Motivational Interviewing and Smoking Cessation: Translating Research into Practice with Fidelity

Special Article – Tobacco and Smoking Cessation

J Fam Med. 2016; 3(3): 1059.

Motivational Interviewing and Smoking Cessation: Translating Research into Practice with Fidelity

Battaglia C1,2*, Peterson J¹, Langner P¹, Whitfield E¹, Nandi A³, Benson SL4, Prochazka AV1,5 and Cook PF6

¹Department of Veterans Affairs, Eastern Colorado Health Care System, USA

²University of Colorado Anschutz Medical Campus, Colorado School of Public Health, USA

³InStil Consulting, USA

4Mental Health Partners, USA

5University of Colorado Anschutz Medical Campus, School of Medicine, USA

6University of Colorado Anschutz Medical Campus, College of Nursing, USA

*Corresponding author: Battaglia C, Department of Veterans Affairs, Eastern Colorado Health Care System 1055 Clermont Street, Research 151, Denver, CO 80220, USA

Received: April 22, 2016; Accepted: May 26, 2016; Published: May 30, 2016


Background: Motivational interviewing (MI) is an evidence based communication style effective for helping patients change their health behaviors such as smoking. We integrated an MI-based smoking cessation intervention into a home telehealth program for patients with Posttraumatic Stress Disorder (PTSD). This study presents treatment fidelity data that can be of use when implementing an MI intervention for individuals with PTSD into clinical practice.

Method: We assessed and monitored the MI-based smoking cessation intervention using a treatment fidelity framework which included five domains (study design, training, treatment delivery, patient receipt and enactment).

Results: Eighty-nine Veterans with PTSD who smoked were enrolled in the intervention arm of this study. Treatment fidelity was established by designing a study that mapped to a stage of change theory to deliver MI-based smoking cessation curricula via home telehealth plus weekly MI counseling calls. Initial and ongoing training by an MI expert ensured treatment delivery fidelity by nurse care managers. On average, participants received 12.25 calls lasting 16.7 minutes. They were satisfied with MI curricula (M= 9.1/11) and nurse counseling (98.5%). A majority (73.1%) of participants stated they wanted to quit. There was a significant difference in stage of change between baseline call and the last call with higher levels of stage of change at the last call (p=<0.0001). Enactment measures revealed participants smoked nine fewer cigarettes over time (p=<0.0001).

Conclusion: Implementing studies in health behavior change necessitates monitoring and demonstrating treatment fidelity. The results of this study can help guide translating smoking cessation research into practice in this hard to treat population.

Keywords: Motivational interviewing; Smoking cessation; Treatment fidelity; Posttraumatic Stress Disorder; Veterans


AARM: Avoid, Alarm, Replace, Mentally Cope; IQR: Interquartile Range; M: Mean; MI: Motivational Interviewing; MIASTEP: Motivational Interviewing Assessment: Supervisory Tools for Enhancing Proficiency; MINT: Motivational Interviewing Network of Trainees; MITI; Motivational Interviewing Treatment Integrity; NIH BCC: National Institutes of Health Behavioral Change Consortium; NRT: Nicotine Replacement Therapy; PTSD: Posttraumatic Stress Disorder; RCT: Randomized Control Trial; R:Q; Reflection to Question Ratio; SD: Standard Deviations; TTM: Transtheoretical Model of Change; VA; Veterans Administration


Motivational interviewing (MI) is a patient-centered, strengthsfocused communication approach to help patients change problematic behaviors [1], which is often used in clinical settings when the patient’s behavior affects treatment outcomes [2]. In 30 randomized controlled trials (RCTs) with over 9,000 total participants, MI has proven efficacious for multiple difficult-to-change behaviors including diet change, exercise, medication adherence, problem gambling, alcohol consumption, and various types of drug use.

The evidence for MI as a smoking cessation intervention is equally strong [3], although in one meta-analysis effect sizes for smoking tobacco were only about half as large as those that have been found for other types of behavior change [4]. One interpretation of this pattern of results is that smoking tobacco is a particularly difficult behavior to change [5]. Despite smaller effect sizes for smoking cessation in some MI studies, tobacco smoking remains a leading cause of mortality in the U. S. [6], and further intervention research is therefore essential. MI works at least as well as any other intervention for this intractable problem and better than many [4], and can be delivered in a relatively brief amount of time compared to other wellstudied patient counseling interventions [7].

MI is a good fit for Veterans with posttraumatic stress disorder (PTSD) who smokes because it is a non-confrontational communication style [8]. Smoking rates are higher among Veterans seen in the Veterans’ Administration (VA) health care system than in the general U. S. population [9], and military service is a frequent time of smoking initiation for patients. Furthermore, Veterans with PTSD have smoking rates of 45% to 66% [10], and PTSD appears to be a risk factor for treatment failure, with lower success rates in smoking cessation intervention among patients with PTSD compared to the general population [11,12]. Based on this combination of risk factors, helping Veterans with PTSD to quit smoking may be one of the most challenging applications of MI for health behavior change. Identifying successes as well as failures of MI in this context therefore may inform the use of MI in other clinical settings where the same type of barriers occur in less extreme forms.

We previously pilot tested the feasibility and treatment fidelity of MI for smoking cessation in a telephone-and-technology-based MI intervention to help this hard to treat population of Veterans with PTSD to reduce their tobacco use [13]. This small scale pilot study examined MI delivered by a single research nurse to a sample of 11 Veterans with PTSD. We then initiated a RCT to evaluate our MIbased smoking cessation intervention, which included a 90-session smoking cessation curriculum integrated into a PTSD home telehealth care management program with 12 weeks of MI telephone counseling to determine if smoking behaviors improved. Outcomes of this study are discussed elsewhere [14]. For this study, we used data obtained from the smoking cessation curricula and the weekly MI counseling treatment to test the treatment fidelity of the intervention delivered by two nurse care managers to Veterans with PTSD enrolled in the intervention arm of RCT.

The usual goal of treatment fidelity studies is to show that an intervention was delivered in a research study according to its original design; in other words, it is a manipulation check on the independent variable of an RCT [15]. Training on MI [16] and behavioral rating of counselors’ MI performance [17] are two elements of treatment fidelity. These factors are usually reported in MI research studies yet they may or may not have any direct impact on treatment outcomes [3]. Nevertheless, treatment fidelity measures also provide crucial information about how well an experimental intervention can be translated into practice settings, including barriers and facilitators that inhibit or enhance its use [18]. The current study therefore presents treatment fidelity data on MI for smoking cessation among Veterans with PTSD, with a focus on how and what these data have to say about the use of MI for smoking cessation in high-risk populations more generally.


Overview of treatment fidelity methods

The treatment fidelity framework [15] suggests five domains for evaluating MI interventions: 1) study design, 2) provider training, 3) treatment delivery by providers, 4) patient receipt of intervention, and 5) patient enactment of treatment. Study design includes elements such as an adequate dose and frequency of intervention [19], as well as considerations about the delivery method and whether these interact appropriately with the intervention content. For example, telehealth methods facilitate more frequent interaction with patients at distant locations [20], but they also might reduce the human connection that has been found particularly important for the success of MI [21]. Study design also captures theory problems such as those that have been noted when a more directive cognitive-behavioral intervention with specific goals for behavior change is combined with an MI intervention for increasing patients’ engagement in care and autonomous decision-making about their own health [22].

The second and third domains of the treatment fidelity framework are the most commonly reported in RCT research, focusing on training and successful delivery of interventions by providers [19]. Researchers report the details of the training procedure, qualifications of the trainers, amount of training, and follow-up supervision procedures, as well as behavioral measures of MI treatment delivery such as the Motivational Interviewing Assessment: Supervisory Tools for Enhancing Proficiency (MIA-STEP) [23] or the Motivational Interviewing Treatment Integrity (MITI) [24]. Although a recent large-scale meta-analysis found no relationship between provider training characteristics or behavioral MI implementation measures and MI treatment outcomes [3], information at these levels nevertheless provides important insights into the process of implementing MI in diverse practice settings.

The last two domains of the framework may require adaptation for studies of motivational methods such as MI in which patients are not necessarily expected to enact specific behaviors. Although the framework is clearly delineated for cognitive-behavioral interventions that rely on knowledge transfer from providers to their patients [19], the focus of MI is on helping patients to make decisions about their health behaviors. In this context, studies have sometimes evaluated patient receipt and enactment of MI interventions via data on the number and type of intervention components actually completed – this is related to the number intended at the study design level, but focused on the actual receipt of interventions by participants rather than what was originally designed. Patient satisfaction is another metric related to receipt of intervention, as in our pilot study [13]. Studies have also looked at patient-provided data on motivation or readiness for change as evidence for enactment in MI [25]. In our previous studies we included quit attempts as a measure of enactment in smoking cessation [13,14].


There were 89 participants who were randomly allocated to a MI telehealth intervention in a study of 178 total Veterans with PTSD who smoked. We enrolled Veterans who regularly smoked at least one cigarette/day whether or not they were ready or trying to quit.


The integrated smoking cessation intervention was designed to be completed over 12weeks with a six-month follow-up period. Participants in the intervention group received 90 daily sessions of the MI-based smoking cessation curricula via a PTSD home telehealth program (the Enhanced PTSD home telehealth program) and weekly individual telephone MI counseling calls from three nurse care managers intended to last for 20 minutes. Our stage-based MI intervention was built upon the core tenets of the transtheoretical model of change (TTM) [26]. TTM provided a conceptual framework of how and why change occurs and MI enhanced personal motivation to change [27]. We “nudged” participants using a MI counseling through stages (precontemplation, contemplation, preparation, action, and maintenance), from no interest in making a change (precontemplation) to sustained change over time (maintenance). We integrated TTM and MI to increase smoking cessation in a high-risk, vulnerable population.

Throughout the smoking cessation telehealth written curricula, participants were asked to think about a topic and if ready, discuss it with their nurse care manager during the next weekly call. For example in a daily addiction curriculum, we addressed prior quit attempts and if the participant had tried to quit twice before the written curriculum stated, So, you’ve practiced twice before to quit smoking. If you haven’t done so already, tell your nurse care manager about it the next time you speak. When the nurse care manager spoke to such a participant she then tailored her MI counseling to his stage of change to encourage the participant to make a behavior change. If the participant was in the contemplation stage of change and had tried to quit twice before, the nurse care manager would explore what happened during those quit attempts, how long they lasted, how the participant felt, or what benefits he noticed. Based on a manualized treatment protocol, the nurse care managers explored how important the participant believed it was to quit smoking, how confident the participant felt that he/she could succeed, and/or what resources the participant needed to be successful.


Study design fidelity: Design fidelity is achieved when the components of an intervention map cleanly to its underlying theory [19]. This can be challenging to achieve in the case of motivational interviewing, which evolved clinically in the absence of a strong theory base [28]. Our assessment of design fidelity included review of the intervention protocol by independent experts in MI [13], a review of the number and type of participant contacts, and a check on the qualifications of providers who delivered the MI intervention [19].

Training fidelity: Training fidelity was measured based on the MI training the nurse care managers completed as well as the qualifications of the trainers, the number and type of training sessions provided, and the type of follow-up training and supervision provided.

Treatment delivery fidelity: Treatment delivery fidelity was measured using several strategies. A treatment manual was used to standardize the intervention and a self-report MI skills checklist was completed by the nurse care managers after each MI session was delivered. The checklist served to document frequency and type of MI consistent skills such as reflection, educating, using elicitprovide- elicit, and open or closed questions, as in our prior study [13]. Additionally, a small subset of sessions was evaluated using a validated behavioral coding system, the MITI [29], completed by an independent rater who is a member of the Motivational Interviewing Network of Trainers (MINT). Participants were surveyed about their satisfaction with the MI counseling phone calls at the end of the intervention period, which captured their opinions regarding how the treatment was delivered by the nurse care managers.

Patient receipt of intervention: Receipt was measured by patient satisfaction scores with MI content.

Additionally, we monitored the participant’s stage of change as evidence that the nurse care managers’ successfully reached the patient for each intended component of the MI-based smoking cessation intervention. We also considered data on the amount and reasons for attrition as evidence related to patient receipt of the intervention; this type of data may provide important information about the feasibility of MI in practice.

Patient enactment: We evaluated enactment based on several MI scales completed by participants via the Enhanced PTSD home telehealth program. These included: readiness to change, importance of change, and confidence level to make a change. Additionally, we assessed answers to specific questions in smoking cessation curricula and stage-of-change on a monthly basis in the written telehealth curricula and during the weekly MI counseling calls. We also assessed number of cigarettes smoked/day over time. Moreover, we tracked the participants’ change talk during each MI counseling call.

Data analysis

Descriptive statistics including means and standard deviations (SDs), median, ranges, and frequency distributions were used to describe baseline characteristics, study completion rates, MI skills checklist data, participant satisfaction surveys, and MI rating scales. Dichotomous variables were summarized with proportions. Measures for evaluating treatment fidelity of the MI-based smoking cessation intervention used the treatment fidelity framework by the National Institutes of Health Behavioral Change Consortium (NIH BCC) which includes five domains of treatment fidelity [19] as described above.

Repeated measures analysis using linear mixed modeling was conducted to evaluate number of cigarettes smoked/day during the intervention and during the follow-up periods. We limited our analysis on number of cigarettes smoked/day to participants who completed the study to ensure a consistent sample size for the repeated measures analysis. During the intervention period, if a participant missed a quit attempt question, we were able to impute with their daily quit attempt responses. However, for the follow-up period, we had eight participants who missed at least one quit attempt question and were removed from the analysis because these patients were not asked quit attempts question daily.

The Wilcoxon signed-rank test was used to test for a difference in stage of change between the first and last calls (movement up or down the range of stages). To test for a difference in the proportion of subjects mentioning change talk in the first call vs. the last call, McNemar’s test for correlated proportions was used. This method was subsequently applied to use of commitment change talk and taking steps change talk individually. Analyses were performed using SAS version 9.4 (SAS Institute, Cary, NC).


We enrolled 89 Veterans with PTSD who smoked into the intervention arm of an integrated care management telehealth and smoking cessation intervention. We only assessed the intervention arm for treatment fidelity. Table 1 shows baseline characteristics for those enrolled in the intervention (n=89). These characteristics did not differ from those randomly assigned to the control group [13,14]. The majority of participants were male (83.1%), white (67.4%), and unemployed (87.6 5%). The average age was 54.6 years with a greater proportion of participants being older than 50 years (74.2%). Slightly over half (57.3%) of participants self-reported being in the contemplation stage of change when asked on the Enhanced PTSD home telehealth program at the beginning of the study.