Evaluation of Dimensionality and Reliability of the Autonomy over Smoking Scale among South Indian Smokers

Abstract

Introduction: In spite of the efforts being directed at reducing the tobacco use among public, it remains a significant concern facing India today. In the quest of providing tobacco cessation counseling, documentation of the tobacco dependence of the individuals is quintessential. This study aims to assess the psychometric properties of the autonomy over smoking scale (AUTOS) among patients seeking oral health care at a teaching dental institution in southern India. Materials and Methods: The study sample constituted 199 subjects who satisfied the eligibility criteria of self-reported current smoking and were willing to participate in the study. Participants’ age, gender, and years of smoking were documented along with their nicotine dependence by administration of Fagerstrom test for nicotine dependence (FTND) and AUTOS by a trained interviewer. The dimensionality of AUTOS was verified by confirmatory factor analysis using the Classical and Bayesian Instrument Development software program. Correlation analysis between FTND and AUTOS subscale scores was performed along with multiple linear regression analyses to identify the predictors for AUTOS subscale scores. Results: The symptom type-wise subscales of AUTOS and the overall scale demonstrated good internal consistency reliability (Cronbach alpha ≥0.758). Significant positive correlation was observed between age, years of smoking, FTND score, and all the AUTOS subscale scores. Linear regression analyses showed that the number of years for which the subject had been smoking was a significant predictor of all the three AUTOS subscale scores. Conclusion: AUTOS was observed to be a very useful tool with good internal consistency reliability that measures tobacco dependence in consistence with FTND among South Indian population, and while doing so, it captures the various forms of tobacco dependence in an independent manner.