Drug use and HIV medication adherence in people living with HIV

Abstract

Background:

Opioid and cocaine use epidemics continue to be substantial in the United States and intersect with the HIV epidemic. Antiretroviral medication (ARV) adherence is critical for optimum HIV outcomes. While previous research explored harm reduction strategies to prevent HIV spread for people who use drugs (PWUD), little is known about strategies used by PWUD living with HIV to maintain ARV adherence.

Methods:

We explored whether PWUD modify their drug use explicitly to maintain ARV adherence, and identified factors associated with this process. We conducted 23 semi-structured interviews. Data were analyzed using a modified framework analysis approach.

Results:

Participants had a mean age of 54 years and were predominantly male (70%) and non-Hispanic black (65%). Most described periods of being able to adhere to ARVs while still using drugs, difficulty adhering to ARVs while using drugs, and abstinence/near abstinence from drug use. In exploring factors that influenced changes in drug use and ARV adherence behaviors, we noted consistent acknowledgement of the roles of family, partners, or providers.

Conclusions:

PWUD living with HIV often modify their drug use to improve ARV adherence. Providers caring for this population might consider family or group education models to encourage harm reduction to improve outcomes.

INTRODUCTION

Despite urgent national attention, opioid and cocaine use continue to be substantial in the United States. Between 2002 and 2016, there was a four-fold increase in overdose deaths involving opioids and a 23-fold increase in overdose deaths from cocaine mixed with non-methadone synthetic opioids.¹ In addition to increased mortality, the risk of both chronic and infectious diseases and their management are worse for people who use opioids or cocaine.^2–5 Because the opioid and cocaine use epidemics intersect with the HIV epidemic, it is crucial to understand and mitigate the impact of drug use on HIV disease.

Adherence to antiretroviral medications (ARVs) is the most important predictor of HIV viral load suppression and survival for people living with HIV.^6,7 It has been commonly believed that perfect ARV adherence requires engaging in drug treatment and abstaining from non-prescribed drug use.⁸ The results of a meta-analysis questions this, though, by showing that adherence rates in studies of people who use drugs (PWUD) are similar to rates in studies that exclude PWUD.⁹ Nonetheless, barriers to achieving perfect ARV adherence and strategies to overcome these barriers are likely to differ between PWUD and those who do not use drugs due to their different social environments.^10–12 Other than engagement in drug treatment, little is known about the strategies that are used to overcome barriers to perfect ARV adherence among PWUD.

Previous research among PWUD has described drug use behaviors in the context of other domains in one’s life including social relationships, work, and health. Drug use has been described in a continuum of functionality, in which one end of this continuum represents PWUD that are able to maintain positive behaviors in all other domains in their life and the other end represents PWUD who are unable to do so.^13–15 Many factors are likely to be associated with choice and ability to regulate one’s own drug use along this continuum.^16,17 Identifying these factors for PWUD living with HIV could be helpful in designing interventions, not only to reduce the spread of HIV, but to improve HIV outcomes.

Harm reduction is an approach that emphasizes, among other things, minimizing negative health and social impacts associated with drug use and other behaviors, and may include regulating drug use.^18,19 Traditionally, when this approach is applied to address the HIV epidemic, it has been for the purpose of HIV prevention.²⁰ Common harm reduction approaches to HIV prevention, including needle exchanges, safe injection sites, and opioid replacement therapy to reduce injection use of drugs, can also be effective to improve outcomes in people living with HIV.^21,22 However, few studies have examined whether PWUD living with HIV apply other harm reduction strategies that regulate drug use behaviors explicitly for the purpose of improving adherence to ARV medications, or to determine factors that might influence the use of such strategies.

In this qualitative study, we explored the relationships among drug use behaviors and ARV adherence from the perspective of PWUD living with HIV who currently use non-prescribed opioids or cocaine. We focused on whether participants felt their drug use was associated with ARV adherence, either positively or negatively, and factors that might influence their ability to modify both drug use behaviors and ARV adherence. We used the Social Ecological Model²³ to organize our exploration, which suggests factors within different levels of social organization may be influential, including individual, relationship, community, and society level factors.

METHODS

We conducted semi-structured interviews with adults living with HIV who recently used non-prescribed opioids or cocaine. The Albert Einstein College of Medicine Institutional Review Board approved the current study.

Study Population

We recruited participants from newspaper advertisements, drug treatment clinics, and primary care clinics of a large hospital system in the Bronx, New York. This study was nested within a parent randomized controlled trial that tested the effect of an abstinence-reinforcing contingency management intervention on HIV outcomes.^24,25 Patients who were screened for eligibility for the parent study were eligible for the current study, irrespective of their participation in the parent study. Additional inclusion criteria for the current study were: aged 18 years or older; infection with HIV by self-report and confirmed by medical records; ARV prescription for at least 16 weeks confirmed by medical records; non-prescribed opioid or cocaine use in the last 30 days by self-report; fewer than three hospitalizations in the past six months, and English language fluency. Participants received $25 compensation.

Data Collection

Sociodemographic data used for this study were collected during baseline interviews for the parent randomized controlled trial using Audio Computer-Assisted Self-Interview technology.

After being advised of the purpose of the study and signing informed consent, one of two interviewers (JJ and a research assistant trained by JJ, both male) conducted a semi-structured interview that included questions about drug use history, current drug use, HIV and other health goals, how living with HIV affects drug use, how drug use affects ARV adherence, and strategies to overcome barriers to ARV adherence. We were particularly interested in how drug use affects ARV adherence, and asked several questions, including “Do you think your using (cocaine or heroin) makes it harder or easier for you to take your HIV medications? Please explain,” and “How important do you think it is for you to stop using (cocaine or heroin) in order to improve HIV outcomes? Why do you think so?” We encouraged participants to talk about behaviors and attitudes intrinsic to themselves and other people, and factors in their social environment that were associated with their drug use and ARV adherence behaviors as organized in the Social Ecological Model.²³ Interviews lasted approximately 45 minutes, and were audio recorded and professionally transcribed.

Analysis

We analyzed the transcripts using a modified framework analysis approach²⁶ in which qualitative data are collected to address specific research questions and evaluate a priori themes. Framework analysis allows for simultaneous data collection and analysis.

We transcribed and coded the first three interviews using a preliminary coding schema that was based on the study research questions and the interview guide. We then discussed the transcripts and modified the interview guide and coding schema. All members of the research team participated in these discussions to ensure the team developed a common understanding of the coding schema.

The two interviewers conducted the remaining 20 interviews using this iterative approach with the coding team. For each transcript, at least two members of the research team applied and discussed codes. After 23 interviews, it was concluded that data saturation was achieved. DS and NS reviewed all transcripts to ensure consistency in coding.

Ongoing comparative analyses were conducted in regular meetings to identify themes from the codes that addressed the research questions. We categorized codes under the main themes that were identified and developed preliminary interpretations of the data. The final table of codes, themes, and interpretations was discussed and agreed upon by the research team. We entered transcripts with their codes into Dedoose software to facilitate analysis and retrieval of quotations from participants.²⁷

RESULTS

As shown in Table 1, among the 23 participants, the average age was 54 years (range 42–73), 16 (70%) of the participants were male, 15 (65%) were non-Hispanic black, six (26%) were Hispanic, and two (9%) were non-Hispanic white. Fourteen (61%) of the participants reported using opioids and 22 (96%) reported using cocaine. Most of the participants (N=19, 83%) reported receiving their HIV diagnosis more than five years ago (not shown in Table 1).

Table 1.

Demographic & clinical characteristics of 23 people who use drugs and have HIV infection

		N(% of 23)
Demographic Characteristics
	Age, years (mean ± s.d.)	54 ± 7
	Male gender	16 (70)
	Race
	Non-Hispanic black	15 (65)
	Hispanic	6 (26)
	Non-Hispanic white	2 ( 9)
Drug Use Behaviors and Treatment, past 30 days
	Opioids	14 (61)
	Cocaine*	22 (100)
	Methadone treatment	6 ( 26)
	Buprenorphine treatment	3 ( 13)
HIV-Related Behaviors
	Taking antiretroviral medications (ARVs)	23 (100)
	Prescribed ARVs for ≥5 years	10 ( 43)

^*N=22

ARV adherence

In our sample, 17 (74%) reported perfect adherence to their ARVs at the time of the interview. All participants reported that perfect ARV adherence is an important goal for achieving good health, and that they monitored their viral load or CD4 count.

Drug use and ARV adherence

All participants stated that either abstinence or reduction in drug use is needed to maintain good health and adhere to ARVs. All participants described goals related to their drug use that ranged from maintaining or achieving abstinence to restricting the amount or frequency of using drugs, and goals related to their health that included perfect adherence to ARVs, attending medical appointments, exercising, and eating healthy foods and/or taking vitamins. When asked, participants consistently expressed stronger confidence in their ability to attain or maintain positive HIV-related health behaviors than their ability to attain or maintain their drug use goals.

We identified three general patterns of ways in which participants described modifying their drug use to facilitate ARV adherence. Most participants reported periods that were consistent with each category at some point in their lives. The three patterns are:

1. Ongoing drug use during the same period as reporting perfect adherence to ARVs

   For example, one participant said, “I would stay clean for a long time, then I go back to doing heroin. But I still take my medication. I’ve shot heroin, take my medication five minutes later.”

2. Ongoing drug use during the same period as reporting poor ARV adherence

   For example, one participant described a previous time period in his life by saying, “…when you’re using drugs, you forget about your meds. I mean, for me, in my case, I do. I forget about taking my meds. I’m so focused on getting high or getting my next high or trying to get the money for my next get high that I forget about taking my meds, even taking a shower.”

3. Abstinence from drug use to achieve or maintain good health

   For example, one participant reported a seven year period of abstinence after becoming very ill as follows:

   Interviewer: Well so what prompted you to stop using and—[start taking AVRs]

   Participant: ‘Cause I wanted to live.

   Interviewer: OK

   Participant: And I, you know, I wanted to just live a better quality of life. And I wanted to live. And I knew that if I kept using these drugs the way I was that I wasn’t going to live, you know.

   Interviewer: So when you started taking your HIV meds—what was that like for you?

   Participant: It was life changing. My numbers went up, you know, like my T cells got higher. And my viral load was undetectable. And everything was good. Everything was good. And I was doing good. No drugs.

Most participants ranked health-related goals, including ARV adherence, higher than their desire to maintain behaviors that facilitated drug use. They told stories about regulating drug use after experiencing health problems in the past, and felt that they could stop using drugs if they were to become sick in the future. For example, when a participant who reported current heroin use was asked what he would do if his HIV viral load became detectable or his health deteriorated, he stated, “I would stop messing with the drugs…That’s something I will have to do if I choose to live.” In response to the same question, another participant said, “If I look in the mirror, when I look in the mirror, and I start to see, like I can see the bones in my face, it just means slow down.”

A few participants who struggled with trying to regulate their drug use to achieve perfect ARV adherence reported that they gave high priority to the aspects of their lives that facilitated drug use. For example, one participant who reported current cocaine use and poor adherence to ARVs talked about his steady girlfriend who was still using drugs. He prioritized that relationship over changing his drug use.

Participant: …it could be a point where she still using, and I’m around and then before you know it, I’m right, I’m right in the, back in the mix with her, because I’m just, like I said, you know, I’m just there, you know, with her. So, she’s like, how can I say it, an enabler, you know what I mean?

Interviewer: Right.

Participant: She’s going to enable, because--and then knowing that I’m still like flirting with it.

Interviewer: And so how confident are you that you can keep this distance?

Participant: I have to work on that.

Others prioritized health over their drug use, but felt they could not control their drug use sufficiently. One participant said, “There has been a lot of circumstances where, you know, drug use has been harmful to me and for whatever reason I just kept using… When it comes to drugs and alcohol it is like they set the rules. They tell you when to use them or not to use them, you know. No matter how sick or healthy I am.” Another participant made a similar argument by saying, “…if you’re being real with yourself and your HIV issues, there’s no room for the drugs. But you always try to make room for the drugs because you always want this comfort right here.”

Those who achieved abstinence from drug use, as opposed to reducing drug use, typically did so in the context of health consciousness. Often this was because participants were unsuccessful in attempts to reduce drug use and found they had to completely abstain from drug use to address other health goals. When asked if there were periods when he tried to modify his drug use, one participant said, “I try. I definitely did try that, but it didn’t work. It didn’t work.”

Factors associated with drug use and ARV adherence

We then explored factors that were associated with drug use and ARV adherence. Six major influences were evident.

1. Seeking approval from family members

Participants who were abstinent from drugs as well as participants who continued to use drugs frequently talked about how approval from family members motivated them to regulate drug use and, at the same time, improve ARV adherence. One participant stated, “My family also plays a part in my trying to be better. I am not 100 percent, but you know, as far as adherence to, you know, taking my drugs, but my family helps, you know. Whenever I talk to my son or my grandkids or my sister, you know, they ask me how I am doing, you know. I like to know that when I respond, I am not lying to them, you know. I am saying yes I am okay, and I am really okay.”

The word “pride” was commonly used. Participants expressed a desire for their family members to be proud of them as a reason to improve their health, take ARVs, and reduce use of drugs. One participant described her goals by saying, “…well basically I just, you know, want to do better. I’m a good person. You know, I have a good heart. I love my family. And I just want them to be proud of me. I want to be proud of myself.”

Sometimes this desire for family approval had negative influences on health and drug use. One participant whose mother forced him to move out of her house when she learned of his HIV diagnosis, talked about using drugs with his mother while attempting to reconcile with her. “So we’re sitting in the kitchen and, now mind you, I already smoked a joint with my family. I already had a drink with my family. But to be sitting in the kitchen doing coke with my mom. Oh, this is definitely a graduation… When [my mother] came back in that day, that day, she asked would I move back into the house. So now she’s going to accept me now.”

2. Drug use as behavior that strengthens relationship with a partner

All participants in this study who continued to use drugs and reported having a significant other, reported using drugs with their significant other. Drug use was described as an activity that defined their relationships. At the same time, in all cases of participants who continued to use drugs with a significant other, the significant other helped the participant to be adherent to ARVs. For example, one participant noted that, “He and I do everything together. If he gets high, I’m getting high. If I’m not getting high, he’s not getting high…” This participant also highlighted her significant other’s role in helping her maintain good health, stating that, “Having a partner is very essential and very helpful because he’s there to push me, you know, when I don’t want to do things, you know. ‘You’ve got to eat so you can take your medication. Did you take your medication?’ You know, so that helps greatly having a partner.”

3. Reliance on people in the health care system as a substitute for lack of other social support

Some participants who did not mention having any other social supports attributed their success in regulating their drug use or maintaining ARV adherence to a health care provider or support group members. For example, one woman who did not mention any family or friends as social support, talked about having a “good support team” to help her with drug use and maintaining perfect ARV adherence. She specifically referred to her nurse and therapist as members of this team. Another participant reported that having a good relationship with his provider helped him to be adherent to ARVs, even though he continued to use drugs. “I think it was more of a struggle--because the doctor was like so bad--until I found a doctor that I could really sync with...” A third participant, who was striving to improve ARV adherence, talked about developing an open relationship with his provider with whom he felt that he could talk about his drug use. “At first, I didn’t tell him, you know. I just told him that I wanted to go on a hiatus ‘cause he didn’t know I was getting high. And he agreed to it, you know, ‘cause my numbers were so good, you know. And then it just became a point where I stopped lying to him, and I got honest with him and told him what I was doing and stuff like that.”

Only one participant talked about support groups as being a negative influence on drug use because being around drug users “triggered” drug use for him. This person had other positive social supports that encouraged her ARV adherence.

4. Rejection of acquaintances that may trigger former lifestyle

Participants in this study typically had narrow social networks that included mainly family, a significant other, or a health care provider. When other acquaintances were mentioned, they were often considered “triggers” for drug use and were avoided by participants in this study. One participant said, “My neighborhood, I just stay away from the neighborhood because it’s like evil you know. I go in there, I go over there, and even if I got no money, drugs come to me for free because they know me, you know. I used to use drugs in that block… So, I got to stay away from that neighborhood. It’s bad news.” The rejection of drug using acquaintances was often mentioned in the context of trying to develop a healthier and more stable life.

5. Stability as a significant health goal

Stability was mentioned as a goal related to regulating drug use and to improving ARV adherence. Stable housing and privacy were common influences on both health and drug use. One participant reported intermittent injection heroin use and perfect ARV adherence, and highlighted the role of housing stability in helping him to maintain good health behaviors and reduced drug use. He went on to say, “I know a couple of friends that I know that are living in the streets right now that I wish I could help, but I can’t bring them home with me… They’re not stable. They have no stability in their life as far as a place to live. Some of them are living with people that they’re miserable with. Or living with their mother. And most have to get out every day. It’s hard for you to take your medication when you have all of these things on your mind, or you’re hustling, or you’re doing drugs and you’re hustling for your next hit or your next bag, and you have to take these meds. You’re not going to take them when you’re supposed to.”

Structure was also commonly mentioned, particularly with regard to regulating drug using behaviors. According to one participant, “I do a lot of functional things now, you know, I wake up early in the morning, I have structure in my life. Because when you don’t have structure everything goes down.” Another participant said, “I have to start thinking of ways of entertaining myself and don’t feel lonely. When I feel lonely, that’s when I start using. So I have to like get a job, do something to try to keep myself busy. Because when I keep myself busy, I don’t use. I don’t use, yes. I’m keeping myself busy now going doing something. Going on trips, working, you know. If I keep myself busy I know I won’t use.”

6. Health conditions other than HIV and aging

All participants in this study were aged older than 40 years and have been living with HIV for more than ten years. It was common for them to report living with other chronic health conditions in addition to HIV, including heart conditions, diabetes, depression, PTSD, and hepatitis B and C, among others. Participants commonly mentioned the need to reduce drug use to improve HIV adherence and other health outcomes because of aging. “I’m more mature now. My health is getting worser.” And, “I think now because I’ve gotten older, I think more about living than dying, you know, and, you know, I cherish life.”

DISCUSSION

We explored how PWUD living with HIV describe their drug use behaviors and ARV adherence, and their perceptions of the influences behind these behaviors. Our study is the first to our knowledge to demonstrate that PWUD commonly report successes in modifying their drug use explicitly to improve ARV adherence. Additionally, we found, despite reporting few social relationships, the most common influence mentioned in regulating drug use and maintaining ARV adherence was either a family member, partner, or someone in the health care provider team. We discuss these two main findings below.

All of the participants wavered between periods of controlled drug use that is associated with positive health behaviors, like taking vitamins, exercising, attending physician appointments and ARV and other medication adherence, and heavier drug use that is associated with less attention to other health behaviors as well. These findings have relevance to strategies for harm reduction education for PWUD, suggesting it may be particularly important to recognize that health needs are constantly changing and must be frequently reassessed. Ongoing screening of drug use and other health behaviors can facilitiate timely brief health promotion educational interventions that are tailored to where people are at in a cycle of drug use that may be associated with other health behaviors. For example, research suggests the value of health promotion education, especially nutritional education, for PWUD,^28,29 however there is a lack of empirical data on effective screening and intervenion approaches for this population.

Certain social relationships were important to maintain healthy behaviors related to both HIV and drug use outcomes. There is a large body of literature exploring how influence of social networks and social support influence both drug use^{13,17,30–35} and ARV adherence,^31,32,35 but the way that social ties facilitate positive health behaviors has not been clarified in the literature, including in PWUD and live with HIV. Our study helps to elucidate these influences. It is possible that providers can draw from this information to enhance treatment approaches. Models of intervention to reduce drug use and improve health outcomes that include families and friends have not been studied extensively, but a few effective models have been documented in the literature including one in adolescents with substance use³⁶ and another in adults.³⁷ Our findings also suggest that involvement of family members in harm reduction interventions could be effective in regulating substance use to maintain ARV adherence in PWUD living with HIV.

It is noteworthy that, in our study, family members and significant others were mostly identified as positive supports, even if they also used drugs, but friends were rarely mentioned as having a positive influence on our participants. This is inconsistent with some research that found no role of having drug-using friends on drug treatment outcomes.³⁸ Our findings may reflect the fact that PWUD often have few close friendship supports either because they feel they overburden friends or they lose friends due to high rates of death and illness in their social networks.³⁹

In the absence of other relationships, some participants in our study considered their HIV provider to be an important form of social support. It is well known that people who trust their providers or who have providers that they can easily communicate with have better HIV outcomes than people who report not trusting their providers.⁴⁰ Social support from health care providers has been described in other chronic health conditions, such as end-stage renal disease resulting in improved health outcomes.⁴¹ The importance of the patient-provider care team has been described elsewhere and has emphasized the importance of holistic care as well as destigmatized care. The pivotal role of providers as the sole source of social support for some PWUD living with HIV was somewhat surprising, and not mentioned in the literature.

Limitations

Our study has several limitations. First, we interviewed a narrowly defined group of people in New York City who live with HIV and use non-prescribed opioids or cocaine. The sample included long term survivors of both HIV and drug use, aged over 40 years, and in contact with the health care system. This group does not represent the diverse population of people living with HIV and drug use in New York City, but a subset that may be growing.

Additionally, some of our participants may have been part of the parent contingency management study, in which patients were given financial incentives to abstain from drug use. Although none of the participants discussed the parent study in these interviews, participation in the parent study may have influenced the way in which participants reported and talked about drug use behaviors. Additionally, although we collected urine samples to test for drug use as part of the parent study, these tests were not conducted for all participants in the current study at the time of these interviews and therefore could not be used to validate self-reports about current drug use. Viral load data taken for the parent study were not used here for the same reason.

Our interviews were not designed to elicit detailed social network information. We identified social relationships as a major theme in the interviews during data analysis and, consequently, were unable to explore directionality between social relationships and drug use or other health behaviors, or reciprocity in social relationships and how this influences health behaviors. Further, our study was not designed to include an in-depth exploration of how the types of social support influenced health behaviors. Traditionally, social support has been described as emotional, instrumental, informational, and appraisal.⁴² Insights from our interviews have guided our development of an interview tool with the goal of exploring these areas of social support as they relate to harm reduction efforts to improve ARV adherence in PWUD.

Finally, our study did not evaluate the role of race/ethnicity, culture, income, language or sexual orientation, all of which are known to be important in HIV-related health behaviors, drug use behavior and HIV outcomes. Future research should define sampling and data collection in such a way to include these factors in analyses.

Conclusions

Since the advent of ARVs, long term survival of people living with HIV is possible,⁴³ but several studies suggest that even with guideline concordant care, they have reduced life expectancy relative to the general population and to people who are not living with HIV.⁴⁴ Furthermore, the prevalence of continued or new drug use in people with HIV is significant.⁴⁵ Patterns of non-prescribed drug use and its impact on ARV adherence and HIV outcomes are important areas to explore. In our study, people living with HIV who also use drugs commonly modify their drug use to address health problems and improve ARV adherence. Family, significant others, and, in the absence of either, health care providers, are influential in this process. Interventions to help improve HIV outcomes in PWUD might consider family models of health education.