Abstract
Background
Chronic pain is prevalent among people living with human immunodeficiency virus (PLWH); managing pain with chronic opioid therapy (COT) is common. Human immunodeficiency virus (HIV) providers often diverge from prescribing guidelines.
Methods
This 2-arm, unblinded, cluster-randomized clinical trial assessed whether the Targeting Effective Analgesia in Clinics for HIV (TEACH) intervention improves guideline-concordant care compared to usual care for PLWH on COT. The trial was implemented from 2015 to 2018 with 12-month follow-up at safety-net hospital–based HIV clinics in Boston and Atlanta. We enrolled 41 providers and their 187 patients on COT. Prescribers were randomized 1:1 to either a 12-month intervention consisting of a nurse care manager with an interactive electronic registry, opioid education, academic detailing, and access to addiction specialists or a control condition consisting of usual care. Two primary outcomes were assessed through electronic medical records: ≥2 urine drug tests and any early COT refills by 12 months. Other outcomes included possible adverse consequences.
Results
At 12 months, the TEACH intervention arm had higher odds of ≥2 urine drug tests than the usual care arm (71% vs 20%; adjusted odds ratio [AOR], 13.38 [95% confidence interval {CI}, 5.85–30.60]; P < .0001). We did not detect a statistically significant difference in early refills (22% vs 30%; AOR, 0.55 [95% CI, .26–1.15]; P = .11), pain severity (6.30 vs 5.76; adjusted mean difference, 0.10 [95% CI, −1.56 to 1.75]; P = .91), or HIV viral load suppression (86.9% vs 82.1%; AOR, 1.21 [95% CI, .47–3.09]; P = .69).
Conclusions
TEACH is a promising intervention to improve adherence to COT guidelines without evident adverse consequences.
Keywords: HIV, chronic pain, chronic opioid therapy, prescription opioid misuse, pain management
In this cluster randomized clinical trial of human immunodeficiency virus (HIV) providers and patients, the TEACH intervention resulted in improved adherence to opioid prescribing guidelines. TEACH is a promising intervention to improve adherence to chronic opioid therapy guidelines for people with HIV.
The overprescribing of opioid medications has contributed to the current opioid overdose epidemic [1–5]. This crisis has compelled greater scrutiny of chronic opioid therapy (COT) for chronic pain [6, 7]. Prevalence estimates of chronic pain in people living with human immunodeficiency virus (PLWH) reported in the last 10 years range from 40% to 58% [8, 9]. Problematic prescription opioid use appears to be common among PLWH [10]. A 2019 study of PLWH receiving COT found that 43% had a high Current Opioid Misuse Measure (COMM) score, suggestive of aberrant use [11, 12]. Thus, pain management is by necessity a major responsibility of physicians who treat PLWH.
Human immunodeficiency virus (HIV) providers have recently reported low confidence in their abilities to treat pain and low satisfaction in delivering pain management [13, 14]. Providers concerned about the potential for opioid misuse may underprescribe opioids to PLWH [15, 16]. Conversely, providers who prioritize engagement in care may prescribe opioids despite risk of misuse [13, 17–21].
Interventions are needed to improve HIV providers’ quality of care regarding opioid prescribing to effectively treat pain and mitigate prescription opioid misuse. Guidelines for opioid prescribing encourage routine monitoring through urine drug tests (UDTs), pill counts, and prescription monitoring programs, as well as assessment for negative consequences such as misuse, diversion, and addiction [7, 21].
Collaborative care interventions, which have potential to improve clinical practice, are multidisciplinary, high-contact, and patient-centered, previously tested in the treatment of chronic diseases [22–26]. Few interventions have been tested to improve opioid prescribing; none, to our knowledge, focus on PLWH [27–29]. Thus, we adapted a previously developed intervention to create the Targeting Effective Analgesia in Clinics for HIV (TEACH) intervention [30]. We conducted a cluster randomized clinical trial to assess whether TEACH, compared to usual care, improves guideline-concordant care for COT among HIV providers and to assess for unintended consequences (eg, pain severity, detectable viral load).
METHODS
Study Objective and Design
The TEACH collaborative care intervention trial assessed the following outcomes: (1) HIV providers’ adherence to COT guidelines; (2) patient-level outcomes as a consequence of the exposure to the TEACH intervention; and (3) virologic control [31]. For HIV provider adherence to COT guidelines, the primary outcome was completion of ≥2 UDTs over 12 months. Secondary outcomes included completion of an opioid treatment agreement, routine use of the prescription monitoring program, and ≥3 primary care visits over 12 months. For patient-level consequences, the primary outcome was any early refills at 12 months. Secondary outcomes included pain severity and pain interference with activities of daily living measured by the Brief Pain Inventory [32], discontinuation of opioid prescriptions, number of early refills, COMM scores ≥9 and ≥13, past 30-day nonmedical opioid and stimulant use, and hazardous alcohol use (Alcohol Use Disorders Identification Test [AUDIT] score ≥8) [12, 33]. Exploratory outcomes concerning virologic control included HIV-1 RNA <200 copies/mL at 12 months and CD4 cell count. The TEACH study protocol was described previously [31], registered with ClinicalTrials.gov (NCT02564341), and approved by the study sites’ institutional review boards.
Provider and Patient Recruitment and Assessment
Recruitment
HIV physicians and advanced practice providers, henceforth referred to as “providers,” were recruited from September 2015 through December 2016 from 2 safety-net hospital–affiliated HIV clinics in Boston and Atlanta (Figure 1). Provider inclusion criteria were (1) being a provider at the HIV clinics and (2) being the main prescriber for 1 or more PLWH receiving COT. Provider exclusion criteria included being a study investigator or planning to leave the clinic within 9 months. Providers’ patients receiving COT, hereafter referred to as “patients,” were enrolled simultaneously with their provider through a waiver of informed consent if the following eligibility criteria were met: (1) age ≥18 years; (2) diagnosis of HIV infection; (3) received COT defined as ≥3 opioid prescriptions ≥21 days apart within a 6-month period in the prior year; and (4) at least 1 HIV clinic visit within the prior 18 months. The waiver of informed consent was pursued as it would not have been logistically possible to obtain informed consent from all patients on COT from a particular provider, and risk to patients was minimal.
Figure 1.
Consolidated Standards for Reporting Trials (CONSORT) diagram for the Targeting Effective Analgesia in Clinics for HIV (TEACH) intervention.
Assessment
Providers completed baseline and 12-month follow-up assessments including demographics, training, and practices for assessing and treating pain, substance use, and managing prescribed opioids [31]. Compensation was a $100 gift card upon each assessment completion. Data were extracted from the medical chart for providers’ patients receiving COT.
Cohort Recruitment and Assessment
Recruitment
Additionally, patients were recruited into an observational cohort of PLWH receiving COT to assess secondary outcomes (eg, pain severity, substance use) from July 2015 through December 2016, hereafter referred to as “observational cohort participants.” A research associate formally screened the patients and assessed additional eligibility criteria: (1) contact information of 2 individuals to assist with follow-up; (2) possession of a telephone; and (3) English speaking. Exclusion criteria included plans to move from the area within 12 months and inability to consent to interviews. Research staff obtained written informed consent.
Assessment
Observational cohort participants underwent 60- to 90-minute assessments by a research associate at baseline and 12 months [12, 31–36]. They were compensated with $35 for baseline and $50 for follow-up assessments.
Randomization
Providers were randomized to either the control group (ie, usual care) or the year-long TEACH intervention. Randomization was stratified by site and COT patient volume (1–2, 3–6, 7–11, and ≥12 eligible patients). To ensure balance with respect to the number of providers in each group, the permuted blocks strategy was used with an allocation ratio of 1:1 and blocks of 2. The random allocation sequence was generated by the study biostatistician and analyst, and was concealed until providers had completed the baseline assessment. Patients received the allocation assigned to their provider. Due to the nature of the intervention, neither providers nor research team were blinded to group assignment. Patients randomized to the intervention were informed that opioid monitoring procedures were being implemented in the clinic.
Control Condition
Providers in the control group received an informational brochure summarizing guidelines for COT and listing a web resource with electronic tools [31].
Intervention
The 12-month TEACH intervention consisted of 3 components: (1) a nurse care manager with an interactive electronic registry to manage patients; (2) opioid education and academic detailing; and (3) facilitated access to addiction specialists (Figure 2). TEACH was modeled on the TOPCARE (Transforming Opioid Prescribing in Primary Care) intervention, which is grounded in the Chronic Care Model for chronic disease management [29].
Figure 2.
Diagram of the Targeting Effective Analgesia in Clinics for HIV (TEACH) intervention. Abbreviations: AD, academic detailing; COT, chronic opioid therapy; NCM, nurse care manager. *An electronic system was used to manage patients.
Didactic Session.
Intervention providers received a 60-minute group didactic session from a study investigator expert on opioid prescribing, which consisted of developing communication skills as well as reviewing resources outlining opioid prescribing guidelines [30].
Collaboration With Nurse Care Manager.
Each site hired a nurse care manager (NCM) with a background in HIV care and interest in addiction. The NCMs collaborated closely with intervention providers to implement essential elements of guideline-driven care. They met with providers to review their COT patients. With provider assent, the NCMs conducted an initial intake assessment for each patient, which included the following: detailed medical, substance use, and social history; COMM score [12]; opioid risk tool [37]; and opioid treatment agreement. The NCM then followed up regularly with patients regarding refills, pain assessments, UDTs, and pill counts, the frequency being determined by patient risk level in collaboration with the provider. The NCM checked in with providers to discuss patient progress and challenges beyond the academic detailing sessions.
Interactive Electronic Registry.
The NCM utilized a HIPAA (Health Insurance Portability and Accountability Act)–compliant, web-based registry specifically built using user-based design methods to record and generate individual or aggregate information in real-time reports on opioid treatment agreements, UDTs, pill counts and checking prescription monitoring programs [38]. The registry enabled the NCM to manage refills, assess risk for opioid misuse, tailor monitoring strategies, and document clinical interactions.
Academic Detailing.
Providers participated in two 30-minute individual academic detailing sessions 2–3 months apart throughout the year-long intervention with an option of a third. At the sessions, providers received registry-generated reports on their COT patients to facilitate self-assessment of prescribing practices relative to patient risk levels [31]. Providers were encouraged to utilize academic detailing sessions to discuss challenging COT patients and develop treatment plans. These sessions helped providers develop effective communication to evaluate co-prescribed medications and to utilize monitoring tools such as UDTs.
Facilitated Access to Addiction Specialist.
The NCM encouraged and arranged referral of patients with active substance use disorders to addiction specialists affiliated with the HIV clinics. The study protocol did not specifically dictate how such patients were managed. The NCMs met separately with study team addiction specialists weekly to obtain input on challenging scenarios to be shared with providers.
Statistical Analysis
Descriptive statistics were calculated for provider-level and patient-level variables at baseline and 12-month follow-up by study arm. This study was conducted under the intention-to-treat principle, including all randomized participants according to their assignment. Randomization and the intervention occurred at the provider level while the unit of observation was either the individual providers’ patients receiving COT or the providers, depending on the outcome. The primary analysis evaluating the effect of the intervention on binary, patient-level outcomes used generalized estimating equation (GEE) logistic regression analyses to account for clustering by providers. The models used an exchangeable working correlation structure and results are reported based on robust empirical standard errors. The models included the randomization group as the main independent variable and also controlled for randomization stratification factors (ie, site and provider volume). Post hoc analyses were conducted using Poisson regression to estimate incidence rate ratios (IRRs). The secondary outcome, number of early refills, was analyzed using a GEE negative binomial regression model. The negative binomial model was selected instead of Poisson regression to allow for overdispersion in the data. We used multiple imputation (using 25 generated complete datasets) to account for the following missing outcome data: ≥2 UDTs (n = 6), any early refills (n = 5), ≥3 primary care visits (n = 4), opioid treatment agreement (n = 7), opioid discontinuation (n = 7), number of early refills (n = 7), CD4 cell count (n = 16), HIV viral load (n = 18), and self-reported outcomes (n = 9). Variables used for imputation of patient-level outcomes included gender, depressive symptoms (Center for Epidemiologic Studies Depression Scale [CES-D] ≥16 [39]), hazardous drinking (AUDIT score ≥8 [33]), any drug use, ethnicity (Hispanic: yes/no), randomization group, study site, provider patient volume stratification group, baseline value of the outcome, race (white, black, other/not available), and the continuous variables age, body mass index, CD4 cell count, HIV viral load, and Charlson comorbidity index [40]. For patient-level outcomes, data were imputed separately by provider and small clusters were combined in some cases due to computational issues. Variables used for imputation of provider-level outcomes included gender, years of practice in HIV care, patient volume stratification group, study site, randomization group, and baseline value of the outcome. Post hoc analyses stratified by study site were conducted for the primary outcomes.
A priori power calculations assumed a 2-sided test, with a significance level of .05. We anticipated that 35 total providers would be enrolled in the study, with an average of 5 eligible patients per provider for a total of 175 patients. We expected that the intraclass correlation coefficient (ICC) would be <0.10 for the outcomes of interest, and conservatively assumed a value of 0.10. The calculated ICCs across all imputed datasets were ≤0.01 for both primary outcomes. Calculations were based on a χ 2 test with continuity correction, with estimates adjusted for clustering based on the inflation factor and assuming 10% loss to follow-up. We calculated that the study would provide 80% power to detect an absolute difference of 28% for the primary outcome ≥2 UDTs and an absolute difference of 27% for the primary outcome any early refills. Analyses were conducted using SAS version 9.4 software (SAS Institute).
RESULTS
Provider Characteristics
Across both sites, 32 physicians and 9 advanced practice providers (n = 41) were eligible, and all chose to participate. Of these, 21 were randomized to the TEACH intervention and 20 to the control group (Table 1). We did not identify any important differences in provider characteristics across study arms. Providers’ characteristics included a mean age of 46 years (standard deviation [SD], 11 years); 63% were female; 63% were white and 10% Hispanic. Among physicians, 16% (5/32) were buprenorphine waivered. The mean number of patients per provider on COT was 5 (SD, 6); 85% had 1–6 patients on COT.
Table 1.
Baseline Provider Characteristics in the Targeting Effective Analgesia in Clinics for HIV (TEACH) Study, by Study Arm and Overall
| Characteristic | Intervention Arm (n = 21) | Control Arm (n = 20) | Overall (N = 41) |
|---|---|---|---|
| Age, y, mean (SD) | 45.0 (11.5) | 46.1 (11.7) | 45.5 (11.5) |
| Female, No. (%) | 12 (57.1) | 14 (70.0) | 26 (63.4) |
| Race/ethnicity, No. (%) | |||
| White | 12 (57.1) | 14 (70.0) | 26 (63.4) |
| African American | 2 (9.5) | 2 (10.0) | 4 (9.8) |
| Asian | 4 (19.0) | 3 (15.0) | 7 (17.1) |
| >1 race | 3 (14.3) | 0 (0.0) | 3 (7.3) |
| Other | 0 (0.0) | 1 (5.0) | 1 (2.4) |
| Hispanic | 2 (9.5) | 2 (10.0) | 4 (9.8) |
| Professional title, No. (%) | |||
| MD | 17 (81.0) | 15 (75.0) | 32 (78.0) |
| Advanced Practice Provider | 4 (19.0) | 5 (25.0) | 9 (22.0) |
| Boston site, No. (%) | 6 (28.6) | 5 (25.0) | 11 (26.8) |
| Buprenorphine waivered, No. (%) | |||
| Yes | 2 (9.5) | 3 (15.0) | 5 (12.2) |
| No | 15 (71.4) | 12 (60.0) | 27 (65.9) |
| Not a physiciana | 4 (19.0) | 5 (25.0) | 9 (22.0) |
| No. of patients on COT, mean (SD) | 4.10 (4.12) | 5.00 (6.69) | 4.54 (5.47) |
| Patients on COT, No. (%) | |||
| 1–2 | 9 (42.9) | 9 (45.0) | 18 (43.9) |
| 3–6 | 9 (42.9) | 8 (40.0) | 17 (41.5) |
| 7–11 | 1 (4.8) | 0 (0.0) | 1 (2.4) |
| ≥12 | 2 (9.5) | 3 (15.0) | 5 (12.2) |
Abbreviations: COT, chronic opioid therapy; SD, standard deviation.
aAt the time of the study, only physicians could obtain a waiver to prescribe buprenorphine.
Patient Characteristics
Providers had a total of 187 patients receiving COT (Table 2). Intervention providers had 87 patients, control providers 100. We did not observe important differences in patient characteristics across study arms. Of the 187 patients, 117 were enrolled in the observational patient cohort, and 114 completed a baseline assessment. Of the 114, 58 were in the intervention arm and 56 in the control arm. Patient characteristics included mean age, 54 years (SD, 9 years); 28% were female; 28% were white and 8% Hispanic. Observational cohort participants had similar demographic characteristics to patients of randomized providers: mean age, 53 years (SD, 8 years); 34% female; and 23% white and 8% Hispanic.
Table 2.
Baseline Patient Characteristics in the Targeting Effective Analgesia in Clinics for HIV (TEACH) Study, by Study Arm and Overall
| Characteristic | Intervention (n = 87) | Control (n = 100) | Overall (N = 187) |
|---|---|---|---|
| Age, y, mean (SD) | 54.4 (8.0) | 53.5 (9.2) | 53.9 (8.7) |
| Female, No. (%) | 25 (28.7) | 28 (28.0) | 53 (28.3) |
| Race/ethnicity, No. (%) | |||
| White | 26 (29.9) | 27 (27.0) | 53 (28.3) |
| Black | 56 (64.4) | 67 (67.0) | 123 (65.8) |
| Other/not available | 5 (5.7) | 6 (6.0) | 11 (5.9) |
| Hispanic, No. (%) | 7 (8.0) | 8 (8.0) | 15 (8.0) |
| Sexual orientationa, No. (%) | |||
| Straight/heterosexual | 38 (65.5) | 37 (66.1) | 75 (65.8) |
| Gay/lesbian/queer/homosexual | 14 (24.1) | 16 (28.6) | 30 (26.3) |
| Bisexual | 5 (8.6) | 3 (5.4) | 8 (7.0) |
| Other | 1 (1.7) | 0 (0.0) | 1 (0.9) |
| High school graduatea, No. (%) | 37 (63.8) | 39 (69.6) | 76 (66.7) |
| Housing, No. (%)a | |||
| Own or rent home/apartment | 47 (81.0) | 49 (87.5) | 96 (84.2) |
| Staying at home of family member(s) | 6 (10.3) | 3 (5.4) | 9 (7.9) |
| In a rooming, boarding, or halfway house | 1 (1.7) | 4 (7.1) | 5 (4.4) |
| Other | 4 (6.9) | 0 (0.0) | 4 (3.5) |
| Jail or prison (past 12 mo)a, No. (%) | 4 (6.9) | 6 (10.7) | 10 (8.8) |
| HIV transmission routea, No. (%) | |||
| MSM/IDU | 3 (5.2) | 3 (5.4) | 6 (5.3) |
| MSM only | 16 (27.6) | 13 (23.2) | 29 (25.4) |
| IDU only | 6 (10.3) | 5 (8.9) | 11 (9.6) |
| Presumed heterosexual + blood/blood products | 5 (8.6) | 6 (10.7) | 11 (9.6) |
| Presumed heterosexual only | 28 (48.3) | 29 (51.8) | 57 (50.0) |
| Undetectable viral load, No. (%) | 81 (95.3) | 85 (86.7) | 166 (90.7) |
| Ever injected drugs b, No. (%) | 14 (24.6) | 9 (18.8) | 23 (21.9) |
Abbreviations: HIV, human immunodeficiency virus; IDU, injection drug use; MSM, men who have sex with men; SD, standard deviation.
aSubset of patients with n = 114 at baseline, n = 105 at follow-up.
bOnly asked at follow-up.
RCT Outcomes
Provider Outcomes
The primary provider-level outcome was completion of ≥2 UDTs in the period between baseline and follow-up. Intervention patients had 13 times the odds of completion of ≥2 UDTs (71% vs 20%; adjusted odds ratio [AOR], 13.38 [95% confidence interval {CI}, 5.85–30.60]) compared to control patients (Table 3). For descriptive purposes, we note that the median (25th, 75th percentiles) for number of UDTs was 3 (1, 5) and 0 (0, 1) for the intervention and control groups, respectively. Post hoc analyses using Poisson regression resulted in an adjusted IRR of 4.76 (95% CI, 2.68–8.48) for having a UDT. The TEACH intervention had higher odds of having the secondary outcome, completion of a signed opioid treatment agreement (AOR, 61.50 [95% CI, 15.30–247.20]). No statistically significant difference was found between arms for routinely using the prescription monitoring program or having ≥3 primary care visits in the study year.
Table 3.
Targeting Effective Analgesia in Clinics for HIV (TEACH) Study Outcomes at 12 Monthsa
| Outcome | Intervention (n = 87) | Control (n = 100) | AOR (95% CI)b |
P Value |
|---|---|---|---|---|
| Provider adherence to COT guidelines | ||||
| ≥2 urine drug tests over 12 moc | 71.0% | 19.9% | 13.38 (5.85–30.60) | <.0001 |
| ≥3 primary care visits over 12 mo | 80.2% | 76.0% | 1.33 (.63–2.84) | .45 |
| Opioid treatment agreement | 75.6% | 12.8% | 61.50 (15.30–247.20) | <.0001 |
| Provider routinely consulted prescription monitoring programd | 71.4% | 45.0% | 3.85 (.99–14.93) | .05 |
| Patient-level outcomes | ||||
| ≥1 early refill over 12 moc | 21.6% | 30.4% | 0.55 (.26–1.15) | .11 |
| No. of early refills over 12 mo, mean (SD) | 0.46 (1.00) | 0.60 (1.14) | 0.64 (.32–1.30)e | .21 |
| COMM score ≥9f | 38.7% | 39.9% | 0.78 (.33–1.88) | .58 |
| COMM score ≥13f | 27.6% | 19.4% | 1.68 (.59–4.77) | .33 |
| Discontinuation of opioid prescriptions (MEDD of 0 in the 60 d prior to visit) | 30.3% | 25.6% | 1.62 (.82–3.22) | .17 |
| Pain severity (BPI), mean (SD)f | 6.30 (2.87) | 5.76 (2.87) | 0.10 (−1.56 to 1.75)g | .91 |
| Pain interference (BPI), mean (SD)f | 5.70 (2.98) | 4.99 (3.58) | 0.30 (−1.34 to 1.95)g | .72 |
| Past 30-d opioid/stimulant used,f | 7.5% | 12.0% | 0.65 (.15–2.79) | .57 |
| Hazardous alcohol use (AUDIT score ≥8)f | 17.9% | 10.4% | 1.03 (.21–5.02) | .97 |
| HIV virologic control | ||||
| Undetectable viral load (<200 copies/mL) | 86.9% | 82.1% | 1.21 (.47–3.09) | .69 |
| CD4 count, cells/μL, mean (SD) | 568.67 (357.43) | 500.48 (316.21) | 70.25 (−46.76 to 187.25)g | .24 |
Abbreviations: AOR, adjusted odds ratio; AUDIT, Alcohol Use Disorders Identification Test; BPI, Brief Pain Inventory; CI, confidence interval; COMM, Current Opioid Misuse Measure; COT, chronic opioid therapy; HIV, human immunodeficiency virus; MEDD, Morphine Equivalent Daily Dose; SD, standard deviation.
a Table presents unadjusted proportions and means.
bAdjusted for stratification variables site (Boston vs Atlanta) and patient volume (1–2, 3–6, 7–11, and ≥12 patients).
cPrimary outcome bolded.
dUnadjusted due to small numbers.
eAdjusted incidence rate ratio.
fSubset of patients, n = 114.
gAdjusted mean difference between arms.
Patient Outcomes
The primary patient-level outcome was 1 or more early refill. We did not detect a statistically significant intervention effect on any early refills (22% vs 30%; AOR, 0.55 [95% CI, .26–1.15]) or number of early refills (secondary). For descriptive purposes, we note that the median (25th, 75th percentiles) for number of early refills was 0 (0, 0) and 0 (0, 1) for the intervention and control groups, respectively. Post hoc analyses using Poisson regression resulted in an adjusted IRR of 0.64 (95% CI, .34–1.22). No significant differences between arms were found for the secondary outcomes: having a COMM score ≥9 or ≥13; pain sensitivity; pain interference; past 30-day opioid or stimulant use; hazardous alcohol use, or discontinuing opioids at 12-month follow-up.
Virologic Control
An exploratory outcome was the intervention’s impact on virologic control among PLWH receiving COT. We found no significant differences in undetectable HIV viral load (<200 copies/mL) (86.9% vs 82.1%; AOR, 1.21 [95% CI, .47–3.09]) or mean CD4 count (569 vs 500 cells/μL; adjusted mean difference, 70.25 [95% CI, −46.76 to 187.25]).
In post hoc analyses stratified by site, conclusions for the primary outcome, completion of ≥2 UDTs, were consistent with the overall analysis (Atlanta: AOR, 12.77 [95% CI, 4.81–33.87]; Boston: AOR, 131.79 [95% CI, 10.09–1721.18]). For the outcome any early refills, the intervention appeared to result in reduced odds of having any early refills in Atlanta (AOR, 0.34 [95% CI, .13–.89]) but not Boston (AOR, 4.03 [95% CI, .78–20.89]).
Discussion
This cluster randomized clinical trial tested the TEACH intervention to improve opioid prescribing among providers treating PLWH on COT. Patients in the TEACH arm had a statistically significant improvement of 2 recommended features of appropriate care for patients on COT: completion of ≥2 UDTs and opioid treatment agreements.
The TEACH intervention significantly improved adherence to CDC and HIV Medicine Association–recommended guidelines [6, 7, 21], and yet did not appear to worsen potential patient adverse unintended consequences: early refills, increased pain levels, more severe substance use, and COMM scores. Of note, we did not detect differences in discontinuation of opioids between groups. By creating a collaborative clinical role for a dedicated NCM, these challenging aspects of achieving guideline-concordant HIV clinical care for patients receiving COT were successfully implemented.
Adhering to CDC guidelines for opioid prescribing requires additional elements of care to be delivered to patients, and this added work can be overwhelming for providers to implement routinely [14]. However, team-based care of chronic pain with an embedded nurse care manager with explicit responsibilities related to COT management can augment safe opioid prescribing without adding major provider tasks. A potential concern regarding TEACH intervention implementation was that increased monitoring might result in adverse patient outcomes. This, however, was not observed among secondary outcomes or the exploratory outcomes of CD4 cell count or viral suppression.
Before this study was implemented, few initiatives to change COT prescribing practices were being implemented at the study sites. However, after the trial’s completion, checking the prescription monitoring program became mandatory in Massachusetts and Georgia.
This study had limitations. Throughout the study period, national attention on the opioid epidemic increased. It is possible that providers in the control group were exposed to an unanticipated level of attention regarding COT and consequently changed their prescribing practices. Such exposure might result in an underestimation of the effect of the TEACH intervention. However, given the marked differences in some outcomes, this is not a major limitation. Due to the lack of blinding, contamination of the control arm as a consequence of communication between enrolled providers may have occurred, which should bias toward the null. Not all patients receiving COT were enrolled in the observational patient cohort, potentially affecting power to detect differences in self-reported outcomes. While examining both Boston and Atlanta provided more generalizability than in a single-site study, they do not represent all HIV care in the United States.
In conclusion, the TEACH intervention was more effective than usual care in improving guideline concordant care for chronic opioid therapy, as measured by completion of ≥2 UDTs and opioid treatment agreements among PLWH. No significant differences were observed for early refills or viral suppression. The TEACH intervention is a promising strategy to address chronic opioid therapy in HIV care.
Notes
Author contributions. J. H. S., C. d. R., J. I. T., M. C. L., C. L. B., J. M. L., A. Y. W., and D. M. C. were responsible for the conception and design of the study. J. H. S., J. I. T., J. M. L., C. L. B., M. C. L., and C. W. S. were responsible for the design of the intervention, and J. M. L., C. W. S., A. Y. W., M. M. S., J. A. C., M. C. L., C. R., K. O., and C. A. implemented the intervention. M. C. L., C. L. B., and C. R. coordinated study activities. K. O., and C. H. collected data, and D. M. C. and L. S. F. managed study data and analyses. J. H. S. and M. C. L. drafted the manuscript. All authors read, revised, and approved the final manuscript. J. H. S. had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Acknowledgments. The TEACH study was approved by the institutional review boards at Boston University Medical Campus and Emory University and the Grady Memorial Hospital Research Oversight Committee. The authors thank the HIV providers and their patients who enrolled in this study. We are grateful to Linda Rosen at Boston Medical Center’s Clinical Data Warehouse and Minh Nguyen and Jeselyn Rhodes of the Emory Center for AIDS Research for their assistance with data extraction, and to Caroline Dames at Boston Medical Center for assistance with editing and formatting the manuscript.
Disclaimer. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute on Drug Abuse (NIDA), the National Institute of Allergy and Infectious Diseases, or the National Institutes of Health (NIH).
Financial support. This work was supported by the National Institutes of Health through (grant number R01DA037768 from the National Institute on Drug Abuse); the Emory Center for AIDS Research (grant number P30AI050409 from the National Institute of Allergy and Infectious Diseases); and the Boston/Providence Center for AIDS Research (grant number P30AI042853 from the National Institute of Allergy and Infectious Diseases).
Potential conflicts of interest. D. M. C. serves on data and safety monitoring boards for Janssen Research and Development. C. L. B. reports an NIH grant to Boston Medical Center, during the conduct of the study. All other authors report no potential conflicts of interest. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
References
- 1.Volkow N, Benveniste H, McLellan AT. Use and misuse of opioids in chronic pain. Annu Rev Med 2018; 69:451–65. [DOI] [PubMed] [Google Scholar]
- 2.Williams AR, Bisaga A. From AIDS to opioids—how to combat an epidemic. N Engl J Med 2016; 375:813–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Stitzer ML, Schwartz RP, Bigelow GE. Prescription opioids: new perspectives and research on their role in chronic pain management and addiction. Drug Alcohol Depend 2017; 173(Suppl 1):1–3. [DOI] [PubMed] [Google Scholar]
- 4.Compton WM, Jones CM, Baldwin GT. Relationship between nonmedical prescription-opioid use and heroin use. N Engl J Med 2016; 374:154–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Substance Abuse and Mental Health Services Administration (SAMHSA). Drug Abuse Warning Network, 2011: national estimates of drug-related emergency department visits. Vol. HHS Publication No. (SMA) 13–4760, DAWN Series D-39. Rockville, MD: SAMHSA, 2013. [Google Scholar]
- 6.Dowell D, Haegerich TM, Chou R. CDC guideline for prescribing opioids for chronic pain—United States, 2016. MMWR Recomm Rep 2016; 65:1–49. [DOI] [PubMed] [Google Scholar]
- 7.Chou R, Fanciullo GJ, Fine PG, et al. Clinical guidelines for the use of chronic opioid therapy in chronic noncancer pain. J Pain 2009; 10: 113– 30.e22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Merlin JS, Cen L, Praestgaard A, et al. Pain and physical and psychological symptoms in ambulatory HIV patients in the current treatment era. J Pain Symptom Manage 2012; 43:638–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Jiao JM, So E, Jebakumar J, George MC, Simpson DM, Robinson-Papp J. Chronic pain disorders in HIV primary care: clinical characteristics and association with healthcare utilization. Pain 2016; 157: 931– 7. [DOI] [PubMed] [Google Scholar]
- 10.Vijayaraghavan M, Penko J, Bangsberg DR, Miaskowski C, Kushel MB. Opioid analgesic misuse in a community-based cohort of HIV-infected indigent adults. JAMA Intern Med 2013; 173:235–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Colasanti J, Lira MC, Cheng DM, et al. Chronic opioid therapy in HIV-infected patients: patients’ perspectives on risks, monitoring, and guidelines. Clin Infect Dis 2019; 68:291–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Butler SF, Budman SH, Fernandez KC, et al. Development and validation of the Current Opioid Misuse Measure. Pain 2007; 130:144–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Tsui JI, Walley AY, Cheng DM, et al. Provider opioid prescribing practices and the belief that opioids keep people living with HIV engaged in care: a cross-sectional study. AIDS Care 2019; 31:1140–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Carroll JJ, Colasanti J, Lira MC, Del Rio C, Samet JH. HIV physicians and chronic opioid therapy: it’s time to raise the bar. AIDS Behav 2019; 23:1057–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Larue F, Fontaine A, Colleau SM. Underestimation and undertreatment of pain in HIV disease: multicentre study. BMJ 1997; 314:23–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Breitbart W, Rosenfeld B, Passik S, Kaim M, Funesti-Esch J, Stein K. A comparison of pain report and adequacy of analgesic therapy in ambulatory AIDS patients with and without a history of substance abuse. Pain 1997; 72:235–43. [DOI] [PubMed] [Google Scholar]
- 17.Vijayaraghavan M, Penko J, Guzman D, Miaskowski C, Kushel MB. Primary care providers’ views on chronic pain management among high-risk patients in safety net settings. Pain Med 2012; 13:1141–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Lum PJ, Little S, Botsko M, et al. , BHIVES Collaborative . Opioid-prescribing practices and provider confidence recognizing opioid analgesic abuse in HIV primary care settings. J Acquir Immune Defic Syndr 2011; 56(Suppl 1):S91–7. [DOI] [PubMed] [Google Scholar]
- 19.Canan CE, Chander G, Monroe AK, et al. , HIV Research Network . High-risk prescription opioid use among people living with HIV. J Acquir Immune Defic Syndr 2018; 78:283–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Gaither JR, Goulet JL, Becker WC, et al. Guideline-concordant management of opioid therapy among human immunodeficiency virus (HIV)-infected and uninfected veterans. J Pain 2014; 15:1130–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Bruce RD, Merlin J, Lum PJ, et al. 2017 HIVMA of IDSA clinical practice guideline for the management of chronic pain in patients living with HIV. Clin Infect Dis 2017; 65:e1–37. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Watkins KE, Ober AJ, Lamp K, et al. Collaborative care for opioid and alcohol use disorders in primary care: the SUMMIT randomized clinical trial. JAMA Intern Med 2017; 177:1480–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Dobscha SK, Corson K, Perrin NA, et al. Collaborative care for chronic pain in primary care: a cluster randomized trial. JAMA 2009; 301:1242–52. [DOI] [PubMed] [Google Scholar]
- 24.Katon WJ, Lin EH, Von Korff M, et al. Collaborative care for patients with depression and chronic illnesses. N Engl J Med 2010; 363:2611–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Saitz R, Cheng DM, Winter M, et al. Chronic care management for dependence on alcohol and other drugs: the AHEAD randomized trial. JAMA 2013; 310:1156–67. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Alford DP, LaBelle CT, Kretsch N, et al. Collaborative care of opioid-addicted patients in primary care using buprenorphine: five-year experience. Arch Intern Med 2011; 171:425–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Wakeland W, Nielsen A, Schmidt TD, et al. Modeling the impact of simulated educational interventions on the use and abuse of pharmaceutical opioids in the United States: a report on initial efforts. Health Educ Behav 2013; 40:74S–86S. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Wiedemer NL, Harden PS, Arndt IO, Gallagher RM. The opioid renewal clinic: a primary care, managed approach to opioid therapy in chronic pain patients at risk for substance abuse. Pain Med 2007; 8:573–84. [DOI] [PubMed] [Google Scholar]
- 29.Bodenheimer T, Wagner EH, Grumbach K. Improving primary care for patients with chronic illness. JAMA 2002; 288:1775–9. [DOI] [PubMed] [Google Scholar]
- 30.Liebschutz JM, Xuan Z, Shanahan CW, et al. Improving adherence to long-term opioid therapy guidelines to reduce opioid misuse in primary care: a cluster-randomized clinical trial. JAMA Intern Med 2017; 177:1265–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Lira MC, Tsui JI, Liebschutz JM, et al. Study protocol for the targeting effective analgesia in clinics for HIV (TEACH) study—a cluster randomized controlled trial and parallel cohort to increase guideline concordant care for long-term opioid therapy among people living with HIV. HIV Res Clin Pract 2019; 20:48–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Keller S, Bann CM, Dodd SL, Schein J, Mendoza TR, Cleeland CS. Validity of the brief pain inventory for use in documenting the outcomes of patients with noncancer pain. Clin J Pain 2004; 20:309–18. [DOI] [PubMed] [Google Scholar]
- 33.Babor TF, Higgins-Biddle JC, Saunders JB, Monteiro MG.. AUDIT: the Alcohol Use Disorders Identification Test: guidelines for use in primary care. 2nd ed. Geneva, Switzerland: World Health Organization, 2001. [Google Scholar]
- 34.McLellan AT, Luborsky L, O’Brien C, Woody GE. An improved diagnostic instrument for substance abuse patients, the Addition Severity Index. J Nerv Ment Dis 1980; 168: 26–33. [DOI] [PubMed] [Google Scholar]
- 35.Institute of Behavioral Research. TCU drug screen V.2014. Fort Worth: Texas Christian University, Institute of Behavioral Research. Available at: ibr.tcu.edu. doi:10.1080/10509674.2018.1549180
- 36.Meltzer EC, Rybin D, Saitz R, et al. Identifying prescription opioid use disorder in primary care: diagnostic characteristics of the Current Opioid Misuse Measure (COMM). Pain 2011; 152:397–402. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Webster LR, Webster RM. Predicting aberrant behaviors in opioid-treated patients: preliminary validation of the opioid risk tool. Pain Med 2005; 6:432–42. [DOI] [PubMed] [Google Scholar]
- 38.Gustafson DH Jr, Maus A, Judkins J, et al. Using the NIATx model to implement user-centered design of technology for older adults. JMIR Hum Factors 2016; 3:e2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Lewinsohn PM, Seeley JR, Roberts RE, Allen NB. Center for Epidemiologic Studies Depression Scale (CES-D) as a screening instrument for depression among community-residing older adults. Psychol Aging 1997; 12:277–87. [DOI] [PubMed] [Google Scholar]
- 40.Charlson ME. Adapting a clinical comorbidity index for use with ICD-9-CM administrative data: a response. J Clin Epidemiol 1993; 46:1083–4. [DOI] [PubMed] [Google Scholar]