HIV care provider perceptions and approaches to managing unhealthy alcohol use in primary HIV care settings: a qualitative study

Background HIV care providers often serve as the specialist and the primary care point-of-contact for persons living with HIV (PLWH) and unhealthy alcohol use. The purpose of the present qualitative study was to understand HIV care provider perceptions and approaches to managing unhealthy alcohol use in HIV primary care settings. Methods Using a semi-structured interview guide, in-depth interviews were conducted among 14 HIV care providers (5 medical doctors, 5 nurse practitioners/physician assistants, 2 medical assistants, 2 clinical administrative staff) in private and public HIV clinics, across urban and rural areas of Florida. Interviews were coded using a grounded theory approach with inter-rater consensus. Results Six themes were identified. In summary, providers reported (1) inconsistent assessment of alcohol consumption, as well as (2) varying levels of confidence in self-report of alcohol use which may be affected by patient provider rapport and trust. While providers (3) acknowledge potential negative impacts of alcohol use on health outcomes and HIV treatment, providers reported (4) inconsistent recommendations regarding alcohol use among their patients. Lastly, providers reported (5) limited resources for patients with unhealthy alcohol use and (6) low confidence in their ability to help patients reduce use. Conclusions Results from our study suggest salient differences in provider approaches to the assessment and management of unhealthy alcohol use in HIV primary care settings. Implementation of care for unhealthy alcohol use in these settings may be facilitated through use of clinically useful, validated alcohol use assessments and use of evidence-based recommendations of alcohol use/non-use among PLWH. Training in brief intervention techniques for alcohol reduction may increase provider confidence and support in the management of unhealthy alcohol use among PLWH.

Despite accumulating evidence suggesting that unhealthy alcohol use has detrimental medical and psychosocial effects among PLWH, the management of this behavior remains a significant challenge. HIV care practitioners often serve as both the HIV specialist and the primary care provider for PLWH living with comorbid substance use, mental health, and chronic illness [28]. In fact, up to 84% of PLWH report that they prefer to receive such integrated care from their HIV care provider, as opposed to having a separate primary care doctor [28]. However, knowledge regarding effective assessment of and intervention for unhealthy alcohol use among HIV-care providers remains a concern. Although HIV care providers report an understanding of the links between alcohol use and HIV transmission (93%), as well as the ability to assess unhealthy alcohol use (85%), providers are unlikely to have received any formal training on how to manage unhealthy alcohol use (41%) and only 54% report an understanding of current recommendations regarding unhealthy alcohol use [29].
While there is great potential to address unhealthy alcohol use in HIV primary care settings, providers often miss unhealthy alcohol use among PLWH, especially in patients whose HIV is well-managed [30]. Only 10% of providers report using a formal screening tool to assess alcohol use [31]. By extension, AUDs are rarely treated in HIV primary care settings, with over two-thirds of providers never/rarely treating AUD and 70% referring patients to specialized treatment [31]. Further, a recent study of over 800,000 patients found that among those with unhealthy alcohol use, PLWH were 15% (RR 0.85, p < .001) and 37% (RR 0.63, p < .001) less likely to receive any evidence-based brief intervention or supportive medicine, respectively, compared to uninfected patients, controlling for confounding factors such as care utilization [32].
General primary care stakeholders, including patients and providers alike, agree that universal screening of substance use, including alcohol, is important to health care management [33]. However, existing qualitative research reveals significant barriers to screening and treatment of unhealthy alcohol use in primary care, including limited resources (access to treatment centers or specialty care) [33][34][35][36], lack of knowledge/experience or training [31, 33-35, 37, 38], attitudes regarding the role of primary care versus specialty care [34,38], and alcohol use stigma felt by patients [33,37,38]. In contrast, facilitators for screening and treatment of unhealthy alcohol use include having tools and support for providers (i.e., behavioral staff, tools to enhance patient motivation) [38,39], screening/treatment goals that are consistent with organization values and/or existing practices [36,39], increased provider self-efficacy [38,39], and specialty care accessibility [36].
Given the general lack of literature regarding treatment for unhealthy alcohol use in HIV care settings, and the high prevalence of unhealthy alcohol use and disparate treatment by HIV status, we sought to gain information on HIV care provider perceptions and approaches to managing unhealthy alcohol use in their HIV primary care clinic.

Participants
Participants were HIV care providers (those providing direct service or facilitation of HIV care services) recruited from private and public HIV primary care settings, in urban and rural areas across the state of Florida between January-May 2016, namely the Florida Departments of Health in Alachua (Gainesville) and Hillsborough (Tampa) Counties, as well as from Jackson Memorial Hospital (Miami) and University of Florida Health Shands Hospital (Gainesville). In order to gain an understanding of the collective perspective on care for unhealthy alcohol use in HIV primary care settings, we sought a diverse group of HIV care providers within all levels of the HIV care cascade, including medical doctors, registered nurse practitioners, physician assistants, medical assistants and clinic administrative staff who have worked with PLWH for at least 1 year. Further, the diversity of care providers serves as a data triangulation method, for which to assess validation of data. A combination of convenience and snowball sampling was used to recruit providers-providers affiliated with the Southern HIV and Alcohol Research Consortium (SHARC), a center located within the University of Florida Department of Epidemiology examining determinants of health outcomes of PLWH within the state for Florida, were contacted with information about the study and asked to suggest alternative providers who may be interested in participating. Initially, the healthcare providers were contacted and given information about the current project, including a Letter to Providers and additional flyers to provide to HIV care staff. If a provider was interested in participating, an interview was scheduled. Upon providing written informed consent on the day of the interview, participants were asked to complete a short demographic questionnaire and the interview started thereafter. While a target sample size of 12-16 was initially proposed, recruitment continued until saturation of themes [40] was reached among the providers in aggregate-at 14 participants.

Procedure
Face-to-face, semi-structured interviews with HIV care providers were conducted for approximately 30-45 min. Participants were provided a lunch gift card (value $15) for their time. An interview guide was used to standardize all interviews, utilizing open-ended questions with prompts in order to elicit a discussion around provider perceptions and care practices related to alcohol consumption (Table 1). While a guide was used, the interviews were semi-structured, allowing for further elicitation of emerging themes and/or clarification of existing themes. Interviews were conducted by the lead and senior authors and were digitally recorded and professionally transcribed by an external transcription service. Transcriptions were spot checked for accuracy; an error rate of greater than 5% prompted re-transcription of interviews. No transcription exceeded this error rate threshold. All audio recordings were stored on a password-protected server and deleted after transcription.
All information was de-identified, with a participant ID replacing the participant's name on all study documentation. The study was approved by the University of Florida Institutional Review Board. Participants provided informed consent prior to participation.

Data analysis
A qualitative data management and analysis program, NVivo 10 ™ (QSR International Pty. Ltd., 2012) was used for all management and analysis of qualitative data. While there is a vast qualitative literature regarding alcohol-related care in general primary care [33][34][35][36][37][38], there is little literature regarding alcohol-related care in HIV care settings [31,39]. Because of this, a Grounded Theory approach [40,41] was utilized-as previous findings in the general primary care literature may not be generalizable to specialty HIV care settings. Initially, one investigator conducted line-by-line analysis to identify initial codes. After initial coding, two additional multidisciplinary (i.e., medicine and public health, epidemiology, and clinical and health psychology) investigators not present at the interviews or previously involved in the project coded and identified themes; all members subsequently discussed the codes and proposed themes, as well as resolved any differences in coding. Codes were then collapsed into final themes presented in this article.

Participant demographics
A total of 14 providers were recruited from private and public primary HIV care clinics, in urban and rural settings across Florida. The sample was 36% (n = 5) non-Hispanic white, 43% (n = 6) Hispanic, and 21% (n = 3) Other race/ethnicity; 28% (n = 4) were aged 25-34 years, 43% (n = 6) 35-44 years, and 28% (n = 4) 45 + years. Providers included medical doctors (MD, n = 5), nurse practitioners (NP)/physician assistants (PA, n = 5), medical assistants (MA, n = 2), and clinical administrative staff (CA, n = 2). Of the sample, an even 50% were recruited from private versus public clinic settings, as well as in rural versus urban areas. Providers had worked in HIV primary care for a mean of 12.2 years (Table 2, median 14, range 1-28 years). On average, providers reported that 51.3% and 42.8% of their patient population experienced issues with mental health and substance use, respectively. On a scale of 0-100% confidence, the providers reported an average of feeling 60% confident that they could help their patients with unhealthy alcohol use.

Emergent themes
(1) Providers reported inconsistent assessment of alcohol consumption Participants were asked how and when alcohol use was assessed in routine care. In terms of frequency of assessments, responses ranged from no assessment to assessment at every visit. Many providers (N = 6; NPs and MDs) reported assessing alcohol consumption periodically, but not at every visit. Two of these providers reported asking patients about alcohol consumption at their first appointment: One provider reported an existing protocol that wasn't always followed: We're supposed to be doing that every year. I don't think we've actually done it every year for the alcohol. " Other providers reported that limited time prevented them from conducting an alcohol use assessment: Of the providers that reported assessing alcohol use (n = 9), most reported asking questions to ascertain the frequency of alcohol use and the quantity (n = 6): The other providers reported asking similar questions, in addition to assessing the type of alcohol that is consumed. (2) There were inconsistent perceptions of self-report accuracy which may be affected by patient-provider rapport and trust When asked about alcohol assessment, several providers (n = 7; PA, NP, MD) indicated that an important barrier to assessment and use of information gained was accuracy of self-report. One provider expressed concerns of under-reporting:  Some providers (n = 3; NP, MD) reported concerns discontinuing ART altogether because of the contraindication with alcohol: Some providers (n = 3, CA, NP, MD) also felt that, aside from HIV-related outcomes, alcohol consumption had a detrimental effect on behavioral/mental health that often affected the level of care that could be provided in the clinic setting: Related, a clinical administrator expressed concern that the primary provider does not engage patients well enough to address the underlying issue. This provider also reported perceptions of stigmatization of patients with unhealthy alcohol use. As mentioned above, most providers did perceive a link between alcohol consumption and poor health outcomes, related to medical non-adherence, viral load, and behavioral issues. However, there was a subset of providers (n = 4, PA, MD) that reported no such link among their patients who engage in light to moderate alcohol use: A couple of providers (n = 2; NP, MD) focus on referrals for those patients that engage in unhealthy alcohol use, as opposed to making quit or reduction plans: Also: Provider 13: "We have psychologists that come here in the afternoon. We have social workers that work with the patient. The patient can be seen the same day, or sometimes they just come and even when they don't have an appointment -we try to get an appointment either with the social worker or the psychologist. " Several providers (n = 6; PA, MD) reported significant barriers to helping patients with unhealthy alcohol use, ranging from lack of resources to lack of patient motivation:

Discussion
Results from our study suggest salient differences in provider perceptions and systemic approaches to managing unhealthy alcohol use among PLWH. This included a lack of uniformity in policies and procedures among HIV primary care providers for assessing unhealthy alcohol use. For example, some providers reported assessing alcohol use at every visit, at the first appointment, or periodically. A similar theme has been demonstrated in general primary care-in which alcohol assessment was not standardized and/or not consistent in assessment period across providers [33,34,37]. Our findings are also consistent with results from a survey of HIV care providers, finding that one-third reported usually asking about alcohol use frequency and quantity [31]. Related, many providers expressed concerns regarding the accuracy of self-report, which may be influenced by patient-provider rapport. Consistent with this finding, previous literature has found the patient-provider relationship as an important determinant of patients' comfort in disclosing behaviors, such as unhealthy alcohol use, in primary care settings [33,37]. While most providers reported that alcohol use deleteriously affects medication adherence, in line with studies showing a significant relationship between alcohol consumption and suboptimal ART adherence [2], providers reported conflicting alcohol use recommendations for patients. Nearly one-third endorsed an abstinence-only approach. In contrast, other providers were approving of light to moderate alcohol consumption, contingent on optimal ART adherence and no signs of adverse health issues, such as liver disease. Further, some providers expressed concern providing particular ART regimens or providing medication at all in the presence of unhealthy alcohol use and potential liver failure. Currently, there are inconsistent guidelines regarding safe levels of alcohol use [41]-which may be reflected in the inconsistent recommendations for alcohol use/non-use reported in the current study.
Many providers in our study also expressed doubt in their ability to effectively treat and refer those with unhealthy alcohol use to appropriate programs. This theme was further exemplified as a lack of formal education/training and resources necessary to intervene. This is consistent with qualitative research in general primary care [33-35, 37, 38] and quantitative research in HIV care [31] settings indicating that providers' perception of experience and training in alcohol use treatment is a major barrier to brief intervention and treatment referrals. This may be compounded by a general lack of knowledge of medications/treatments for unhealthy alcohol use that are consistently effective with minimal side effects or contraindications [42], which was reported as a barrier in providing pharmacological treatment for unhealthy alcohol use in HIV care settings [31].
Our findings are consistent with previous work by Strauss et al. [43,44], describing the state of alcohol reduction support offered by HIV care providers in hospital-based HIV/AIDS centers utilizing Screening and Brief Intervention (SBI) procedures. Strauss et al. [44] found that barriers to implementation of SBI components included inaccurate patient self-report of alcohol use, inconsistent provider perspective on alcohol use, and provider specialization that discourages treatment of comorbid unhealthy alcohol use. Additionally, five providers in our study reported low confidence in their ability to effectively manage patient alcohol abuse. Along these lines, previous studies have shown that providers with limited confidence in their ability to provide assistance related to unhealthy alcohol use among PLWH were less likely to have a high level of role legitimacy (i.e., extent to which healthcare providers believe that treating substance abuse issues is their responsibility [45]). Ultimately, patients with unhealthy alcohol use are less likely to be satisfied with their care and overall patient-provider communication regarding their unhealthy alcohol use [46,47].
Limited resources for unhealthy alcohol use interventions and conflicting provider reports about the effect of alcohol use on HIV-related health outcomes complicate the recommendations and treatment of this behavior [48]. Given the negative implications of alcohol use on ART adherence, in conjunction with self-doubt expressed by providers to effectively treat unhealthy alcohol use, these findings support calls for integrated substance use and mental health treatment providers in infectious disease and primary care clinics frequented by PLWH. Specifically, identification of unhealthy alcohol use and providing appropriate referrals for intervention may improve HIV-related outcomes, especially given findings from previous studies showing decreased alcohol consumption and improved ART adherence with components of brief-intervention [49,50]. Additionally, HIVcare providers may benefit from training in motivational interviewing, a goal-oriented therapy approach for eliciting behavior change, as this style of point-of-care therapy has been associated with reduction in alcohol use quantity and frequency among PLWH [50,51].
There are limitations to this study that readers should consider, as well as notable strengths. First, while we aimed to recruit until saturation of themes was achieved, we recognize that the current sample is small and further qualitative research should be conducted with larger samples of diverse providers. Second, self-selection bias among our provider participants is possible. The providers that agreed to participate may represent a subgroup of providers that are highly motivated to address alcohol use in their clinics. We made great effort to engage providers at multiple levels of care for depth and breadth of experience and opinion. Further, we utilized flyers, provider letters, and word of mouth to ensure that all qualifying providers would have an opportunity to participate. Third, patient perspectives on alcohol related issues and harm reduction services within HIV primary care were not examined. Therefore, further qualitative investigation is needed in order to understand the patient perspectives on alcohol use and preferred treatment strategies to reduce unhealthy alcohol use. Despite these limitations, notable and consistent themes emerged from a heterogeneous group of HIV-care providers across the state of Florida.

Conclusions
With 43% of PLWH reporting unmet healthcare needs related to drug or alcohol use [52], support for addressing alcohol use within a primary care setting is of great importance and studies have demonstrated the usefulness of brief intervention to reduce unhealthy alcohol use [53]. As HIV-primary care providers continue to serve as both the specialist and the primary care pointof-contact for PLWH, it is recommended that clinicians assess all patients aged 18 years and older for unhealthy alcohol use, and to provide support to reduce unhealthy alcohol consumption [54]. Use of standard definitions and diagnostic measures will provide uniformity in the assessment of alcohol use [55], and will help to inform specific substance abuse treatments tailored to PLWH. Future practice steps should include educating patients on the adverse effects of consuming varying quantities of alcohol in relation to ART use, HIV disease progression, and the development of comorbid illnesses. Additionally, implementation of more readily accessible and effective alcohol treatment resources and intervention programs may strengthen provider self-efficacy in providing appropriate treatment and referrals for unhealthy alcohol use among PLWH.