WhatsApp Embedded in Routine Service Delivery for Smoking Cessation

Trial design

The study was designed as a RCT including two parallel arms: an intervention group receiving messages through WhatsApp Messenger operating on smart phones and a control group not receiving these messages. Both groups received the standard outpatient care of the clinic including aproximately 45 minutes face-to-face individual counseling at first contact ending with the decision of treatment and quit date, and a support booklet on quitting and subsequent follow-ups by the same clinician at 1st, 2nd, 4th weeks and 3rd month after quit date, either face-to-face at the clinic or by telephone if the patient does not come to the clinic for follow-up. The study was carried out in the smoking cessation clinic of Ege University Medical School's Public Health Department.

Sample size

An a priori power analysis was conducted with OpenEpi, Version 3 by selecting a two-sided test to compare 30% success in the control group at 1st month versus 60% success rate in the intervention group with an error margin of 5%, a power of 80% and an allocation ratio of 1:2 which yielded a minimum sample size of 36 in the intervention and 72 in the control groups. For possible loss to follow-up, the sample sizes were increased by 20% to 43 and 86, respectively.

Randomization

Among the 132 participants included in the study, 44 were randomly allocated to the intervention arm and 88 were randomly allocated to the control group. Among the intervention group , there were no losses to follow-up at 1st and 3rd months, but two participants have requested the cessation of the intervention after their 1st month follow-up.

Blinding

The study was single-blind as blinding of the helthcare providers to intervention assignment was achieved and the physicians conducted usual care for smoking cessation counseling.The physicians were blind throughout the follow ups as well. However, participants and the researcher who sent the messages were not blind.

Variables

Primary outcome variables: Quitting success rates at the end of the 1st, 3rd and 6th month of follow-up in the intervention and control groups. Quitting success rate at 1st month was calculated with point prevalence. History of cessation was based on self report and those who declared not smoked even a single puff on a cigarette at all in the past two weeks were considered as "successful" in the quitting attempt. The success rate at 3rd month was calculated with point prevalence. In this point, not smoking at all in total in the past ten weeks was considered "successful" in the quitting attempt. The success rate at 6th month was calculated with point prevalence. In this point, not smoking at all in total in the past 24 weeks was considered "successful" in the quitting attempt.

Secondary outcome variables: The total number of follow-ups in the 1st, 3rd and 6th month. Contacts were divided into two categories: face-to face contacts and telephone calls. The total number of routine follow-ups was classified as adequate when 3 or more in the 1st month and a total of 4 follow-ups was considered as adequate at 3rd month. At 1st month, at least one of these follow-ups was required to be face-to-face for adequacy. The continuity to drug/NRT therapy was categorized as <1 month or ≥1 month. Any change in weight was categorized as; has weight gain, no weight gain

Data collection Data collection was conducted at the Smoking Cessation Clinic of the Public Health Department at Ege University Medical School Hospital between March 2017-March 2018. The intended number of participants was achieved in 4.5 months and with the completion of the last follow-up, the study was finalized.

Analysis

The analyses were conducted according to the Intention-To-Treat (ITT) principle. The participants lost to follow-up were considered unsuccessful in quitting, as non-response could be expected in relapsing individuals, thus Last Observation Carried Forward (LOCF) was not used.

Smoking cessation successes in the groups were compared with the incidences and relative efficacy of the intervention. Chi-square test, Student's t test in independent groups, single and multivariate logistic regression (enter method) were used for analyses. The multivariate analysis of factors associated with smoking cessation success was performed by using logistic regression analysis with two models. In the first model the investigators controlled for age and gender, in the second model controlling was for age, gender and all the other associated variables in univariate analyses. Analyses were made separately for the 1st, 3rd and 6th month. Statistical significance was set at p<0.05.

Ethical issues

The study was approved by Ege University Medical School's institutional review board (decision no.16-12.1/11 on January 6, 2017). Written informed consent of every participant to enroll and receive WhatsApp messages were also obtained with confidentiality for name and address. With the blinding procedures described above, the delivery of routine service for all participants regardless of the allocation to intervention or control group was ensured.