OUP user menu

Effects of Smoking Status on Immediate Treatment Outcomes of Multidisciplinary Pain Rehabilitation

W. Michael Hooten MD, Cynthia O. Townsend PhD, Barbara K. Bruce PhD, John E. Schmidt PhD, Jennifer L. Kerkvliet MA, Christi A. Patten PhD, David O. Warner MD
DOI: http://dx.doi.org/10.1111/j.1526-4637.2008.00494.x 347-355 First published online: 1 March 2009

Abstract

Objective. The primary aim of this study was to determine the effects of smoking on treatment outcomes following multidisciplinary pain rehabilitation. A secondary aim was to assess the tobacco use characteristics of smokers with chronic pain.

Design. A prospective, nonrandomized, repeated measures design.

Setting. Multidisciplinary pain rehabilitation center at a tertiary referral medical center.

Patients. All patients (N = 193) consecutively admitted from August 2005 through February 2006.

Interventions. A 3-week outpatient multidisciplinary pain rehabilitation program.

Outcome Measures. The Multidimensional Pain Inventory (MPI), Short Form-36 Health Status Questionnaire (SF-36), Center for Epidemiologic Studies-Depression scale (CES-D), Pain Anxiety Symptom Scale (PASS-20) and Pain Catastrophizing Scale (PCS) were administered at admission and dismissal.

Results. The study involved 49 (83% women) smokers and 144 (83% women) nonsmokers. The mean number of cigarettes smoked daily was 17.5 (SD 11.1) and the mean duration of smoking was 21.5 years (SD 12.1). After adjusting for demographic differences, repeated measures of analysis of covariance showed a main effect of smoking status for MPI affective distress (P = 0.008), CES-D (P = 0.001) and PCS (P = 0.011). An interaction of smoking status and time was found for the PASS-20 (P = 0.028), MPI affective distress (P = 0.033), MPI life control (P = 0.003) and SF-36 role-emotional (P = 0.004) subscale. While the majority of smokers were ready to consider smoking abstinence, 43% declined a brief smoking cessation intervention.

Conclusions. In this series of patients undergoing multidisciplinary treatment for chronic pain, immediate treatment effects for a variety of outcome measures were similar or significantly better in smokers compared with nonsmokers.

  • Chronic Pain
  • Smoking
  • Pain Rehabilitation

Introduction

Cigarette smoking remains prevalent in the United States. In 2005, 45.1 million adults (20.9%) in the United States were current smokers [1]. Estimates of the prevalence of smoking among adults with chronic pain range from 17% to 31% [2–5]. In population-based postal survey studies, smokers report more severe pain, depression, anxiety, and pain-related impairments in physical functioning compared to nonsmokers [5–9]. In many, but not all studies, smokers are also more likely to report diffuse widespread pain and use of analgesic medications [6,7]. Long-term pain-related outcomes may also be worse for smokers compared to nonsmokers. For example, adults with chronic pain who smoke were more likely to be hospitalized for pain-related disorders, more likely to experience exacerbations of pain symptoms, and were more likely to develop chronically impaired physical and mental functioning [10–12].

The treatment of chronic pain is challenging, but techniques incorporating a multidisciplinary treatment approach have proven effective [13]. Although available studies are limited, one study suggests that responses to such therapy may be impaired in smokers [14]. Fishbain et al. [14] found that smokers with chronic low back pain were less likely to return to work 1 month following multidisciplinary pain treatment. Current smokers were also less likely than nonsmokers to return to work at 6-, 12-, and 24-month follow-up and greater pain intensity among smokers was predictive of unemployment during the 2-year follow-up period [14]. If this result was confirmed in other treatment settings and other painful conditions, it could have profound effects on the approach to treatment in these patients. For example, multidisciplinary treatment approaches also provide an opportunity to incorporate tobacco interventions. However, there have been few attempts to systematically provide tobacco use interventions to patients with chronic pain, and it is not known how abstinence might affect long-term responses to pain treatment.

The primary aim of this prospective observational study was to determine the effects of smoking on immediate post-treatment outcomes in a consecutive series of patients undergoing multidisciplinary treatment for chronic pain. A secondary aim was to assess the tobacco use characteristics of smokers with chronic pain, including readiness to consider smoking cessation and willingness to undergo a brief smoking cessation intervention.

Materials and Methods

This study was approved by the Mayo Foundation Institutional Review Board.

Study Design

We used a prospective, nonrandomized, repeated measures design.

Participant

All patients >18 years-of-age consecutively admitted to the Mayo Comprehensive Pain Rehabilitation Center for chronic noncancer pain of greater than 6-month duration from September 2005 through February 2006 were eligible for study participation. During this time period, 209 individuals with chronic pain were admitted. Nine patients declined to participate, and seven provided informed consent but then later requested that their medical records not be used for research purposes. Thus, the study cohort included 193 participants (women 159, men 34).

Baseline demographic and clinical data were collected at admission including age, pain duration, principle pain site, marital status, years of education, and medication use. Pain-related outcome measures and smoking status were assessed at admission and at program completion. Self-report smoking status at 6-month follow-up was assessed by mailed questionnaire.

Multidisciplinary Pain Rehabilitation

The outpatient pain rehabilitation program is of 3 weeks duration. A cognitive-behavioral model served as the basis for treatment and incorporated daily physical reconditioning, occupational therapy, and group educational sessions as previously described [15]. The primary goal of treatment was improved physical and psychosocial functioning. Admissions to the rehabilitation program occurred on a revolving basis and individuals attended 8-hours daily for 15 consecutive working days. The 3-week program has been previously completed by 86% to 89% of all individuals [16,17]. Daily opioid intake at admission was converted to oral morphine equivalents for analysis. This is the established method of calculating oral opioid equivalents at the rehabilitation center [16–19].

Brief Smoking Cessation Intervention

At the time of admission, all smokers were referred for a single, one-hour smoking cessation intervention that was conducted by a specialty trained master's level provider from the Mayo Nicotine Dependence Center. The brief intervention involved development of a detailed treatment plan, including use of pharmacotherapy, and development of a relapse prevention plan that included follow-up telephone calls [20,21]. Patients are provided recommendations for use of prescription medications. This intervention is the standard of care for a brief smoking cessation treatment at our institution. The components of the brief intervention are consistent with evidence-based practice guidelines that are in widespread clinical use [22]. The choice was left to the smoker as to whether to take part in the intervention or not. No other intervention or self-help materials were provided as an integral part of the outpatient pain rehabilitation program. Other than the single session, no other tobacco intervention was offered or provided.

Measurements

Smoking Status

Smoking status was assessed upon admission using the techniques employed by the Behavioral Risk Factor Surveillance System [23]. All study participants completed the following series of questions: 1) Have you ever smoked a cigarette, even a puff? (Yes/No); 2) Have you smoked at least 100 cigarettes in your entire life? (Yes/No/Not sure); 3) Do you now smoke cigarettes every day, some days, or not at all? (Everyday/Some days/Not at all). At 6-month follow-up, the self-reported 7-day point prevalence smoking abstinence was assessed. Those who did not complete the follow-up or had missing data were counted as smokers. This definition of abstinence is based on consensus statements from the Society for Research on Nicotine and Tobacco Subcommittee on Abstinence Measures that recommend using a 7-day period for point prevalence estimates of smoking abstinence [24].

Smoking History

Four components of tobacco use history were assessed at admission including number of cigarettes smoked daily, duration of smoking at least one cigarette daily, number of prior quit attempts that lasted at least 24 hours, and use of other forms of tobacco (snuff, chewing tobacco, pipe, cigars, bidis, and clove cigarettes). The severity of nicotine dependence was assessed using the 6-item Fagerström Test for Nicotine Dependence (FTND); a cut-off score of ≥6 indicates severe dependence [25]. The validated Contemplation Ladder was administered as a continuous assessment of motivation to quit smoking, with a possible range of 0 (no thought of quitting) to 10 (taking action to quit) [26].

Pain Outcomes

The Multidimensional Pain Inventory (MPI) measures the psychosocial impact of chronic pain [27]. Raw scores from the 52-item self-report questionnaire contains 12 subscales and responses to each item are scored by the patient on a 7-point Likert scale. The MPI has proven reliability and construct validity [28]. The raw scores are converted to standardized T-scores with a normative value of 50 (range 0–100) and a standard deviation of 10 [29]. For purposes of the current study, four MPI subscales were used including affective distress, life interference, life control, and pain severity. The affective distress subscale provides a measure of overall mood, irritability, and anxiety. The life interference subscale assesses the interference of pain in relationships and daily activities whereas the life control subscale provides a measure of problem management and control related to chronic pain. The pain severity subscale quantifies pain intensity and pain-related suffering. Lower scores on the MPI affective distress, life interference, and pain severity subscales indicate less severe symptoms whereas higher scores on the life control subscale indicate improved levels of functioning. These four subscales have been previously used to measure psychosocial functioning among patients undergoing multidisciplinary pain treatment at our institution [16–19].

The Short Form-36 Health Status Questionnaire (SF-36) was developed to assess health attributes during the past month [30–32]. The self-administered 36-item questionnaire contains eight subscales and each item is scored using a Likert scale. The raw scores were converted to T-scores with a normative value of 50 and a standard deviation of 10. These standardized T-scores were calculated with the use of published age- and gender-specific mean scores and standard deviations for the SF-36 in the general United States population [33]. Higher scores reflect a more favorable health status. For purposes of the current study, three subscales were used including physical functioning, role-physical functioning, and role-emotional functioning. Individuals involved in vigorous physical activities would score high on the physical functioning subscale. Similarly, individuals who are able to work and perform other daily activities without limitations due to physical or emotional problems would score high on both the role-physical and role-emotional subscales. These subscales have been previously used to measure health attributes in patients undergoing pain rehabilitation at our institution [16–19].

The Center for Epidemiologic Studies-Depression (CES-D) scale provides a validated measure of depressive symptoms [34]. Scores from the 20-item self-report questionnaire range from 0 to 60 where higher scores indicate greater levels of depression. Among individuals with chronic pain, cutoff scores of 16 and 27 indicate minor and major depressive symptoms, respectively [35,36]. The short version of the Pain Anxiety Symptoms Scale (PASS-20) was used to measure pain-related anxiety and fear [37]. The 20-item self-report questionnaire has a score range of 0 to 100 where higher scores indicate greater levels of anxiety. The Pain Catastrophizing scale (PCS) is a self-report questionnaire that assesses negative cognitions and emotions associated with actual or anticipated pain experiences [38]. Higher scores indicate negative expectancies regarding the capacity to cope with pain.

Statistical Analyses

Demographic and clinical characteristics of smokers and nonsmokers were compared using t-tests for continuous variables and the chi-square test for categorical variables. For smokers and nonsmokers, the mean admission and dismissal scores of the MPI, SF-36, CES-D, PASS-20, and PCS were compared using t-tests. Further analyses involving the mean admission and dismissal scores of all outcome measures were conducted using repeated measures analysis of variance (anova) including main effects of time, smoking status, and the interaction of time and smoking status. A post hoc repeated measures of analysis of covariance (ancova) was conducted using group differences in demographic and clinical characteristics as covariates. Two-sided tests were used and the level of significance for all statistical tests was set at P < 0.05. All analyses were completed using Statistical Package for the Social Sciences, version 15.0 (SPSS, Inc., Chicago, IL).

Results

Baseline Demographics and Smoking Characteristics

Of the 193 participants, 49 (25%) were smokers and 144 (75%) were nonsmokers. The 3-week program was completed by 90% (n = 44) of smokers and 89% (n = 128) of nonsmokers. Table 1 contains a summary of the demographic and clinical characteristics for all patients. Compared with nonsmokers, smokers were significantly younger (P = 0.010), less likely to be married (P = 0.025), had fewer years of education (P = 0.002) and were more likely to report low back or abdominal pain as their primary site of pain (P = 0.048).

View this table:
Table 1

Differences between smokers and nonsmokers on baseline demographics and clinical characteristics

CharacteristicSmokers (N = 49)Nonsmokers (N = 144)*P value
NMean (SD) or %NMean (SD) or %
Age4942.7 (11.6)14448.5 (14.1)0.010
Gender0.832
  Male918%2517%
  Female4082%11983%
Race0.583
  Caucasian4898%13493%
  Other12%107%
Marital status0.025
  Single/Divorced/Widow2857%5337%
  Married2143%9163%
Years of education4913.3 (2.5)14414.7 (2.7)0.002
Pain duration (months)48109.2 (110.4)143124.8 (134.5)0.469
BMI4929.3 (7.1)14430.1 (8.2)0.543
Morphine equivalent (mg/d)3473.9 (113.4)7974.0 (108.2)0.996
Primary pain site0.048
  Fibroymalgia612%3524%
  Low back1837%3323%
  Headache612%1410%
  Generalized612%118%
  Abdominal612%96%
  Upper extremity36%64%
  Pelvic24%75%
  Other24%2920%
  • * t-test for continuous variables and chi-square test for categorical variables.

Among nonsmokers, 39% were former smokers, 33% were never smokers. Among smokers, 95% reported daily cigarette use and 4% reported current nondaily use. The mean number of cigarettes smoked per day was 17.5 (SD 11.1) and the mean duration of smoking was 21.5 years (SD 12.1). Smokers reported a mean of 2.6 (SD 2.5) previous quit attempts of at least 24 hours duration. Smokers reported using no other forms of tobacco except cigarettes. With respect to the FTND score, 43% had a score ≥6, indicating severe dependence. The mean score on the Contemplation Ladder was 6.4 (SD 2.5), indicating that many smokers were considering quitting or reducing cigarette consumption.

Pain Outcomes

At admission, smokers had significantly worse physical and emotional functioning compared to nonsmokers (P < 0.05) on all assessed outcome measures except the MPI pain severity and SF-36 physical function subscales (Table 2). At dismissal, smokers still had significantly worse scores on the CES-D (P = 0.012), PCS (P = 0.036) and SF-36 role-emotional (P = 0.029) subscale (Table 2). No other significant differences were found among the remaining outcome measures at dismissal (P > 0.1).

View this table:
Table 2

Mean pre- and posttreatment scores for the MPI, SF-36, CES-D, PASS-20 and PCS

ScalePretreatmentBetween Subjects PretreatmentPosttreatmentBetween Subjects Posttreatment
Smokers (N = 49)Nonsmokers (N = 144)Smokers (N = 44)Nonsmokers (N = 128)
Mean (SD)Mean (SD)P value*Mean (SD)Mean (SD)P value*
MPI affective distress54.2 (7.6)49.0 (10.0)0.00139.8 (10.0)37.7 (9.4)NS**
MPI life control43.2 (6.8)47.2 (7.5)0.00257.0 (7.6)57.1 (6.8)NS
MPI life interference53.3 (6.9)50.0 (9.7)0.03637.5 (12.5)35.9 (12.1)NS
MPI pain severity50.4 (7.8)47.9 (9.9)NS38.6 (12.0)38.3 (12.8)NS
SF-36 physical function27.4 (13.1)31.1 (13.5)NS39.1 (12.0)41.6 (11.6)NS
SF-36 role-physical26.9 (7.0)30.3 (10.8)0.03941.3 (12.1)43.7 (13.6)NS
SF-36 role-emotional32.3 (10.3)38.4 (13.2)0.00444.6 (12.6)48.9 (10.2)0.029
CES-D33.3 (11.0)26.2 (13.0)0.00118.0 (11.7)13.4 (9.8)0.012
PASS-2055.3 (18.4)46.0 (21.0)0.00722.7 (18.6)22.2 (18.1)NS
PCS32.1 (9.8)26.0 (12.8)0.00314.8 (11.9)11.0 (9.6)0.036
  • * t-test.

  • ** P > 0.1.

  • MPI = Multidimensional Pain Inventory; SF-36 = Short Form-36 Health Status Questionnaire; CES-D = Centers for Epidemiologic Studies-Depression scale; PASS-20 = Pain Anxiety Symptoms Scale short version; PCS = Pain Catastrophizing Scale; NS = not significant.

From admission to program dismissal, a main effect of time (anova) was demonstrated for all outcome measures (P < 0.001), indicating the multidisciplinary program was effective (Table 3). A main effect of smoking status was identified for affective distress (P = 0.003), life control (P = 0.022), role-emotional functioning (P = 0.002), CES-D (P < 0.001), and PCS (P = 0.003; Table 3). A significant interaction of time and smoking status was found for affective distress (P = 0.025), life control (P = 0.003), and PASS-20 (P = 0.01), indicating smokers improved more on these measures compared to nonsmokers (Table 3).

View this table:
Table 3

Effects of smoking status on treatment outcomes following pain rehabilitation

ScaleMain Effect of TimeMain Effect of SmokingInteraction of Time and Smoking
F value (df)*P valueF value (df)P valueF value (df)P value
MPI affective distress196.6 (1, 169)<0.001 9.1 (1, 169) 0.003 5.1 (1, 169)0.025
MPI life control278.0 (1, 168)<0.001 5.3 (1, 168) 0.022 9.1 (1, 168)0.003
MPI life interference198.6 (1, 169)<0.001 2.1 (1, 169) 0.150 0.4 (1, 169)0.509
MPI-pain severity 86.3 (1, 169)<0.001 0.1 (1, 169) 0.322 1.2 (1, 169)0.267
SF-36 physical function147.8 (1, 163)<0.001 2.0 (1, 163) 0.157<0.1 (1, 163)0.772
SF-36 role-physical134.0 (1, 164)<0.001 3.1 (1, 164) 0.082 0.2 (1, 164)0.648
SF-36 role-emotional 79.0 (1, 164)<0.00110.2 (1, 164) 0.002 0.6 (1, 164)0.443
CES-D195.8 (1, 169)<0.00113.9 (1, 169)<0.001 3.1 (1, 169)0.080
PASS-20244.3 (1, 163)<0.001 3.2 (1, 163) 0.076 6.8 (1, 163)0.010
PCS223.3 (1, 168)<0.001 9.1 (1, 168) 0.003 1.3 (1, 168)0.250
  • * Repeated measures of anova.

  • Note: Data not adjusted for differences in demographic characteristics.

  • MPI = Multidimensional Pain Inventory; SF-36 = Short Form-36 Health Status Questionnaire; CES-D = Centers for Epidemiologic Studies-Depression scale; PASS-20 = Pain Anxiety Symptoms Scale short version; PCS = Pain Catastrophizing Scale.

Post hoc repeated measures ancova were conducted using demographic characteristics that were significantly different between the two groups at admission as covariates (Figure 1). The covariates included age, years of education, marital status, and pain site. In this adjusted model, the main effect of time remained significant for all outcome measures except MPI pain severity (F[1,165] = 1.6, P = 0.211) and SF-36 role-emotional functioning (F[1,160] = 0.5, P = 0.475). The main effect of smoking status remained significant for MPI affective distress (P = 0.008), CES-D (P = 0.001), and PCS (P = 0.011; Figure 1). However, the main effect of smoking was not significant for MPI life control (F[1,164] = 3.1, P = 0.080). The interaction of time and smoking status remained significant for MPI affective distress (P = 0.033), MPI life control (P = 0.003), SF-36 role-emotional (P = 0.004); and PASS-20 (P = 0.028; Figure 1).

Figure 1

Effects of smoking status on treatment outcomes following pain rehabilitation after adjusting for differences in age, years of education, marital status and pane site.

Smoking Outcomes

Upon admission to the pain rehabilitation program, all smokers were referred for a brief smoking cessation intervention and 67% (n = 33/49) accepted and completed the referral. The pain rehabilitation program was completed by 95% of smokers who refused the smoking cessation intervention and 86% of smokers who accepted referral for the brief intervention (P > 0.1). All those who smoked at baseline continued to smoke at the time of program completion and at 6-month follow-up.

Discussion

The main finding of this study was that responses to multidisciplinary rehabilitation for chronic pain were not less in smokers compared with nonsmokers. In fact, for measures of affective distress, life control, role-emotional functioning, and anxiety, smokers experienced significantly greater improvement compared to nonsmokers. However, despite these improvements, smokers had significantly worse measures of role-emotional functioning, depression, and pain catastrophizing at both program admission and completion.

There are few prior studies of how smoking status affects multidisciplinary treatment of chronic pain. Fishbain et al. [14,39] analyzed the effect of smoking status on outcomes in a cohort of patients treated for chronic low back pain. Outcomes included return to work and pain up to 2 years of follow-up. They found that these outcomes were worse in smokers compared with nonsmokers throughout their follow-up period. It is not possible to directly compare our results, given that we did not measure the same outcomes, patients received a different intervention, and that we examined a more diverse group of pain conditions. Any or all of these differences may explain why we did not observe a similar effect of smoking status on treatment-related changes in several outcomes.

Consistent with prior studies, before treatment smokers reported increased levels of several measures of psychological distress compared with nonsmokers [5–9]. Fishbain et al. [39] also found that depression and anxiety scores were higher among smokers compared to nonsmokers at program initiation (P < 0.03), albeit not significantly so given that they chose a level of P < 0.01 to define statistical significance. However, measures of pain severity and daily morphine equivalent use were similar for smokers and nonsmokers. In contrast, population-based studies generally find that smokers report greater pain intensity and more frequent use of analgesic medications compared to nonsmokers [6,7]. This may reflect differences between patients with sufficient chronic pain to seek a relatively intensive course of multidisciplinary therapy compared to a more general population of patients reporting pain. It suggests that studies of the interaction between smoking status and pain must carefully define the population under study.

We cannot explain why the treatment effects were significantly greater for some outcomes in smokers. One possibility is that smokers simply had more to gain on these measures as the admission scores were worse compared to nonsmokers. However, improvement was not greater in smokers for every measure that was elevated before treatment. For example, smokers had significantly elevated depression and PCS scores at admission and dismissal but no favorable interaction between smoking status and time was observed. An alternative explanation, also consistent with our observation that smokers had worse depression and pain catastrophizing, pertains to the cognitive coping skills of smokers. Specifically, smokers with a history of depression are in some respect less functional compared to smokers without a history of depression [40]. For example, smokers with a history of depression have worse cognitive coping skills compared to smokers without a history of depression [41]. These findings suggest that smokers with a history of depression lack important cognitive coping skills and may especially benefit from cognitive-behavioral based treatments. These findings may provide the basis for future research directed toward delineating the independent effects of depression on treatment outcomes of smokers with chronic pain.

We also examined tobacco use characteristics among these individuals undergoing multidisciplinary treatment for chronic pain. The proportion of patients who were identified as former or never smokers was similar to the rates observed in other groups of smokers [5–9]. The mean Contemplation Ladder score was 6.4, indicating that many patients were willing to consider smoking abstinence. This is consistent with prior work suggesting that the presence of pain symptoms does not adversely impact smokers' readiness to consider smoking abstinence [42]. Referral for the smoking cessation intervention was accepted by 67% of smokers which is consistent with the acceptance rates reported for other groups of smokers [43,44]. No patients quit smoking, which is consistent with anecdotal observations that it may be very difficult for patients undergoing multidisciplinary therapy to quit smoking [14]. The 6-month abstinence rate for a similar smoking cessation intervention at our institution has been reported to be 26% [21]. It is likely that smokers with chronic pain will require more than the single brief intervention offered in the present study. Some prior work suggests that smokers with chronic pain can respond to tobacco use interventions. For example, smokers with chronic spinal pain who received smoking cessation education and support from a spine specialty practitioner were more likely to quit or reduce cigarette consumption compared to smokers who did not receive support from a spine specialist [45]. Future studies should evaluate whether smokers with chronic pain may benefit from the direct integration of smoking cessation services into the delivery of ongoing pain-related healthcare [14].

Our study has several limitations. All study participants were specifically referred for pain rehabilitation and had health care resources and motivation to participate in the 3-week outpatient program. Thus, the study results may not be applicable to all smokers with chronic pain. We also performed limited follow-up of smoking behavior or assessment of compliance with the relapse prevention plan. Although none of the patients quit smoking, we did not assess for reduced cigarette consumption or rate of pharmacotherapy use at program completion or during the 6-month follow-up period. We also did not assess the frequency of quit attempts during the 6-month follow-up period; smokers may have quit after dismissal from the program but relapsed prior to 6-month follow-up. Finally, the multiple comparisons reported for this sample could increase the risk of a Type I error. However, these findings provide important observational data that could help guide the focus of future investigations regarding the effects of smoking status on treatment outcomes of chronic pain.

In summary, in this series of patients undergoing multidisciplinary treatment for chronic pain, immediate treatment effects for a variety of outcome measures were similar or significantly better in smokers compared with nonsmokers. Although many of these patients are amenable to tobacco use interventions, the current approach of offering a single tobacco use intervention does not appear to be effective in modifying smoking behavior. Incorporation of effective tobacco use interventions as a component of multidisciplinary pain rehabilitation would provide not only health benefits to the patients, but an opportunity to further investigate how smoking cessation might affect chronic pain.

References

View Abstract