451597000_2020.12.13.20247254v3.6",
"unstructured": "Ahmad FS , Ali L , Raza Ul M , Khattak HA , Hameed T , Wajahat I , et al. A hybrid machine learning framework to predict mortality in paralytic ileus patients using electronic health records (EHRs). Journal of Ambient Intelligence and Humanized Computing. 2020."
},
{
"DOI": "10.1177/2150132720963634",
"doi-asserted-by": "crossref",
"key": "2021083112451597000_2020.12.13.20247254v3.7",
"unstructured": "Naseem M , Akhund R , Arshad H , Ibrahim MT . Exploring the Potential of Artificial Intelligence and Machine Learning to Combat COVID-19 and Existing Opportunities for LMIC: A Scoping Review. J Prim Care Community Health. 2020;11:2150132720963634."
},
{
"key": "2021083112451597000_2020.12.13.20247254v3.8",
"unstructured": "Pathologist CoA. Guidance for COVID-19 Testing for CAP-Accredited Laboratories. CAP Guidelines, caporg. 2020."
},
{
"key": "2021083112451597000_2020.12.13.20247254v3.9",
"unstructured": "Prevention CfDCa. CDC’s Diagnostic Test for COVID-19 Only and Supplies. CDC Guidelines, cdcgov. 2020."
},
{
"article-title": "SMOTE: Synthetic Minority Over-sampling Technique",
"first-page": "321",
"issue": "1",
"journal-title": "J Artif Int Res",
"key": "2021083112451597000_2020.12.13.20247254v3.10",
"volume": "16",
"year": "2002"
},
{
"DOI": "10.3389/fpubh.2020.00216",
"article-title": "COVID-19: Emergence, Spread, Possible Treatments, and Global Burden",
"doi-asserted-by": "crossref",
"first-page": "216",
"journal-title": "Front Public Health",
"key": "2021083112451597000_2020.12.13.20247254v3.11",
"volume": "8",
"year": "2020"
},
{
"DOI": "10.1016/S0140-6736(20)30845-X",
"doi-asserted-by": "publisher",
"key": "2021083112451597000_2020.12.13.20247254v3.12"
},
{
"key": "2021083112451597000_2020.12.13.20247254v3.13",
"unstructured": "(CDC) CfDCaP. CDC COVID Data Tracker: Maps, charts, and data provided by the CDC. http://www.cdcgov. 2020."
},
{
"key": "2021083112451597000_2020.12.13.20247254v3.14",
"unstructured": "(FDA) FDA. COVID-19 Vaccines: FDA has rigorous scientific and regulatory processes in place to facilitate development and ensure the safety, effectiveness and quality of COVID-19 vaccines. Online. 2020."
},
{
"DOI": "10.1001/jamanetworkopen.2019.15997",
"article-title": "Machine Learning Approaches to Predict 6-Month Mortality Among Patients With Cancer",
"doi-asserted-by": "crossref",
"first-page": "e1915997",
"issue": "10",
"journal-title": "JAMA Netw Open",
"key": "2021083112451597000_2020.12.13.20247254v3.15",
"volume": "2",
"year": "2019"
},
{
"article-title": "Machine learning for prediction of all-cause mortality in patients with suspected coronary artery disease: a 5-year multicentre prospective registry analysis",
"first-page": "500",
"issue": "7",
"journal-title": "Eur Heart J",
"key": "2021083112451597000_2020.12.13.20247254v3.16",
"volume": "38",
"year": "2017"
},
{
"DOI": "10.1101/2020.02.04.20020222",
"doi-asserted-by": "crossref",
"key": "2021083112451597000_2020.12.13.20247254v3.17",
"unstructured": "Ahmed FS , Raza-Ul-Mustafa, Ali L , Imadud-Deen, Hameed T , Ikram A , et al. Machine Learning Can Predict Deaths in Patients with Diverticulitis During their Hospital Stay. medRxiv. 2020:2020.02.04.20020222."
},
{
"DOI": "10.5588/ijtld.20.0278",
"article-title": "Effect of chronic obstructive pulmonary disease and smoking on the outcome of COVID-19",
"doi-asserted-by": "crossref",
"first-page": "838",
"issue": "8",
"journal-title": "Int J Tuberc Lung Dis",
"key": "2021083112451597000_2020.12.13.20247254v3.18",
"volume": "24",
"year": "2020"
},
{
"DOI": "10.1093/ndt/gfaa271",
"article-title": "Mortality analysis of COVID-19 infection in chronic kidney disease, haemodialysis and renal transplant patients compared with patients without kidney disease: a nationwide analysis from Turkey",
"doi-asserted-by": "crossref",
"first-page": "2083",
"issue": "12",
"journal-title": "Nephrol Dial Transplant",
"key": "2021083112451597000_2020.12.13.20247254v3.19",
"volume": "35",
"year": "2020"
},
{
"DOI": "10.1161/CIRCULATIONAHA.120.047349",
"doi-asserted-by": "publisher",
"key": "2021083112451597000_2020.12.13.20247254v3.20"
},
{
"DOI": "10.1016/j.hrtlng.2020.10.002",
"doi-asserted-by": "crossref",
"key": "2021083112451597000_2020.12.13.20247254v3.21",
"unstructured": "Wang XH , Duan J , Han X , Liu X , Zhou J , Wang X , et al. High incidence and mortality of pneumothorax in critically Ill patients with COVID-19. Heart Lung. 2020."
},
{
"DOI": "10.1186/s13054-020-02911-9",
"doi-asserted-by": "publisher",
"key": "2021083112451597000_2020.12.13.20247254v3.22"
},
{
"DOI": "10.1016/j.chest.2020.07.023",
"article-title": "Coronavirus Disease 2019 Sepsis: A Nudge Toward Antibiotic Stewardship",
"doi-asserted-by": "crossref",
"first-page": "1833",
"issue": "5",
"journal-title": "Chest",
"key": "2021083112451597000_2020.12.13.20247254v3.23",
"volume": "158",
"year": "2020"
},
{
"DOI": "10.1016/j.jinf.2020.06.055",
"doi-asserted-by": "publisher",
"key": "2021083112451597000_2020.12.13.20247254v3.24"
},
{
"DOI": "10.1371/journal.pone.0236618",
"article-title": "Prediction model and risk scores of ICU admission and mortality in COVID-19",
"doi-asserted-by": "crossref",
"first-page": "e0236618",
"issue": "7",
"journal-title": "PLoS One",
"key": "2021083112451597000_2020.12.13.20247254v3.25",
"volume": "15",
"year": "2020"
},
{
"DOI": "10.1186/s13054-020-03223-8",
"doi-asserted-by": "publisher",
"key": "2021083112451597000_2020.12.13.20247254v3.26"
},
{
"DOI": "10.1515/cclm-2020-0369",
"doi-asserted-by": "publisher",
"key": "2021083112451597000_2020.12.13.20247254v3.27"
},
{
"DOI": "10.1016/j.kint.2020.03.005",
"doi-asserted-by": "publisher",
"key": "2021083112451597000_2020.12.13.20247254v3.28"
},
{
"DOI": "10.1038/s41598-020-73962-9",
"doi-asserted-by": "publisher",
"key": "2021083112451597000_2020.12.13.20247254v3.29"
},
{
"DOI": "10.1016/S2213-2600(20)30312-X",
"article-title": "Mortality in COVID-19 is not merely a question of resource availability",
"doi-asserted-by": "crossref",
"first-page": "832",
"issue": "9",
"journal-title": "The Lancet Respiratory medicine",
"key": "2021083112451597000_2020.12.13.20247254v3.30",
"volume": "8",
"year": "2020"
},
{
"article-title": "Mortality of Patients With Severe COVID-19 in the Intensive Care Unit: An Observational Study From a Major COVID-19 Receiving Hospital",
"first-page": "e10906",
"issue": "10",
"journal-title": "Cureus",
"key": "2021083112451597000_2020.12.13.20247254v3.31",
"volume": "12",
"year": "2020"
},
{
"DOI": "10.1016/j.mad.2020.111255",
"doi-asserted-by": "publisher",
"key": "2021083112451597000_2020.12.13.20247254v3.32"
},
{
"DOI": "10.1001/jama.2020.17023",
"doi-asserted-by": "publisher",
"key": "2021083112451597000_2020.12.13.20247254v3.33"
},
{
"DOI": "10.1002/emp2.12205",
"doi-asserted-by": "publisher",
"key": "2021083112451597000_2020.12.13.20247254v3.34"
}
],
"reference-count": 34,
"references-count": 34,
"relation": {
"is-preprint-of": [
{
"asserted-by": "subject",
"id": "10.1016/j.ijmedinf.2021.104556",
"id-type": "doi"
}
]
},
"resource": {
"primary": {
"URL": "http://medrxiv.org/lookup/doi/10.1101/2020.12.13.20247254"
}
},
"score": 1,
"short-title": [],
"source": "Crossref",
"subtitle": [],
"subtype": "preprint",
"title": "Predicting mortality in SARS-COV-2 (COVID-19) positive patients in the inpatient setting using a Novel Deep Neural Network",
"type": "posted-content"
}
