Effectiveness of an individual acceptance and commitment therapy for smoking cessation, delivered face-to-face and by telephone to adults recruited in primary health care settings: a randomized controlled trial

Acceptance and Commitment Therapy (ACT) is a psychological intervention that has been used for smoking cessations [18]. According to the study by Baker et al. [19], avoidance of distressing smoking-related thoughts, feelings, and bodily sensations is the principal motive for addictive behaviors such as smoking. The goal in ACT is to increase one’s acceptance of these aversive internal experiences by increasing psychological flexibility through targeting six major processes, namely “acceptance,” “defusion,” “self-as-context,” “the present moment,” “values,” and “committed action” [20]. In the context of smoking cessation, the acceptance component focuses on helping individuals to recognize and increase their willingness to experience the internal triggers of smoking without trying to control or avoid them, whereas the committed action component focuses on getting them to make commitments to engage in adaptive behavioral change guided by their values related to quitting [21].

Promising results have been found in several studies on the application of ACT to smoking cessation in general populations, namely, a greater abstinence rate than when the following therapies were used: Nicotine Replacement Therapy (NRT) [21]; Bupropion (front-line pharmacotherapy) [22], Cognitive Behavioral Therapy (CBT) [23, 24]; standard web-based smoking cessation interventions [25], and smartphone applications for smoking cessation [26]. To date, only three studies have been conducted on telephone-based ACT for smoking cessation: one with individuals with bipolar disorder [27] and two with general populations [24, 28]. These studies found that it is feasible to deliver ACT by phone, and that highly promising quit rates were obtained compared with standard CBT and quit line counseling. As these studies were conducted in the U.S. through a reactive approach (i.e., the participants were self-motived to quit smoking as they were recruited via Quitline or through an advertisement), it is unclear whether proactively recruiting individuals from primary care settings to an individual, telephone-based ACT would be more effective than a control intervention to achieve smoking cessation among a Chinese population, particularly among a Hong Kong (HK) Chinese population, as HK people tend to have busy lifestyles [29]. Rather than delivering ACT over an exclusively telephone-based approach, the ACT intervention was first initiated with a face-to-face meeting in a health care setting to increase the participants’ engagement in subsequent telephone sessions. The aim of this study was to determine the effectiveness of this ACT intervention approach as compared to the usual treatment at primary health care settings, by providing a very brief smoking cessation reminder and standardized self-help printed cessation materials, as well as to identify baseline predictors of self-reported quitting.

The protocol for the present study has previously been described [30]. Briefly this study was a prospective, randomized trial with two parallel groups. Participants were recruited from six primary health care centers in Hong Kong. Attendees of the center were approached and smokers were identified through a preliminary screening with the help of staff of the clinic. Individuals who smoked were then invited to complete a baseline questionnaire and to undergo an exhaled carbon monoxide (CO) test administered by clinic staff or by the research assistants, who conducted the screening after the individuals had provided their written consent to be screened. Those who were interested in participating in the smoking cessation program and who met the criteria for inclusion were further invited to provide their written informed consent to take part in the study. The eligibility criteria were: (1) aged 18 years or above; (2) smoked at least one cigarette per day in the past 30 days; (3) not currently taking part in any other smoking cessation programs; (4) able to communicate in Cantonese; (5) a Hong Kong resident; (6) currently residing in Hong Kong and expecting to continue to do so for the next 6 months; and (7) have access to a telephone. The sample size calculations were based on the primary outcome of smoking cessation – the 7-day point prevalence quit rate as measured at the 12-month follow-up. Reference is made to the benchmark study by Bricker et al. [28] that used ACT by telephone approach where 29% of smokers quit smoking, compared with the 10% of smokers who quit with self-help materials [12]. To achieve a power of 80% at a 5% significance level for a chi-square test to detect a difference of 19% in the quit rate in 12 months (29% vs. 10%), a total of 136 subjects (68 per group) were required.

One hundred and forty-four individuals who met the criteria for inclusion and gave their written consent to participate in the study completed the baseline questionnaire. They were then randomized into either the intervention (ACT) group (n = 70) or the control group (n = 74). The process of randomization was based on computer-generated, block randomization with random block sizes, which were placed in sealed opaque envelopes. The randomization procedure was undertaken by another research assistant who was not directly involved in the study. The randomized study assignments were concealed from participants after eligibility were determined and consent to join this trial was obtained. Both groups received self-help materials on smoking cessation. Those in the ACT group also underwent an initial face-to-face session and two telephone ACT sessions at 1 week and 1 month following the first session. After 3, 6, and 12 months, they were re-contacted through telephone and were asked to complete the follow-up assessments of their smoking status, attempts to quit, and ability to deal with physical sensations, emotions, and thoughts without smoking. Those who reported as having quit at 6 and 12 months were then invited to undergo carbon monoxide (CO) and urinary cotinine tests. Participants of both groups also completed an evaluation survey through telephone at the 6- and 12-month follow-up sessions. From randomization through follow-up assessments, all research staff were blinded to the group allocations. The overall design of the study was illustrated in a published study protocol [30].

The primary outcome was self-reported 7-day point prevalence abstinence at the 12-month follow-up session. Secondary outcomes included: (i) a readiness to quit smoking; (ii) the average number of cigarettes consumed daily in the past 12 months; (iii) the number of attempts to quit in the past 12 months; (iv-vi) the difficulty of quitting, confidence in quitting, and the importance of quitting; and (vii) the level of psychological flexibility.

Statistical analyses were carried out using the Statistical Package for the Social Sciences (SPSS) version 25.0. All tests were two tailed with a significance level of p < 0.05. An intention-to-treat (ITT) principle was applied to all the analyses whenever applicable. The participants’ demographics, smoking and quitting information, number of sessions attended, attrition rate, and satisfaction with the program were examined by means of descriptive statistics. Baseline data were analyzed to check the comparability of the intervention and control groups. Chi-square tests were performed for the categorical variables. Two-tailed t-tests were performed for continuous variables that were normally distributed and homoscedastic, while Mann-whiney U tests were performed for those were not, respectively. Primary evidence of the efficacy of ACT was assessed using risk ratio for the primary outcome (self-reported 7-day point prevalence abstinence at 12 months) and mean differences for the continuous secondary outcomes between the intervention and the control groups. Their associated 95% confidence intervals (CI) were also reported. A chi-square test was used to examine the differences between the two groups in stage of readiness to quit smoking at the 12-month follow-up session. Comparisons were also made between the two groups with regard to changes in the secondary outcomes (stage of readiness to quit smoking, importance of quitting, difficulty in quitting, confidence in quitting, and psychological flexibility) from baseline to the 12-month follow-up, using independent t-tests for continuous variables and chi-square tests for the categorical variables. Hochberg’s set-up procedure was used to control the overall type 1 error as 5% for testing 11 hypotheses in the study. Missing data were imputed by the last observation carried forward approach. Two sets of sensitivity analyses were performed with (1) missing data replaced by the baseline score, and (2) complete cases only. Significant variables in the bivariate analysis and treatment group were included as independent variables in the logistic regression model to identify the predictors of quitting smoking at 12 months. The goodness-of-fit for the model was assessed by the Hosmer-Lermershow test, where p > 0.5 indicates an agreeable model fit. The estimated adjusted odds ratios and 95% confidence intervals (CI) were reported.

This study adheres to CONSORT guidelines and includes a completed CONSORT checklist as an addendum to this manuscript. Figure 1 shows the flow of participants through the phases of the RCT. A total of 3890 smokers were identified. Of the 423 smokers who agreed to complete the baseline questionnaires, 144 (34%) were eligible and agreed to be randomized into the study. The reasons given for not taking part in the study and the predictors of participation in the trial were reported eleswhere [42]. Seventy participants (48.6%) were allocated to the ACT group and 74 (51.4%) to the control group. There was no statistical significant difference in the retention rate between the two groups at the 3-month (intervention vs control: 65.7% vs 60.8%, p = 0.54), 6-month (50% vs 52.7%, p = 0.75) and 12-month follow-up (50% vs 41.9%, p = 0.33) respectively.

Table 1 shows the demographic and other baseline variables of the participants in each group. At baseline, one-fourth of the participants were female in the ACT group and one-third were female in the control group, while the mean age was 3 years older in the ACT group than in the contol group. The mean number of years of smoking was 19 years in the ACT group and 18 years in the control group. More than 70% of the participants smoked at home, and more than 60% in both groups had made a previous attempt to quit smoking. There were no significant differences between the intervention and control groups in any other sociodemographic variables, smoking-related variables, intention to quit, and perceived health status.

In Table 5, a multiple logistic regression analysis was performed to test all of the above significant variables. Several independent variables were found that significantly contributed to success at quitting smoking. They were: age (p < 0.01), monthly household income (p < 0.01), number of cigarettes consumed in the past month at the baseline assessment, and those who perceived themselves to be more confident about quitting (p < 0.001). Participants who underwent ACT were 96% more likely to quit smoking than those who had been allocated to the control group. However, this effect was not statistically significant.

An ACT counselor conducted all of the ACT sessions. This counselor is a professional member of the Association of Contextual Behavioral Science (ACBS) and has worked as a private ACT practitioner helping people with psychological difficulties or substance misuse. Throughout the study, his ACT skills were supervised by the principal investigator of the study and the first author of this report (YWM), a professional member of the ACBS and experienced ACT researcher. After each ACT session, the counselor completed the ACT competency checklist [20]. In addition, all completed ACT sessions of 18 out of 70 participants from the ACT group (the assessment rate was 24%) were audio-recorded and evaluated independently by YWM and another peer counselor. Fidelity findings indicated that the ACT counselor generally adhered well to the principles of ACT. The well-performed items (80–90% adhered to the principles – very satisfactory) were therapeutic stance, undermining cognitive defusion, and building patterns of committed action. Two principles were adhered to at a moderately satisfactory level, at between 60 and 70%. The two items were: used the concept of “workability” in interacting with the participants, and used appropriate exercises or metaphors to show willingness (acceptance/openness) as an effective alternative to avoiding difficult internal experiences or helping the participants notice cravings and urges (thoughts, feelings, and bodily sensations), while also contacting a self who chooses and behaves with these experiences, rather than for these experiences. During their weekly meetings, the counselor, YWM, and the peer counselor discussed any noteworthy discrepancies between the ACT protocol and the ACT sessions, and worked out strategies on how to adhere to protocol in the coming sessions.

Of the 70 participants who joined the intervention group, 50 (71.4%) completed all three sessions. On a scale of 1 (did not practice at all) to 5 (had regular practice), the participants in the intervention group were asked to rate their regularity in practicing or applying the techniques that they had learned for dealing with smoking cravings; the mean rating given by the ten participants who responded to this item was 2.20 (SD = 1.23).