Development of Machine Learning Models for Prediction of Smoking Cessation Outcome

Abstract

Predictors for success in smoking cessation have been studied, but a prediction model capable of providing a success rate for each patient attempting to quit smoking is still lacking. The aim of this study is to develop prediction models using machine learning algorithms to predict the outcome of smoking cessation. Data was acquired from patients underwent smoking cessation program at one medical center in Northern Taiwan. A total of 4875 enrollments fulfilled our inclusion criteria. Models with artificial neural network (ANN), support vector machine (SVM), random forest (RF), logistic regression (LoR), k-nearest neighbor (KNN), classification and regression tree (CART), and naïve Bayes (NB) were trained to predict the final smoking status of the patients in a six-month period. Sensitivity, specificity, accuracy, and area under receiver operating characteristic (ROC) curve (AUC or ROC value) were used to determine the performance of the models. We adopted the ANN model which reached a slightly better performance, with a sensitivity of 0.704, a specificity of 0.567, an accuracy of 0.640, and an ROC value of 0.660 (95% confidence interval (CI): 0.617–0.702) for prediction in smoking cessation outcome. A predictive model for smoking cessation was constructed. The model could aid in providing the predicted success rate for all smokers. It also had the potential to achieve personalized and precision medicine for treatment of smoking cessation.

Exploring the Correlation Between the COVID-19 Pandemic and Increased Daily Cigarette Consumption in Yogyakarta, Indonesia: A Machine Learning Approach

“…DT approach resembles the human thought process, where humans rely on their experiences to make decisions [16]. The decision tree algorithm breaks down data by calculating information gain or using input information in the form of features, in this study we input 27 features in connection with smoking behaviour and the COVID-19 pandemic, while LoR was used to predict the probability of an event occurring [19], in this case predicting the probability of an increase in cigarette consumption during the pandemic, this model predicts outcomes based on selected features, LoR has good performance because it is well-suited for predicting the probability of an event, such as increased cigarette consumption. It uses independent variables to be linearly related to the log odds of the outcome, a relationship between the input variables and the outcome, allowing it to capture the relationship and make accurate predictions.…”

Classification of acute poisoning exposures with machine learning models derived from the National Poison Data System

“…15 Dong et al investigated ML to predict opioid overdose, and found high recall using a random forest model and high accuracy with deep learning models. 22 Other studies have leveraged ML to predict paraquat poisoning prognosis, 23,24 seizures from tramadol poisoning, 25 adverse drug events in elderly patients, 26 smoking cessation treatment outcome, 27 lead poisoning in children, 28 pesticide ototoxicity 29 and inadequate medication responses in the emergency department. 30 In recent years, ML in medicine has garnered considerable interest, from anticipated cost-effectiveness, speed, and accuracy.…”

Who responds to a multi‐component treatment for cannabis use disorder? Using multivariable and machine learning models to classify treatment responders and non‐responders

“…Artificial intelligence in medical decision-making [14], including machine learning methods, can integrate numerous data points to predict clinical outcomes, often outperforming logistic regression models and other methods with limited predictors [12,15]. Similar models have been developed to predict alcohol [11,[16][17][18][19][20], tobacco [21,22] and general substance use disorder [12] treatment outcomes, as well as response to a web-based intervention for CUD [23].…”