A Machine Learning approach to improve smoking cessation trials

Lead Investigator: Riccardo Polosa, Center of Excellence for the Acceleration of Harm Reduction (CoEHAR)
Title of Proposal Research: A Machine Learning approach to improve smoking cessation trials
Vivli Data Request: 7498
Funding Source: None
Potential Conflicts of Interest: None

Summary of the Proposed Research:

8144 current smokers at baseline from EAGLES dataset will be included in our study.
Identifying individual characteristics that can predict relapse and quit success in smoking cessation trials could help shape a strategy help to optimize health care resources and refine state funded tobacco policies.

Delivering on these aspirations requires an improved understanding of success factors in well characterized cohorts of smokers. Currently, no accurate determinants of smoking cessation have been found. Socio-demographic features, psychological factors, smoking history, severity of nicotine dependence, past quit attempts, and measures of motivation to stop have been extensively investigated both in aided and unaided smoking cessation, but results are not clear-cut. Most of the studies have focused on the prediction model of smoking cessation using standard statistical methods such as bivariate analysis and multivariate logistic regression applied to traditional widely studied risk factors. Their findings have been generally unhelpful to optimize smoking cessation studies and to improve success rate outcomes.

Predictors of smoking cessation are derived from very different populations, factors usually explored may not be the significant ones, and study outcomes are not always clearly defined.

Without this knowledge, improved behavioural classification of smokers with a strong desire to quit and adopting a successful personalized management plan for their tobacco dependence will remain just a dream. A novel innovative research approach identifying stronger predictors of smoking relapse and quit success is the only recipe for success.

Acquisition and analysis of extensive datasets is, today, a central tool in most research fields. Machine learning provides powerful methods to obtain descriptive and predictive models for the data in many applications. The acquisition of quality information is fundamental for the reliability and accuracy of predictive and classification models increasingly used in various applications. A correct and adequate use of Artificial Intelligence (A.I.) models allows to extend and overcome the classical statistical methods, thus helping experts and professionals of different fields in decisions and policy-making.

Requested Studies:

A Phase 4, Randomized, Double-blind, Active And Placebo-controlled, Multicenter Study Evaluating The Neuropsychiatric Safety And Efficacy Of 12 Weeks Varenicline Tartrate 1mg Bid And Bupropion Hydrochloride 150mg Bid For Smoking Cessation In Subjects With And Without A History Of Psychiatric Disorders
Data Contributor: Pfizer Inc.
Study ID: NCT01456936
Sponsor ID: A3051123

Update: This data request was withdrawn on 13-Jul-2023 by the researcher.