|Home | About | Journals | Submit | Contact Us | Français|
Screening and brief intervention (SBI) for alcohol reduction is an important health promoting strategy for patients with HIV, and HIV care providers are optimally situated to support their patients' reduction efforts. We report results from analyses that use data collected from providers (n=115) in 7 hospital-based HIV care centers in the New York City metropolitan area in 2007 concerning their routine use of 11 alcohol SBI components with their patients. Providers routinely implemented 5 or more of these alcohol SBI components if they (1) had a specific caseload (and were therefore responsible for a smaller number of patients), (2) had greater exposure to information about alcohol's effect on HIV, (3) had been in their present positions for at least 1 year, and (4) had greater self efficacy to support patients' alcohol reduction efforts. Findings suggest the importance of educating all HIV care providers about both the negative impact of excessive alcohol use on patients with HIV and the importance and value of alcohol SBIs. Findings also suggest the value of promoting increased self efficacy for at least some providers in implementing alcohol SBI components, especially through targeted alcohol SBI training.
Although both the morbidity and mortality of patients with HIV have decreased considerably in the past decade with the introduction of antiretroviral (ARV) therapy, many individuals with HIV continue to seriously jeopardize their health by using and abusing alcohol.1 These individuals experience more rapid disease progression and HIV-related complications as a result of alcohol use.2–4 HIV-infected people who use alcohol excessively may also engage in risky sexual behaviors while under its influence, exposing both themselves and their partners to sexually transmitted infections.5–7 In addition, studies have shown that the use of various substances can have an effect on HIV medication adherence.8–10 While not all studies examining the relationship between alcohol use and medication adherence have found a negative association between the two,11 some studies have determined that people consuming alcohol while undergoing ARV treatment prescribed are considerably less likely to be adherent to the medication.12–14 Of especially great concern is the high prevalence of coinfection of HIV and hepatitis C virus (HCV), with the consequence that end-stage liver disease, accelerated as a result of alcohol use among those coinfected, has become a leading cause of illness and death among these individuals.15,16 Furthermore, HCV treatment is less effective in people with HIV/HCV coinfection, and its effectiveness is limited even more by ongoing alcohol use.17
Importantly, HIV care providers have great potential to serve as patients' advocates and counselors regarding alcohol reduction, both to prevent the development of serious health consequences, and to limit transmission of the virus to others.18,19 In fact, some providers view alcohol reduction counseling as consistent with patients' health promotion.20 However, other providers and their patients have come to expect that the provider will facilitate access to services and respond to the patient's expressed needs, rather than focus on the patient's alcohol use practices.21 Some providers may therefore resist implementing alcohol reduction counseling. This may especially be the case if they feel that it will shift priorities away from providing more traditional care and support, and if it threatens to harm their good relationships with patients.20,22 Even if they are comfortable in their roles as alcohol reduction supporters, experienced HIV care providers may still have difficulty assisting in their patients' alcohol reduction efforts. They may have (1) limited exposure to comprehensive and current information regarding the identification of at-risk drinkers and the negative impact of alcohol use on HIV patients, (2) limited time to conduct patient risk reduction counseling (often due to their responsibility for a large numbers of patients), (3) limited self efficacy to implement alcohol reduction support, and (4) limited training to facilitate addressing patients' alcohol abuse issues.2,19,22–24 Less experienced providers (who constitute a substantial proportion of the HIV care workforce as a result of high provider turnover rates25,26) may find it especially challenging to incorporate addiction counseling into their interactions with patients if they lack the skills, confidence, experience, and organizational support to do so.21,27,28 Regardless of their experience with HIV patients, providers' personal alcohol use may be an additional unspoken barrier. The importance of alcohol reduction among HIV patients, however, argues for the need for its support by HIV care providers.
To support alcohol reduction among a variety of populations, alcohol screening and brief intervention (SBI) has been shown to be an effective approach.29,30 Alcohol screening involves assessing patients for alcohol use, frequency, abuse, and dependence, and for problems caused by this use.31,32 The brief intervention involves the implementation of a variety of components (e.g., assessing patients about their readiness to cut down their use, providing patients with suggestions about alcohol reduction, creating actual plans with patients about reducing their drinking). It typically incorporates patient-centered, motivational, and interactive counseling techniques that increase patients' readiness to change harmful behaviors.33 Regretfully, although they are effective, SBIs are frequently underutilized and often not incorporated by health providers as a standard component of regular counseling activities and strategies.34,35 Given the serious consequences of alcohol use for patients with HIV, it would be especially unfortunate if this were the case among HIV care providers.
At present, little is known about the extent to which HIV care providers, including physicians, physicians' assistants, nurses, nurse practitioners, case managers, social workers, and others, routinely offer alcohol reduction support to their patients in the form of SBI components. This paper therefore describes the alcohol reduction support offered by HIV care providers (n=115) in hospital-based HIV/AIDS centers in the New York City metropolitan area. In addition, it presents the results of a logistic regression analysis that examines the salient correlates that differentiate providers who routinely offer more than the median number of alcohol SBI components in this HIV care provider sample and those who do not.
Study participants included HIV care providers who delivered direct patient care in 7 designated AIDS centers (DACs) in the New York City metropolitan area in 2007. DACs are comprehensive, hospital-based, state-licensed HIV treatment centers providing both inpatient and outpatient care. They utilize interdisciplinary teams and provide case management services, emphasizing quality improvement in order to provide a high level of clinical and support services. Reflecting differences in the sizes of their patient populations and in the corresponding sizes of their direct care staff, the number of participating staff in each DAC varied from 6 to 21, totaling 115 participants. An average of 7.3 staff and a median of 8 staff participated in each DAC, with only 1 DAC accounting for fewer than 12 staff participants.
The data used in the current research were collected from these 115 HIV care providers as part of a larger study funded by the National Institute on Alcohol Abuse and Alcoholism (NIAAA). This larger study is intended to evaluate a state-of-the-art training on supporting alcohol reduction in HIV patients. The 3-hour training is an adaptation for HIV care providers of NIAAA's alcohol screening and brief intervention protocol (as described in NIAAA's Clinician's Guide).36 Analyses in the current research use data collected from providers at each of the DACs before the training took place.
After receiving assurances regarding the voluntary nature of the research and the confidentiality of responses, providers who consented to participate in each of the 7 DACs completed a 20-minute self-administered survey. The survey items had been pretested with a group of 20 community-based HIV care providers. Five of these community-based providers participated in cognitive interviews to ensure that the wording of the items was clear, and that the items measured what they were intended to measure. Following the pretesting phase, project investigators and consultants modified some items and eliminated those that were redundant.
Eleven of the survey items assessed the extent to which components of alcohol SBIs were routinely practiced by providers in the past month. The items were created using information gathered in informal discussions with community-based HIV care providers and DAC administrators, and using past literature regarding the implementation of SBI components.37–42 The 11 items contained within the survey reflect many of the alcohol SBI components described in NIAAA's Clinician's Guide.36 We first asked respondents to indicate whether they asked none, a few, some, most, or all of their HIV patients about their alcohol use. We then asked respondents whether they implemented 10 specific components of brief alcohol interventions with none, a few, some, most, or all of their HIV patients who drank alcohol. For the purposes of this research, responses to each of the 11 items were dichotomized. They differentiate “routine” implementation of the alcohol screening or alcohol reduction component (i.e., implementation in the past month with most or all patients), from “nonroutine” implementation (i.e., implementation in the past month with none, a few, or some of these patients). For each participating provider, the number of components routinely provided was then tallied. The median of these tallied components among the 115 HIV care providers was computed. The dependent variable for the research was then created by dichotomizing the group of providers into those who routinely implemented no more than the median number of these components in the past month and those who implemented more than the median number.
In addition to their demographic characteristics (e.g., gender, race, ethnicity), providers indicated whether or not they were medically credentialed (i.e., were physicians, physicians' assistants, nurses, or nurse practitioners). We asked about providers' experience with patients with HIV, including the length of time they had worked with these patients, how long they were employed in their current positions, and whether they worked full-time or part-time. As those who did not have specific caseloads worked with an especially large number of patients, we asked whether providers had specific caseloads. To understand providers' past exposure to relevant training and information regarding alcohol reduction support, respondents also indicated whether they had participated in workshops on motivational interviewing and/or brief interventions within the past year. They were also asked whether or not they had ever participated in a workshop that specifically covered brief interventions for alcohol reduction. In addition, they indicated the way(s) in which they had learned about alcohol's impact on HIV (i.e., workshops/trainings; books/pamphlets; videos; internet; medical people; another way). The number of different ways was then tallied for each provider as an indicator of the extent of their exposure to the relationship between alcohol and HIV infection.
Respondents also completed an eight item Brief Intervention Knowledge Assessment and an 8-item Alcohol Reduction Support Self-Efficacy Scale. In consultation with experts in the field, items were created for the Knowledge Assessment and Self-Efficacy Scale that adapted those that were in the literature.38,43–47 We obtained a total score for each respondent on the Brief Intervention Knowledge Assessment by determining the number of items that were correctly endorsed. In the case of each of the items in the Alcohol Reduction Support Self-Efficacy Scale, respondents gave numerical ratings (from 0 to 10) regarding the degree to which they felt confident in their ability to provide the specific type of support assessed. A total score on the Alcohol Reduction Support Self-Efficacy Scale was obtained for each respondent by summing the scores on each of the individual items.
Using SPSS 15.0 (SPSS Inc., Chicago, IL), we first examined the range of responses on the Brief Intervention Knowledge Assessment and the reliability, factor structure, range, and variability of the Alcohol Reduction Support Self-Efficacy Scale. We then performed a series of bivariate logistic regression analyses. These analyses determined which of the Assessment and Scale scores and which of the other independent variables should be considered for inclusion in a multiple logistic regression model. This model was intended to explain variation in a provider's routinely high level of implementation of the alcohol SBI components (i.e., more than the median number of components). Variables that were significant at the 0.20 level or less in the bivariate analyses were included in the initial multiple logistic regression model.
To determine the final, most parsimonious multiple logistic regression model, variables were eliminated from the initial model, one at a time. A variable was dropped from the model if it did not significantly contribute to explaining variation in the dependent variable, if the coefficients of the remaining variables changed only minimally, and if the difference in the log-likelihoods of the models with and without the variable was not significant. After the elimination of a variable, the model was reexamined to determine if other variables should be eliminated.
Because participants were drawn from 7 DACs, we also investigated the possibility of site differences. To do so, we used STATA to fit a generalized estimating equation (GEE) model with logit link and binomial distribution using the same model structure as our final multiple logistic regression model, and with the addition of a working correlation matrix to capture within program correlations.
The majority (80%) of the participating providers were female. Approximately half (56.7%) were white, approximately one third (32.0%) were black, and the remainder were of other and mixed races. One in five of the participants (20.9%) were Hispanic/Latino. Close to half (42.6%) reported that they had medical degrees (including physicians, physicians' assistants, nurses, and nurse practitioners). The vast majority (91.1%) worked full time at their DACs, with most (86.6%) in their current positions for 1 year or more. A great majority (83.8%) had at least 3 years experience working with patients with HIV. Approximately three quarters (73.0%) of the providers had a specific caseload.
Exposure to information and training regarding alcohol reduction support varied among the participating providers. Two in five (40.0%) had attended a workshop in the past year on brief interventions, and the same proportion had attended a training on motivational interviewing during this time period. Approximately one third (37.2%) indicated that they had attended a workshop specifically on brief interventions on alcohol reduction some time in the past. On average, providers indicated that they had learned about the effects of alcohol use on patients with HIV in 2.6 of the following different ways: workshops/trainings; books/pamphlets; videos; Internet; medical people; another way.
For each of the eight items on the Brief Intervention Knowledge Assessment (Table 1), respondents indicated if the item was true or false or if they did not know. A total score on the Assessment was obtained for each respondent by determining the number of items that was correctly endorsed. Each individual's score could therefore range between 0 and 8. Participants scored 5.7 of 8, on average.
The vast majority of providers knew that screening for current alcohol use and its consequences is an essential component of a brief alcohol intervention, and that brief interventions generally need to be specifically tailored to patients' readiness to change harmful behaviors. More than half of the respondents, however, incorrectly believed that brief interventions rarely promote significant, lasting reductions in drinking levels in at-risk drinkers.
Each of the eight items on the Alcohol Reduction Support Self-Efficacy Scale (Table 2) asked respondents to rate their confidence with regard to supporting their patients with HIV in reducing their alcohol use. The 11-point rating ranged from 0=not confident to 10=very confident. Item scores were summed to obtain a total score with a possible range from 0 to 80, with higher scores exhibiting a greater sense of overall self efficacy. The scale exhibited excellent reliability (α=0.92). A principal components factor analysis identified one factor, with an eigenvalue of 5.3, accounting for 65.8% of the variance. Total scores on the scale ranged from 0 to 80, with a median of 58, and a mean and standard deviation of 56.1 and 15.2, respectively.
Providers' ratings indicated that they were most confident in their ability to bring up the subject of alcohol use with their patients with HIV, to help them understand the health risks related to their drinking, and to screen these patients for alcohol use. They gave their lowest rating to their confidence about knowing the appropriate questions to ask HIV patients when providing alcohol reduction counseling.
As can be seen in Table 3, three quarters (77.4%) of the participating providers routinely asked patients with HIV about alcohol use. Approximately half of the providers also routinely educated patients with HIV who drink regarding the risks of alcohol use, advised most of these patients about sensible drinking, and encouraged them to talk about alcohol reduction. However, only about one third of the providers routinely acknowledged patients' challenges about changing drinking patterns, assessed their readiness to cut down on alcohol use, asked about their alcohol reduction progress in subsequent meetings, provided suggestions to them about reducing their drinking, or encouraged or arranged follow-up alcohol reduction support for them. In addition, only approximately 1 in 10 of the providers routinely created an alcohol reduction plan with their patients with HIV who drink, or provided them with alcohol reduction literature.
Of the 11 components of SBI for alcohol reduction that we asked about, providers indicated that they routinely provided 4.4 of these components, on average, with a median of 4 components. A total of 61 participants indicated that they routinely provided 4 or fewer of these components, and the remaining 54 participants indicated that they provided 5 or more of them on a routine basis.
We conducted a multiple logistic regression analysis to determine the variables that predict having routinely provided more than the median number of alcohol SBI components (n=4) with patients with HIV in the past month. We considered a number of variables including: (1) demographics (gender, race, ethnicity); (2) professional characteristics (medical training, having worked with HIV patients for at least 3 years, having been in their present positions for at least 1 year, full-time work, having a specific caseload); (3) past exposure to brief intervention techniques and/or alcohol's impact on patients with HIV (the number of different ways the provider learned about HIV and alcohol, attendance at a workshop on motivational interviewing in the past year, past year attendance at a workshop on brief interventions, having ever attended a workshop specifically on brief interventions for alcohol reduction); (4) the score on the Brief Intervention Knowledge Assessment; and (5) the score on the Alcohol Reduction Support Self-Efficacy Scale. In bivariate logistic regression analyses, with each variable entered separately in the model, it was determined that providers were significantly (p<0.05) more likely to have routinely provided more than the median number of alcohol SBI components with patients with HIV in the past month if they: (1) had a specific caseload, (2) had a larger number of different ways in which they had learned about HIV and alcohol, (3) had a higher score on the Brief Intervention Knowledge Assessment, and (4) had a higher score on the Alcohol Reduction Support Self-Efficacy Scale. Providers who (1) had worked with HIV patients for at least 3 years, (2) were in their current positions for at least 1 year, and (3) had attended a workshop on motivational interviewing in the past year, tended (p<0.20) to provide this higher level of service provision.
All of the individual variables found to be significant at the 0.20 level or less when considered separately as predictors of having routinely provided 5 or more alcohol SBI components with HIV patients in the past month were entered into a multiple logistic regression model. We eliminated three variables from the final model: worked with HIV patients for at least 3 years, attended a workshop on motivational interviewing in the past year, and score on the Brief Intervention Knowledge Assessment. None contributed significantly in the final model to explaining variation in the dependent variable, the coefficients of the remaining variables in the model changed only minimally, and the change in −2 log-likelihood (5.986 with 3 degrees of freedom) between the initial model and the model with the 3 variables removed was not significant (p>0.1). In addition, because the GEE model (created to examine the possibility of site bias) produced results almost identical to the simple logistic regression model, we present this latter model in Table 4 for simplicity. The final model indicates that providers were significantly more likely to routinely implement more than the median number of alcohol SBI components with HIV patients in the past month if they (1) had a specific caseload, (2) had a larger number of different ways in which they learned about HIV and alcohol, (3) were in their current positions for at least 1 year, and (4) had a higher score on the Alcohol Reduction Support Self-Efficacy Scale.
In view of the importance of supporting alcohol reduction among patients with HIV, it is of considerable concern that the median number of routinely implemented components of alcohol SBIs among DAC providers was 4 of the 11 that we asked about. Certainly, it is gratifying that three quarters of the providers routinely asked their HIV-infected patients about alcohol use, and had a high level of confidence in their ability to bring up the subject of alcohol use and screen patients for this use. Only a minority of providers, however, followed through with their patients who drink in supporting alcohol reduction efforts. In particular, only one third of the providers routinely asked about their patients' alcohol reduction progress or routinely provided them with suggestions about cutting down on their drinking, and only 1 in 10 routinely created an alcohol reduction plan with their patients who drink. Without such alcohol reduction support, many of these patients may continue to endanger their health by persisting in their drinking. Because, as our data indicate, many providers are unaware that brief interventions can promote significant and lasting reductions in drinking levels in at-risk drinkers, providers may not be aware of the value and importance of these alcohol reduction support efforts.
Providers who routinely implemented more than the median number of components of alcohol reduction support were more likely to have had greater exposure to information about HIV and alcohol. Greater exposure to relevant health information has also been shown to influence behavior in other patient care contexts (e.g., more frequent prescribing of naltrexone in drug treatment programs48). This suggests the importance of educating HIV care providers about the negative effects of drinking among patients with HIV, education that may lead to an understanding of the value and importance of alcohol SBIs.
We also found that providers implemented a greater number of alcohol reduction components if they had been in their current positions for at least 1 year. These providers likely felt more integrated into their organizations than those who had only been there for a very short time, perhaps generating more comfort in their roles as patients' alcohol reduction supporters. Providers also implemented more alcohol reduction components if they had more self efficacy in their ability to support patients' alcohol reduction efforts. Consistent with the current findings, self efficacy has been shown to be an important predictor of performance in a great variety of occupational fields,49 including HIV/AIDS.50 Fortunately, training in SBI can increase some providers' confidence in performing alcohol screening procedures, and in conducting brief alcohol reduction interventions.37
Our analyses also indicate that those providers without specific caseloads routinely implemented fewer alcohol reduction components. Providers with no specific caseloads (including many of the participating DAC primary care physicians and nurses) generally serve very large numbers of patients. Especially for these providers, patient volume limits the amount of time that can be spent with each patient. This, together with the complexity of many patients' issues and the attention that these issues require, present substantial barriers to finding the time to implement alcohol reduction support during patients' visits.51–53 Thus, from a practical standpoint, these providers may need to use a “specialist” approach,19 referring patients to other professionals in their programs who have more time and expertise to effectively counsel patients regarding their alcohol reduction needs.54
Several limitations of the research need to be acknowledged. First, as is true with all self report questionnaires, social desirability may have biased some of the responses to our questions. This is likely to be mitigated, at least in part, by the fact that respondents did not identify themselves by name. In addition, although the data were gathered in 7 DACs in the New York City metropolitan area, it is unclear to what extent these results are generalizable to staff in other HIV programs in New York City and elsewhere. Finally, routine implementation of a higher level of alcohol reduction support was measured relative to the 11 SBI components that we asked about. Providers may have implemented other types of alcohol reduction support that were not assessed in our research.
In spite of its limitations, this research suggests the importance of supporting HIV providers in their alcohol reduction efforts with patients. They especially need to be exposed to information about the dangers of alcohol use among these patients, and the effectiveness of alcohol SBIs. Furthermore, training in the use of SBIs for HIV care providers who view alcohol reduction support as consistent with their roles should increase their self efficacy in implementing them with their patients. This, in turn, should result in more frequent use of SBI components. Even for those providers with especially large caseloads who may find that implementing SBIs is impractical for them personally, increasing their awareness of SBIs' effectiveness may encourage them to refer patients to other clinicians in the HIV care practice. In this way, HIV patients who drink can be supported in reducing their alcohol use in order to preserve their health and limit the spread of the virus to others.
This research was supported by grant #1R21AA016743 from the National Institute on Alcohol Abuse and Alcoholism. We also gratefully acknowledge the support of the Muriel and Virginia Pless Center for Nursing Research at the New York University College of Nursing.
No competing financial interests exist.