Smoking Behavior and Motivation to Quit Among Chronic Pain Patients Initiating Multidisciplinary Pain Treatment: A Prospective Study

Abstract

Objective

The primary aim of this study was to assess smoking characteristics and cessation motivation prior to, and after initiation of multidisciplinary chronic pain treatment. A secondary aim was to identify predictors of cessation motivation among smokers initiating treatment for chronic pain.

Design

A prospective, nonrandomized, repeated measures design.

Setting

Multidisciplinary specialty pain treatment program at a veterans hospital.

Patients

Smokers (N = 90) referred to a multidisciplinary pain program for the treatment of chronic pain.

Measures

Patients completed questionnaires assessing pain-related and smoking-related factors prior to (baseline) and 8 weeks post (follow-up) specialty pain treatment initiation. Primary outcome measures were the Contemplation Ladder and the Stages of Change (SOC) algorithm.

Results

At baseline, patients reported moderate levels of cessation motivation, and 69% were in the contemplation stage or higher on the SOC. Motivation to quit smoking was higher at follow-up compared to baseline on both continuous, t (89) = 2.11, p<.05, and stage-based, z = 3.69, p<.01 measures. At follow-up, participants reported greater interest in receiving cessation interventions, and 7.8% of patients had quit smoking. Pain-related predictors of motivation (e.g., pain intensity) were subsumed by more general predictors (e.g., nicotine dependence).

Conclusions

Patients in this sample were more motivated to quit smoking a few weeks after, as compared to before initiating specialty pain treatment. Future research into pain-specific predictors of cessation motivation is warranted to inform the development of interventions that address pain patients’ unique needs.

Introduction

Tobacco smoking is a frequently co-occurring health risk factor among individuals with chronic pain [1]. Smoking prevalence rates of up to 54% have been found in clinical populations of chronic pain patients [2,3,4]. Additionally, smokers with pain report smoking more heavily as compared to those without pain [5], placing them at an increased risk for smoking-attributable morbidity and mortality [6].

In addition to an increased risk for smoking-related illness, a growing body of literature indicates that smoking is associated with increased risk for, and exacerbation of, a variety of pain disorders [7,8,9]. For instance, results from meta-analyses for cross-sectional and prospective cohort studies showed that both current and former smokers have a higher prevalence and incidence of low back pain than never smokers, with odds ratios ranging between 1.3 and 2.1 [10]. Research with patients evaluated at pain treatment facilities indicates that smokers with chronic pain experience significantly greater pain intensity levels and functional impairment [2,11,12]. In addition, smokers have been found to require significantly more post-operative opioid medication [13,14] than non-smokers, and in a large population-based study, current and former smokers were more likely to use analgesic drugs than never smokers [15]. There is also evidence that smoking is negatively associated with pain-related treatment outcomes [16,17] and that very few smokers undergoing pain treatment successfully quit smoking despite receiving a tobacco intervention [18]. Although the smoking-pain relationship is not well understood, some research suggests that smoking may reduce a person’s pain perception or help them cope with the pain [19]. At least one study found that pain was a potent motivator of smoking [20]. In a literature review, Ditre et al. concluded that smoking and pain may exacerbate each other via a positive feedback loop [21].

Depression may also play an important role in the smoking-pain relationship. Compared with the general population, prevalence of depression is significantly higher among both smokers [22,23] and individuals with chronic pain [24,25,26,27]. Depression is also associated with greater pain severity [28] and poor smoking cessation outcomes [29,30,31]. Hence, smokers with chronic pain may be less responsive to standard cessation interventions due to the burden associated with multiple co-morbidities, and, may be influenced by unique determinants of cessation. A better understanding of the characteristics of chronic pain smokers and the relationship between pain-specific factors and smoking may lead to improved tobacco cessation interventions for this population. However, there is a knowledge gap with regard to factors that may inform the development of interventions targeted to the unique needs of chronic pain patients. Another challenge to helping the more resistant smokers is to reach them. Because chronic pain is a common presenting complaint within the healthcare system [32,33], treatment for smokers’ pain may also present an opportunity to provide effective smoking cessation interventions.

The first aim of this prospective study was to assess smoking and cessation-related characteristics of patients undergoing treatment for chronic pain. The second aim was to assess cessation motivation and interest in smoking interventions prior to and following specialty chronic pain treatment initiation. We hypothesized that patients would be more motivated to quit smoking several weeks after pain treatment initiation, when pain treatment demands and the need to self-medicate pain-related stress are expected to be lower relative to start of pain treatment. The third aim was to identify factors predictive of increased cessation motivation. In addition to the known predictors, such as negative affect and nicotine dependence, we examined potential pain-specific factors, including pain intensity, disability, catastrophizing, as well as pain-related smoking expectancies, smoking temptation situations, and barriers to quitting.

Methods

Participants

Patients referred for treatment of non-cancer pain at a Veterans Affairs chronic pain specialty clinics were recruited into the study. Treatments provided within the pain clinics included: medication management, including antidepressant medications; trigger point injections; epidural steroid, nerve block, and other spine injections; physical and/or occupational therapy; or combinations of the above. Patients undergoing treatment in the outpatient pain specialty clinics usually average between 2 to 3 visits over a period of 4 to 8 weeks. Pain clinic treatment did not include any smoking cessation interventions or treatment by a pain psychiatrist or psychologist, although some patients may have been referred to those services outside of their pain clinic treatment. Study inclusion criteria were: a) at least 18 years of age; b) smoked at least 10 cigarettes per day for at least one year; c) breath carbon monoxide (CO) of at least 8 ppm, the cut-off level commonly used to identify smoking status [34]; d) initiating a new course of treatment for chronic pain (i.e. have not been treated in pain specialty clinics during the previous 6 months); e) able to read and write English; and f) able to give informed consent (e.g., no documented or apparent cognitive impairment). Participants were paid $20 for completing each of the assessments and received a bonus payment of $10 for completing both assessments.

Procedure

The study was approved by the University of South Florida Institutional Review Board. Potential participants were identified by reviewing medical records of patients who were referred to and scheduled for outpatient pain specialty clinics for the following week. Of the 2255 medical records that were screened during the course of the study, 786 patients (34.8%) met the smoking criteria, and 444 of the identified smokers met the remaining eligibility criteria. A majority of patients (68%) failed to attend their scheduled pain clinic appointments. Research assistants approached potential participants (i.e., those who came to their appointment) in the waiting room (n=144) and asked whether they were willing to be screened for additional eligibility criteria to be in a study. Eligible participants were provided with study information and asked to sign an informed consent form and to agree to complete 2 self-report assessments, baseline and follow-up. Of those approached, 103 (71.5%) agreed to participate and completed the baseline assessment, and 90 (87.4%) of those who agreed to participate completed the follow-up assessment.

The baseline assessment was completed at the time of the initial pain treatment evaluation (i.e., prior to pain treatment initiation). The follow-up assessment was completed between 8 and 10 weeks after the baseline assessment via telephone or mailed questionnaire. The 8-week timeframe was selected based on the notion that it may take 6 to 8 weeks for the pain treatment to reach optimal effects. For instance, it may take patients a few weeks to adjust to their new medication, or the dose may need to be adjusted. Other treatments, such as steroid injections, or physical therapy, require multiple treatments before optimal effects are reached. Therefore, this timeframe corresponds to when pain treatment demands and the need to self-medicate pain-related stress are expected to be lower relative to when pain treatment is initiated. Moreover, the selected follow-up point is within the timeframe during which the majority of patients remain in contact with the healthcare system, which is necessary for the delivery of smoking cessation interventions within the context of chronic pain treatment. This design allowed us to assess whether motivation to quit and interest in cessation interventions are higher prior to or following pain treatment initiation.

Measures

Pain-related measures

Demographic and pain history information, including pain site, duration, and use of medication, was collected using the Pain Outcomes Questionnaire, developed and validated at the veteran hospital’s chronic pain program [35]. Pain intensity and pain unpleasantness were measured using single-item numerical rating scales (NRS) that ask patients to rate their pain from 0 to 10, where 0 represents one end of the continuum (i.e., no pain at all, not unpleasant at all) and 10 represents the other extreme (i.e. worst possible pain, the most unpleasant feeling possible for me) [36]. The validity and sensitivity of NRSs have been well documented [37,38]. The Pain Disability Index (PDI) was used to measure the level of interference with, and disruption of, daily activities [39]. The PDI consists of 7 items on a 0 to 10 scale that measure the degree to which patients believe that their pain interferes with functioning in the areas of family/home responsibilities, recreation, social activities, occupation, sexual behavior, self-care, and life-support activity. Studies support the reliability and validity of the PDI as a brief measure of pain-related disability [40]. The Pain Catastrophizing Scale (PCS) assesses negative cognitions and emotions associated with actual or anticipated pain experiences (e.g., “It’s terrible and I think it’s never going to get any better”) [41]. Higher scores indicate negative expectancies regarding the ability to cope with pain.

Smoking-related measures

Smoking Status

The Smoking Status Questionnaire developed for this study was used to assess smoking history and baseline smoking-related variables including: amount and duration smoked, quantity and duration of quit attempts, and current and past use of smoking cessation intervention. Current interest in smoking cessation interventions and preferences for intervention modalities was assessed by asking participants to check all of the treatments they prefer to receive from a list of options (e.g., individual, group, self-help, nicotine patches, gum, bupropion). Nicotine dependence was assessed using the well established and validated 6-item measure, Fagerström Test for Nicotine Dependence (FTND) [42]. The FTND includes item content such as number of cigarettes smoked daily, time to first cigarette of the day, and difficulty refraining from smoking where it is forbidden. The total score is on a scale of 0 – 10, with 8–10 signifying very high dependence. The Smoking Follow-up Questionnaire developed for this study assessed the same smoking and cessation behaviors as the baseline questionnaire, with reference to the timeframe between baseline and follow-up assessment. Smoking status was based on self-reported 7-day point-prevalence abstinence (i.e., did not smoke in the past 7 days) [43].

Cessation Motivation

The Contemplation Ladder is an 11-point single item scale with good psychometric properties and discriminant validity [44] that asks smokers to rate their intention to quit smoking. The Stages of Change (SOC) algorithm [45] has been shown to be predictive of cessation [46,47]. It places smokers into precontemplation, contemplation, preparation, and action stages. As a binary measure of cessation motivation, we assessed whether participants had initiated any smoking cessation treatment (i.e., nicotine replacement therapy (NRT), prescription medication, etc…), by follow-up.

Cognitive Smoking Measures

The Smoking Consequences Questionnaire-Adult (SCQ-A) is a validated measure of smoking expectancies that has been found to be associated with nicotine dependence and smoking status [48]. Three scales of the SCQ-A that are most relevant to pain (negative affect reduction, stimulation/state enhancement, and health risks) were used, resulting in 20 total items. A 5-item Pain and Smoking Expectancies (PSE) questionnaire was used to assess pain-specific expectancies of the effect of smoking on pain [20]. High scores on the PSE indicate a greater belief that smoking helps to diminish and/or cope with pain. PSE items correlated with SCQ-A(negative affect), r(87) = .58, p < .001 and with SCQ-A(stimulation), r(87) = .78, p < .001 subscales of the original SCQ-A instrument. Temptation to smoke was assessed using the Situational Temptation Inventory (STI) of the self-efficacy scale [49] which has also been found to be associated with nicotine dependence and smoking status. The 3 subscales of the STI include: positive affect/social situations (STI-POS), negative affect situations (STI-NEG), and habitual/craving situations (STI-CRAVE). Six additional questions designed to assess participants’ temptation to smoke in the presence of pain (e.g. “How tempted are you to smoke when you experience a lot of pain”) comprised a separate pain-specific temptation score (STI-pain). The 6 STI-pain items demonstrated high internal consistency reliability (α = .97), suggesting that the items are measuring a unitary construct. STI-pain correlated with STI-POS, r(89) = .50, p < .001; STI-NEG, r(89) = .59, p < .001; and STI-CRAVE, r(89) = .70, p < .001 subscales of the original STI instrument. The Barriers to Cessation Scale (BCS) is a 19-item questionnaire that measures perceived stresses associated with smoking cessation and has been shown to differentiate between abstainers and relapsers [50]. Modeled on questions developed by Asher et al. [51], 7 additional items designed to assess barriers related to chronic pain (e.g., concerns that it will be harder to manage pain if they quit smoking), comprised a separate pain-specific barriers score (BCS-pain). The 7 BCS-pain items demonstrated high internal consistency reliability (α = .93), and correlated with the original BCS measure, r(89) = .64, p < .001.

Negative affect measures

The Center for Epidemiologic Studies Depression scale (CES-D) [52] is a 20-item instrument widely used to measure depressive symptomatology and has previously been used with chronic pain patients [53]. The State-Trait Anxiety Inventory – State subscale (STAI) is a 20-item questionnaire used to measure current symptoms of anxiety [54].

Data Analytic Strategy

Smoking and clinical characteristics at intake and follow-up were compared using parametric (paired samples t-tests) and non-parametric tests (the sign test and the Wilcoxon signed-rank test), depending on the characteristics of the variable being tested. Changes in the interest in smoking cessation interventions were assessed using the sign test. To identify predictors of cessation motivation at follow-up assessment, we first conducted Pearson and Spearman correlational analyses using baseline variables, including nicotine dependence (FTND), cognitive smoking measures (SCQ, PSE, STI, BCS), pain-related measures (pain intensity, pain unpleasantness, PDI, PCS), and negative affect measures (CES-D, STAI). This was followed by backward, stepwise regression analyses that included all of the above predictor variables that were correlated with the outcome variables at p < .10. Backward stepwise regressions with the more liberal p-value of < .10 for inclusion eligibility were used because of the exploratory nature of the study, with no previous research on pain-relevant predictors of motivation to quit smoking. Cessation motivation included the Contemplation Ladder, the Stages of Change, and smoking cessation treatment initiation (yes/no) since baseline assessment.

Results

Sample Characteristics

Data from the 90 participants who completed the baseline and follow-up assessments were used for analyses. There were no significant differences between the 90 completers and the 13 who did not complete the follow-up assessment on any demographic or baseline smoking variables. The sample was predominantly male, Caucasian, unemployed, and moderately dependent on nicotine (see Table 1 for baseline demographic, smoking, and clinical characteristics).

Table 1.

Demographic, Smoking, and Clinical Characteristics (N=90)

	M (SD)/%
Age	52.9 (8.6)
Male	91
Race/Ethnicity
Caucasian	86
African American	10
Other	4
Hispanic	3
Marital status
Single/never married	8
Married/living with partner	57
Divorced/separated/widowed	34
Years of education	12.9 (1.9)
Employment status
Employed (FT/PT)	16
Unemployed, disabled	56
Unemployed, looking	11
Retired	13
Other	3
FTND	5.64 (2.3)
Number of cigarettes smoked per day	22.3 (11.9)
Years of daily smoking	33.9 (10.9)
Lifetime quit attempts (median)	3.0 (0–50)
One or more quit attempt past year	46
CO ppm	30.8 (14.2)
Primary pain site
Low back	51
Lower extremity	14
Head	8
Neck	8
Upper extremity	10
Other	8
Pain duration, years (median)	6.0 (0–46)
Daily opioid use	43

Note. FTND = Fagerstrom Test for Nicotine Dependence.

Changes in Cessation Motivation, Smoking-Related and Clinical Characteristics

Cessation motivation pre and post pain treatment initiation

The average score on the Contemplation Ladder was higher at follow-up (M = 6.82, SD = 3.1) compared to baseline (M = 5.90, SD = 3.1) assessment; t (89) = 2.11, p < .05. These scores reflect a moderate level of motivation to quit smoking. Similarly, on the Stages of Change measure of cessation motivation, the Wilcoxon Signed Rank Test revealed a general shift toward higher stages (i.e., greater readiness to quit smoking) from baseline to follow-up, z = 3.69, p < .01 (see Table 2). Nearly 70% of patients prior to pain treatment initiation, and 79% after pain treatment initiation were in the Contemplation stage or higher.

Table 2.

Smoking and Clinical Characteristics Pre-Post Pain Treatment Initiation (N=90)

	Pre pain treatment initiation	Post pain treatment initiation	p value
Smoking	M (SD)/%	M (SD)/%
Contemplation Ladder 0–10	5.90 (3.11)	6.82 (3.11)	.010
Stages of Change (%)
Precontemplation	30.7	21.1	< .001*
Contemplation	48.9	42.2
Preparation	20.5	28.9
Action	0	7.8
# cigs smoke daily	22.39 (10.6)	19.13 (10.5)	.001
% receiving cessation Tx	17.8	25.6	.386
Cognitive Smoking Measures
SCQ-A-negative affect reduction	58.52 (17.7)	_	_
SCQ-A-stimulation	28.88 (15.3)
SCQ-A-health risks	31.19 (7.7)
PSE	21.78 (13.6)
STI-positive affect/social	24.13 (4.4)	_	_
STI-negative affect	24.56 (5.7)
STI-habitual/craving	22.74 (5.2)
STI-pain	21.53 (7.9)
BCS	29.28 (11.1)	23.92 (12.2)	< .001
BCS-pain	12.4 (6.7)	11.38 (6.7)	.097
Pain and negative affect
Pain intensity 0–10	7.46 (1.6)	6.67 (1.83)	< .001
Pain unpleasantness 0–10	7.27 (2.02)	6.88 (2.02)	.086
Pain Disability Index	43.42 (16.30)	45.68 (15.29)	.168
Pain Catastrophizing Scale	32.44 (12.6)	34.68 (11.04)	.073
Current daily opiod use	42.7%	46.1%	< .001
CES-D (depression)	25.37 (6.63)	24.04 (7.28)	.038
STAI – State (anxiety)	47.79 (13.14)	47.99 (15.11)	.888

Note.

^*Stage of Change p–value reflects overall shift toward higher stages based on Wilcoxon test results;

CES-D = Center for Epidemiologic Studies Depression; STAI = State-Trait Anxiety Inventory; SCQ-A = Smoking Consequences Questionnaire-Adult; PSE = Pain and Smoking Expectancies; STI = Situational Temptation Inventory; BCS = Barriers to Cessation Scale; SCQ, PSE, STI measures were administered at baseline only.

Smoking and clinical characteristics pre and post pain treatment initiation

As can be seen in Table 2, there were significant changes in smoking and clinical characteristics from pre- to post-pain treatment initiation, including reduced number of cigarettes smoked daily, reduced pain intensity, and a reduction in depressive symptomatology. At follow-up, 7.8% of the sample (n = 7) had quit smoking.

With regard to interest in smoking cessation interventions, a larger proportion of patients were interested in obtaining smoking cessation services post-pain treatment initiation relative to pre-pain treatment initiation, including telephone quitline services, internet-based interventions, and alternative medicine (see Table 3). At follow-up, patients expressed the greatest interest in prescription medication, nicotine replacement therapy (NRT), and alternative medicine.

Table 3.

Interest in Cessation Intervention Modalities Pre and Post Pain Treatment Initiation

	Pre pain treatment %	Post pain treatment %	p value
NRT (e.g., nicotine patch, gum, lozenge)	33.3	42.2	ns
Prescription medication (e.g., Chantix, Zyban)	35.6	43.3	ns
Counseling/support (group or individual)	14.4	20.0	ns
Telephone Quitline	2.2	10.0	p < .05
Self-help/educational materials	6.7	16.7	ns
Internet-based	1.1	12.2	p < .01
Alternative medicine (e.g., acupuncture, hypnosis)	16.7	36.7	p < .01
Interested in any intervention modality	75.6	70.0	ns

Note. p–values reflect differences based on the sign test results; responders were asked to check all that apply.

Predictors of Post-Pain Treatment Cessation Motivation

Table 4 presents correlations between baseline predictor variables and cessation motivation measures at follow-up. Analyses revealed significant negative correlations between the Contemplation Ladder and nicotine dependence, pain intensity, the stimulation/state enhancement subscale of the Smoking Consequences Questionnaire, the Barriers to Cessation Scale, and Pain-specific Smoking Expectancies. In addition, there was a trend for an inverse relationship between the Contemplation Ladder and pain-specific barriers to quitting. The Stages of Change measure was negatively correlated with depression, anxiety, Barriers to Cessation Scale, and Pain-specific Smoking Expectancies. There were also trends toward negative correlations between Stages of Change and nicotine dependence as well as pain-specific quitting barriers. Nicotine dependence was the only significant correlate of treatment initiation, r (89) = −.251, p = .018.

Table 4.

Correlations Between Baseline Predictors and Follow-up Cessation Motivation Measures

Baseline Predictors	Contemplation Ladder (at follow-up)	Stages of Change (at follow-up)
Fagerstrom Test for Nicotine Dependence (FTND)	−0.307**	−0.177+
Center for Epidemiologic Studies Depression scale(CES-D)	−0.113	−0.226*
State-Trait Anxiety Inventory – State (STAI)	−0.157	−0.237*
Pain intensity	−0.246*	−0.164
Smoking Consequences Questionnaire (SCQ) – Stimulation/State enhancement subscale	−0.306**	−0.165
Barriers to Cessation Scale (BCS)	−0.267*	−0.286**
Pain and Smoking Expectancies (PSE)	−0.250*	−0.233*
BCS - Pain subscale	−0.194+	−0.191+

Note.

⁺p < .10,

^*p < .05,

^**p < .01;

Significant predictors in regression models are in bold; Contemplation Ladder model R² = .19, F(2,84) = 9.76, p < .001.

Next, backward elimination regression models were conducted with the Contemplation Ladder and the Stages of Change as dependent variables. The initial set of predictor variables were those correlated with the outcome variables at p < .10 (as presented in Table 4). The predictor with the highest p-value was eliminated at each step until all predictors had p-values less than .10. The final model for the Contemplation Ladder included nicotine dependence (β = −.28, t (84) = −2.77, p < .01) and the Smoking Consequences Questionnaire (SCQ-A) -Stimulation/state enhancement subscale (β = −.27, t (84) = −2.62, p = .01). Lower nicotine dependence and SCQ-A-stimulation scores at baseline were related to higher motivation to quit at follow-up. The model explained a significant proportion of variance in the Contemplation Ladder scores, R² = .19, F(2, 84) = 9.76, p < .001. The final model for the Stages of Change resulted in one predictor: Barriers to Cessation Scale (BCS), r (89) = −.286, p = .001.

Discussion

As the smoking prevalence in the general population declines, those who continue to smoke are less responsive to generic public health approaches due to multiple co-morbidities and other complicating factors [55,56]. A better understanding of the unique characteristics of the difficult to treat smokers may lead to improved tobacco cessation interventions. This study was designed to gather preliminary data on smoking and cessation-related characteristics unique to patients undergoing specialty chronic pain treatment. To date, few studies have examined cessation-related factors in this population, characterized by high rates of smoking and psychiatric comorbidities. With a goal of contributing to the development of a targeted intervention for chronic pain patients, cessation motivation was assessed prior to and following specialty chronic pain treatment initiation, and pain-relevant predictors of cessation motivation were tested.

In the current study, motivation to quit smoking was higher several weeks after, as compared to before, pain treatment initiation. As hypothesized, this finding is consistent with the self-medication theory of drug use [57] and the notion that chronic pain patients may be more motivated and interested in quitting smoking when the need to self-medicate pain-related stress and the demands of pain treatment are lower, as is commonly the case several weeks after pain treatment has been initiated. Indeed, in the current study, more patients were using opioid medication daily, and reported lower levels of pain and depression, 8 weeks following pain treatment initiation. In addition, more patients were interested in cessation interventions at follow-up as compared to baseline. Taken together, these findings may help to inform the timing for smoking cessation interventions among this population. It is important to note, however, that although these data suggest that patients may be more motivated to quit smoking several weeks after initiating pain treatment, motivation to quit is not a consistent predictor of actual cessation [58]. Research is needed to test whether cessation, and the efficacy of smoking cessation interventions are affected by the timing of treatment initiation. Future longitudinal research is also needed to examine changes in pain intensity as a function of smoking status changes. Although research in this area is beginning to shed light on the poorly understood relationship between smoking and pain [59], it is not clear how quitting smoking may affect chronic pain. In the short term, abstinence may remove a mechanism that smokers perceive as useful in coping with pain and distress [60,61]. In the long term, however, smoking cessation may improve responses to pain interventions and overall health, and should, perhaps be considered an important part of a comprehensive approach to optimal pain management treatment. This approach may be particularly important given the high prevalence of smoking in the chronic pain population. Consistent with previous estimates [2,3,4], data from the current study showed that 35% of patients seeking treatment for a chronic pain condition were smokers. Provision of cessation interventions is also warranted, given that smokers with chronic pain appear to be motivated to quit smoking [62]. In the current study, on average, patients reported moderate levels of cessation motivation, and over 70% were in the contemplation stage of readiness to quit or higher, indicating greater overall motivation levels compared with those of the general population of smokers [63,64].

With regard to predictors of cessation motivation, lower nicotine dependence, fewer positive smoking expectancies and barriers to quitting were related to higher levels of motivation and readiness to quit. These findings are consistent with previous studies that showed an association between the above factors and smoking status [30,48,49,50].

The current study also tested several predictors of cessation motivation unique to this population. In univariate correlation analyses, lower pain intensity, pain-specific smoking expectancies and pain-specific barriers to quitting were related to higher cessation motivation. That is, individuals with fewer beliefs that smoking helps them cope with pain and concerns that quitting will negatively impact their pain, reported higher motivation to quit smoking. When tested in models that included all significant predictors, only lower nicotine dependence, smoking expectancies, and barriers to quitting remained significant predictors of cessation motivation. That is, the pain-specific predictors were subsumed within the more general predictors of cessation motivation. The current study did not involve any cessation interventions, and although many patients had sought out and received treatment, the sample was not large enough to test predictors of cessation. Future larger-scale studies can assess predictors of cessation, and examine whether changes in pain-specific factors (e.g., pain intensity) may lead to changes in motivation to quit and to successful quit attempts via their impact on the more general predictors of cessation motivation.

This study has several limitations. Due to the preliminary nature of this research, we did not attempt to identify or account for specific pain treatment modalities that may be related to cessation motivation. Rather, the main objective was to assess whether motivation changed as a function of multidisciplinary pain treatment, regardless of treatment modality. Our results set the stage for future research focusing on a more fine-grained analysis by assessing which specific pain treatments (e.g., antidepressants, opoids) may contribute to increased motivation, and ultimately successful cessation. Another limitation is that the sample consisted of a veteran population of predominantly unemployed males initiating specialty pain treatment. In addition, the majority of the sample had primary pain sites in the lower back or lower extremities. There was a high no-show rate, resulting in relatively low enrollment and participation in the study among the eligible participants identified via medical record reviews. Hence, the generalizability of the current findings is unknown. Future research is needed to examine potential gender differences, assess predictors of cessation motivation, and replicate current findings in a larger, more heterogeneous sample of individuals receiving chronic pain treatment in different clinical settings.

In summary, the current study has been fruitful in identifying potential targets for intervention. Results indicated that a large proportion of smokers with chronic pain were motivated to quit smoking, and that cessation motivation was higher several weeks after specialty pain treatment initiation, while patients had continued contact with the healthcare system. Hence, these findings suggest that it is feasible to reach and provide treatment to this population in the context of their primary presenting pain problem. This study also assessed a number of predictors of cessation motivation unique to chronic pain patients, and has illuminated areas for future research and development of interventions for an understudied, at-risk population.