Self-medication of pain and discomfort with alcohol and other substances by people with HIV infection and substance use disorder: preliminary findings from a secondary analysis

Abstract

There is a limited literature regarding factors associated with self-medication of pain and discomfort using alcohol, nonprescription substances or overuse of prescription medications among people living with Human Immunodeficiency Virus (HIV). This cross-sectional analysis used data from the Boston ARCH Cohort among participants with HIV infection and a history of alcohol or other substance use. Among 248 participants, 37% were female, 50% Black, 25% Latinx; 36% reported fair to poor health and 89% had CD4 cell counts >200/mm³. Half reported self-medication and of those, 8.8% reported doing so only with alcohol, 48.8% only with other substances and 42.4% with both alcohol and other substances. Those reporting self-medication were significantly (p<.05) younger (mean 47 vs 50 years), less employed (11% vs 21%), and less likely to have HIV viral suppression (60% vs. 80%). Depression, anxiety, and HIV symptoms were associated with significantly greater odds of self-medicating, as were substance dependence, recent injection substance use, heavy alcohol use, cocaine use, opioid use, sedative use, and cannabis use. Self-medication, highly prevalent and associated with worse mental health symptoms, greater substance use, and lesser HIV disease control, should be explored by HIV clinicians caring for people who use substances.

Introduction

The life expectancy gap has substantially narrowed between people living with HIV/AIDS (PLWH) who have full access to HIV care and the general population (J. L. Marcus et al., 2016). With longer lifespans, comorbid chronic conditions, such as pain, have become increasingly recognized as clinical challenges by clinicians. Many, if not most PLWH report persistent pain (K. S. Marcus et al., 2000; Merlin et al., 2015; Miaskowski et al., 2011). Even among those with well-controlled HIV disease, pain is prevalent and frequently under-treated (Basu et al., 2007; Klepp et al., 2023; Larue et al., 1997; Parker et al., 2014; Tsao et al., 2007).

Under-treatment or ineffective treatment of chronic pain in PLWH is particularly problematic. Chronic pain increases functional impairment (Merlin et al., 2013), lowering quality of life. Further, chronic pain is associated with lower retention in HIV primary care, which will necessarily affect long-term HIV-related health outcomes (Merlin et al., 2012).

Pain is debilitating, and comorbid with psychologically distressing conditions such as depression, anxiety and insomnia (Adams et al., 2016; Orlando et al., 2002). Substance use disorder is also highly prevalent among PLWH (Hartzler et al., 2017; Malee et al., 2014). As illicit opioid use and prescription opioid misuse are common among PLWH, both clinicians and patients are reluctant to treat pain with opioids (Lum et al., 2011; Önen et al., 2012; Vijayaraghavan et al., 2011) and nearly half of patients with moderate to severe pain do not receive any pain treatment (Larue et al., 1997). Without effective treatment for physical and emotional pain, patients may develop alternative coping mechanisms such as the use of illicit substances.

Self-medication occurs when a person takes a substance (alcohol and/or substance, including misused prescribed medications) to relieve an illness symptom without the guidance of a medical prescriber. The stress-coping and the self-medication models of substance misuse, in which pain and/or stress drives individuals to use substances to cope and/or self-medicate psychological suffering (Khantzian, 1997), may help explain the use of a variety of substances to relieve both acute and chronic physical pain (Alford, 2016), even among persons with substance use disorders.

Among pain-self management strategies, the continued use of substances to regulate pain among those with substance use disorder is common, but under-explored (Merlin et al., 2015, 2018; Nicholas et al., 2007). In the general population, self-medication is a common practice to treat the physical and mental etiologies that drive the patient experience of pain, yet such self-treatment is associated with poorly controlled pain as noted, as well as depression and anxiety symptoms (Brennan et al., 2005; Khantzian, 1997). Alcohol, the most prevalent substance used in the US (Kacha-Ochana et al., 2022), is ingested by patients for pain relief despite dubious anti-nociceptive effect (James et al., 1978); alcohol can also cause pain through neuropathy (Julian et al., 2019), falls, accidents and injury (Cawthon et al., 2006; Cherpitel et al., 2015). Cannabis too is commonly used to self-treat pain (Romero-Sandoval et al., 2018) as are illicit opioids, and even cocaine (Alford et al., 2016; McCabe et al., 2007; Weiss et al., 1992). The use of alcohol and other substances, often at greater amounts than the general population reports, is well-documented among PLWH (Galvan et al., 2002; Martinez et al., 2019; Samet et al., 2004). However there is a limited literature regarding factors associated with self-medication using alcohol, cannabis and other substances among PLWH (Dansak, 1997; Greenwald et al., 2023; Tsui et al., 2014; Woolridge et al., 2005).

Here, we studied the prevalence of and factors associated with use of alcohol and other substances to self-medicate discomfort or pain among people with HIV and substance use disorder or a lifetime history of any injection substance use.

Methods

Participants

This study is a cross-sectional analysis of data collected from adults living with HIV who have either past year alcohol or substance dependence, based on DSM-IV criteria, or who have ever injected substances. All were participants in the Boston Alcohol Research Collaboration on HIV/AIDS (ARCH) Cohort, a prospective cohort study of alcohol use and bone density. A detailed description of the inclusion and exclusion criteria of the parent study, and description of the original sample are described elsewhere (Saitz et al., 2018; Thakarar et al., 2020; Ventura et al., 2017). The Institutional Review Board at Boston University Medical Campus reviewed and approved this study; all participants provided written informed consent.

Recruitment occurred in two infectious disease clinics, one at an urban, academic medical center and the second at a community health center serving the housing insecure. Study recruitment occurred from December 2012 to November 2014. Of the 250 participants enrolled, 248 had non-missing responses to questions about self-medication.

Measures

Baseline research interviews assessed demographics, substance use, pain, physical and mental health, and self-medication. Demographic data included age, sex, race/ethnicity, employment and housing status. Smoking status was dichotomous, with current smokers defined as smoking at least 100 cigarettes in a lifetime and now smoking “everyday” or “most days.”

Heavy drinking was defined as consuming four or more standard drinks (14 grams of alcohol) for women and five or more for men on one or more day of the prior 30 days, using the 30-day Timeline Follow-Back (TLFB) (Sobell & Sobell, 1992) method. Substance use was assessed using the Addiction Severity Index (McLellan et al., 1980) to capture past 30-day use of cannabis, cocaine, tranquilizers or sedatives, heroin, opioid medication, along with route of administration. Alcohol and other substance dependence were assessed using the Mini-International Neuropsychiatric Interview (MINI) (Sheehan et al., 1998).

The risk of five-year all-cause mortality was assessed using the Veterans Aging Cohort Study (VACS) index obtained from the electronic medical record that includes age, HIV severity markers, and measures of organ system injury (Justice et al., 2013; Tate et al., 2013). HIV-related symptoms, such as fatigue, memory loss, and skin problems, and how bothersome they were, were assessed with the HIV symptom index (Justice et al., 2013). HIV disease markers (CD4 cell count and HIV viral load (with HIV viral suppression defined as <200 copies/mL)) and current prescribed medications (opioids, sedatives/tranquilizers) were obtained from electronic medical records.

Self-reported general health status was assessed using a single item from the VR-12 (Kazis et al., 2004). Symptoms consistent with depression were assessed using the Patient Health Questionnaire (PHQ-2) (Kroenke et al., 2003); a score of three or greater was considered consistent with depression. Anxiety was assessed with the Overall Anxiety Severity and Impairment Scale (OASIS) (Campbell-Sills et al., 2009; Norman et al., 2006); a score of eight or greater was considered consistent with anxiety.

Pain was assessed using the Brief Pain Inventory (BPI) (Cleeland, 1991), which asks about pain experienced by the participant other than “everyday kinds of pain” such as such as minor headaches, sprains, and toothaches. Those who reported pain in the past seven days were asked duration, severity (rated 0–10 for the worst) and interference with activity, mood, walking, work, relations, sleep and enjoyment of life (also rated 0–10). Pain severity and pain interference (average across seven domains) was categorized into four groups based on published cutoffs (Li et al., 2007): none (0), mild (1–4), moderate (5–6), and severe (7–10). Pain duration was categorized as pain present: only in past week, more than one week but less than three months, three to 24 months, and more than two years.

Primary Outcome

Self-medication was intended to capture the breadth of participant self-medication across a wide range of symptoms, and was assessed using the following question, “In the past three months, have you used alcohol to treat pain or discomfort?” This question was then repeated asking about “nonprescription drugs (like heroin, cocaine or marijuana)” “prescription medications taken without a doctor’s prescription (e.g., narcotic pain medications, sedatives/benzos, or Ritalin/amphetamine)” and “prescription medications taken in greater amounts than prescribed,” replacing “alcohol.” Self-medication was defined as an affirmative response to any of those queries.

Analytical Methods

We summarized all variables of interest using proportions, with patient age and number of HIV bothersome symptoms categorized at the sample median. We then tested unadjusted associations between participant characteristics and self-medication using chi-square and t-tests. Self-medication was explored among all study participants, not only those reporting pain or discomfort because pain may be have eliminated in some by self-medication. To account for possible confounding due to demographic characteristics, we performed separate multiple logistic regression analyses predicting self-medication from each substance or health characteristic, controlling for demographic characteristics that were significantly associated with self-medication. Results are presented as adjusted odds ratios and 95% confidence intervals. All analyses were completed using SAS, Version 9.3.

Results

Prevalence of self-medication with alcohol and/or other substances

Of the 248 participants, 134 (54%) reported pain in the past week. Of these, 61% had pain for at least three months. Among participants, 125 (50%) reported past three-month self-medication with at least one substance and 123 reported no self-medication. In the total cohort, 61 (25%) reported self-medication with alcohol, 93 (38%) reported self-medication with nonprescription substances, 37 (15%) reported prescription medications without a prescription and 26 (11%) reported prescription medications in greater amounts than prescribed. Of those who reported self-medication, 11 (9%) reported doing so only with alcohol, 61 (49%) reported doing so only with other substances and 53 (42%) reported self-medicating with both alcohol and other substances. Overall, 51% had drug and alcohol dependence, 21% had drug dependence only, 9% had alcohol dependence only and 19% had no dependence. Among the more than three in four participants who reported any substance dependence, they most frequently reported cocaine (54%), opioids (29%), and cannabis (26%) and benzodiazepine (11%). One-third of participants used medication for OUD (MOUD) and there was no significant difference in self-medication between those with and without MOUD.

Patient factors associated with self-medication with alcohol and other substances

Self-medication was not associated with sex, race or homelessness (Table 1). However, prevalence of self-medication was significantly higher for younger patients (59% vs 40% for those 50 years or younger vs. over 50 years, p<.01), and patients who were unemployed vs employed (53% vs 35%, p=0.03). Substance use was associated with self-medication, with higher odds of self-medication for current smokers vs non-smokers (57% vs 28%; aOR 2.80 (1.40, 5.57)), as well as for those with vs without past 30 day heavy drinking (66% vs 35%; aOR 3.42 (2.00, 5.84)), cannabis use (67% vs 37%; aOR 3.49 (2.02, 6.04)), and illicit prescription opioid use (83% vs 45%; aOR 5.85 (2.31, 14.83)). Past 30-day cocaine use, heroin use, illicit tranquilizer/sedative use, and injection substance use were all significantly associated with higher odds of self-medication, with adjusted odds ratios ranging from 3.25 to 4.69 (Table 1). Those with current substance dependence or substance and alcohol dependence were also significantly more likely to self-medicate (aOR 2.97 (1.19, 7.40); aOR 5.68 (2.48, 12.99) respectively).

Table 1:

Self-medication over the past three-months, by demographic, substance use, and health characteristics among HIV-infected adults with (current) substance dependence or injection substance use (ever)^a

	Any self-medication %, (n/N)	p-valuek	aOR (95% CI)l,m
Overall Sample	50 (125/248)	--	--
	Demographics
Age  50 years or less	59 (80/136)	<0.01	--
Greater than 50 years	40 (45/112)
Sex  Female	53 (49/92)	0.49	--
Male	49 (76/156)
Race/ethnicity  Black	49 (60/123)
White	53 (27/51)	0.65	--
Latinx	48 (30/62)
Other	67 (8/12)
Employed  Employed	35 (14/40)	0.03	--
Not employed	53 (111/208)
Homeless statusa  Homeless	59 (37/63)	0.13	--
Housed	48 (88/185)
	Substance Use
Smoking statusd  Current smoker	57 (110/194)	0.0002	2.80 (1.40, 5.57)
Current non-smoker	28 (15/53)		Ref
Heavy drinking day in past 30 daysb  1+ heavy drinking days	66 (82/125)	<0.0001	3.42 (2.00, 5.84)
0 heavy drinking days	35 (43/123)		Ref
Cocaine use in past 30 daysc  Any cocaine use	71 (53/75)	<0.0001	3.25 (1.80, 5.89)
No cocaine use	42 (72/173)		Ref
Heroin use in past 30 daysc  Any heroin use	76 (29/38)	<0.001	3.31 (1.47, 7.44)
No heroin use	46 (96/210)		Ref
Illicit prescription opioid use in past 30 daysc  Any illicit use	83 (30/36)	<0.0001	5.84 (2.30, 14.82)
No illicit use	45 (95/212)		Ref
Cannabis use in past 30 daysc  Any cannabis use	67 (74/110)	<0.0001	3.49 (2.02, 6.04)
No cannabis use	37 (51/138)		Ref
Illicit tranquilizer/sedative use in past 30 daysc  Any illicit use	77 (17/22)	<0.01	3.25 (1.13, 9.34)
No illicit use	48 (108/226)		Ref
Injection substance use in past 30 daysc  Any injection use	81 (22/27)	<0.01	4.69 (1.69, 12.96)
No injection use	47 (103/221)		Ref
Substance dependence, current  No dependence	21 (10/47)		Ref
Alcohol dependence only	39 (9/23)	<0.0001	2.29 (0.76, 6.94)
Substance dependence only	48 (25/52)		2.97 (1.19, 7.40)
Both alcohol and substance dependence	64 (81/126)		5.68 (2.48, 12.99)
Currently prescribed opioids  Prescribed opioids	59 (50/85)	0.08	1.58 (0.91, 2.76)
No opioid prescription	47 (65/139)		Ref
Currently prescribed tranquilizer/sedative  Prescribed tranquilizer/sedative	50 (20/40)	0.85	0.87 (0.43, 1.75)
No tranquilizer/sedative prescription	52 (95/184)		Ref
	Pain, Physical and Mental Health
VACS Indexe (risk of all-cause mortality)  0–9%	52 (87/168)	0.64	Ref
10–19%	47 (16/34)		0.86 (0.40, 1.83)
20–29%	41 (12/29)		0.78 (0.34, 1.80)
30–39%	59 (10/17)		1.76 (0.61, 5.11)
Self-reported health statusf  Excellent	26 (5/19)	0.13	0.33 (0.11, 1.03)
Very Good	55 (27/49)		1.06 (0.51, 2.18)
Good	50 (45/90)		Ref
Fair	51 (40/79)		1.01 (0.54, 1.87)
Poor	73 (8/11)		2.77 (0.67, 11.47)
Depressioni  Depression	62 (46/74)	0.02	2.07 (1.17, 3.68)
No Depression	46 (79/173)		Ref
Anxietyj  Anxiety	63 (71/112)	<0.01	2.34 (1.37, 3.97)
No Anxiety	40 (54/136)		Ref
CD4 cells/mm3  <200 cells/mm3	50 (111/221)	0.87	0.94 (0.42, 2.12)
> 200 cells/mm3	52(14/27)		Ref
HIV viral load  On ART, <200 copies/mL	43 (75/174)	0.002	Ref
On ART, 200+ copies/mL	71 (32/45)		3.00 (1.45, 6.19)
Not on ART	62 (18/29)		1.86 (0.81, 4.28)
Bothersome HIV symptomsh  9 or fewer symptoms	38 (47/125)	<0.0001	Ref
10 or more symptoms	63 (78/123)		2.90 (1.71, 4.94)
Pain Severityg  none (0/10)	44 (51/116)		Ref
mild (1–4/10)	53 (24/45)	0.28	1.57 (0.77, 3.21)
moderate (5–6/10)	59 (27/46)		2.28 (1.11, 4.71)
severe (7–10/10)	56 (23/41)		2.06 (0.97, 4.36)
Pain Interferenceg  none (0/10)	42 (49/116)		Ref
mild (1–4/ 10)	47 (20/43)	0.03	1.40 (0.67 2.90)
moderate (5–6/10)	58 (15/26)		2.46 (0.99, 6.09)
severe (7–10/10)	65 (41/63)		2.96 (1.53, 5.74)
Pain Durationg  No pain past 7 days	43 (49/114)		Ref
< 3 months	56 (29/52)	0.15	1.89 (0.95, 3.75)
3 months – 2 years	50 (10/20)		1.69 (0.63, 4.55)
> 2 years	60 (37/62)		2.45 (1.26, 4.73)

^a1 or more nights on the street or in a shelter

^bHeavy drinking day: four or more standard drinks (14 grams of alcohol) for women and five or more for men, assessed by Timeline Follow-Back for past 30 days

^cAssessed by Addiction Severity Index alcohol and other substance items assessing number of days of use in the past 30 days

^dCenters for Disease Control definition: ≥100 cigarettes in lifetime and now smoking “every day” or “some days”

^eVeterans Aging Cohort Study (VACS) index for all-cause mortality, calculated from age, HIV severity markers, and measures of organ system injury

^fVeterans-Rand 12 (VR-12)

^gBrief Pain Index (BPI), pain interference measured on 0–10 scale, with higher score indicating greater pain interference.

^hHIV Symptom Index

ⁱPatient Health Questionnaire (PHQ-2) score ≥3

^jOverall Anxiety Severity and Impairment Scale (OASIS) score ≥8

^kp-value for unadjusted comparison of those with vs. without self-medication

^ladjusted Odds Ratio, controlling for age and employment, comparing odds of row characteristic for those self-medicating vs not self-medicating

^mConfidence Interval

Patients with mental health conditions—depression and anxiety—were significantly more likely to report self-medication (62% vs 46%; aOR 2.07 (1.17, 3.68 for depression, 63% vs 40%; aOR 2.34 (1.37, 3.97) for anxiety).

Self-medication was higher among those without HIV viral suppression (71% vs 43%), although there were no significant associations between CD4 cell count, self-report general health status, or the risk of all-cause mortality. Severe pain interference was associated with greater odds of self-medication (aOR 2.96 (1.53, 5.74)).

Discussion

In this cohort of people with HIV and substance use disorders or a history of injection substance use, self-medication of pain or discomfort with substances was reported by over half. Consistent with overall increase in polysubstance use, self-medication appears to also involve multiple substances (Aouizerat et al., 2010; Hartzler et al., 2017; Larue et al., 1997; Uebelacker et al., 2015). Self-medication most commonly involved non-prescription substances, followed by alcohol use, prescription medications without a prescription, and finally prescription medications used in greater amounts than indicated by their prescriber. Self-medication is higher among patients with symptoms of depression and anxiety and in those with more HIV symptoms, which also may be psychologically distressing. While these data were collected nearly a decade ago, the overdose epidemic caused by opioid use (a primary substance used to self-treat pain) and the increase in alcohol use during the era of COVID-19, make the findings of continuing relevance to clinicians who treat PLWH in primary care settings (Pytell et al., 2022; Roberts et al., 2021).

Of note, among those self-medicating with a substance, about half also report self-medicating with alcohol. One clinical implication is that knowing about a patient who is self-medicating with a substance should prompt questions about self-medicating with alcohol, which often is unrecognized in patients using illicit substances. Conversely, if a patient uses alcohol for self-medicating (approximately half the sample), the likelihood that they are also using another substance for self-medication is relatively high.

Notably, self-medicating was not associated with the overall level of pain severity, suggesting that self-medication (nearly half of which involves polysubstance use) may be reducing pain from a level that was higher before self-medicating to a level similar to people not self-medicating. Worse pain interference scores suggest the former. Self-medication with pain medication, particularly with opioids, can worsen pain by producing hyperalgesia (Mercadante et al., 2019). Alcohol may also exacerbate pre-existing pain among PLWH (Kuerbis et al., 2019), consistent with a strong positive association between chronic pain and heavy alcohol use in a general population of older adults (Brennan et al., 2005). Interestingly, about half of the participants who self-medicated in the past three months did not have pain in the past seven days.

One limitation of our cross-sectional study was our inability to determine direction of causality. Still, HIV in association with self-medicating substance use is a multi-morbid state that might benefit more from integrated behavioral health, addiction, and pain management. The integration of healthcare services and psycho-socio-behavioral therapy in this context has been advocated previously (Nguyen et al., 2023). Health system policies should support the integration of treatment for unhealthy alcohol and drug use into current HIV care and interventions (Nguyen et al., 2023). Separate management of symptoms and substance by patients or providers may assure progression through this cycle (Manhapra & Becker, 2018).

Self-medication was associated with depression and anxiety, an important novel finding in analyses of pain among PLWH. This finding is consistent with previous studies with other populations that show a relationship between increased pain and negative mood (Merlin et al., 2015), sleep disturbance (Aǧargün et al., 1999; Cappelleri et al., 2009), lack of exercise (Dzierzewski et al., 2014; Geneen et al., 2017), and low self-efficacy coping with pain (Jensen et al., 1991). Notably, pain, depression and anxiety are often an inter-related, co-occurring cluster of symptoms. While we examine associations between each of these individual symptoms and self- medication, we do not attempt to separate out the effects of these individual symptoms. Our results provide important information about potential avenues for intervention. Given that cognitive behavioral therapy is already an empirically supported treatment for both chronic pain and substance use (Lunde et al., 2009; Monti, 2002; Walters & Rotgers, 2011), our results provide further support for broadening the use of CBT to target a wider variety of behaviors with this group to reduce both pain and substance use. Under-treated pain, withdrawal, depression, or anxiety may each be a catalyst for substance use or risky levels of use (Giovanniello, 2016; Vallerand et al., 2005).

Who self-medicates? In this case, persons who are younger, have greater unemployment, and are more likely to smoke cigarettes, drink alcohol heavily and use illicit substances. Perhaps these groups, seen within the health care system, may have had inadequate response to standard medications, inattention by clinicians for the source of their pain or discomfort, or less access to pharmacological interventions or to cognitive and/or behavioral treatment altogether. Why substance use is intertwined with pain or discomfort for some patients with substance use disorders but not others remains an open question but may involve patient characteristics such as greater psychiatric comorbidity, past trauma, or worse functioning.

There are additional limitations to this analysis. First, generalizability is limited to adults living with HIV and substance use disorder or a history of injection substance use. While we use the term substance use disorder, we enrolled persons based, in part on DSM-IV criteria, which may have lower diagnostic validity than the more recent DSM-V. Second, self-medication was assessed by a question that has not been validated and did not specifically inquire about the symptom or symptoms participants were self-medicating. Still, the self-medication question has face validity and is similar to those in validated pain scales that ask about discomfort from physical or mental health symptoms. Self-medication could have been for physical pain, but it also could have been for mental health symptoms sometimes perceived as painful, or to cope with withdrawal symptoms and/or drug cravings. Similarly, we do not know if any pain/discomfort symptoms were HIV-related, for instance due to neuropathy. While the question we used to assess self-medication for this particular investigation was intended to capture a broad range of symptoms, the inability to distinguish specific symptom(s) in our analysis is a limitation. Future research could examine which specific symptoms PLWH are self-medicating. Third, as with any patient self-report, there is the possibility for recall bias, and/or an under-estimate of the prevalence of self-medication. Fourth, the question posed to participants to assess self-medication grouped illicit substances together (i.e., cannabis, heroin, cocaine, methamphetamine) precluding analysis of use by specific substance. Respondents were not asked if they used medically sanctioned or illicit cannabis. Of course, since the years of the current study, cannabis use has followed the changing legal and political landscape with an increasing body of work suggesting analgesic benefits along with other possible clinical harms (Manthey et al., 2023). Fifth, our analyses examined the main effects of factors on self-medication. Our sample size does not support analyses of potential synergistic effects of multiple types of substance use on self-medication, or modification of substance use effects on self-medication by individual characteristics such as age or depression. Finally, which “prescription medications” were used to self-medicate remain undefined and could include medications as diverse as gabapentin or benzodiazepines.

The significance of our findings lies in the high prevalence of self-medication and substance use even among a cohort of individuals engaged in HIV primary care. We are aware that pain, particularly among people with HIV, may have increased during COVID-19 due to interruptions in HIV care or to COVID-19 and its complications, and such increases may produce increased self-medication using substances (Clauw et al., 2020; Drożdżal et al., 2020). Similar to the association that we report in this decade-old HIV cohort, substance use has increased in general population samples during COVID-19, particularly among persons with psychological conditions (Cerezo et al., 2023; Schieber et al., 2023).

Healthy self-care behaviors for pain are frequently initiated by patients and include physical activity (or its avoidance), cognitive strategies, spiritual practice, and distraction by spending time with family and friends (Merlin et al., 2015). But the effectiveness of such strategies may be limited. Substance misuse, as commonly reported in this cohort can lead to overdose, injuries and accidents, and worse HIV outcomes; importantly its effect on pain is often short-lived. Collaborative patient-provider care requires informed providers.

Thus clinicians should assess for self-medication behaviors. When identified, inquire about multiple types of substances used, screen for psychiatric symptoms, and provide support for HIV antiretroviral medication for viral suppression. Whether offering safer alternatives such as cognitive behavioral therapy, physical therapy or mindfulness techniques to address their patients’ pain, mental health and physical symptoms results in less self-medicating with substances is worthy of investigation.

Availability of data and material:

Not applicable