Practices, attitudes, and confidence related to tobacco treatment interventions in HIV clinics: a multisite cross-sectional survey

Abstract

Tobacco use disorder (TUD) is a major threat to health among people with HIV (PWH), but it is often untreated. Among HIV clinicians and staff, we sought to characterize practices, attitudes, and confidence addressing TUD among PWH to identify potential opportunities to enhance provision of care. Cross-sectional deidentified, web-based surveys were administered from November 4, 2020 through December 15, 2020 in HIV clinics in three health systems in the United States Northeast. Surveys assessed provider characteristics and experience, reported practices addressing tobacco use, and knowledge and attitudes regarding medications for TUD. Chi-square tests or Fisher’s exact tests were used to examine differences in responses between clinicians and staff who were prescribers versus nonprescribers and to examine factors associated with frequency of prescribing TUD medications. Among 118 survey respondents (56% prescribers), only 50% reported receiving prior training on brief smoking cessation interventions. Examining reported practices identified gaps in the delivery of TUD care, including counseling patients on the impact of smoking on HIV, knowledge of clinical practice guidelines, and implementation of assessment and brief interventions for smoking. Among prescribers, first-line medications for TUD were infrequently prescribed and concerns about medication side effects and interaction with antiretroviral treatments were associated with low frequency of prescribing. HIV clinicians and staff reported addressable gaps in their knowledge, understanding, and practices related to tobacco treatment. Additional work is needed to identify ways to ensure adequate training for providers to enhance the delivery of TUD treatment in HIV clinic settings.

Clinicians and staff who work with patients living with HIV reported addressable gaps in knowledge, understanding, and practices related to treating tobacco use disorder.

Implications.

Practice: Addressing identified gaps in provider knowledge, understanding, and practices related to tobacco treatment is important for improving health outcomes for people living with HIV.

Policy: Healthcare systems should consider ways to increase the delivery of evidence-based treatment for tobacco use among people living with HIV to enhance provision of care.

Research: Future research is needed to identify optimal practices for training providers to enhance the delivery of TUD treatment in HIV clinic settings.

Introduction

Cigarette smoking is a major health threat to people with HIV (PWH) across the globe, with prevalence estimates of cigarette smoking ranging from 10% to 45% [1, 2]. In the USA, estimates indicate over one-third of PWH report current cigarette smoking [3]. In addition to the health consequences of cigarette smoking, including lung cancer and cardiovascular disease [4], tobacco use is also associated with worse control of HIV infection [2]. Tobacco use is the leading preventable cause of morbidity and mortality among PWH [5], and is responsible for more lost life years than treated HIV [6]. Importantly, stopping smoking improves quality of life and decreases symptom burden, morbidity, and mortality risk [7–10]. Given the demonstrated safety and efficacy of both behavioral and medication treatments to address tobacco use disorder (TUD) among PWH [11–14], these treatments are widely recommended by leading international and national organizations [15, 16]. Further, because models integrating TUD and HIV care are both feasible and beneficial, there is an ongoing effort to promote delivery of TUD treatment in HIV clinical settings [16].

Despite the existence of these guidelines and data demonstrating that many PWH who smoke do want to stop [17–20], many patients do not receive indicated TUD treatments beyond brief counseling [21–23]. For example, data from a national sample of Veterans with HIV who smoke demonstrated that less than one third received any medications for TUD and that treatment rates were significantly lower compared with a matched sample of Veterans without HIV (29% vs. 36%) [24]. To further understand reasons for the gaps in provision of TUD treatment among PWH, we surveyed clinicians and staff working in one of three health systems in the United States Northeast in the context of a patient-focused randomized clinical trial. Given that HIV clinical settings typically provide multidisciplinary, team-based care with a range of individuals who are potentially well positioned to address TUD [25–32], we surveyed both clinicians and staff who were involved in treating patients in the HIV clinics and sought to: (a) characterize reported practices in addressing tobacco use, (b) examine attitudes regarding medications for tobacco use, and (c) examine factors associated with reported frequency of prescribing first-line medications, nicotine replacement therapy (NRT) and varenicline, to identify potential opportunities to enhance the provision of care to PWH who smoke. This work is important for identifying actionable opportunities to enhance provision of TUD treatment in HIV treatment settings.

Methods

Study overview

As previously described [33], ASMARTApproach toTreatingTobacco Use Disorder in Persons with HIV (SMARTTT) is a multisite randomized trial originally designed to identify the optimal adaptive approach involving clinical pharmacist-delivered first-line tobacco treatment medications (i.e., NRT and varenicline) and contingency management to promote exhaled carbon monoxide-confirmed smoking abstinence and its impact on HIV-specific outcomes. Consistent with a hybrid type 1 effectiveness–implementation design [34], we conducted a survey of clinicians and staff at participating sites within the first 6 months of initiation of patient recruitment to understand factors that may impact future implementation efforts of our intervention packages in real-world settings. The current study reports the results of this baseline survey conducted from November 4, 2020 through December 15, 2020. The survey was designed to provide insights on factors that may impact adoption and implementation of the intervention components and packages, including perspectives on the evidence for these treatments and context for delivering these treatments in HIV treatment settings.

Study context and participants

The multisite study is being conducted in the context of Yale’s Center for Interdisciplinary Research on AIDS (CIRA)-supported New England HIV Implementation Science Network, whose mission includes stimulating and supporting research and evaluating collaborations across New England and neighboring regions and promoting implementation science. The coordinating center is located at Yale School of Medicine, in New Haven, CT; the Yale Center for Analytical Sciences (YCAS) coordinates data management and statistical support. The three participating health systems include: (a) Yale New Haven Hospital (YNHH)’s Nathan Smith Clinic and Haelen Center, New Haven, CT; (b) the State University of New York (SUNY) Downstate’s STAR Health Center, Brooklyn, NY; and (c) Mount Sinai’s Institute for Advanced Medicine’s Jack Martin Clinic and Morningside Clinic, New York, NY. These sites were chosen given an identified need for additional patient-level interventions to address tobacco use [35] as well as ongoing commitment to address tobacco-related harms [36–38]. The Institutional Review Board (IRB) at Icahn School of Medicine at Mount Sinai serves as the single IRB for this protocol and approved this study for all sites and per reliance agreements with Yale School of Medicine and SUNY Downstate. The study is registered on Clinicaltrials.gov NCT04490057.

Study procedures

To obtain a broad perspective on healthcare provider experience providing tobacco treatment to patients in HIV clinics, we invited clinicians and staff who were involved in treating patients to participate in a confidential, online survey. All clinicians and staff were invited by email to participate after an introductory email was sent out by clinic leadership and potentially eligible clinicians and staff were identified by clinic leadership/site-PIs. Survey data were collected using a secure online REDCap-based survey system hosted at Yale University [39, 40].

Survey

Items included in this survey were drawn from the existing literature assessing provider practices and attitudes [22, 35, 41, 42], based upon validated tools as possible [43], and then piloted and refined as needed by the investigative team. The full survey has been previously published [33].

Clinician and staff characteristics

Clinician and staff characteristics collected included their current role, the number of patients they typically treat, their training and experience providing tobacco interventions, and their own tobacco use status. Respondents were categorized as prescribers if they reported being a physician, registered nurse, nurse practitioner, or physician assistant; otherwise they were categorized as nonprescribers (e.g., patient navigator, social worker, case manager, medical assistant).

Clinician and staff reported practices in addressing tobacco use

Respondents were asked to rate their frequency measured on a five-point Likert scale in addressing tobacco use on ­several recommended practices, including: how often they ask patients about their smoking, provide specific evidence-based smoking cessation interventions (e.g., motivational interviewing), provide referrals for smoking cessation services (e.g., Quitline), and counsel patients on the impact of smoking on their health. This included reported adherence to clinical practice guidelines for brief TUD interventions in healthcare settings for patients who are ready to quit (5As: Ask, Advise, Assess, Assist, Arrange) or currently unwilling to quit (5Rs: Relevance, Risks, Rewards, Roadblocks, Repetition) and involves asking patients about smoking and assisting with services and referrals [44].

Among prescribers, the survey also assessed how often they prescribe specific medications for TUD including NRT, varenicline, and bupropion. Responses to reported practices addressing tobacco use and prescribing medications were dichotomized as high frequency (“always,” “most of the time”) versus low frequency (“sometimes,” “rarely,” “never”).

Clinician and staff attitudes on medications for TUD

Self-rated knowledge of best practices and potential concerns with NRT and varenicline for treating TUD were measured on a five-point Likert scale (“strongly agree,” “somewhat agree,” “neither agree nor disagree,” “somewhat disagree,” and “strongly disagree”). Responses were dichotomized as “strongly agree” or “somewhat agree” versus “neither agree nor disagree,” “somewhat disagree,” or “strongly disagree.”

Data analysis

First, we used descriptive statistics to characterize self-reported practices addressing TUD, prior training experience for treating TUD, and attitudes regarding medications for TUD. Additionally, we compared responses between prescribers and nonprescribers using chi-square tests and Fisher’s exact tests for categorical variables. Next, among prescribers, we used descriptive statistics to examine self-reported frequency of prescribing medications for TUD. Lastly, among prescribers, we used chi-square tests or Fisher’s exact tests to examine factors associated with high frequency versus low frequency of prescribing NRT and varenicline. p value less than .05 was considered statistically significant. Given the nature of these exploratory analyses, corrections for multiple comparisons were not made. Analyses were conducted using SAS 9.4 (SAS/STAT software, Cary, NC).

Results

Participant characteristics

Across 3 New England healthcare systems, 188 HIV clinicians and staff who treat patients were invited to complete the survey, among whom 140 (74%) initiated the survey and 118 (63%) provided valid responses to the variables of interest. Among the 118 included in the analytic sample, 56% (N = 66) were prescribers and 44% (N = 52) were nonprescribers (Table 1). On average, respondents worked in a setting caring for patients with HIV for almost 10 years. HIV clinicians and staff reported treating many patients each month who have TUD (prescribers median 20 patients per month, interquartile range = 6–50; nonprescribers 10 patients per month, interquartile range = 2–20). Overall, 50% of HIV clinicians and staff reported ever receiving training on providing brief counseling for smoking cessation, 42% reported receiving training on motivational interviewing, and only 9% reported receiving training on 5As/5Rs.

Table 1.

Healthcare provider characteristics and experience treating tobacco use disorder (TUD) by prescriber status

Variable	Prescriber (N = 66)	Nonprescriber (N = 52)	Total (N = 118)
No. of years working in a setting caring for patients with HIV, mean (SD)	10.6 (9.3)	9.6 (8.8)	10.1 (9.1)
No. of patients with HIV typically treated in 1 month, median (IQRa)	30 (15–100)	30 (20–75)	30 (15–85)
No. of patients treated in past 30 days who have TUD, median (IQR)	20 (6–50)	10 (2–20)	12 (5–40)
No. referred to a healthcare provider or smoking cessation program in past 30 days, median (IQR)	0 (0–5)	4 (0–12)	1 (0–10)
No. of patients you provided counseling for TUD in past 30 days, median (IQR)	15 (5–40)	3 (0–10)	10 (2–30)
No. of patients you prescribed medications for TUD in past 30 days, median (IQR)	5 (0–20)	0 (0–0)	0 (0–10)
Received training on providing brief counseling for smoking cessation, N (%)	36 (54.6%)	23 (44.2%)	59 (50.0%)
Received training on 5As/5Rs, N (%)	6 (9.1%)	4 (7.7%)	10 (8.5%)
Received training on motivational interviewing (MI), N (%)	30 (45.5%)	19 (36.5%)	49 (41.5%)
Smoking status of the provider, N (%) current use	2 (3.0%)	5 (9.6%)	7 (5.9%)

Interquartile range.

Reported practices addressing tobacco use among prescribers and nonprescribers

Figure 1 compares reported practices providing treatment, referring patients for smoking cessation services, and counseling patients on the impact of smoking on their health by prescriber versus nonprescriber status. Overall, 24.1% of prescribers and 18.8% of nonprescribers responded in agreement that due to competing priorities, it is not ­practical to address smoking cessation in routine healthcare visits for patients with HIV. However, over 98% of prescribers and over 67% of nonprescribers reported asking patients about smoking always or most of the time (Fig. 1). Prescribers were more likely than nonprescribers to report a high frequency of providing TUD treatment in several domains (i.e., ask patients about their smoking, advise to stop or cut down, discuss medications about smoking, discuss ways to overcome barriers to stopping). Additionally, prescribers reported high frequencies of counseling patients on the impact of smoking on their health in areas outside of HIV (i.e., lung disease, cardiovascular disease, and cancer risk, 79.7%–84.5% vs. 44.9%–53.1% among nonprescribers). However, similar proportions of prescribers and nonprescribers reported counseling patients on the impact of smoking on HIV (54.2% vs. 46.9%). Furthermore, nonprescribers were more likely than prescribers to refer patients elsewhere for smoking cessation services (i.e., nurse specialist, counselor, clinical pharmacist).

Fig 1.

Characterizing providers’ reported practices in addressing tobacco use. Frequency of providing specific tobacco treatment services by prescriber versus nonprescriber status. Values indicate the proportion of providers reporting providing treatment services frequently (i.e., reported as “always” or “most of the time”), *p < .05.

Attitudes on medications for TUD

Figure 2 displays attitudes, confidence, and concerns recommending NRT and varenicline for treating tobacco use by prescriber versus nonprescriber status. Prescribers were more likely than nonprescribers to correctly endorse that NRT (70.2% vs. 42.6%) and varenicline (65.5% vs. 21.7%) are recommended by clinical practice guidelines to help patients with HIV stop smoking. However, over 1/3 of prescribers and almost half of nonprescribers reported they did not know enough about clinical practice guidelines for these medications. Prescribers were also more likely than nonprescribers (44.8% vs. 21.7%) to endorse concerns about psychiatric effects of varenicline in patients.

Fig 2.

Providers’ attitudes, confidence, and concerns recommending medications to treat tobacco use by prescriber status. Reported attitudes, concerns, and confidence regarding nicotine replacement therapy (NRT) and varenicline by prescriber status. Values indicate the proportion of each groups agreement with the statement (as indicated by “strongly agree” or “agree” responses), *p < .05.

Prescribing medications for TUD

Figure 3 shows the proportions of self-reported prescribing of various smoking cessation medications. Prescribing any NRT was the most common with 44.1% reporting prescribing it always or most of the time. However, fewer reported prescribing first-line treatment options of dual NRT (22.1%) or varenicline (27.1%) always or most of the time. Many prescribers reported rarely or never prescribing these medications (27.1% NRT, 37.3% varenicline, 61.0% dual NRT), and most rarely or never switched pharmacotherapy for patients from NRT to varenicline for nonresponse (60.4%).

Fig 3.

Self-reported frequency of prescribing medications for TUD among HIV healthcare providers treating patients who smoke. TUD tobacco use disorder.

Factors associated with frequency of prescribing NRT and varenicline

There were no significant differences in attitudes, confidence, and concerns about NRT between prescribers who reported high versus low frequency of prescribing NRT. In general, prescribers had a low frequency of endorsing concerns about prescribing NRT (≤25%, see Table 2). However, specific concerns about varenicline were significantly associated with frequency of prescribing. Specifically, almost half (47.6%) of the providers who reported low frequency of prescribing varenicline endorsed that they did not know enough about best practices for varenicline prescribing, compared with 12.5% among those who reported high frequency of prescribing varenicline (p = .014). Additionally, those with low frequency of prescribing varenicline rated stronger agreement endorsing specific concerns about cardiovascular disease effects, psychiatric effects, interactions with antiretroviral therapy, and concerns that the benefit of varenicline does not outweigh the risks of adding additional pills to the patient’s regimen (all p values ≤.05).

Table 2.

Attitudes, confidence, and concerns prescribing NRT and varenicline and the association with frequency of prescribing these medications

	High-frequencya prescribing	Low-frequencyb prescribing	p
Nicotine replacement therapy (NRT)
NRT is recommended by clinical practice guidelines to help patients with HIV stop smoking	80.8%	61.3%	.11
I do not know enough about current best practices for NRT prescribing	34.6%	34.4%	.98
I am concerned about side effects of NRT in my patients	11.5%	25.0%	.31
I am concerned about interactions of NRT and antiretroviral therapy in patients	4.0%	18.8%	.12
I am concerned about prescribing NRT to patients who are still smoking	7.8%	15.6%	.44
Varenicline (Chantix)
Varenicline is recommended by clinical practice guidelines to help patients with HIV stop smoking	75.0%	61.9%	.35
I do not know enough about current best practices for varenicline prescribing	12.5%	47.6%	.014
I am concerned about cardiovascular disease effects of varenicline in my patients	0.0%	26.2%	.025
I am concerned about psychiatric effects of varenicline in my patients	18.8%	54.8%	.014
I am concerned that the benefits of varenicline do not outweigh risks of adding pills to patient’s regimen	0.0%	23.8%	.04
I am concerned about interactions of varenicline and antiretroviral therapy in patients	0.0%	21.4%	.05
I am concerned about prescribing varenicline when patients are still smoking	0.0%	14.3%	.17

Values indicate rated agreement with statements and concerns about NRT and varenicline between high- and low-frequency prescribers. p values indicate statistical significance for chi-square tests or Fisher’s exact tests evaluating differences in the proportions endorsing specific attitudes, confidence, and concerns prescribing medications by frequency of prescribing.

Bolded values indicate p < .05.

High frequency: defined as reported prescribing always or most of the time (N = 26 NRT, N = 16 varenicline).

Low frequency: defined as reported prescribing sometimes, rarely, or never (N = 33 NRT, N = 43 varenicline).

Discussion

The current study provides survey data from clinicians and staff treating patients in three HIV health systems in the USA to examine practices, attitudes, and confidence related to delivering care for tobacco treatment for PWH in order to inform ways to enhance the provision of care for TUD in HIV treatment settings. Our study reveals several important findings. Despite clinical guidance [44] and coordinated efforts from large national and international organizations emphasizing the importance of addressing TUD among PWH [16, 45], many clinicians and staff treating PWH reported never receiving training to provide treatment for TUD, first-line medications for TUD were infrequently prescribed, and specific concerns related to medications for TUD were associated with low frequency of prescribing. Thus, opportunities exist for increasing knowledge and providing training to HIV clinicians and staff to optimize the delivery of TUD treatment.

Our findings build on prior work emphasizing the importance of multidisciplinary HIV treatment teams that allow for multiple opportunities for substance use assessment and intervention [25–32], by providing new information about the experience and practices treating TUD among clinicians and staff who are prescribers and nonprescribers. Overall, while both prescribers and nonprescribers reported treating patients in the HIV clinics with TUD, prescribers (compared with nonprescribers) more frequently reported asking patients about their smoking, advising them to stop or cut down, and counseling patients on the impact of smoking on their health (including lung disease, cardiovascular disease, and cancer risk). By comparison, only around two-thirds of nonprescribers reported asking patients about smoking and advising them to stop or cut down, indicating this is an important area for increasing discussion and engagement to treat TUD among PWH. Providing training in brief TUD interventions in accordance with clinical practice guidelines [44] for all providers who interact with patients may be one way to increase the delivery of TUD care for PWH, since only 50% of ­clinicians and staff overall reported receiving training on providing brief TUD counseling and less than 10% received training on 5As/5Rs. Furthermore, only about half of the clinicians and staff, regardless of prescriber versus nonprescriber status, reported frequently counseling patients about the impact of their smoking on their HIV status. While there are limits on time during clinical visits and multiple competing priorities when treating PWH [35], given that tobacco use is also associated with worse HIV control [2] and is responsible for more lost life years than treated HIV [6], counseling PWH on the impact of smoking on HIV may be an important way to increase motivation and engagement with stopping smoking to reduce morbidity and mortality. Such efforts should be coupled with clinician and staff education regarding factors that drive tobacco use among PWH, such as HIV-related discrimination [46], to help inform patient-centered care.

Additionally, our findings build on earlier research that indicates readiness to provide tobacco treatment medications in HIV clinics [35] by investigating knowledge, attitudes, and practices prescribing TUD medications. Overall, prescribers were significantly more likely than nonprescribers to correctly endorse that NRT and varenicline are recommended by clinical practice guidelines to help patients with HIV stop smoking. However, over 1/3 of prescribers and almost half of nonprescribers reported they did not know enough about clinical practice guidelines for these medications. Among prescribers, single NRT was the most commonly prescribed medication, and most prescribers reported infrequently prescribing the most effective first-line treatment options of dual NRT (i.e., long-acting nicotine patch plus short-acting ad libitum NRT such as gum or lozenge) or varenicline [44]. The low frequency of routinely prescribing medication for TUD has been reported elsewhere [24, 47], and the current study findings provide additional information about concerns that are associated with low frequency of prescribing varenicline, such as concerns about psychiatric side effects or concerns about the interaction with antiretroviral therapy, despite reassuring data about the safety and tolerability of varenicline [11, 12]. These findings identify addressable gaps such as providing updated training and information on the benefits of medication use and safety of TUD medications alongside antiretroviral therapy to clinicians and staff that may be important to ensure that patients are receiving care in line with current best practices.

Limitations

Study results should be considered in the context of specific limitations. Data were collected from three sites in the ­northeastern part of the USA, so additional research is needed to confirm whether similar patterns are present in other geographic areas. Additionally, data on reported practices providing TUD treatment and prescribing medication were based on self-report and not cross-validated based on medical record data. While we assessed self-reported attitudes, knowledge, and practices regarding TUD treatment, we did not explicitly assess clinician and staff sense of preparedness to administer TUD treatment.

Conclusions

Addressing TUD among PWH is an important public health priority. HIV clinicians and staff in the present study reported addressable gaps in their knowledge, understanding, and practices related to tobacco treatment. Additional work is needed to identify ways to ensure adequate training for providers and models of care to enhance the delivery of TUD treatment in HIV clinic settings.

Funding

This work is funded by the National Cancer Institute (R01CA243910). K.W.B. received funding from the National Institute of Drug Abuse (K12DA000167) during the conduct of this work. The funder had no role in the design, conduct, analysis, or reporting of the data. This content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH or NCI.

Compliance with Ethical Standards

Conflict of Interest: The authors have no conflicts of interest or disclosures to report.

Human Rights: All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. The study was approved by the Institutional Review Board (IRB) at Icahn School of Medicine at Mount Sinai, which serves as the single IRB for this protocol and approved this study for all sites and per reliance agreements with Yale School of Medicine and SUNY Downstate.

Informed Consent: Informed consent was obtained from all individual participants included in the study.

Welfare of Animals: This article does not contain any studies with animals performed by any of the authors.

Transparency Statements

Study Registration: The study was preregistered in clinicaltrials.gov NCT04490057 and we have published details of the overall protocol in Contemporary Clinical Trials (PMID: 33794354).

Analytic Plan Preregistration: The analysis plan was not formally preregistered.

Analytic Code Availability: Analytic code used to conduct the analyses presented in this study are not available in a public archive. They may be available by emailing the corresponding author.

Materials Availability: Materials used to conduct the study are published as an appendix in the protocol paper in Contemporary Clinical Trials (PMID: 33794354).

Data Availability: Deidentified data from this study are not available in a public archive. Deidentified data from this study will be made available (as allowable according to institutional IRB standards) by emailing the corresponding author and upon approval by the investigative team.