Using Machine Learning to Predict the Smoking Cessation Success Plan in Taiwan

Introduction:
Tobacco use is a major public health issue all over the world. It is one of the main risk
factors for amount of chronic diseases such as cancer, lung diseases, and cardiovascular
diseases. The government implemented the “Outpatient Smoking Cessation Services
(OSCS) program” since 2002 We intended to use machine learning to develop a
prediction model which can be used in providing appropriate strategy on smoking
cessation.

1. Methods:
The data collected from January 2009 to December 2010 in two teaching hospitals in
different districts in Taipei, Taiwan. We used data from one hospital as the training set,
and the other hospital dataset for validation. We collected 18 input variables from
quitters’ demographic information, Fagerstrom Test for Nicotine Dependence (FTND)
scores, medical history, medication, smoking status and whether they have had
withdrawal symptoms. Three models (A, B, C) were designated based on the use of
different input variables. Each model was fitted to the data using ANN and Logistic
Regression (LR). .

2. Result:
The overall smoking cessation success rate was 20.38%. The AUROC of ANN for the
three models were 0.86, 0.83, 0.77 (model A, B, C) and the AUROC of LR for the
three models were 0.85, 0.84, 0.70 (model A, B, C) respectively Both ANN and LR can
predict smoking cessation with good accuracy with an AUROC ranging from 0.86 to
0.70. The performance of ANN was slightly better than LR without statistical
significance. We believe that this prediction model can be used for better management
of national-level smoking cessation program.