The performance of data mining technique for prediction of delirium tremens

Abstract

Delirium tremens (DTs) is the most severe form of alcohol withdrawal syndrome that can lead to multiple physical complications, such as electrolyte imbalances, cardiovascular collapse. Untreated DTs carries a mortality rate of approximately 20 to 40%. Developing models for predicting DTs occurrence in hospitals could enable early diagnosis and timely treatment, significantly reducing the mortality rate to around 1-4%. To develop and compare the performance of models for predicting DTs in inpatients with alcohol dependence. A predictive model for DTs was developed using 10 data mining techniques. The data is collected from hospitalized patients diagnosed with alcohol dependence at Pakchong Nana Hospital. A training and testing set was created from a retrospective review of 1,960 electronic medical records, containing 11 features. The model's performance was evaluated using 10-Fold Cross Validation.The most effective predictive model was Gradient Boosting Trees with feature selection, achieving an accuracy of 74.14%, precision of 72.56%, sensitivity of 78.02%, and an F-measure of 75.09%. The Gradient Boosting Trees with feature selection was found to be the most optimal predictive model when compared to other algorithms. This model could be further developed into a clinical decision support system (CDSS) for predicting delirium tremens in hospitalized patients diagnosed as alcohol dependence.

References

A. D. Kaye et al., “Delirium Tremens: A Review of Clinical Studies,” Cureus, vol. 16, no. 4, p. e57601, doi: 10.7759/cureus.57601.

A. Rahman and M. Paul, “Delirium Tremens,” in StatPearls, Treasure Island (FL): StatPearls Publishing, 2024. Accessed: Sep. 14, 2024. [Online]. Available: http://www.ncbi.nlm.nih.gov/books/ NBK482134/

“Mortality and alcohol‐related morbidity in patients with delirium tremens, alcohol withdrawal state or alcohol dependence in Norway: A register‐based prospective cohort study - Bramness - 2023 - Addiction - Wiley Online Library.” Accessed: Sep. 14, 2024. [Online]. Available: https://onlinelibrary. wiley.com/doi/10.1111/add.16297

J. Han and M. Kamber, Data mining: concepts and techniques, 2nd ed. in The Morgan Kaufmann series in data management systems. Amsterdam ; Boston : San Francisco, CA: Elsevier ; Morgan Kaufmann, 2006.

G. Burkhardt et al., “A machine learning approach to risk assessment for alcohol withdrawal syndrome,” Eur. Neuropsychopharmacol., vol. 35, pp. 61–70, Jun. 2020, doi: 10.1016/j.euroneuro.2020.03.016.
M. L. Kolling et al., “Data Mining in Healthcare: Applying Strategic Intelligence Techniques to Depict 25 Years of Research Development,” Int. J. Environ. Res. Public. Health, vol. 18, no. 6, p. 3099, Mar. 2021, doi: 10.3390/ijerph18063099.

Kawasaki, Y, “Trends and Challenges in Data Mining Research in Southeast Asia,” Asian J. Data Min., vol. 3, no. 2, pp. 45–49, 2021.

D. C. To et al., “Alcohol Withdrawal Severity Measures for Identifying Patients Requiring High-In tensity Care,” Crit. Care Explor., vol. 6, no. 3, p. e1066, Mar. 2024, doi: 10.1097/CCE.000000000 0001066.

J. R. Maldonado, “Delirium in the acute care setting: characteristics, diagnosis and treatment,” Crit. Care Clin., vol. 24, no. 4, pp. 657–722, vii, Oct. 2008, doi: 10.1016/j.ccc.2008.05.008.

S. Grover and A. Ghosh, “Delirium Tremens: Assessment and Management,” J. Clin. Exp. Hepatol., vol. 8, no. 4, pp. 460–470, Dec. 2018, doi: 10.1016/j.jceh.2018.04.012.

“Ensuring the Robustness and Reliability of Data-Driven Knowledge Discovery Models in Production and Manufacturing - PMC.” Accessed: Sep. 14, 2024. [Online]. Available: https://www. ncbi.nlm.nih.gov/pmc/articles/PMC8236533/

L. B. Iantovics, “Comparative Study of Random Forest, Gradient Boosted Trees, Feedforward Neural Networks and Convolutional Neural Networks using Fingerprints and Molecular Descriptors for Adverse Drug Reaction Prediction,” Jan. 2024.

S. Gupta, K. Saluja, A. Goyal, A. Vajpayee, and V. Tiwari, “Comparing the performance of machine learning algorithms using estimated accuracy,” Meas. Sens., vol. 24, p. 100432, Dec. 2022, doi: 10.1016/j. measen.2022.100432.

Gupta et al., “Personalized treatment of alcohol withdrawal syndrome using machine learning: A case study,” Artif. Intell. Med., vol. 127, no. 102163, 2022.

Weinreb et al., “Machine learning to predict risk of alcohol withdrawal syndrome in hospitalized patients,” J. Subst. Abuse Treat., vol. 120, no. 108201, 2021.

“Predisposing and Precipitating Factors Associated With Delirium: A Systematic Review | Neurology | JAMA Network Open | JAMA Network.” Accessed: Sep. 15, 2024. [Online]. Available: https:// jamanetwork.com/journals/jamanetworkopen/ fullarticle/2800112

S. Sarkar, S. Choudhury, G. Ezhumalai, and J. Konthoujam, “Risk factors for the development of delirium in alcohol dependence syndrome: Clinical and neurobiological implications,” Indian J. Psychiatry, vol. 59, no. 3, pp. 300–305, 2017, doi: 10.4103/psychiatry.IndianJPsychiatry_67_17.

Malec et al., “Machine learning in the prediction of alcohol withdrawal syndrome severity,” J. Med. Toxicol., vol. 16, no. 2, pp. 165–171, 2020.

A. Mohammed and R. Kora, “A comprehensive review on ensemble deep learning: Opportunities and challenges,” J. King Saud Univ. - Comput. Inf. Sci., vol. 35, no. 2, pp. 757–774, Feb. 2023, doi: 10.1016/j.jksuci.2023.01.014.

T. Parr, J. Hamrick, and J. D. Wilson, “Nonpa rametric feature impact and importance,” Inf. Sci., vol. 653, p. 119563, Jan. 2024, doi: 10.1016/j .ins.2023.119563.