A risk estimation system to predict postpartum cigarette smoking relapse

Postpartum relapse to cigarette smoking (PRS) rate has not substantially improved for more than two decades. Over 55% of women successfully quit smoking during pregnancy; however, half (50%) return to smoking within three months of childbirth and 90% relapse within a year. The identification of effective PRS prevention interventions are needed, especially since factors related to PRS risk factors vary by person, time, and context. In this paper, a prototype risk estimation system using daily ecological momentary assessment data is proposed to develop an adaptive intervention system which will consider multiple risk factors. The risk estimator is designed using a hierarchical fuzzy inference system design scheme to capture human knowledge. A particle swarm optimization scheme is also applied. The simulation results show the feasibility of the proposed estimator for the PRS prevention intervention system.