Abstract
Background:
The prevalence of tobacco smoking is high among patients living with HIV, supporting the need for effective targeted interventions.
Materials and Methods:
All current smokers at our outpatient HIV clinic were invited to participate in a smoking cessation program.
Results:
Of the 716 patients living with HIV, 280 (39%) reported active smoking and were younger, more recently HIV infected and more frequently infected due to intravenous drug use (IDU). One hundred forty-seven (53%) smokers agreed to participate in the smoking cessation program and had a higher Fagerström score and were less likely IDU. During follow-up, 41 (28%) smokers withdrew from the program. After 6 months, 60 (57%) of the 106 smokers who completed the intervention had stopped tobacco smoking and were more likely to use varenicline, adjusting for a history of depression.
Conclusion:
Our smoking cessation program was feasible. However, strategies to reach and retain in smoking cessation program specific groups such as IDU are needed to improve the smoking cessation cascade.
Introduction
Contemporary antiretroviral therapy to control HIV infection has never been so simple, effective, and well tolerated. As a consequence, life expectancy of HIV-infected patients dramatically increased. 1 Non-AIDS defining malignancies such as lung cancer, non-AIDS bacterial infections, coronary artery disease, and myocardial infarction now account for 30% of all deaths among treated HIV-infected patients. 2,3
Tobacco smoking is a major modifiable risk factor for all these conditions and is prevalent at much higher levels among HIV-infected patients (around 50%) as compared to the general population of the same age and gender. 4 In addition, around 60% of HIV-infected smokers have severe nicotine dependence, 60% experience depressive symptoms, though 40% report being significantly motivated to quit smoking. 5 In a large cohort study of 2108 HIV-infected patients, the adjusted hazard ratio for all-cause and cause-specific mortality was 1.8 (95% confidence interval [CI]: 1.3-2.3) for smokers compared to never smokers. 6 In another study, the population attributable risk of death associated with smoking was 61.5% among HIV-positive patients and 34.2% among controls. 7 It is therefore of utmost importance to implement adequate smoking cessation programs in HIV-infected patients. How these programs should be implemented in this context is however unclear. HIV specialist physicians may not be adequately qualified to provide a holistic smoking cessation program during regular HIV consultations. 8,9 In addition, several factors associated with both, HIV infection and smoking, have a negative impact on the chances of successful smoking cessation, such as psychiatric disorders, lack of information and health education, social precariousness limiting access to specialists of smoking cessation, and the cost of drug cessation and counseling. 10,11
We conducted systematic screening of cigarette consumption among HIV-infected patients referring to our local outpatient clinic followed by a prospective pilot smoking cessation program provided by a smoking cessation specialist including free access to nicotine replacement therapy. The objectives of this study were to describe smoking habits and explore risk factors associated with smoking, acceptance of the program and predictors of 6-month smoking cessation.
Materials and Methods
Design and Population
Between September 2013 and September 2014, we conducted a cross-sectional evaluation of tobacco smoking among all consecutive HIV-infected patients in our outpatient consultation. The infectious disease department at the Caen University Hospital care for about 800 HIV-infected patients in Normandy, France. The medical staff includes 8 HIV physician specialists, none of them had experience in smoking cessation. In collaboration with the Department of Health and Smoking Cessation at the Caen University Hospital, we developed a pilot quasi-experimental intervention named “Tabacologie et Suivi en Maladies Infectieuses” (TASMI) for HIV-infected adult smokers who agreed to participate in the smoking cessation program. The partner of the HIV-infected patient was also offered the intervention but was not included in the analysis.
Intervention
The research staff conducted a systematic screening for tobacco smoking among all HIV-infected patients at the outpatient consultation before or after the medical consultation or during blood sampling by asking the question: “Do you smoke?” If yes, we measured expired carbon monoxide and provided a questionnaire to obtain information on sociodemographics, HIV-related variables, dealing with smoking such as the Fagerström test which quantified the degree of dependence on tobacco, 12 hospital anxiety and depression scale, 13 and the history of depression. The physician specialized in smoking cessation subsequently contacted all patients who were interested in tobacco cessation by phone and delivered the smoking cessation program independently of HIV physicians. The smoking cessation program incorporated several sessions including a baseline assessment and regular follow-up at 2 weeks, 4 weeks, 3 months, and 6 months. Individual consultations discussed the importance of quitting or reducing cigarette consumption, techniques for cutting down, and strategies for relapse prevention. In addition, nicotine replacement therapies (patch and gum) and varenicline were provided free of charge in various combination at the discretion of the physician specialized in smoking cessation. Baseline characteristics and follow-up data were prospectively monitored including the modality of therapy they used during the 6-month follow-up, which were prescribed according to the physician’s discretion. The study was approved by the ethical committee at the Caen University.
Outcomes
The primary end point was the expired carbon monoxide confirmed the rate of smoking cessation after 6 months. The prevalence of smoking and the smoking cessation program acceptance rate were the secondary end points.
Statistical Analysis
Qualitative variables were reported as numbers (percentages) and 95% CIs, quantitative variables were reported as mean (standard deviation [SD]) or median (interquartile range), according to their distribution. The statistical analysis plan had 3 main objectives: (1) compare current smokers versus others, (2) compare smokers who accepted to participate in the study and others, and (3) compared smokers who succeeded to stop smoking and others. The comparisons between groups were conducted by univariate analysis with χ2, student t test, or Wilcoxon tests as appropriate followed by multivariate analysis with logistic regression models. Variables associated with a P value < .2 in univariate analysis were included in a backward elimination automated procedure to select the most parsimonious multivariate model. A P value < .05 was considered to denote statistical significance. All analyses were conducted with SAS version 9.4 and STATA version 12.3.
Results
A total of 716 HIV-infected patients were actively followed up by an HIV specialist at the Caen University Hospital, France, during the study period (Figure 1). Among them, 280 were current smokers (39.1%, 95% CI: 35.5-42.7). Figure 1 shows the disposition of the 280 active smokers including 147 smokers who had accepted to participate in TASMI study and 106 smokers had completed the study.
Flowchart.
Baseline Characteristics of Smokers and Nonsmokers
The sociodemographics and HIV-related characteristics of the entire cohort overall and by smoking status are presented in Table 1. The mean (SD) age in the smoker group was 47.5 (9.3) and 49.6 (12.7) years in the nonsmoker group (P < .0001). Overall, 462 (64.6%) were men; 590 (82.4%) were HIV infected through sexual contact and 126 (17.6%) through intravenous drug use (IDU). The mean duration since HIV diagnosis was 156 (9.2) months, 144.3 (93.8) in the smokers group and 163.0 (98.0) in the nonsmokers (P = .01). The stage of disease in this population was mainly asymptomatic with a high of CD4 count.
Characteristics of Current Smokers versus Nonsmokers.
Abbreviations: CDC, US Centers for Disease Control and Prevention; IDU, intravenous drug use; OR, odds ratio; SD, standard deviation.
aQuanlitative variables were compared with chi-square test and quantitative variables were compared with Student t or Mann-Whitney test, as appropriate.
In multivariate analysis, 3 risk factors were independently associated with smoking: younger age (odds ratio [OR] = 0.96, 95% CI: 0.95-0.98, P < .001), shorter duration of HIV infection (OR = 0.998, 95% CI: 0.996-0.999, P < .03), and IDU as the route for HIV infection (OR = 3.7, 95% CI: 2.3-5.9, P < .001).
Baseline Characteristics of TASMI Participants and Smokers Who Declined to Participate
Of the 280 smokers screened, 147 (52.5%, 95% CI: 46.7%-58.3%) accepted. The sociodemographics and HIV-related characteristics are presented in Table 2. The mean age was similar between groups: 45.5 (9.0) in the participant group and 45.9 (9.6) in the nonparticipant group (P = .73). One hundred eighty-nine (67.5%) smokers were men and the difference between the 2 groups was not significant (P = .09), 200 (71.4%) were HIV infected through sexual contact, 80 (28.6%) through injecting drug and differed significantly between groups (P = .05), the mean duration since HIV diagnosis was 144.3 (93.8) months overall and was significantly shorter in the participant group (128 [89.3] versus 162.0 [95.7] in the nonparticipant group [P = .0025]). The stage of disease in this population was mainly asymptomatic with a high rate of CD4 count, and the 2 variables were not significantly different between groups. The number of cigarettes smoked per day was also similar between the 2 groups. The participant group had a higher Fagerström score (5.2 [2.7] versus 2.1 [1.7] in the nonparticipant group, P < .0001) indicating that participant had a stronger dependence on tobacco. The anxiety and depression scores were significantly higher in the participant group compared to the nonparticipant group (P = .04).
Characteristics of Smoking Patients Who Agreed and Declined to Participate in TASMI.
Abbreviations: CDC, US Centers for Disease Control and Prevention; IDU, intravenous drug use; OR, odds ratio; SD, standard deviation; TASMI, Tabacologie et Suivi en Maladies Infectieuses.
aQuanlitative variables were compared with chi-square test and quantitative variables were compared with Student t or Mann-Whitney test, as appropriate.
bFagerström score varies between 0 and 6, the score increases when addiction to tobacco increases.
cThe anxiety score is a score that goes from 0 to 21, the score increases when anxiousness increases.
dThe depression score is a score that goes from 0 to 21, the score increases when depression increases.
In multivariate analysis (Table 2), the 2 variables that were independently associated with the acceptance of TASMI participation were higher Fagerström score (OR = 1.75, 95% CI: 1.48-2.02, P < .001) and IDU (OR = 0.35, 95% CI: 0.14-0.89, P = .02).
Efficacy of TASMI Intervention
Of the 147 smokers who accepted to participate in the TASMI intervention, 106 (72%) completed the study and 41 (28%) withdrew (Figure 1). The intent-to-treat expired carbon monoxide confirmed rate of smoking cessation after 6 months was 60 of 147 (40.8%, 95% CI: 32.9-48.8). The per-protocol expired carbon monoxide–confirmed rate of smoking cessation after 6 months was 60 of 106 (56.6%, 95% CI: 47.2-66.0). The sociodemographics and HIV-related characteristics are presented in Table 3. The Fagerström score was nonsignificantly lower among those who quit smoking compared to those who experienced a failure (5.8 [2.3] versus 6.3 [2.3], P = .17). As shown in Table 3, 2 factors were independently associated with the TASMI intervention outcome: history of depression (OR = 0.28, 95% CI: 0.11-0.71, P = .007) and the use of varenicline (OR = 2.45, 95% CI: 1.0-5.8, P = .04).
Characteristics of Participants Who Succeeded in Quitting Smoking after 6 Months.
Abbreviations: CDC, US Centers for Disease Control and Prevention; CO, carbon monoxide; CVA, cerebrovascular accident; HTA, hypertension; IDU, intravenous drug use; MI, myocardial infarction; OR, odds ratio; SD, standard deviation.
aQuanlitative variables were compared with chi-square test and quantitative variables were compared with Student t or Mann-Whitney test, as appropriate.
bOne patient had missing data.
Discussion
As expected, the overall prevalence of active smoking was high in our population of HIV-infected patients. Although the smoking cessation program was clearly effective among participants who complete the intervention (56%), several smokers responded poorly to smoking cessation efforts, with rather disappointing overall success rates, as only 1 of the 5 active smokers was able to quit at 6 months, if we consider all potential active smokers (Figure 2). The use of varenicline was independently associated with success.
The smoking cessation cascade.
Two hundred eighty (39%) of the HIV-infected patients reported active smoking in our sample. This prevalence is higher than in the French general population (31% 14 ). Smoking rates tend to be higher among persons of lower socioeconomic status, 15 drug-abusing populations, 16 and certain ethnic minority groups. 17 Given that HIV infection tends to cluster among individuals in these risk groups, smoking prevalence among HIV-infected persons who generally exhibit these sociodemographic or risk profiles is higher than that in the general population. A study conducted in Switzerland between 2000 and 2002 found a smoking prevalence of 61%. 18 The smoking prevalence in our sample was, however, similar to the settings in France. In a cross-sectional study, 14 255 (43%) of the 593 patients reported active smoking. In the Agence Nationale de Recherche contre le SIDA (ANRS) CO3 Aquitaine Cohort of HIV-infected patients, 5 257 (50%) of 509 reported active smoking.
Consistent with prior studies, 14,19 smokers were younger and more frequently infected by IDU. Contrary to a prior study conducted in 2002 which found a positive association between smoking and duration of HIV infection, 19 we identified a negative association between smoking and duration of HIV infection that remained statistically significant in multivariate analysis. Among the potential explanations, the fact that HIV-infected patients are treated earlier, regardless of the levels of CD4 count, and therefore may feel less concerned with the risks of health associated with cigarette smoking. This figure may also reflect a surviving bias in our cohort.
Of the 280 smokers identified through the screening, 133 (47%) declined to participate in the intervention and 41 (15%) additional patients withdrew from the intervention. Numerous studies have attempted to identify factors that may predict the success of smoking cessation, but relatively less attention has been paid to factors predicting the initiation of a cessation program to quit tobacco smoking and even less completion of the program. In a smoking cessation trial conducted in the general population, 20 471 (63%) of the 753 eligible patients were not enrolled. In a large study conducted in Spain among HIV-infected smokers, 21 1137 (68%) of the 1681 patients accepted to initiate, but the rate of completers was not reported. In another randomized trial 22 comparing a smoking cessation program to standard care among HIV-infected smokers in the United States including financial incentives, 43 (52%) of the 83 eligible patients were not included and 12 (15%) additional patients did not complete the study. Taken together, the potential efficacy of any smoking cessation program is therefore hampered by the low acceptance rate. We identified patients infected due to IDU and those who had a low dependence on smoking as being independently less prone to participate.
Sixty active smokers reached the 6-month smoking cessation end point, which represents 57% in the per-protocol data set and 41% in the intent-to-treat data set. Of note, the expired carbon monoxide was an objective end point, as patients self-report has been shown to overestimate the success of tobacco cessation interventions. In a systematic review of randomized and quasi-experimental trials, 23 the rate of smoking cessation varied between 39% and 50% among 4 intervention participants after the intervention phase. 18,24 -26 Of the 5 randomized controlled trials, 4 reported significantly 27 -30 and 1 nonsignificantly 31 increased smoking cessation rates among participants in the intervention groups compared with those in the control groups. Of note, the design and follow-up of the intervention and control groups were heterogeneous between studies. Nevertheless, previous research supports our results that smoking cessation program is effective in the HIV population as well. Interestingly, the use of varenicline as a part of the nicotine replacement therapy combination was independently associated with the success of the program. Varenicline binds with high affinity and selectivity at α4β2 neuronal nicotinic acetylcholine receptors. The efficacy of varenicline in smoking cessation is believed to be the result of varenicline’s activity at α4β2 subtype of the nicotinic receptor where its binding produces agonist activity, while simultaneously preventing nicotine binding to these receptors. To the best of our knowledge, the benefit of its use as a part of smoking cessation programs among HIV-infected patients has been reported only in 2 previous reports. 24,32
Our model included a separate referral for smoking cessation, which might have led to reduced acceptance of the intervention. Colocation and integration of services within a patient-centered care site may improve acceptance. On the other hand, it allowed the management of smokers with a highly specialized care.
We are aware of limitations. First, our smoking cessation program had no control group, smokers who declined the program were not followed for the primary end point, and therefore, the effectiveness of TASMI cannot be compared with the standard care. However, a prior study found that only 15% of the HIV-infected smokers quit by standard care alone. 18 Second, this study was conducted in a single center, which may limit the generalizability of the results regarding our estimated prevalence of end points and risk factors associated with them, although as described above, our finding was generally consistent with prior research on this topic. Third, the follow-up was limited to 6 months. Finally, the use of nicotine replacement therapy was not standardized, free of charge, and was left to the discretion of the smoking cessation specialist. Therefore, residual confounding may remain despite the use of multivariate analysis.
More research is needed to identify strategies to improve the smoking cessation cascade, by increasing screening and retention in tobacco cessation care, particularly in specific groups such as intravenous drug users. Financial incentives may be a promising tool that deserves further studies in these populations. Our results may also have potential implication for clinical practice in settings similar to ours. In the context of HIV-infected patients, a systematic tobacco screening and linkage to a tobacco cessation specialist outside the routine outpatient HIV consultation is feasible and leads to a substantial number of successes. As a first step, this organization alleviates the limited awareness and experience of HIV specialists in smoking cessation program. Second, as HIV physicians see more and more successful tobacco quitters, together with the health benefit, among their HIV-positive patients, they may want to increase their awareness and skills to start their own smoking cessation program.
Footnotes
Acknowledgments
The authors thank Claudine Hecquard and Pascale Goubin for their technical support.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The study was supported by an academic grant DEPREV-11-124 from INCa (Institut National du Cancer) and by Ligue Contre le Cancer from Calvados, France.