Machine Learning to Predict Disease Severity and Progression in Hospitalized COVID-19 Patients Using Laboratory Data on Admission

Abstract

Background: Herein, we aimed to develop and test machine learning (ML) models to predict disease severity and/or progression in hospitalized COVID-19 patients through baseline laboratory features. Methods: In this retrospective study of hospitalized COVID-19 patients admitted to a tertiary care center, we evaluated routine admission data to determine the accuracy rates of different ML algorithms: k-nearest neighbor classifier, bagging classifier, random forest (RF), and decision tree. These models were compared over three outcomes: those who needed oxygen supplementation vs. who did not on admission (Analysis 1, n: 180), those who later developed oxygen requirement vs. those who did not (Analysis 2, n: 112), and those who needed invasive mechanical ventilation vs. those who did not during hospitalization (Analysis 3, n: 164). Results: The median age of the patients was 55 (44-68) years, with males constituting 47.2% of the subjects. At admission, 37.8% of the patients required oxygen supplementation. During hospitalization, 17.5% needed mechanical ventilation, and 8.3% died. For all analyses, RF had the highest accuracy in classifying the need for oxygen supplementation on admission (89.4%) or during hospitalization (91.1%) and for invasive mechanical ventilation (92.2%). These were followed by a bagging classifier for Analysis 1 (88.3%) and Analysis 3 (91.0%) and by a decision tree for Analysis 2 (88.4%). C-reactive protein, monocyte distribution width, and high-sensitive troponin-T were the most crucial laboratory contributors to Analysis 1, Analysis 2, and Analysis 3, respectively. Conclusion: Our study showed that ML algorithms could predict the need for oxygen supplementation and mechanical ventilation during hospitalization using baseline laboratory data, suggesting a slight superiority of RF, among others.

Keywords

• accuracy
• classifiers
• COVID-19
• inpatient
• oxygen supplementation
• random forest

References

1. WHO. WHO Coronavirus (COVID-19) Dashboard 2020, 2022. Available at: www.covid19.who.int Accessed April 1, 2024.
2. Fontanarosa PB, Bauchner H. COVID-19-looking beyond tomorrow for health care and society. JAMA. 2020 May 19;323(19):1907-8. doi: 10.1001/jama.2020.6582.
3. Huang C, Wang Y, Li X, Ren L, Zhao J, Hu Y, Zhang L, Fan G, Xu J, Gu X, Cheng Z, Yu T, Xia J, Wei Y, Wu W, Xie X, Yin W, Li H, Liu M, Xiao Y, Gao H, Guo L, Xie J, Wang G, Jiang R, Gao Z, Jin Q, Wang J, Cao B. Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China. Lancet. 2020 Feb 15;395(10223):497-506. doi: 10.1016/S0140-6736(20)30183-5.
4. Wang D, Hu B, Hu C, Zhu F, Liu X, Zhang J, Wang B, Xiang H, Cheng Z, Xiong Y, Zhao Y, Li Y, Wang X, Peng Z. Clinical characteristics of 138 hospitalized patients with 2019 novel coronavirus-infected pneumonia in Wuhan, China. JAMA. 2020 Mar 17;323(11):1061-9. doi: 10.1001/jama.2020.1585.
5. Wolff J, Pauling J, Keck A, Baumbach J. The economic impact of artificial intelligence in health care: Systematic review. J Med Internet Res. 2020 Feb 20;22(2):e16866. doi: 10.2196/16866.
6. WHO working group on the clinical characterisation and management of COVID-19 infection. A minimal common outcome measure set for COVID-19 clinical research. Lancet Infect Dis. 2020 Aug;20(8):e192-e197. doi: 10.1016/S1473-3099(20)30483-7.
7. Su Y, Ju MJ, Xie RC, Yu SJ, Zheng JL, Ma GG, Liu K, Ma JF, Yu KH, Tu GW, Luo Z. Prognostic accuracy of early warning scores for clinical deterioration in patients with COVID-19. Front Med (Lausanne). 2021 Feb 1;7:624255. doi: 10.3389/fmed.2020.624255.
8. Myrstad M, Ihle-Hansen H, Tveita AA, Andersen EL, Nygård S, Tveit A, Berge T. National Early Warning Score 2 (NEWS2) on admission predicts severe disease and in-hospital mortality from Covid-19 - a prospective cohort study. Scand J Trauma Resusc Emerg Med. 2020 Jul 13;28(1):66. doi: 10.1186/s13049-020-00764-3.

9. Gidari A, De Socio GV, Sabbatini S, Francisci D. Predictive value of National Early Warning Score 2 (NEWS2) for intensive care unit admission in patients with SARS-CoV-2 infection. Infect Dis (Lond). 2020 Oct;52(10):698-704. doi: 10.1080/23744235.2020.1784457.
10. Brajkovic M, Vukcevic M, Nikolic S, Dukic M, Brankovic M, Sekulic A, Popadic V, Stjepanovic M, Radojevic A, Markovic-Denic L, Rajovic N, Milic N, Tanasilovic S, Todorovic Z, Zdravkovic M. The predictive value of risk factors and prognostic scores in hospitalized COVID-19 patients. Diagnostics (Basel). 2023 Aug 11;13(16):2653. doi: 10.3390/diagnostics13162653.
11. Veldhuis L, Ridderikhof ML, Schinkel M, van den Bergh J, Beudel M, Dormans T, Douma R, Gritters van den Oever N, de Haan L, Koopman K, de Kruif MD, Noordzij P, Reidinga A, de Ruijter W, Simsek S, Wyers C, Nanayakkara PW, Hollmann M. Early warning scores to assess the probability of critical illness in patients with COVID-19. Emerg Med J. 2021 Dec;38(12):901-5. doi: 10.1136/emermed-2020-211054.
12. Kamran F, Tang S, Otles E, McEvoy DS, Saleh SN, Gong J, Li BY, Dutta S, Liu X, Medford RJ, Valley TS, West LR, Singh K, Blumberg S, Donnelly JP, Shenoy ES, Ayanian JZ, Nallamothu BK, Sjoding MW, Wiens J. Early identification of patients admitted to hospital for covid-19 at risk of clinical deterioration: model development and multisite external validation study. BMJ. 2022 Feb 17;376:e068576. doi: 10.1136/bmj-2021-068576.
13. Rauseo M, Perrini M, Gallo C, Mirabella L, Mariano K, Ferrara G, Santoro F, Tullo L, La Bella D, Vetuschi P, Cinnella G. Machine learning and predictive models: 2 years of Sars-CoV-2 pandemic in a single-center retrospective analysis. J Anesth Analg Crit Care. 2022 Oct 14;2(1):42. doi: 10.1186/s44158-022-00071-6.
14. Kourou K, Exarchos TP, Exarchos KP, Karamouzis MV, Fotiadis DI. Machine learning applications in cancer prognosis and prediction. Comput Struct Biotechnol J. 2014 Nov 15;13:8-17. doi: 10.1016/j.csbj.2014.11.005.
15. Pan L, Liu G, Lin F, Zhong S, Xia H, Sun X, Liang H. Machine learning applications for prediction of relapse in childhood acute lymphoblastic leukemia. Sci Rep. 2017 Aug 7;7(1):7402. doi: 10.1038/s41598-017-07408-0.
16. Karthikeyan A, Garg A, Vinod PK, Priyakumar UD. machine learning based clinical decision support system for early COVID-19 mortality prediction. Front Public Health. 2021 May 12;9:626697. doi: 10.3389/fpubh.2021.626697.
17. Bian Y, Han Q, Zheng Y, Yao Y, Fan X, Lv R, Pang J, Xu F, Chen Y. SUPER score contributes to warning and management in early-stage COVID-19. Infect Med (Beijing). 2023 Oct 19;2(4):308-14. doi: 10.1016/j.imj.2023.09.003.
18. Datta D, George Dalmida S, Martinez L, Newman D, Hashemi J, Khoshgoftaar TM, Shorten C, Sareli C, Eckardt P. Using machine learning to identify patient characteristics to predict mortality of in-patients with COVID-19 in south Florida. Front Digit Health. 2023 Jul 28;5:1193467. doi: 10.3389/fdgth.2023.1193467.
19. Yu Z, Li X, Zhao J, Sun S. Identification of hospitalized mortality of patients with COVID-19 by machine learning models based on blood inflammatory cytokines. Front Public Health. 2022 Nov 17;10:1001340. doi: 10.3389/fpubh.2022.1001340.
20. Remeseiro B, Bolon-Canedo V. A review of feature selection methods in medical applications. Comput Biol Med. 2019 Sep;112:103375. doi: 10.1016/j.compbiomed.2019.103375.
21. Pereira RB, Plastino A, Zadrozny B, Merschmann LH. Categorizing feature selection methods for multi-label classification. Artif Intell Rev. 2018;49(1):57-78. doi: 10.1007/s10462-016-9516-4.
22. Bhargava N, Sharma S, Purohit R, Rathore PS. Prediction of recurrence cancer using J48 algorithm. Presented at: 2nd International Conference on Communication and Electronics Systems (ICCES); 19–20 October, 2017; Coimbatore, India. doi: 10.1109/CESYS.2017.8321306.
23. Yadav S, Shukla S. Analysis of k-fold cross-validation over hold-out validation on colossal datasets for quality classification. Presented at: IEEE 6th International conference on advanced computing (IACC); 27-28 February, 2016; Bhimavaram, India. doi: 10.1109/IACC.2016.25.
24. Breiman L. Bagging predictors. Mach Learn. 1996;24(2):123-40. doi: 10.1007/BF00058655.
25. Sun Y, Zhang H, Zhao T, Zou Z, Shen B, Yang L. A new convolutional neural network with random forest method for hydrogen sensor fault diagnosis. IEEE Access. 2020;8:85421-30. doi: 10.1109/ACCESS.2020.2992231.
26. Song YY, Lu Y. Decision tree methods: applications for classification and prediction. Shanghai Arch Psychiatry. 2015 Apr 25;27(2):130-5. doi: 10.11919/j.issn.1002-0829.215044.
27. Gao Y, Cai GY, Fang W, Li HY, Wang SY, Chen L, Yu Y, Liu D, Xu S, Cui PF, Zeng SQ, Feng XX, Yu RD, Wang Y, Yuan Y, Jiao XF, Chi JH, Liu JH, Li RY, Zheng X, Song CY, Jin N, Gong WJ, Liu XY, Huang L, Tian X, Li L, Xing H, Ma D, Li CR, Ye F, Gao QL. Machine learning based early warning system enables accurate mortality risk prediction for COVID-19. Nat Commun. 2020 Oct 6;11(1):5033. doi: 10.1038/s41467-020-18684-2.
28. Huang I, Pranata R, Lim MA, Oehadian A, Alisjahbana B. C-reactive protein, procalcitonin, D-dimer, and ferritin in severe coronavirus disease-2019: a meta-analysis. Ther Adv Respir Dis. 2020 Jan-Dec;14:1753466620937175. doi: 10.1177/1753466620937175.
29. Sun Q, Qiu H, Huang M, Yang Y. Lower mortality of COVID-19 by early recognition and intervention: experience from Jiangsu Province. Ann Intensive Care. 2020 Mar 18;10(1):33. doi: 10.1186/s13613-020-00650-2.
30. Hu H, Yao N, Qiu Y. Predictive value of 5 early warning scores for critical COVID-19 patients. Disaster Med Public Health Prep. 2022 Feb;16(1):232-239. doi: 10.1017/dmp.2020.324.
31. Fan G, Tu C, Zhou F, Liu Z, Wang Y, Song B, Gu X, Wang Y, Wei Y, Li H, Wu X, Xu J, Tu S, Zhang Y, Wu W, Cao B. Comparison of severity scores for COVID-19 patients with pneumonia: a retrospective study. Eur Respir J. 2020 Sep 10;56(3):2002113. doi: 10.1183/13993003.02113-2020.
32. Elhazmi A, Al-Omari A, Sallam H, Mufti HN, Rabie AA, Alshahrani M, Mady A, Alghamdi A, Altalaq A, Azzam MH, Sindi A, Kharaba A, Al-Aseri ZA, Almekhlafi GA, Tashkandi W, Alajmi SA, Faqihi F, Alharthy A, Al-Tawfiq JA, Melibari RG, Al-Hazzani W, Arabi YM. Machine learning decision tree algorithm role for predicting mortality in critically ill adult COVID-19 patients admitted to the ICU. J Infect Public Health. 2022 Jul;15(7):826-34. doi: 10.1016/j.jiph.2022.06.008.
33. Liu H, Wang J, Geng Y, Li K, Wu H, Chen J, Chai X, Li S, Zheng D. Fine-grained assessment of COVID-19 severity based on clinico-radiological data using machine learning. Int J Environ Res Public Health. 2022 Aug 26;19(17):10665. doi: 10.3390/ijerph191710665.
34. Kocadagli O, Baygul A, Gokmen N, Incir S, Aktan C. Clinical prognosis evaluation of COVID-19 patients: An interpretable hybrid machine learning approach. Curr Res Transl Med. 2022 Jan;70(1):103319. doi: 10.1016/j.retram.2021.103319.
35. Yamanaka S, Morikawa K, Azuma H, Yamanaka M, Shimada Y, Wada T, Matano H, Yamada N, Yamamura O, Hayashi H. Machine-learning approaches for predicting the need of oxygen therapy in early-stage COVID-19 in Japan: Multicenter retrospective observational study. Front Med (Lausanne). 2022 Feb 23;9:846525. doi: 10.3389/fmed.2022.846525.
36. Bohn MK, Lippi G, Horvath A, Sethi S, Koch D, Ferrari M, Wang CB, Mancini N, Steele S, Adeli K. Molecular, serological, and biochemical diagnosis and monitoring of COVID-19: IFCC taskforce evaluation of the latest evidence. Clin Chem Lab Med. 2020 Jun 25;58(7):1037-52. doi: 10.1515/cclm-2020-0722.
37. Polilli E, Sozio F, Frattari A, Persichitti L, Sensi M, Posata R, Di Gregorio M, Sciacca A, Flacco ME, Manzoli L, Di Iorio G, Parruti G. Comparison of monocyte distribution width (MDW) and procalcitonin for early recognition of sepsis. PLoS One. 2020 Jan 10;15(1):e0227300. doi: 10.1371/journal.pone.0227300.
38. Agnello L, Vidali M, Lo Sasso B, Giglio RV, Gambino CM, Scazzone C, Ciaccio AM, Bivona G, Ciaccio M. Monocyte distribution width (MDW) as a screening tool for early detecting sepsis: a systematic review and meta-analysis. Clin Chem Lab Med. 2022 Feb 15;60(5):786-792. doi: 10.1515/cclm-2021-1331.
39. Piva E, Zuin J, Pelloso M, Tosato F, Fogar P, Plebani M. Monocyte distribution width (MDW) parameter as a sepsis indicator in intensive care units. Clin Chem Lab Med. 2021 Mar 5;59(7):1307-14. doi: 10.1515/cclm-2021-0192.
40. Li CH, Seak CJ, Chaou CH, Su TH, Gao SY, Chien CY, Ng CJ. Comparison of the diagnostic accuracy of monocyte distribution width and procalcitonin in sepsis cases in the emergency department: a prospective cohort study. BMC Infect Dis. 2022 Jan 4;22(1):26. doi: 10.1186/s12879-021-06999-4.
41. Kim SW, Lee H, Lee SH, Jo SJ, Lee J, Lim J. Usefulness of monocyte distribution width and presepsin for early assessment of disease severity in COVID-19 patients. Medicine (Baltimore). 2022 Jul 8;101(27):e29592. doi: 10.1097/MD.0000000000029592.
42. Daorattanachai K, Hirunrut C, Pirompanich P, Weschawalit S, Srivilaithon W. Association of monocyte distribution width with the need for respiratory support in hospitalized COVID-19 patients. Indian J Crit Care Med. 2023 May;27(5):352-7. doi: 10.5005/jp-journals-10071-24447.
43. Cordeanu EM, Duthil N, Severac F, Lambach H, Tousch J, Jambert L, Mirea C, Delatte A, Younes W, Frantz AS, Merdji H, Schini-Kerth V, Bilbault P, Ohlmann P, Andres E, Stephan D. Prognostic value of troponin elevation in COVID-19 hospitalized patients. J Clin Med. 2020 Dec 17;9(12):4078. doi: 10.3390/jcm9124078.
44. Alhindi T, Awad H, Alfaraj D, Elabdein Salih S, Abdelmoaty M, Muammar A. Troponin levels and the severity of COVID-19 pneumonia. Cureus. 2022 Mar 15;14(3):e23193. doi: 10.7759/cureus.23193.
45. Zimmerman A, Kalra D. Usefulness of machine learning in COVID-19 for the detection and prognosis of cardiovascular complications. Rev Cardiovasc Med. 2020 Sep 30;21(3):345-52. doi: 10.31083/j.rcm.2020.03.120.