Retention predictors of a smoking treatment provided by a public psychosocial unit in Brazil.

Psychosocial units in Brazil (CAPS) provide access to mental health and addiction patients, who are not routinely treated for nicotine dependence. The present study analysed predictors of retention of a 6-week treatment provided by a CAPS unit to 367 smokers with a high rate of psychiatric disorders and addictions for the period 2007-2010. Several baseline variables were collected. Retention was defined as the presence of the individual in all four medical consultations and six group sessions. Multivariate discrete time Cox survival regression models were used to test for the outcome of interest. Timetables were used to explore in which moment of the treatment each predictor was important. Time to smoking the first cigarette (TTFC) 5 min or later after waking and nicotine patch use (nicotine replacement therapy, NRT) were associated with retention. The present study supports the importance of the variables TTFC and NRT when used in treatment retention for a sample with a high rate of psychiatric and alcohol disorders. NRT seems to be very important in the beginning of the treatment, probably because of withdrawal symptoms. Individuals currently undergoing psychiatric treatment and with alcohol problems had good retention rates comparable to the other individuals.