If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Homewood Research Institute, Guelph, ON, CanadaPeter Boris Centre for Addiction Research, St. Joseph's Healthcare Hamilton, Hamilton, ON, CanadaDepartment of Psychiatry and Behavioural Neurosciences, McMaster University, Hamilton, ON, Canada
High rates of tobacco use within an inpatient addiction treatment population.
•
Inpatient tobacco use decreased among group exposed to a tobacco-free environment.
•
Comprehensive approaches (combination of cessation program and tobacco ban) most effective.
•
Reductions also observed in tobacco use frequency and amount used.
•
Some patient characteristics (e.g., sex, age) may promote tobacco cessation.
Abstract
Purpose
Tobacco use is highly prevalent within addiction treatment settings, despite the potential benefits that cessation may provide to patients' psychosocial functioning and overall sobriety. Moreover, tobacco cessation is often insufficiently addressed in addiction treatment programs, although evidence suggests that tobacco control policies, such as access to evidence-based treatment or tobacco-free policies, may be effective. The objective of our study was to evaluate the impact of the implementation of these two tobacco control policies in an inpatient addiction treatment centre.
Methods
Using a 3-group quasi-experimental design, we examined how the implementation of the Ottawa Model for Smoking Cessation (OMSC) and a subsequent campus-wide tobacco ban influenced patients' overall smoking status, frequency, amount used per day, and quit attempts during treatment, compared to usual care. Participants (N = 397) responded to a comprehensive questionnaire upon admission and discharge from an addiction treatment program. We used generalized linear mixed modelling to measure changes over time while accounting for relevant sociodemographic covariates.
Results
Patients exposed to a more comprehensive tobacco control environment (i.e., the OMSC, plus complete tobacco ban) were over 80% less likely to report having used tobacco during treatment, compared to patients exposed to usual care (AOR = 0.17, 95% CI [0.05–0.63]). Receiving treatment in this setting also contributed to a 35% decrease in the average number of days patients used tobacco compared to usual care (AOR = 0.65, 95% CI [0.53–0.98]), and a 27% decrease in the average number of cigarettes used per day compared to usual care (AOR = 0.73, 95% CI [0.58–0.93]).
Conclusion
Comprehensive tobacco control policy interventions within inpatient addiction treatment hospitals promote tobacco cessation. Such interventions should include a combination of evidence-based treatment for patients and environmental restrictions to discourage tobacco use. The results of our study suggest that, within inpatient addiction treatment settings, use of the OMSC in combination with a campus-wide tobacco ban may be more effective than usual care or the OMSC alone.
Tobacco use among the general population has decreased substantially over the past few decades; recent estimates point towards a smoking prevalence of 15% in Canada (Reid et al., n.d.) and 14% in the United States (
Wang, T. W., Asman, K., Gentzke, A. S., Cullen, K. A., Holder-Hayes, E., Reyes-Guzman, C., … & King, B. A. (2018). Tobacco product use among adults—United States, 2017. Morbidity and Mortality Weekly Report, 67(44), 1225.
). However, smoking rates remain high among people seeking treatment for substance use disorders (SUDs) – well above rates among persons with SUDs in the general population (
). The negative consequences of tobacco use for SUD patients are substantial; increased morbidity and mortality observed among those with SUD are largely correlated with preventable tobacco-related causes (
). Perhaps then it is not surprising that clinicians often anecdotally report some patients take up tobacco while in inpatient SUD settings. However, contrary to many of these beliefs, previous research has found that reduction or cessation of tobacco is associated with improvements of psychiatric symptoms, and supports better SUD treatment outcomes (
Comprehensive tobacco control efforts can have a positive impact on tobacco use rates, even among “high risk” populations (e.g., SUD patients) and in hospital settings (
). A growing body of literature points towards improved psychosocial and treatment outcomes among individuals with SUD who quit smoking. For example, research shows simultaneous treatment of nicotine addiction during inpatient SUD treatment is correlated with increased likelihood of maintaining abstinence from alcohol and/or drugs over time (
Clinical practice guidelines recommend clinicians systematically identify and treat tobacco use among individuals with mental illness and/or SUD as well as individuals within hospital settings (
Canadian smoking cessation clinical practice guideline.
Canadian Action Network for the Advancement, Dissemination and Adoption of Practice-informed Tobacco Treatment, Centre for Addiction and Mental Health,
Toronto, Canada2011
). This includes systematically screening patients for tobacco use, providing advice and supports to quit (i.e., pharmacotherapy, behavioural counselling), and co-ordinating longer-term follow-up and counselling (
Canadian smoking cessation clinical practice guideline.
Canadian Action Network for the Advancement, Dissemination and Adoption of Practice-informed Tobacco Treatment, Centre for Addiction and Mental Health,
Toronto, Canada2011
). Hospital-based smoking cessation interventions that incorporate these recommendations, including the Ottawa Model for Smoking Cessation (OMSC), have been shown to effectively increase tobacco abstinence rates within inpatient settings (
). In addition, tobacco control policies that restrict use of tobacco within indoor and outdoor environments have proven to be effective in lowering population smoking rates across a variety of settings, including general hospitals (
). As a result, some jurisdictions have implemented legislation requiring all public and private hospitals and psychiatric facilities (and grounds) to be entirely smoke-free, including the Smoke-Free Ontario Act—a provincial-level legislation in Canada. Collectively, however, there remains a paucity of research addressing the impact that comprehensive tobacco control measures may have within inpatient SUD treatment settings, specifically.
Moreover, few studies have evaluated the impact tobacco-free policies have on patient-level indicators and outcomes within SUD populations. While a variety of sociodemographic characteristics – including age, gender, educational attainment, employment status, and marital status – may be associated with SUD (
), there are few studies investigating the role these factors may have on tobacco use and cessation during inpatient SUD treatment. As such, more research is needed to help understand the impact tobacco control interventions may have on tobacco use and cessation among individuals receiving SUD treatment within an inpatient setting.
Using a three-group, quasi-experimental design with repeated measures, we evaluated the impact of two tobacco control interventions implemented within an inpatient SUD treatment setting located in a larger, mental health and addiction (MHA) hospital in Ontario, Canada. In this study, we took advantage of a natural experiment which first saw the implementation of a hospital-wide tobacco cessation intervention (i.e., OMSC;
), followed by the implementation of a complete tobacco ban on hospital campus. We used data from the Recovery Journey Project (RJP) – a study that aims to better understand the recovery process among those who receive SUD treatment (
) – to address whether exposure to these tobacco control interventions impacted patients': 1) tobacco use status, and 2) tobacco use behaviours, over the course of their SUD treatment. Given the demonstrated effectiveness of tobacco bans in other clinical settings, as well as evidence supporting the benefits of tobacco cessation for SUD patients, we hypothesized that exposure to a comprehensive tobacco-free environment would have a positive effect on patients' tobacco use during treatment. More specifically, we predicted exposure to comprehensive tobacco control interventions, compared to no exposure, would be associated with: 1) a reduction in tobacco use prevalence, frequency, and amount used, and; 2) an increase in self-reported quit attempts during treatment (i.e., between admission and discharge).
2. Methods and Materials
2.1 Setting
This study took place in large, 105-bed, inpatient SUD treatment program located in Guelph, Ontario, Canada where the provincial smoking prevalence was 13% (
). The program is embedded within a larger MHA hospital, situated on a 50-acre campus, that offers a range of inpatient services (e.g., 300 inpatient beds), elective, and community-based programs designed for people with a range of complex, intensive behavioural health needs. At the time of the study, the program offered 35-day, group-based treatment to adults aged 18+ with alcohol and/or other substance use disorders accommodating approximately 1050 patients each year. Patients reside within the inpatient setting for the duration of their treatment with access to full hospital grounds; but, may leave campus to access to the surrounding residential area when given permission from their clinical team. The program used an abstinence-based approach to SUD recovery, informed by 12-step facilitation, and other evidence-based modalities such as cognitive behaviour and dialectical behaviour therapies. Treatment is led by multidisciplinary teams including registered addiction counsellors, addiction medicine certified physicians, registered nurses, occupational therapists, social workers, recreation therapists, and a tobacco cessation specialist. Specialized 56-day treatment, integrated within the SUD program, is offered for patients with co-occurring symptoms of post-traumatic stress disorder (PTSD). Treatment is paid for through private (e.g. out-of-pocket expenses), semi-private (e.g. private health insurance), and provincial (public) funding.
2.2 Tobacco cessation treatment and policy
Policy-level efforts to reduce tobacco use resulted in the tobacco cessation interventions described here. Prior to introduction of these interventions, patients who used tobacco were provided with usual care which included advice to quit tobacco use and access to nicotine replacement therapy (NRT), if desired. At this time, patients were permitted use of a designated outdoor smoking area on the hospital's campus. On April 4, 2016 an evidence-based cessation program, the OMSC (
), was introduced in addition to the usual clinical care described above. The OMSC presented clinical staff with a method of collecting patients' tobacco use history, so as to systematically identify those who use tobacco, advise them to quit, and offer cessation treatment and follow-up post-discharge.
On September 6, 2016 a campus-wide Tobacco-Free Policy (TFP) was introduced that prohibited all use and possession of tobacco products by patients, staff, and visitors on the hospital's campus. Incoming patients were made aware of the TFP and the expectation that they were not to use or possess tobacco while in treatment. The designated smoking area on hospital property was removed, and the use or possession of tobacco during treatment resulted in early intervention to support cessation, with the potential of discharge from the program for non-compliance. In addition, treatment using the OMSC continued to be in effect post-policy implementation. Fig. 1 provides a more detailed description of the three distinct intervention periods within the treatment setting.
Fig. 1Descriptions of tobacco cessation treatment and Tobacco-Free Policy interventions within the AMS treatment setting, comprising three quasi-experimental exposure groups.
Data presented in the current study were derived from the larger RJP host study – an ongoing, longitudinal cohort study and research platform designed to evaluate post-discharge recovery outcomes and generate evidence about the recovery process among individuals receiving inpatient treatment for SUDs (
). The RJP is one of the only ongoing longitudinal studies in Canada designed to generate new knowledge about the SUD recovery process, both during and following treatment. The RJP consists of repeated measurement of various indicators of recovery at two time-points during treatment (admission, discharge) as well as 1-, 3-, 6-, and 12-months post-discharge. The RJP received ethics clearance from the Regional Centre for Excellence in Ethics, Research Ethics Board in Guelph, Ontario, Canada. More information about the RJP can be found in print (
The present study used a quasi-experimental design to compare three distinct groups of participants: (1) those who received ‘usual care’ and were not exposed to the TFP; (2) those who received an evidence-based tobacco cessation program and were not exposed to the TFP; and, (3) those who received an evidence-based cessation program and were exposed to the TFP. Among these three groups, we used admission and discharge data from the RJP to detect: 1) changes in tobacco use status over time among the general sample of patients, and 2) changes in tobacco use and cessation behaviours over the course of treatment among those who indicated having used tobacco 90 days prior to admission.
2.4 Procedures
All patients admitted to the program, for 2 days or longer (i.e., those who had completed the full hospital intake process), between April 1, 2015 and March 31, 2017 were eligible to participate in the host study. Within the first week of admission, patients were scheduled to attend a session facilitated by research staff, who introduced the RJP and informed them of what their participation in the project would entail. Patients provided informed, written consent to participate in the project and completed the self-administered baseline Recovery Questionnaire (RQ) using a computer tablet. Participants attended a second group session within the last week of their program stay and completed a second RQ. The RQ is comprised of set of psychometrically valid scales and measures that assess a variety of recovery outcomes directly related to alcohol and drug use treatment, as well as tobacco use and cessation behaviours (
Between April 1, 2015 and March 31, 2017, 808 patients consented to participate in the host study and completed the baseline RQ. Participants were assigned to one of three policy exposure groups based on admission and discharge dates (as described below). Of the baseline sample included in the analyses (n = 714), 91 were discharged early from treatment (i.e., discharged home unplanned or signed out against medical advice). Rates of early discharge did not differ significantly across the three exposure groups (results not shown). However, participants who completed treatment differed from those who were discharged prematurely by age (42.2 years vs. 38.2 years, p < .01) and baseline tobacco use (59% vs. 76%, p < .01). No other significant differences were detected between groups in key demographics or other baseline measures under study (results not shown).
To address our first research objective, our analytic sample was restricted to those who completed treatment (i.e., those who were discharged as planned) and who completed the RQ again at the discharge time-point (n = 397). Compared to those who only participated at baseline, participants who completed the RQ at both time points were, on average, older (43 years vs. 40 years, p < .01); proportionally more had completed at least some university (47% vs. 35%, p < .01) and were married or partnered (59% vs. 45%, p < .01); and, fewer had reported using tobacco in the past 90-days (56% vs. 70%, p < .01), respectively. No other significant differences were detected between groups in key demographics or other baseline measures under study (results not shown). To address our second research objective, we further restricted our analytic sample to those who indicated having used any tobacco 90 days prior to admission (n = 218). Refer to Supplementary File A for a consort diagram depicting sample participation and retention.
2.6 Measures
2.6.1 Independent variables
2.6.1.1 Policy exposure
Participants were categorized into one of three exposure groups based on their admission and discharge dates. Those admitted and discharged between April 1, 2015 and April 3, 2016 represented the ‘usual care’ referent group (Group 1). Patients admitted on or after April 4, 2016 and discharged before September 6, 2016 were labelled as ‘Group 2,’ as they had received the evidence-based OMSC program. Finally, those admitted after September 6, 2016, who were exposed to the evidence-based cessation intervention as well as the TFP, were labelled as ‘Group 3’. Group membership was mutually exclusive; that is, no participant was exposed to more than one condition. Each group is described further in Fig. 1.
2.6.1.2 Demographics
The RQ asked participants about several demographic characteristics (
). At baseline, participants reported their sex (male, female), and age (in years), which was categorized into ‘19 to 29 years,’ ‘30 to 39 years,’ ‘40 to 49 years,’ or ‘50 years and older’ for descriptive purposes. In addition, participants were asked about their education (categorized as: ‘completed high school or less,’ ‘at least some college,’ or ‘at least some university’), employment (categorized as: ‘employed’ or ‘not employed’), and marital status (categorized as: ‘married or partnered’ or ‘not married or partnered’).
2.6.1.3 Treatment characteristics
Treatment was characterized by noting the stream participants completed (i.e., ‘addiction only’ or ‘addiction and PTSD’). Participants also responded to several questions on the RQ about their use of substances prior to being admitted to the program (
). Specifically, at baseline, participants were asked to report their use of substances during the 90 days prior to admission. They responded with “yes” or “no” to a list of substances before being prompted for more details about their use of each specific substance. Substances were recategorized as such: alcohol; cannabis; opioids (including heroin or heroin mixed with any drugs, non-prescription methadone, or non-prescription opioids other than heroin); cocaine and stimulants; and other drugs (including non-prescription sedatives, non-prescription steroids, inhalants, and any other drugs used for the purpose of getting high or for a use other than was intended). Categories are not mutually exclusive as it was possible for participants to report using substances from more than one category. Patients were also categorized according to whether their discharge event was coded as planned (i.e., completed treatment) or unplanned.
2.6.2 Tobacco use and cessation behaviours.
The RQ included questions about patients' tobacco use and tobacco cessation behaviours (
). Baseline tobacco use was captured among all participants using the question: “during the past 90 days… have you used any tobacco products? (e.g., cigarettes, pipes, cigars, chewing tobacco, snuff or snus).” Response options were “yes” (i.e., used tobacco) or “no” (i.e., did not use tobacco). Those who reported using tobacco were then asked a series of follow-up questions pertaining to their use, including: “on approximately how many days during this period did you use tobacco products,” “on the days you used tobacco, about how much did you use, on average? [cigarettes smoked per day],” and “have you stopped using tobacco for one day or longer because you were trying to quit?” These questions were repeated at discharge using a 30-day timeframe to capture participants' time in treatment. Any tobacco use was used as the primary outcome to address the first research objective. To address the second objective, indicators of tobacco use (i.e., number of days of tobacco use
The RQ also allowed participants to report the number of “pipes/cigars smoked per day,” and “chew/pouches/other used per day”; however, due to cell count constraints within these measures, we were only able to analyze the self-reported number of cigarettes smoked per day among respondents.
) and cessation behaviours (i.e., made quit attempt vs. no attempt) were examined.
2.7 Data analysis
Baseline characteristics of the general patient sample (N= 397) were summarized and compared across the policy exposure groups (Group 1: usual care, no ban; Group 2: evidence-based program, no ban; Group 3: evidence-based program + ban). Differences between groups on demographic characteristics, program stream, substance use, and smoking behaviours were compared using Pearson's chi-square tests for categorical variables and one-way analysis of variance tests for continuous variables. These tests were repeated among a subsample of participants who reported tobacco use at baseline (n = 218). A generalized linear mixed-modelling (GLMM) approach was then used to predict tobacco use status between admission and discharge time-points among the general sample, while testing for policy exposure, age (as continuous), sex, education, and marital status (Model 1). The same GLMM approach was repeated to predict tobacco use outcomes over time (number of days used, cigarettes smoked per day, quit attempt vs. not; Models 2–4) among the subset of participants. The Laplace approximation method was used for maximum likelihood estimation across all models. We specified the logit link function for Models 1 and 4 with binary outcomes and assumed Poisson distribution for Models 2 and 3, where number of days used and number of cigarettes per day were treated as count variables. Adjusted odds ratios are reported alongside 95% confidence intervals (α = 0.05). SAS software (version 9.4) was used for all analyses (
On average, among the general sample, participants were aged 42.7 years, 67.4% were male, and most identified as white (92.8%). The majority of participants (81.8%) had at least started or completed post-secondary school (college, university), and over 80% were employed. Over half reported being married or partnered (56.7%) and 13.9% were enrolled in the co-occurring PTSD-symptom program stream. Over half of participants reported using tobacco in the 90 days prior to treatment (56.0%) and there was no significant difference between policy exposure groups in the proportion of participants who reported using tobacco at baseline. There were no significant differences between participants in each group on the demographic, treatment, or tobacco use variables examined. Refer to Supplementary File B for a table summarizing the baseline characteristics of the study's full sample, collected at the admission time-point.
Among the subsample of participants who reported baseline tobacco use (n = 218), 62.7% reported daily use. This subsample also reported using tobacco on an average of 25.1 (SD = 9.4) days out of the past 30 days, and reported using 14.8 (SD = 9.0) cigarettes per day on average. Approximately one in three (30.5%) reported making at least one quit attempt in the 90-days prior to treatment. Among the subsample of those who reported using tobacco at baseline, there were no significant differences between participants in each group on the variables examined, except for use of ‘other drugs’ (X2 = 7.5, p < .05) (Table 1).
Table 1Baseline characteristics of participants who reported using tobacco (n = 218), collected at the admission time-point.
Refers to number of days participants reported using tobacco at baseline, standardized to the past 30 days before admission.
25.1 (9.4)
25.1 (9.4)
25.0 (9.7)
25.1 (9.3)
0.0 (0.998)
Cigarettes per day
Mean (SD), cigarettes
14.8 (9.0)
14.0 (8.2)
13.8 (7.7)
16.4 (10.4)
1.7 (0.187)
Tobacco quit attempt
No attempt
146 (69.5)
73 (70.2)
24 (70.6)
49 (68.1)
0.1 (0.945)
Made a quit attempt
64 (30.5)
31 (29.8)
10 (29.4)
23 (31.9)
Note: The full study sample size is N = 397. This table only represents those who reported using tobacco at baseline (n = 218, 56%). There was no significant difference in overall tobacco use among patients in Groups 1, 2, or 3 at baseline. Supplementary File B further summarizes baseline characteristics of the entire sample.
Note: Bolded values represent statistical significance at α < 0.05.
1 Refer to Fig. 1 for description of intervention status groups.
2 Cumulative percentages do not add up to 100 as these groups are not mutually exclusive.
3 Refers to number of days participants reported using tobacco at baseline, standardized to the past 30 days before admission.
Results from the GLMM analyses are presented in Table 2. Model 1 predicts the log-odds of all participants' self-reported tobacco use, between admission and discharge time-points. Participants who were exposed to the evidence-based program and the tobacco-ban (Group 3) had significantly lower odds of using tobacco over the 30-day period prior to discharge compared to participants who received usual care and no policy (Group 1). We observed no significant difference in the odds of tobacco use over time among those in Group 3 versus Group 2 or Group 2 versus Group 1. Participants who completed at least some university education, and who were married or partnered, had lower odds of overall tobacco use at discharge compared to participants who completed high school or less and who were not married or partnered (Model 1). On average, there was a 15% decrease in the likelihood of using tobacco by discharge per year increase in patient age (Model 1). Refer to Fig. 2 for a visualization of trends in overall tobacco use over time, from baseline to discharge, for participants in each policy exposure group.
Table 2Generalized linear mixed-model results predicting changes in tobacco use outcomes over time, between admission and discharge time-points.
Refer to Fig. 1 for description of intervention status groups.
Group 2 (ref = Group 1)
0.47 (0.11–2.06)
0.97 (0.75–1.26)
0.99 (0.74–1.33)
0.90 (0.42–1.92)
Group 3 (ref = Group 2)
0.36 (0.08–1.70)
0.67 (0.51–0.89)
0.74 (0.54–1.01)
1.90 (0.83–4.33)
Group 3 (ref = Group 1)
0.17 (0.05–0.63)
0.65 (0.53–0.98)
0.73 (0.58–0.93)
1.71 (0.91–3.21)
Model 1: Predicts the log-odds of reporting using tobacco versus not using tobacco over time, during treatment.
Model 2: Predicts the log-odds of a one-day increase in number of days tobacco was used (treated as a count variable), standardized to the past 30 days.
Model 3: Predicts the log-odds of a one-cigarette increase in number of cigarettes used per day (treated as a count variable).
Model 4: Predicts the log-odds of at least one quit attempt versus no quit attempts over time.
Note: Model 1 represents all participants (N = 397); Models 2–4 represent the subsample of participants who reported tobacco use at baseline (n = 218).
Note: Bolded values represent statistical significance at α < 0.05.
1 Refer to Fig. 1 for description of intervention status groups.
Fig. 2Average trends in tobacco use and tobacco use behaviours over time, between baseline and discharge time points, by tobacco policy exposure group.
Models 2–4 (Table 2) were used to predict tobacco use behaviours among the subsample of participants who reported baseline tobacco use, between the baseline and discharge time points. There was a significant decrease in average number of days tobacco was used among those in Group 3 versus Group 1 (Model 2); this was also true among those in Group 3 compared to those in Group 2. In Model 3, there was nearly a 30% decrease over time in the number of cigarettes participants reported using per day among those in Group 3 compared to Group 1. There were no differences among exposure groups for quit attempts made during the 30-days prior to discharge (Model 4). Increasing patient age (in years) was associated with a decrease in average number of days used (Model 2), and being female was associated with fewer reported cigarettes used per day (Model 3). Fig. 2 depicts trends in tobacco use behaviours for which statistical significance was detected (number of days used, and cigarettes per day), by policy exposure group.
3.4 Further exploratory analyses
Although rates of early discharge among the baseline sample did not differ significantly across the three exposure groups, when we combined Groups 1 and 2 (i.e., those not exposed to the tobacco ban) and compared them to Group 3 (i.e., those exposed to the ban), we detected a slightly higher rate of early discharge among those exposed to the ban versus those not exposed to the ban (17% vs. 11%, p = .02). Furthermore, among participants who reported using tobacco at baseline, those in Group 3 had slightly higher rates of early discharges compared to those in Groups 1 and 2, collectively (21% vs. 13%, p < .04). Refer to Supplementary File C for tables summarizing these exploratory analyses.
4. Discussion
Tobacco use was highly prevalent within our sample of patients receiving inpatient SUD treatment. More than half reported using tobacco upon admission, and of those individuals, the majority used tobacco daily and heavily. Our results are consistent with previous literature demonstrating exceedingly high rates of tobacco use within the SUD treatment population (
). We found comprehensive tobacco control interventions may be effective in promoting tobacco cessation and preventing uptake during patients' inpatient stay; introduction of a full tobacco ban contributed to an 83% reduction in the proportion of self-reported tobacco use over time. Where approximately six million Canadian adults meet criteria for SUD (
), these findings highlight a large yet overlooked population in need of comprehensive tobacco control interventions to promote and support tobacco cessation.
Our study also provides evidence that SUD patients who are exposed to more comprehensive tobacco control interventions will be less likely to use tobacco during treatment. We found that compared to standard clinical care, use of a complete tobacco ban was effective in: 1) collectively reducing overall smoking prevalence by >80% among patients; 2) decreasing the average number of days tobacco was used by 35%, and; 3) decreasing the average number of cigarettes patients reported using per day by almost one third during treatment. Our analyses demonstrate the potential effectiveness of comprehensive tobacco cessation strategies that include campus-wide tobacco bans, and which restrict tobacco use beyond the mandate of clinical interventions such as the OMSC.
Notably, despite these results, we also find no indication that exposure to comprehensive tobacco control interventions is associated with attempts to quit using tobacco. This null finding warrants further research as to whether patients receiving SUD treatment in an inpatient facility perceive their reductions in tobacco use as quit attempts, or, whether they attribute their cessation behaviour simply as compliance with program policies and rules. Our findings are consistent with existing tobacco control literature. Previous research shows that tobacco bans alone may have little or no effect when these policies fail to include additional cessation supports and strategies for patients in SUD treatment settings (
). Conversely, as we have demonstrated here, tobacco cessation interventions in inpatient SUD treatment centres have been found to be most effective when comprehensive and systematic (i.e., strategies with integrated tobacco bans and evidence-based cessation treatment) (
). Within our own analysis, exposure to the OMSC alone without a campus-wide tobacco ban represents a relatively small timeframe (April 4, 2016 to September 5, 2016) in which early implementation may not have been fully adherent to all components of the model. In practice, the OMSC may not have been drastically different to what we classify here as usual care. Nonetheless, we found the most consistently effective policy intervention to be one in which evidence-based programming (the OMSC) was offered within the context of a campus-wide tobacco ban.
While the use of an evidence-based cessation intervention alongside a complete tobacco ban was shown to be beneficial in reducing tobacco use among an inpatient SUD treatment population, the results of our additional exploratory analyses suggest that treatment during the campus-wide tobacco ban may be associated with early discharge among patients who use tobacco. Perceived concern over patients' reluctance to complete SUD treatment has been previously described in the literature as a barrier to tobacco ban implementation (
Do total smoking bans affect the recruitment and retention of adolescents in inpatient substance abuse treatment programs? A 5-year medical chart review, 2001–2005.
Journal of Substance Abuse Treatment.2007; 33: 279-285
). Nonetheless, our data suggest hospitals with inpatient SUD treatment programs instituting similar campus-wide tobacco bans may anticipate a modest decrease (6%) in retention among tobacco users; however, it remains the case that they may also expect a significant reduction (35% fewer days used and 27% fewer cigarettes per day) in tobacco use behaviours among tobacco users who complete treatment. In light of the current study and existing literature, further investigations are required. Uncertainty as to whether the observed phenomenon is a direct consequence of the tobacco-free policy remains a caveat of the present findings.
Beyond exposure to tobacco control interventions, our study identified several characteristics of patients that may be associated with tobacco use during inpatient SUD treatment. Our results show that reductions in overall smoking status and cigarettes per day were associated with increased patient age, higher education, and being married or partnered. These sociodemographic correlates are largely consistent with observed tobacco use patterns among the general population (
), pointing to the potential success of targeted tobacco cessation efforts even within SUD treatment settings, including youth, those with lower educational attainment and/or those who are not married or partnered. This information may be useful for clinicians, counsellors, and tobacco cessation specialists when developing individual- and/or group-based programs for SUD patients, ideally within treatment settings that have also implemented comprehensive campus-wide tobacco bans.
4.1 Strengths and limitations
The adoption of natural experimentation to evaluate the effectiveness of two tobacco control policies (i.e., an evidence-based cessation program and a complete tobacco ban) on inpatient tobacco use, compared to usual care, is a core strength of the present study. In contrast to a randomized control trial design, the use of a three-group quasi-experimental design capitalized on the external implementation of tobacco control interventions that were not manipulated in any way by the research team. Our findings represent the effectiveness of such policies under real-world conditions (
). This study provides timely and robust evidence to help further justify the use of comprehensive tobacco control interventions within similar treatment settings. A secondary strength of this study is the inclusion of multiple indicators for tobacco use behaviours. These additional outcome measures (e.g., number of cigarettes used per day) provide a multidimensional perspective to tobacco use beyond classifying patients as those who use tobacco versus those who do not use tobacco.
Despite its strengths, the present study should be interpreted in light of several limitations. First, although patients were assured that their answers to the RQ were kept private from clinical staff, some patients may have under-reported their tobacco use. There are a variety of likely explanations for this possibility, ranging from social desirability bias to fear of being reprimanded (despite RJP protocols that assure patient confidentiality). Moreover, we were not able to capture treatment dropout attributable to the TFP. Although existing evidence demonstrates the validity of self-report tobacco use measures within the general population (
). Our study results are also subject to potential recall bias, particularly due to the RQ's long (i.e., 90-day) timeframe prior to admission. This bias may have been carried forward by our attempt to standardize the admission and discharge timeframe (i.e., by dividing baseline estimates by three). Next, findings represent tobacco use captured only during patients' inpatient treatment stay. Some patients may very well resume their tobacco use after discharge; a possibility which warrants further research. It should be noted as well that despite our best efforts to control for confounding sociodemographic factors, the results of this study may not be generalizable to other inpatient SUD treatment settings within or outside of Canada. For example, participant responses indicate that our sample is largely white, male, older, and well-educated. It is possible that our participant sample reflects a greater access to SUD treatment services; further research is warranted within more diverse contexts.
Finally, since the interventions evaluated here were not experimentally designed or implemented by the researchers, there are limitations with respect to unbalanced sample sizes and times-to-effect. Despite moderately large analytic samples, Group 2 was notably smaller than Groups 1 and 3 and as a result, may have been underpowered. It is possible that within the relatively shorter timeframe of the OMSC-only period (i.e., Group 2), any practical implementation delays may not have been captured. These limitations may partially explain why significant differences were not detected between Groups 1 and 2. One notable caveat of natural experiments is the timing and implementation of policy interventions, which are usually beyond the researcher's control (
). The present study made use of a natural experimental research design alongside best available data to evaluate a set of comprehensive tobacco control policies implemented within an inpatient SUD setting. Nonetheless, future research that aims to evaluate the impact of such policies should consider addressing the limitations noted above.
4.2 Recommendations and future directions
The knowledge generated from this study highlights the benefits of comprehensive cessation and TFP interventions within inpatient SUD treatment hospitals; these findings serve to inform clinical practice and hospital policies. Strategies to address tobacco use within SUD patient populations should include a combination of approaches that address tobacco use prevention and cessation, including systematic identification of those who use tobacco, access to cessation medication and NRTs, support groups, cessation resources, and importantly, a full tobacco ban. Given the available literature and findings from this study, such efforts constitute the most effective approach to evidence-based tobacco cessation interventions in clinical settings. The OMSC itself has been implemented within 12 other MHA settings across Canada (of a total of 350 general health care sites); as such, we recommend a need for future evaluations of the program's effectiveness among other SUD and mental health patient populations in Canada, and in combination with other tobacco control efforts. Notwithstanding these recommendations, exploration of the challenges and barriers related to implementation of tobacco control policies within inpatient SUD settings is also required. Issues surrounding ongoing enforcement of such policies (e.g., administrative burden), and the impact that disciplinary actions may have on patients' treatment process (e.g., therapeutic alliance, treatment completion), warrant further evaluation.
While similar studies have been conducted in the U.S. (
), this is the first Canadian study of its kind, as well as the first to include three distinct cohorts of participants with contrasting intervention exposure statuses. This is also the first study to evaluate a hospital tobacco ban using a comprehensive questionnaire (specifically, the RJP RQ) administered at admission and discharge time-points, capturing patients' full inpatient stay. While the effectiveness of the ban post-discharge remains a question, the RJP can be used to evaluate the longer-term impacts that policies may have on patients' tobacco use while in recovery from SUDs. Current research remains limited, especially within Canadian contexts; however,
found that a state-wide tobacco ban across hospitals in New York contributed to sustained tobacco cessation up to one-year post-discharge. Future directions for research should also include evaluations of the impact that tobacco control policies may have on SUD patients' overall recovery after treatment, particularly on alcohol/drug use and mental wellbeing over time. The RJP is well-positioned to answer these research questions using similar quasi-experimental study designs.
5. Conclusions
This study sought to evaluate the impact of a comprehensive tobacco cessation intervention and campus-wide TFP on patient tobacco use within an inpatient SUD treatment population. Results indicate that within our patient sample, those who were exposed to evidence-based tobacco cessation intervention alongside the tobacco ban were less likely to use tobacco during treatment. Additionally, the cessation intervention combined with the tobacco ban was effective in reducing the number of days tobacco was used and the number of cigarettes smoked per day over the course of treatment. This study constitutes important evidence which supports the public health movement towards tobacco-free clinical environments – including inpatient SUD treatment settings, which have been largely excluded from such efforts despite the apparent benefits.
The following are the supplementary data related to this article.
Homewood Research Institute is an independent charitable organization funded through a variety of sources including community stakeholders, corporations, private foundations and philanthropic support from the Schlegel family. The Schlegel family owns Homewood Health.
Acknowledgements
Special thanks to Homewood Health Centre staff for their support in implementation of the Recovery Journey Project and to their patients for their ongoing participation.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
References
Asher M.K.
Martin R.A.
Rohsenow D.J.
MacKinnon S.V.
Traficante R.
Monti P.M.
Perceived barriers to quitting smoking among alcohol dependent patients in treatment.
Journal of Substance Abuse Treatment.2003; 24: 169-174
Do total smoking bans affect the recruitment and retention of adolescents in inpatient substance abuse treatment programs? A 5-year medical chart review, 2001–2005.
Journal of Substance Abuse Treatment.2007; 33: 279-285
Canadian smoking cessation clinical practice guideline.
Canadian Action Network for the Advancement, Dissemination and Adoption of Practice-informed Tobacco Treatment, Centre for Addiction and Mental Health,
Toronto, Canada2011
Wang, T. W., Asman, K., Gentzke, A. S., Cullen, K. A., Holder-Hayes, E., Reyes-Guzman, C., … & King, B. A. (2018). Tobacco product use among adults—United States, 2017. Morbidity and Mortality Weekly Report, 67(44), 1225.
30045-5&id=gr1.jpg){kind=link}
30045-5&id=gr2.jpg){kind=link}