Persons living with HIV (PLWH) smoke at approximately triple the rate of the US general population,1 and tobacco use has emerged as a leading cause of mortality in this group.2,3 Although modeling studies have demonstrated the significant impact that effective cessation interventions could have on morbidity and mortality among PLWH,3,4 research conducted to date has shown weak evidence of short-term efficacy and no evidence of long-term efficacy, in quit rates, among the range of tobacco treatments studied.5 In our randomized controlled trial of positively Smoke Free (PSF) group therapy, a social cognitive theory–based intensive behavioral intervention, we found increased rates of biochemically confirmed cessation at three months with loss of the cessation advantage at six-months.6 A six- or twelve-month abstinence is the standard primary outcomes in tobacco treatment trials,7 whereas a few studies have reported long-term follow-up of 8 years or more, and demonstrated sustained effects of behavioral treatments on smoking activities beyond the 12-month mark in other populations.8,9 To the best of our knowledge, no tobacco treatment trials studying PLWH smokers have documented cessation outcomes beyond twelve-months. Herein, we report extended follow-up of our cohort at a mean of approximately 3 years after trial enrollment.
The trial of PSF group therapy, described in a separate publication,6 enrolled at 3 sites, 2 in the Bronx, New York, and one in Washington DC, between May 2014 and August 2017. The PSF intervention consisted of eight 90-minute group therapy sessions delivered over 6 weeks, co-led by a masters- or doctoral-level social worker or psychologist and an ex-smoker living with HIV peer (all leaders completed Tobacco Treatment Specialist training), with content targeting the PLWH audience and guided by the social cognitive theory behavioral model.10 In the original study, participants in the intervention condition completed a mean of 3.8 sessions (4.7 sessions excluding the 43 participants who never attended a single group therapy session).6 The late follow-up study was limited to the Bronx participants, and enrolled from June 2018 to September 2019. The only eligibility criteria for enrollment into the follow-up study were (1) having had at least 365 days elapse since enrollment into the original study and (2) prior enrollment at one of the Bronx sites in the original study (which included PLWH adults aged ≥18 years who had smoked combusted tobacco in the past 5 days, endorsed some motivation to quit, and had no contraindications to nicotine patch use).
Participants in the original study were randomly allocated in a 1:1 ratio to PSF or to a control condition consisting of one session of standardized brief advice to quit and a self-help brochure. All participants were offered a 12-week course of nicotine patches. In the PSF intervention, quit day was scheduled for 28 days after the first group therapy session, and therefore, the three-month assessment for abstinence occurred on or about 120 days postenrollment and the 6-month assessment on or about 210 days postenrollment. The cessation endpoints were seven-day point-prevalence abstinence confirmed by exhaled carbon monoxide (ECO) < 10 parts per million (ppm) at 3 and 6 months. In the current study, an attempt was made to contact by phone, letter, and/or email everyone who enrolled in the original study in the Bronx, and they were invited to return, participate in a new informed consent procedure, complete an audio computer-assisted self-interview (ACASI, QDS software) and an ECO measurement. Participants received travel fare and a $50 gift card.
Consistent with the 3- and 6-month abstinence outcomes in the original study, the biochemically confirmed, 7-day point-prevalence abstinence endpoint at the late follow-up visit used a worst-case scenario, intent-to-treat analytic strategy, ie, participants from the original study who did not complete the current study were presumed to be continued smokers unless they were known to be deceased (in which case they were excluded). We also collected and analyzed 12-month point-prevalence abstinence at the late follow-up visit. Further details about the study's measures and outcomes may be found in Supplemental Table 1, Supplemental Digital Content, https://links.lww.com/QAI/B442. We evaluated baseline factors for their associations with the 3 abstinence outcomes, biochemically confirmed 7-day point-prevalence abstinence at late follow-up, biochemically confirmed 12-month point-prevalence abstinence at late follow-up, and biochemically confirmed abstinence at any of the study timepoints (3-months, 6-months, and late follow-up). Comparisons of proportions were done using χ2 or Fisher exact analyses. Comparisons of means were done using Student t test or the Mann–Whitney U test where appropriate. Longitudinal linear mixed effects modeling adjusting for baseline measurements was used to compare self-efficacy scores between intervention and control groups. A missing at random strategy was used for the latter analyses. All analyses were completed using the SPSS V25.0 and R (R Core Team 2018).
The original and the late follow-up study were reviewed and approved by the institutional review board of the Albert Einstein College of Medicine.
There were 342 participants enrolled in the original study at the Bronx sites. Eleven were known to be deceased since completing the original trial (2 intervention participants and 9 controls, P = 0.03, Fisher exact). Of the remaining 331, 194 (58.6%) were successfully recalled to complete the late follow-up study. Twelve declined participation and 125 could not be contacted despite numerous attempts. We compared baseline characteristics of those who participated to those who were lost to follow-up or declined participation. Similar proportions of intervention participants vs. controls participated (61.8% vs. 54.8%, P = 0.20). The only significant differences noted were that participants who completed the late follow-up study were less likely to be men 45.2% vs. 63.4% [odds ratio (OR) = 0.48 (0.30–0.75), P = 0.001], have educational attainment ≥ high school graduation 58.8% vs. 70.5% [OR = 0.60 (0.38–94), P = 0.03], and had a higher mean baseline ECO level 13.4 vs. 11.3 ppm, P = 0.02. The remaining baseline characteristics were similar between those who completed the late follow-up study and those who did not.
Participants completed their late follow-up visits a mean of 38.1 ± 12.7 months (range = 13.2–62.0 months) after study enrollment, with no significant difference between the intervention and control groups. There were no significant differences between intervention and control participants in reported use of nicotine replacement therapy, bupropion, or varenicline since completing the original study. Twenty-one of 165 (12.7%) intervention participants vs. 11/166 (6.6%) control participants had biochemically confirmed 7-day point-prevalence abstinence at late follow-up, OR = 2.06 (95% CI: 0.96–4.41), P = 0.06. Seventeen of 165 (10.3%) intervention participants had biochemically confirmed 12-month point-prevalence abstinence (ie, denied smoking a cigarette, even a puff, in the past 12-months) vs. 7/166 controls (4.2%) at late follow-up, OR = 2.61 (95% CI: 1.05–6.47), P = 0.03. Forty-two of 167 (25.1%) intervention participants vs. 28/175 (16.0%) of control participants had biochemically-confirmed seven-day point-prevalence abstinence at any of the study follow-up visits, OR = 1.76 (95% CI: 1.03–3.01), P = 0.04. The primary outcome of the parent study was biochemically confirmed 7-day point-prevalence abstinence at 6-months. Of the 43 participants at the Bronx sites who achieved this outcome, 18 (41.9%) were confirmed abstinent at late follow-up [9/23 (39.1%) of the group therapy participants vs. 9/20 (45.0%) of controls by ITT, OR = 0.79 (95% CI: 0.23–42.65), P = 0.70]. Of those who were not confirmed abstinent at 6-months, group therapy participants were significantly more likely to be confirmed abstinent at late follow-up compared with controls [12/140 (8.6%) vs. 2/146 (1.4%), OR = 6.75 (95% CI: 1.48–30.7), P = 0.005]. Finally, among participants assigned to the intervention, we compared group therapy session attendance between those who achieved the abstinence endpoints and those who did not. For the 7-day late follow-up point-prevalence, 12-month late follow-up point-prevalence, and the 7-day point-prevalence at any study follow-up abstinence outcomes, abstinent participants attended more sessions than nonabstinent ones, but these differences did not achieve statistical significance (4.3 vs. 3.6, P = 0.24; 4.6 vs. 3.7, P = 0.21; and 4.2 vs. 3.6, P = 0.17, respectively).
Table 1 summarizes study variables and their associations with the abstinence outcomes. Baseline factors that predicted seven-day point-prevalence abstinence at late follow-up included female gender, nondaily smoking, current marijuana use, denial of current cocaine use, lower nicotine dependence, and undetectable HIV viral load. Baseline factors that predicted 12-month point-prevalence abstinence at late follow-up included allocation to the group therapy intervention and lower nicotine dependence. Baseline factors that predicted the achievement of abstinence at any of the study timepoints included allocation to the group therapy intervention, non-Latinx ethnicity, lower daily cigarette consumption, nondaily smoking, not reporting current cocaine use, lower nicotine dependence, and undetectable HIV viral load.
A major behavioral target of the intervention was the promotion of self-efficacy, so we completed analyses comparing self-efficacy between those allocated to the group therapy condition and controls. The results of these analyses are depicted in Figure 1. It is important to note that lower scores on the Self Efficacy/Temptations Scale indicate higher levels of self-efficacy. Participants in the group therapy condition had slightly higher self-efficacy at baseline. Self-efficacy remained higher in this group compared with controls throughout late follow-up, and in the linear mixed effects model adjusted for baseline measures, the difference in self-efficacy between the intervention and control groups was highly significant (P = 0.002).
Among the most salient and discouraging findings of the largest meta-analysis conducted to date of tobacco treatments for PLWH smokers is that there has been no evidence of efficacy, in successful cessation, at 6 months or later.5 Similar to prior trials of intensive behavioral therapy combined with an offer of nicotine replacement therapy in PLWH, our trial of PSF group therapy demonstrated a significant cessation advantage among intervention participants at 3 months, but not at 6 months.11 In our original trial, there was a doubling of abstinence rates in control participants between the 3 and 6 month visits in the absence of any protocol-driven tobacco treatment, that we were unable to explain. One additional placebo-controlled trial of varenicline plus counseling for PLWH smokers published after the meta-analysis also showed significant efficacy at 12 weeks, but not at 24 weeks.11
In the present study, we undertook a long-term follow-up assessment of PLWH who participated in a cessation trial conducted at 2 sites in the Bronx, New York. We were successful in recalling 58.6% of the original cohort, a mean of more than 3 years after initial enrollment. Reporting of long-term cessation outcomes of this duration is rare in the medical literature. In their systematic review of abstinence outcomes in 28 tobacco treatment trials, Hughes et al12 found that only 2 (7%) reported outcomes beyond 12 months. We are not aware of any prospective trials of tobacco treatment for PLWH that have reported cessation outcomes beyond 12 months. The current study suggests that follow-up at later timepoints may be important.
We found that at a mean of more than 3 years postenrollment, more group therapy participants vs. controls had biochemically verified 7-day point-prevalence abstinence (P = 0.06), significantly more had biochemically verified 12-month point-prevalence abstinence (P = 0.03), and significantly more had biochemically verified point-prevalence abstinence at any of the 3 study timepoints at which it was measured (P = 0.04). Among those who did not succeed in quitting at 6-months, group therapy participants were much more likely to quit after 6-months (P = 0.005). We also found that the initial improvements in abstinence self-efficacy, a key behavioral target of the group therapy, seemed to be sustained at late follow-up in the PSF intervention group.
The determination of baseline factors that are associated with abstinence at long-term follow-up may be useful in stratifying PLWH smokers into prognostic categories for potential cessation, and it may also allow for the development of improved cessation strategies. In our cohort, we found that allocation to our group therapy intervention, female gender, nondaily smoking, current marijuana use, denial of current cocaine use, lower nicotine dependence, and undetectable HIV viral load were associated with the abstinence outcomes at late follow-up.
Our study had several limitations that require mention. First, although we successfully recalled almost 60% of the cohort after they completed the original study, our loss to follow-up rate was sizeable and may have introduced important biases. The intention to treat/lost to follow-up = nonabstinent strategy, however, applies a “worst case scenario” to the lost subsample, and using this strategy, the abstinence rate of 12.7% at late follow-up is respectable in this challenging target group, and is within the meta-analytically derived range of estimated abstinence rates after group therapy delivered to other populations, ie, 13.9% (95% CI: 11.6%–16.1%).13 The timing of the late follow-up assessment in relation to original study enrollment spanned a large time period, ranging from 13 to 62 months, and our cessation rates were too low to assess time as a variable in our cohort. Our inability to include the Washington DC site in this project not only reduced the size of the study population, but also constrained it to a single urban center, and results may not be generalizable to other geographic locations or to sociodemographically different groups of PLWH. The most recent tobacco treatment guidelines encourage combination long- and short-acting nicotine replacement therapy as first line pharmacotherapy,14 so the lack of NRT effect may have been partially attributable to suboptimal usage. In addition, just-published guidelines recommend lowering the ECO cutoff for nonsmoking status to 5–6 ppm (although they still recommend a cutoff of 10 ppm in certain settings),15 so it is possible that a small number of smokers may have been misclassified in our analyses. In our post-hoc exploration of this matter, we found that intervention participants were less likely to have ECO levels in the intermediate 5–9 ppm range than controls. Finally, results are from post-hoc analyses that were not included in the trial's original plans and are subject to the potential biases inherent in projects of this type.
In conclusion, we completed late follow-up assessments in a group of 194 PLWH who enrolled in a randomized controlled trial of PSF group therapy vs. brief advice to quit a mean of more than 3 years after their baseline visits. In this first study of prolonged follow-up of PLWH smokers who were exposed to tobacco treatment, we found that those participants who were allocated to the group therapy condition enjoyed higher abstinence rates approximately 3 years after the behavioral intervention. It is possible that group therapy participants acquired skills, beliefs, and behaviors in the course of their treatment that translated into clinical benefit several years later. These findings challenge the view that tobacco treatments in PLWH smokers produce short-term but not long-term cessation benefits,5 and suggest a role for later assessment timepoints in future studies of tobacco treatments for PLWH.
1. Mdodo R, Frazier EL, Dube SR, et al. Cigarette smoking prevalence among adults with HIV compared with the general adult population in the United States: cross-sectional surveys. Ann Intern Med. 2015;162:335–344.
2. Helleberg M, Afzal S, Kronborg G, et al. Mortality ttributable to smoking among HIV-1-infected individuals: a nationwide, population-based cohort study. Clin Infect Dis. 2013;56:727–734.
3. Reddy KP, Parker RA, Losina E, et al. Impact of cigarette smoking and smoking cessation on life expectancy among people with HIV: a US-based modeling study. J Infect Dis. 2016;214:1672–1681.
4. Reddy KP, Kong CY, Hyle EP, et al. Lung cancer mortality associated with smoking and smoking cessation among people living with HIV in the United States. JAMA Intern Med. 2017;177:1613–1621.
5. Pool ER, Dogar O, Lindsay RP, et al. Interventions for tobacco use cessation in people living with HIV and AIDS. Cochrane Database Syst Rev. 2016:CD011120.
6. Stanton CA, Kumar PN, Moadel AB, et al. A multicenter randomized controlled trial of intensive group therapy for tobacco treatment in HIV-infected cigarette smokers. J Acquir Immune Defic Syndr. 2020;83:405–414.
7. Hughes JR, Keely JP, Niaura RS, et al. Measures of abstinence in clinical trials: issues and recommendations. Nicotine Tob Res. 2003;5:13–25.
8. Murray RP, Connett JE, Rand CS, et al. Persistence of the effect of the Lung Health Study (LHS) smoking intervention over eleven years. Prev Med. 2002;35:314–319.
9. Nohlert E, Ohrvik J, Tegelberg A, et al. Long-term
follow-up of a high- and a low-intensity smoking cessation intervention in a dental setting—a randomized trial. BMC Public Health. 2013;13:592.
10. Bandura A. Social cognitive theory: an agentic perspective. Annu Rev Psychol. 2001;52:1–26.
11. Ashare RL, Thompson M, Serrano K, et al. Placebo-controlled randomized clinical trial testing the efficacy and safety of varenicline for smokers with HIV. Drug Alcohol Depend. 2019;200:26–33.
12. Hughes JR, Carpenter MJ, Naud S. Do point prevalence and prolonged abstinence measures produce similar results in smoking cessation studies? A systematic review. Nicotine Tob Res. 2010;12:756–762.
13. Fiore MC, Bailey BW, Cohen SJ, et al. Treating Tobacco Use and dependence. Clinical Practice Guideline. Rockville, MD: U.S. Department of Health and Human Services. Public Healh Service; 2000.
14. Writing Committee, Barua RS, Rigotti NA, et al. 2018 ACC expert consensus decision pathway on tobacco cessation treatment: a report of the American College of Cardiology Task Force on Clinical Expert Consensus Documents. J Am Coll Cardiol. 2018;72:3332–3365.
15. Benowitz NL, Bernert JT, Foulds J, et al. Biochemical verification of tobacco use and abstinence: 2019 update. Nicotine Tob Res. 2019. doi: 10.1093/ntr/ntz132.