The Design and Baseline Characteristics for the HOPE Consortium Trial to Reduce Pain and Opioid Use in Hemodialysis

Abstract

The HOPE Consortium Trial to Reduce Pain and Opioid Use in Hemodialysis (HOPE Trial) is a multicenter randomized trial addressing chronic pain among patients receiving maintenance hemodialysis for end-stage kidney disease. The trial uses a sequential, multiple assignment design with a randomized component for all participants (Phase 1) and a non-randomized component for a subset of participants (Phase 2). During Phase 1, participants are randomized to Pain Coping Skills Training (PCST), an intervention designed to increase self-efficacy for managing pain, or Usual Care. PCST consists of weekly, live, coach-led cognitive behavioral therapy sessions delivered by video- or tele-conferencing for 12 weeks followed by daily interactive voice response sessions delivered by telephone for an additional 12 weeks. At 24 weeks (Phase 2), participants in both the PCST and Usual Care groups taking prescription opioid medications at an average dose of ≥20 morphine milligram equivalents per day are offered buprenorphine, a partial opioid agonist with a more favorable safety profile than full-agonist opioids. All participants are followed for 36 weeks. The primary outcome is pain interference ascertained at 12 weeks for the primary analysis. Secondary outcomes include additional patient-reported measures and clinical outcomes including falls, hospitalizations, and death. Exploratory outcomes include acceptability, tolerability, and efficacy of buprenorphine. The enrollment target of 640 participants was met 27 months after trial initiation. The findings of the trial will inform the management of chronic pain, a common and challenging issue for patients treated with maintenance hemodialysis. NCT04571619

Introduction

People with end-stage kidney disease (ESKD) undergoing treatment with maintenance hemodialysis have a high burden of symptoms related to kidney failure, comorbid conditions, and the dialysis procedure itself. Pain is among the most common of these symptoms with approximately 60% of patients reporting that they have pain, and many describing the pain as moderate or severe¹. Altered pharmacokinetic and pharmacodynamic properties of analgesic agents in the setting of kidney failure make pain management challenging for this patient population^2,3. Because adverse effects, accumulation of toxic metabolites, and sensitivity to sedating effects with analgesic medications are more significant in the presence of kidney failure, medications commonly used for management of chronic pain in other populations are either contraindicated or need to be administered at markedly reduced dosages among people receiving maintenance hemodialysis. In part because of limited analgesic options, clinicians often prescribe opioid medications for patients with kidney failure. An analysis of data from 2010 indicated that 20% of patients treated with maintenance hemodialysis were receiving long-term opioid therapy (LTOT) defined as prescribed opioids for 90 days or longer⁴. Among hemodialysis patients, LTOT is associated with higher rates of falls, hip fractures, hospitalizations, dialysis withdrawal, and death^4,5. Additionally, patients treated with hemodialysis generally have low levels of physical activity and high rates of insomnia and depression^4,6–8, each of which can exacerbate chronic pain perception, complicate its management, and be potentiated by LTOT⁹.

Non-pharmacologic approaches for management of chronic pain such as cognitive behavioral therapy, mindfulness, and acceptance and commitment therapy have demonstrated efficacy for chronic pain in the general population but have not been adequately studied in ESKD^10,11. Buprenorphine, a partial opioid agonist with a better safety profile compared with other opioids, is an effective analgesic agent that has been used to support discontinuation of full agonist opioids^12–14, but there is limited experience with this drug in the ESKD population². Thus, the primary objective of the HOPE Consortium Trial to Reduce Pain and Opioid Use in Hemodialysis (HOPE trial) is to evaluate the effectiveness of a cognitive behavioral therapy-based intervention called Pain Coping Skills Training (PCST) for reducing pain interference and other pain-related outcomes among people receiving maintenance hemodialysis who have chronic pain. A secondary objective is to explore acceptability, tolerability, and efficacy of switching from full agonist opioid medication to buprenorphine among patients with LTOT. This paper describes the design of the trial and provides the baseline characteristics of the randomized participants.

Methods

Hypothesis and Design

The primary hypothesis for the trial is that PCST will improve pain outcomes compared with usual clinical care. The design, as depicted in Figure 1, is a multicenter, sequential, multiple assignment trial with a randomized component (Phase 1) and a non-randomized component (Phase 2). In Phase 1 (Weeks 1 – 24), participants are randomized in equal proportions to the PCST group or the Usual Care group. The PCST intervention has two sequential components, each taking place over 12 weeks: 1) weekly video- or telephone-conferencing sessions with a PCST coach, followed by 2) daily interactive voice response (IVR) sessions via telephone. The primary outcome is pain interference ascertained at Week 12 which coincides with the end of the PCST weekly coaching sessions. For Phase 2, participants in both randomized groups are assessed at Week 24 for eligibility for a second intervention, buprenorphine, as an approach to addressing physical dependence on prescribed opioids. The principal determinant of eligibility for the buprenorphine intervention is use of prescription full agonist opioid medication at an average dose of at least 20 morphine milligram equivalents (MME) per day. Participants eligible for buprenorphine are encouraged to switch from their full agonist opioid medication to buprenorphine. The initial design of the buprenorphine phase of the trial included randomization of eligible participants to either a Buprenorphine Group in which switching from the full agonist opioid to buprenorphine was recommended, or to a No Buprenorphine Group in which buprenorphine was not offered. However, because the number of participants meeting the full agonist opioid MME threshold for eligibility was lower than anticipated, approximately one year into the conduct of the trial, and with the approval of the Data and Safety Monitoring Board (DSMB), randomization was eliminated for the Phase 2 component such that all subsequent participants meeting the eligibility criteria would be given the opportunity to switch to buprenorphine.

Figure 1. Trial Design.

The trial uses a sequential multiple assignment design with a randomized component (Phase 1) and a non-randomized component (Phase 2). Eligible patients are randomized in equal proportions to the Phase 1 intervention: Pain Coping Skills Training (PCST) or Usual Care for 24 weeks. The PCST intervention has two components each taking place over 12-week periods: 1) weekly telehealth sessions with a PCST coach, and 2) daily interactive voice response (IVR) sessions via telephone. At Week 24, participants in both the PCST and Usual Care groups who are taking opioids at an average dose of ≥20 morphine milligram equivalents (MME)/day are assessed for eligibility for the Phase 2 intervention. Participants who meet eligibility criteria for the Phase 2 intervention are encouraged to switch from their current full agonist opioid medication to the partial opioid agonist, buprenorphine. Participants who do not meet the Phase 2 eligibility criteria are not offered buprenorphine. Follow-up from Week 24 until Week 36 continues for all participants. Outcomes are ascertained at: 1) Week 12 to determine the effect of the coach-led component of PCST (primary analysis), 2) Week 24 to determine whether IVR enhances the effect of coach-led component of PCST, and 3) Week 36 to assess the durability of PCST and to explore the effects of buprenorphine.

Setting

The trial is being conducted by research teams at 8 Clinical Centers comprising a total of 16 enrolling sites, and a Scientific and Data Research Center (SDRC) as shown in Figure 2. Participants were enrolled from 103 outpatient dialysis facilities affiliated with, or in geographic proximity to, the enrolling sites. The dialysis facilities are operated by national or regional dialysis provider organizations, the Veterans Health Administration, academic medical centers, or independent nephrology practices, each of which provided approval for conduct of the trial within their facilities.

Figure 2. HOPE Consortium.

The consortium consists of 8 Clinical Centers that include 16 enrolling sites and multiple affiliated outpatient dialysis units, the Scientific and Data Research Center, the Patient Advisors, and the NIDDK project scientists. Oversight is provided by two institutional review boards, an independent data and safety monitoring board, and the Food and Drug Administration. Abbreviations: SDRC, Scientific and Data Research Center; IRB, institutional review board; VA, Veterans Administration, FDA, United States Food and Drug Administration.

Involvement of Patients in the Design and Conduct of the Trial

A patient advisory group composed of 11 individuals with either current treatment with maintenance hemodialysis or previous experience with hemodialysis before receiving a kidney transplant was assembled by the Clinical Centers and the SDRC prior to protocol development. The patient advisors and investigators collaboratively developed the protocol, visual materials for the PCST intervention, informed consent documents, and recruitment materials including a recruitment video¹⁵. The patient advisory group has continued to be highly engaged during the conduct of the study with representation on consortium committees including the Steering Committee, Executive Committee, Recruitment and Retention Committee, and Quality Control Committee. Additionally, the patient advisors meet monthly and participate in the meetings with the DSMB.

Eligibility Criteria

The eligibility criteria for trial participation are provided in Table 1. The requirement for receipt of maintenance hemodialysis for at least 90 days is intended to ensure that participants have irreversible kidney failure rather than acute kidney injury and to avoid the medical and emotional complexity that often characterizes the early period after dialysis initiation¹⁶. Incorporation of score threshold of ≥4 for the 3-item Pain, Enjoyment of Life and General Activity (PEG) scale reduces the likelihood of a floor effect for the primary outcome of pain interference¹⁷. Excluding individuals with substance use disorder is consistent with the focus of the trial on chronic pain rather than on opioid use disorder. The trial excludes individuals being treated with home dialysis in order to evaluate the intervention in the setting in which it will most often be applied clinically. Eligibility for the Phase 2 buprenorphine intervention (Table 1) includes a minimum average daily dose of a full agonist opioid of 20 MME because the potential benefits of converting to buprenorphine are less clear with use of lower doses of full agonist opioids^18,19.

Table 1.

Eligibility Criteria

Trial Eligibility Criteria
Inclusion Criteria	Exclusion Criteria
Age ≥18 years	Current opioid use disorder or other non-tobacco substance use disorder
Undergoing in-center maintenance hemodialysis for ≥90 days	Current use of heroin
English- or Spanish-speaking	Current use of methadone, buprenorphine, or naltrexone for opioid use disorder
Chronic pain1	Current receipt of hospice care
PEG score ≥4	Cognitive impairment precludes trial participation
Willing to provide informed consent	Active suicidal intent2
Willing to allow research team to obtain opioid pharmacy refill data	Unstable bipolar disorder, schizophrenia, post-traumatic stress disorder, or other psychotic disorder
Willing to allow research team to contact and work with their opioid prescriber if taking opioids	Life expectancy <6 months
	Anticipated kidney transplantation, transfer to another dialysis facility, or transition to home dialysis within 6 months
	Current incarceration
	Any other condition that the investigator considers as precluding participation
Buprenorphine Eligibility Criteria
Inclusion Criteria	Exclusion Criteria
Current use of prescribed opioids at an average dosage of ≥20 MME/day	Current active addiction to substances other than tobacco, defined as probable moderate-severe substance use disorder for which evidence-based addiction treatment is indicated
For participants of childbearing potential, a negative serum pregnancy test and willingness to use an effective form of contraception or remain abstinent from sexual intercourse with partners of the opposite sex during the buprenorphine administration period	ALT, AST, or total bilirubin >2X the upper limit of normal
	History of long QT syndrome, family history of long QT syndrome, or use of Class IA or Class III anti-arrhythmic medications3
	Known allergy or prior intolerance to buprenorphine
	Current use of buprenorphine

¹Chronic pain is defined as a response of “most days” or “every day” to the following question: “In the past 3 months, how often have you had pain?” Answer options: never, some days, most days, every day

²Based on an initial screening with the PHQ-9 question #9 followed by positive findings on further evaluation

³Class 1A anti-arrhythmic medications include quinidine, procainamide, and disopyramide. Class III anti-arrhythmic medications include amiodarone, sotalol, dofetilide, and dronedarone

Abbreviations: PEG, Pain, Activity and Enjoyment of Life Scale; MME, morphine milligram equivalents; ALT, alanine transaminase; AST, aspartate aminotransferase.

Approach to Enrolling Participants

To leverage the efficiencies of enrolling patients in the outpatient dialysis setting, all patients being treated at a participating dialysis facility were invited to complete a brief pre-screening survey to identify those who should be approached for eligibility screening. The survey, administered after obtaining verbal consent, includes one question about the frequency of pain and the 3-item PEG scale which assesses pain intensity and interference over the preceding week¹⁷. Patients who met the pre-screening pain frequency and PEG scale criteria were approached for their willingness to undergo screening activities for the trial; however, completion of the pre-screening survey was not required to proceed with screening.

Randomization to Treatment Assignment

Randomization was performed at the level of the individual participant using random permuted blocks of 2, 4, and 6, with stratification by enrolling site and by the presence or absence of opioid use. Treatment assignments are not masked to participants or local research team members because of 1) impracticability, and 2) an interest in implementing the interventions as they would be outside of a trial setting. Potential bias that could occur from lack of masking is minimized by collecting baseline data before assignment to the treatment group, concealing to research teams the allocation sequence, and masking the individuals ascertaining the patient-reported outcomes to participants’ treatment assignments.

Interventions

Educational Information for Participants Randomized to Either PCST or Usual Care

At the baseline visit, participants in both the PCST and Usual Care groups are given a brochure with educational material about chronic pain and opioid medications, and information about available resources for treatment. Information about the trial and the participant’s randomized assignment is provided to the treating clinicians for each participant.

Pain Coping Skills Training Intervention

The PCST intervention is designed to provide participants with skills for self-management of pain, with the primary goal of reducing pain interference and other associated factors such as negative emotions and sleep disturbance. The intervention is based on cognitive behavioral therapy and includes elements of acceptance and commitment therapy, motivational interviewing, and mindfulness aiming to teach and practice skills, and increase self-efficacy for controlling pain. For participants who have recent or current opioid use, the motivational interviewing elements are aimed at reducing opioid use. For participants who do not have recent or current opioid use, motivational interviewing skills are focused on identified adherence goals or value-based skills of interest to the participant. The PCST intervention is 24 weeks in duration. During Weeks 1 – 12, a PCST coach conducts weekly sessions, 45–50 minutes in duration, with a participant by video conferencing or, if the participant prefers, by telephone. During Weeks 13 – 24, daily interactive voice response (IVR) sessions designed to enhance the training that took place during the initial 12 weeks replace the live PCST sessions. The IVR sessions, which typically last 5 minutes, consist of recorded telephone messages with a series of questions that participants answer using the telephone keypad. The participant’s PCST coach provides customized suggestions for the subsequent week based on the keypad responses. The IVR platform also enables participants to select skill refreshers that can be used during or after the call. Both the coach-led and IVR components of the PCST intervention are available in English and Spanish. For participants who have the coach-led PCST sessions during their dialysis treatments, privacy is maintained by using headphones with highly sensitive microphones allowing participants to speak quietly and be heard by the coach.

Intervention Fidelity

The individuals serving as PCST coaches have masters-level training in a healthcare service delivery field with prior coursework and clinical experience in health behavior coaching. Health psychologist and nephrologist members of the research group conducted training for both the coach-led and IVR components of the intervention before initiation of the interventions, and provide ongoing supervision throughout the course of the trial. An independent monitoring group reviews approximately 10% of audio recordings from the coach-led PCST sessions to assess intervention fidelity.

Buprenorphine Intervention

At Week 24, all participants are assessed for eligibility for the buprenorphine intervention, regardless of randomized group assignment. Participants who meet the eligibility criteria are encouraged to switch from their prescription opioid analgesic to buprenorphine, either the buccal buprenorphine or sublingual buprenorphine/naloxone formulation, depending on average daily pre-switch full-agonist dosage. A crossover approach to switching from the full agonist opioid to buprenorphine is used to prevent withdrawal symptoms that commonly occur with the traditional approach to switching to buprenorphine²⁰. The buprenorphine is prescribed by trial investigators using the dosing regimen shown in Table 2.

Table 2.

Approach for Switching to Buprenorphine Based on Current Dose of Full-Agonist Opioid

Day	Buprenorphine Formulation and Dose	Full-Agonist Dose
Full Agonist Opioid Dose: 20–25 MME/day
1	Belbuca 150 mcg BID	None
2 and on	Belbuca 300 mcg BID	None
Full Agonist Opioid Dose: 26–30 MME/day
1	Belbuca 150 mcg BID	None
2	Belbuca 300 mcg BID	None
3 and on	Belbuca 450 mcg BID	None
Full Agonist Opioid Dose: 31–35 MME/day
1	Belbuca 150 mcg BID	None
2	Belbuca 300 mcg BID	None
3 and on	Belbuca 450 mcg BID1	None
Full Agonist Opioid Dose: 36–40 MME/day
1	Belbuca 150 mcg BID	36–40 MME/day
2	Belbuca 300 mcg BID	36–40 MME/day
3	Belbuca 450 mcg BID	36–40 MME/day
4 and on	Buprenorphine/naloxone 2/0.5 mg BID	None
Full Agonist Opioid Dose: 41–50 MME/day
1	Belbuca 300 mcg BID	41–50 MME/day
2	Belbuca 450 mcg BID	41–50 MME/day
3 and on	Buprenorphine/naloxone 2/0.5 mg BID2	None
Full Agonist Opioid Dose: 51–60 MME/day
1	Belbuca 300 mcg BID	51–60 MME/day
2	Belbuca 450 mcg BID	51–60 MME/day
3	Buprenorphine/naloxone 2/0.5 mg BID	51–60 MME/day
4 and on	Buprenorphine/naloxone 2/0.5 mg BID-TID	None
Full Agonist Opioid Dose: 61–80 MME/day
1	Belbuca 300 mcg BID	61–80 MME/day
2	Belbuca 450 mcg BID	61–80 MME/day
3	Buprenorphine/naloxone 2/0.5 mg BID	61–80 MME/day
4 and on	Buprenorphine/naloxone 2/0.5 mg TID	None
Full Agonist Opioid Dose: 81–120 MME/day
1	Belbuca 300 mcg BID	81–120 MME/day
2	Belbuca 450 mcg BID	81–120 MME/day
3	Buprenorphine/naloxone 2/0.5 mg BID	81–120 MME/day
4 and on	#2 Buprenorphine/naloxone TID	None
Full Agonist Opioid Dose: 81–120 MME/day
1	Belbuca 300 mcg BID	81–120 MME/day
2	Belbuca 450 mcg BID	81–120 MME/day
3	Buprenorphine/naloxone 2/0.5 mg BID	81–120 MME/day
4 and on	#2 Buprenorphine/naloxone TID	none
Full Agonist Opioid Dose: 121–150 MME/day
1	Belbuca 300 mcg BID	121–150 MME/day
2	Belbuca 450 mcg BID	121–150 MME/day
3	Buprenorphine/naloxone 2/0.5 mg BID	121–150 MME/day
4	Buprenorphine/naloxone 2/0.5 mg BID	120 MME/day
5 and on	Buprenorphine/naloxone 2/0.5 mg TID	None
Full Agonist Opioid Dose: 151–200 MME/day
1	Belbuca 300 mcg BID	151–200 MME/day
2	Belbuca 450 mcg BID	151–200 MME/day
3	Buprenorphine/naloxone 2/0.5 mg BID	151–200 MME/day
4	Buprenorphine/naloxone 2/0.5 mg BID	100–150 MME/day
5 and on	Buprenorphine/naloxone 8/2 mg BID	None
Full Agonist Opioid Dose: >200 MME/day
1	Belbuca 300 mcg BID	>200 MME/day
2	Belbuca 450 mcg BID	>200 MME/day
3	Buprenorphine/naloxone 2/0.5 mg BID	>200 MME/day
4	Buprenorphine/naloxone 2/0.5 mg BID	100–150 MME/day
5	Buprenorphine/naloxone 8/2 mg BID	None
6 and on	Increase Buprenorphine/naloxone as needed to up to 8/2 mg TID	None

Management of Non-Study Medications

Initiation or modifications in dosing of medications to treat pain, depression, or anxiety by the participant’s treating clinicians is permitted. During Weeks 24–36, for participants who switch from a full-agonist opioid to buprenorphrine as a trial intervention, resumption of the full agonist opioid by a treating clinician requires discussion and coordination with the research team.

Outcomes

The trial outcomes are shown in Table 3. The primary outcome is pain interference, meaning the degree to which pain interferes with daily function and quality of life. Pain interference rather than pain intensity has been recommended by expert consensus groups as a key outcome for chronic pain trials because of its greater salience for patients and generalizability across populations ^21–25. The outcome measure for pain interference used in the trial is the Brief Pain Inventory Interference subscale (BPI Interference) with a 7-day lookback period²². The BPI is a multidimensional measure of pain intensity and interference developed for cancer pain but subsequently validated in diverse populations^22,26. The BPI is composed of the Severity subscale, which consists of four pain intensity items, and the Interference subscale, which consists of seven pain interference items. The BPI Interference subscale has demonstrated responsiveness to change in numerous clinical trials^27–29.

Table 3.

Outcomes

Domain	Instrument or Data Source	Look-Back Period	Score Range	Interpretation of Higher Score	Ascertainment Weeks
Primary Outcome
Pain interference	Brief Pain Inventory (BPI) Interference	1 week	0–10	More interference	0, 12, 24, 36
Secondary Outcomes – Pain and Opioid Use
Pain intensity	Brief Pain Inventory (BPI) Severity	1 week	0–10	More severe	0, 12, 24, 36
Pain catastrophizing	Pain Catastrophizing Scale – SF 6	None	0–24	More catastrophizing	0, 12, 24, 36
Opioid use	Timeline Followback	2 weeks	N/A	--	0, 12, 24, 36
Composite of pain and opioid use	BPI Interference / Timeline Followback	1 week / 2 weeks	N/A	--	0, 12, 24, 36
Secondary Outcomes – Conditions or Symptoms Associated with Pain or Opioid Use
Quality of life	Single-Item QOL Scale – McGill Quality of Life	2 days	0–10	Better quality of life	0, 12, 24, 36
Depression	Patient Health Questionnaire (PHQ)-9	2 weeks	0–27	More depression	0, 12, 24, 36
Anxiety	Generalized Anxiety Disorder (GAD)-7	2 weeks	0–21	More anxiety	0, 12, 24, 36
Coping	Coping Strategies Questionnaire: 24-item	None	0–42 per component	More coping	0
Coping	Coping Strategies Questionnaire: Single item	None	0–6	More coping	12, 24, 36
Self-efficacy Pain	PROMIS Self-Efficacy for Managing Chronic Conditions – Managing Symptoms:  • Short Form 8A  • Single item	None	8–40 1–10	More self-efficacy	0, 12, 24, 36
Sleep quality	PROMIS Sleep Disturbance 6a + Sleep Duration Question	7 days	6–30	Worse sleep	0, 12, 24, 36
Fatigue	PROMIS Fatigue SF 6a	7 days	6–30	Worse fatigue	0, 12, 24, 36
Physical functioning	PROMIS Physical Functioning SF 6b	7 days	6–30	Better physical function	0, 12, 24, 36
Social support	Multidimensional Scale of Perceived Social Support	None	1–7	More support	0, 12, 24, 36
Dialysis-associated symptoms	Dialysis Symptom Index  • Number  • Severity	1 week	0–30 0–120	More symptoms More severe symptoms	0, 12, 24, 36
Satisfaction with treatment	Patient Global Impression of Change	Since Baseline	0–5	Less improvement	0, 12, 24, 36
Family intrusion	PROMIS Satisfaction with Social Roles and Activities	7 days	8–40	More satisfaction	0, 12, 24, 36
Discrimination	Everyday Discrimination Scale	None	0–25	More discrimination	0, 36
Secondary Outcomes – Clinical Events
Falls	Self-report, medical records	N/A	N/A	--	0 through 36
Hospitalizations	Self-report, medical records	N/A	N/A	--	0 through 36
Death	Dialysis facility personnel, family, medical records	N/A	N/A	--	0 through 36
Exploratory Outcomes
Acceptability of buprenorphine	Proportion of eligible participants who agree to switch to buprenorphine	N/A	N/A	--	24
Tolerability of buprenorphine	Proportion of participants who remain on buprenorphine for the full treatment period	N/A	N/A	--	36
Efficacy of buprenorphine	Brief Pain Inventory (BPI) Interference	1 week	0–10	More interference	24, 36

The secondary outcomes fall into 3 categories: 1) patient-reported indicators of pain or opioid use, 2) patient-reported indicators of conditions or symptoms that often accompany pain or opioid use, and 3) clinical outcomes. The outcomes for the buprenorphine intervention are considered exploratory and include acceptability, defined as willingness to try buprenorphine, and tolerability, defined as continued use of buprenorphine for the full intervention period.

Ascertaining Patient-Reported Outcomes

Patient-reported outcomes are ascertained in English or Spanish using computer-assisted telephone interviewing (CATI) administered by a centralized team, the members of which are unaware of the participant’s treatment assignment. Use of a centralized approach to ascertaining patient-reported outcomes provides standardization in delivery of the instruments and avoids potential biases introduced when members of the local research team, with whom a participant has regular encounters, are administering the questionnaires³⁰. Additionally, use of telephone-based administration rather than electronic or paper forms completed by participants allows inclusion of individuals with impairment in vision or poor manual dexterity, both of which are common among the dialysis patient population.

Adverse Event Monitoring

Serious adverse events are reported to the SDRC within one calendar day of awareness by the research team at the clinical site. Adverse events of interest include suicidality, adverse effects of medications initiated for pain, depression, or anxiety, opioid withdrawal symptoms, development of opioid use disorder, and development of non-opioid substance use disorder. In January, 2021, in response to a new FDA warning, dental events were added as an adverse event of interest for participants who switch to buprenorphine³¹.

Analyses

The primary analysis will compare the change from baseline to week 12 in the BPI Interference scores between the PCST and Usual Care groups using a linear mixed effects model and an intent to treat approach in which all available data on all randomized participants are included³². The model will include fixed effects of intervention, time, and interactions between intervention and time, accounting for repeated measures and the 16 enrolling sites via random effects. Similar analyses will be performed using the week 24 BPI Interference score to determine whether IVR enhances effects of the coach-led component of the PCST and using the week 36 BPI Interference score to determine the durability of the effects of PCST. The same approach will be used for continuous secondary outcomes. For binary secondary outcomes, such as the composite of pain interference and opioid use, generalized linear mixed effects models with a binomial distribution and a log link adjusting for baseline measures will be used. For recurrent endpoints such as hospitalizations or falls, generalized linear mixed effects models with a Poisson distribution, an offset of follow-up time, and a log link adjusting for baseline measures will be used to compare event rates. For comparisons between the PCST and Usual Care groups that use outcomes at week 36, additional adjustment for buprenorphine eligibility and for buprenorphine initiation will be incorporated. Methods to address missing outcome data will include adjustment by factors associated with missingness, and inverse weighting based on propensity of the missingness. Sensitivity analysis, such as pattern mixture models, will be conducted under missing not at random assumptions to examine the robustness of the model-based results³³. The overall level of significance will be set at 0.05 with adjustment for multiple comparisons where appropriate. All analyses will be adjusted for the randomization stratification factors.

Sample Size Determination

The sample size of 640 participants is based on the primary objective of comparing the 12-week change in the BPI Interference score between the PCST and Usual Care groups. With 640 participants the trial should have 90% power at a two-sided alpha level of 0.05 to detect minimum between-group differences in the change in BPI Interference that range from 0.62 to 1.17 points assuming a standard deviation of change of 2.0 – 3.0, an intra-site correlation coefficient of up to 0.02, and a single interim efficacy/futility analysis at 50% information time. A change of 1 point in the BPI Interference scale is generally considered to be clinically important at the individual person level³⁴.

Adaptations Related to the COVID-19 Pandemic

To address the COVID-19 pandemic, regulatory approval was granted to allow informed consent to be obtained by telephone or video conferencing, and for electronic documentation of consent. For in-person interactions with patients, research team members followed the approaches to use of personal protective equipment instituted by the dialysis units. Because the PCST and buprenorphine interventions had been designed to be implemented through remote interactions, substantial modifications to the intervention delivery approach were not needed.

Funding and Oversight

The trial is funded by the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) through the National Institutes of Health’s Helping to End Addiction Long-term (HEAL) Initiative. An independent DSMB appointed by the NIDDK oversees the safety, quality, and progress of the trial. Approval of the protocol was provided by each of the dialysis provider organizations operating the dialysis facilities administering care to patients approached for enrollment in the trial and by the leadership of the individual dialysis facilities. The Institutional Review Board (IRB) at the University of Pennsylvania serves as the IRB of record for the 11 non-Veterans Affairs (VA) sites, and the VA Central IRB serves as the IRB of record for the 5 VA sites. Because buprenorphine/naloxone, one of the two study formulations, is approved for opioid use disorder but does not have marketing approval for the treatment of pain, the trial is being conducted under an investigational new drug (IND) application (15134). The trial is registered at clinicaltrials.gov (NCT04571619).

Results

Enrollment and Baseline Characteristics of the Trial Cohort

Prescreening was initiated on December 28, 2020, and the first participant was randomized on January 22, 2021. The target date for competing randomization of the full cohort of 640 participations was June 1, 2023, and the target was met 2 months ahead of schedule with 640 participants randomized by March 29, 2023. Three additional participants who had consented to the trial before enrollment was closed were randomized by April 7, 2023. The baseline characteristics of the participants are shown in Table 4. Approximately 45% of the participants are female, nearly half identify as Black, and 19% identify as Hispanic. Approximately 60% of the participants have diabetes. Long-term use of prescribed opioids, defined as use for at least 3 of the previous 6 months, was reported by 22.5% of participants. Baseline BPI-Interference and BPI-Intensity scores are 6.6 and 6.0, respectively, both using scales ranging from 0 to 10 with higher scores indicating worse pain.

Table 4.

Baseline Characteristics

Demographic Characteristics
Age, years	60.2 (12.8)
Female	288 (44.8%)
American Indian or Alaskan Native	22 (3.4%)
Asian	6 (0.9%)
Black	308 (47.9%)
Native Hawaiian or Other Pacific Islander	6 (0.9%)
White	209 (32.5%)
Multiple Races	11 (1.7%)
Race not reported	81 (12.6%)
Hispanic or Latino	119 (18.5%)
Ethnicity not reported	7 (1.1%)
Clinical Characteristics
Post-dialysis BMI, kg/m2	31.3 (12.7)
Pre-dialysis systolic BP, mm Hg	144.2 (25.4)
Pre-dialysis diastolic BP, mm Hg	76.8 (15.7)
Duration of dialysis treatment, years	4.5 (4.7)
Comorbidities
Diabetes mellitus	380 (59.1%)
Coronary artery disease	175 (27.2%)
Heart failure	169 (26.3%)
Atrial fibrillation or flutter	113 (17.6%)
Ventricular arrhythmia	33 (5.1%)
Heart valve replacement or repair	35 (5.4%)
Stroke or transient ischemic attack	128 (19.9%)
Peripheral vascular disease	102 (15.9%)
Cancer	106 (16.5%)
Substance Use
Opioid use during 3 of the last 6 months	144 (22.4%)
Opioid use during the last 14 days	161 (25.0%)
Average MME/day for those with opioid use during the last 14 days	6.4 (0.0 – 22.5)
Current or former tobacco use	321 (49.9%)
Current alcohol use	64 (10.0%)
Laboratory Values
Blood urea nitrogen, mg/dL	54.0 (22.5)
Creatinine, mg/dL	9.3 (3.9)
Albumin, g/dL	3.9 (0.4)
Hemoglobin, g/dL	10.9 (1.7)
Bicarbonate, mEq/L	24.3 (4.1)
Kt/V	1.6 (0.3)
Pain
BPI Interference2	6.57 (5.14 – 7.86)
BPI Severity2	6.00 (4.50 – 7.50)

¹Values are means (SD), n (%), or median (25^th-75^th percentile)

²Possible range is 0 – 10 with higher scores indicating more pain interference or more severe pain, respectively Abbreviations: BMI, body mass index; BP, blood pressure; MME, morphine milligram equivalents; BPI, Brief Pain Inventory

Discussion

Several attributes of the HOPE trial evaluating behavioral and pharmacologic approaches to chronic pain among patients treated with maintenance hemodialysis warrant discussion: 1) the consortium comprises a multidisciplinary group of investigators and highly engaged patient advisors; 2) the PCST intervention was adapted to accommodate both the unique clinical characteristics of the patient population and the care delivery setting; 3) buprenorphine as an alternative analgesic to full agonist opioids has not previously been evaluated for the treatment of pain among patients with ESKD; 4) the extensive set of patient-reported outcome measures will allow characterization of the effects of treatment on psychosocial factors that accompany and potentially influence the experience of chronic pain in this patient population; 5) the trial design allows for multiple questions to be addressed within a single clinical trial; and 6) the trial cohort of 643 participants is demographically diverse and was enrolled ahead of schedule.

As required by the funding opportunity, the research teams at the clinical centers and the SDRC include nephrologists, pain specialists, health psychologists, clinical trialists, and a bioethicist, allowing for multidisciplinary input into the design and implementation of the trial. Consistent with goals of the NIDDK and the NIH HEAL initiative, the consortium also includes patients with end-stage kidney disease treated with hemodialysis, all of whom have experience with chronic pain³⁵. The patients served as advisors while the trial was being designed and have continued to serve in this capacity during the trial conduct. The advisors provided important perspectives about proposed interventions, selection of outcomes, and burden of the survey instruments, and they helped create the recruitment materials including a video shown to potential participants, visual materials used for the PCST intervention, and the informed consent document. Views expressed by the patient advisors contributed to the decision during the design phase to encourage rather than require a switch to buprenorphine for participants eligible for the buprenorphine intervention.

The PCST intervention was adapted from existing pain coping skills programs to address the high comorbidity and treatment burden that characterizes dialysis-dependent kidney disease. The sessions focus principally on pain interference but also address, if relevant to the participant, pain-related symptoms of depression, anxiety, stress, and sleep disturbance, all of which are common in this patient population. The delivery of the coach-led component of the PCST intervention via telehealth allows the location and timing of interactions to be based on the preference of participants. Because patients spend approximately 12 hours each week receiving dialysis treatments, the study group felt it was important to offer the PCST intervention during dialysis sessions if that was preferred by participants. Delivering the intervention during dialysis allows research coordinators to assist, if necessary, with establishing the electronic tablet-based connection with the PCST coach for participants with eHealth hesitancy or other barriers. The delivery of the intervention by a centralized team of coaches is cost-efficient since the workload of the coaches can be based on overall rather than site-specific enrollment, and the approach enables enrollment of Spanish-speaking participants without requiring Spanish-speaking coaches at every enrolling site. Use of a centralized delivery approach during the trial is also anticipated to facilitate translation of the trial intervention to the clinical care setting and make the intervention more accessible to patients in underserved areas since uptake would not be dependent on availability of local personnel. The IVR component of PCST is intended to solidify, and potentially enhance, the skills acquired during the coach-led sessions. While previously evaluated in a few studies of PCST^36,37, IVR is a relatively novel aspect of the intervention, and may be particularly important for the ESKD patient population because of the high rate of hospitalizations and other inter-current events that might interrupt or decrease completion of the full set of coach-led PCST sessions.

Because of limited experience with buprenorphine as a treatment for pain in the ESKD population and the expectation that only a subset of trial participants would meet the eligibility criteria for buprenorphine, the pharmacologic component of the HOPE trial was designed to generate preliminary estimates and signals about acceptability, tolerability, and efficacy of buprenorphine rather than to serve as a definitive efficacy evaluation. Because of uncertainties about the willingness of the patients to participate in a trial of buprenorphine, willingness to switch to buprenorphine was not an inclusion criterion for participation in the HOPE trial. The pharmacologic component of the trial is expected to inform the feasibility and design of future trials of buprenorphine as an option for treating chronic pain in the setting of ESKD.

The serial ascertainment of a large set of patient-reported outcomes addressing not only pain but also many other psychosocial domains will enable the pre-specified assessment of the effects of the interventions on outcomes other than pain that are important to patients, as well as identification of factors associated with responsiveness to the interventions. Additionally, some of the domains addressed by the patient-reported outcomes have not been previously characterized among a large, demographically diverse group of patients with dialysis-dependent kidney disease. Ascertainment of the outcomes at multiple time points not only protects against missing data but also allows multiple questions to be answered by the trial. For example, in secondary analyses, the questions of 1) whether IVR provides benefit beyond the coach-led component of PCST and 2) whether the effect of PCST is sustained after it has been completed will be answered using the Week 24 and Week 36 outcomes, respectively. Week 36 outcomes will also be used to address, in an exploratory manner, the questions related to buprenorphine. All analyses that include data beyond Week 24 will incorporate both buprenorphine eligibility and buprenorphine use in the models.

Meeting enrollment targets on schedule is one of the most common challenges when conducting clinical trials. The HOPE trial had an ambitious enrollment target that was met ahead of schedule likely, in part, because the trial addresses something felt by patients to be important. It is also likely that guidance from the highly engaged patient advisors, who designed the recruitment materials and were featured in the recruitment video, contributed to interest by patients in participating in the trial. The development of recruitment materials by patients may have also helped with generating a demographically diverse cohort with enrollment of Hispanic patients, including monolingual Spanish speakers, in a proportion matching that of the overall US dialysis population and an over-representation of Black patients relative to the US dialysis patient population.

Conclusion

The HOPE trial addresses an important unmet need for patients receiving maintenance hemodialysis using a rigorous trial design that allows multiple questions to be answered. The success with participant enrollment suggests that interventions are appealing to patients and that the outcomes are perceived as salient. The trial findings are expected to inform the management of chronic pain for this patient population.

HOPE Consortium, Cores, and Data and Safety Monitoring Board Members

Hennepin Healthcare Clinical Center and Enrolling Site: Kirsten Johansen (PI), Gavin Bart, Erin Krebs, James Wetmore, Maria Pacheco-Hernandez, Ursula Munet, Rudy Qamhiyeh, Mike Wambua; Massachusetts General Hospital Clinical Center and Enrolling Site: Sahir Kalim (PI), Sagar Nigwekar, Daniel Weiner, Kome Ekor, Beza Mengesha, Shananssa Percy; New York University Langone Health Clinical Center and Enrolling Site: David Charytan (PI), Keith Goldfeld, Joshua Lee, Nobuyuki Miyawaki, Jennifer Scherer, Amanda Shallcross, Miri Cazes, Sobaata Chaudry, Paula Dutka, Yasmine Flores, Daniela Fraticelli Ortiz, Candace Grant, Colin Keane, Pragna Krishnamurthy, Ashley Macina, Angela McCarthy, Kathleen Rice, Grace Robinson, Dalila Varela, Javaughn Ways; University of Illinois Chicago Clinical Center and Enrolling Site: Michael Fischer (PI), Ardith Doorenbos, Christopher Holden, James Lash, Mark Lockwood, Alana Steffen, Cheryl Gilmartin, Amanda Goldstein, Monya Meinel, Kimberly Silva, Guillermo Zamora; University of Pittsburgh Clinical Center and Enrolling Site: Manisha Jhamb (PI), Hailey Bulls, Meghan Hamm, Sanjana Kamat, Jane Liebschutz, Jennifer Steel, Jonathan Yabes, Precious Lacey, Donna Olejniczak, Mary Schopp, Melissa Weimer, Vincent Wood; University of Pennsylvania Enrolling Site: Nwamaka Eneanya (PI), Sarah Schrauben (PI), Stephany Almonte-Then, Chigozie Amonu, Nicholas Bishop, Diane Park, Taylor Stallings; University of Washington Clinical Center and Enrolling Site: Rajnish Mehrotra R (PI), Nisha Bansal, Elenore Bhatraju, Steve Weisbord, Lisa Anderson, Sydney Johnson, Kels Laszlo, Lori Linke; University of New Mexico Enrolling Site: Mark Unruh (PI), Christos Argyropoulos, Shane Pankratz, Davin Quinn, George Garcia, Monica Bajana Meza, Monica Cardona, Tammy Seaman Weidner, Greg Trejo; Rogosin Institute Enrolling Site: Daniel Cukor (PI), Nathaniel Berman, Nelson Chen, Ines Chicos, Stephanie Donahue, Anna Gong; Vanderbilt University Medical Center C Clinical Center and Enrolling Site: Kerri Cavanaugh (PI), Carrie Brintz, David Edwards, Alp Ikizler, Puneet Mishra, Thomas Stewart, DeVitra Berry, Don Merrimon, Samuel Opeke, Sarah Pleasant, Hadassah Pegues, Christopher Roach, Sonya Williams; West Virginia University Enrolling Site: Bethany Pellegrino (PI), Daniel McNeil, Alvin Moss, Rebecca Schmidt, Cheryl Dalton, Kristy O’Connell, Maryanne Wilkinson; Yale School of Medicine/VA Connecticut Healthcare System Clinical Center and Enrolling Site: William Becker (PI), Justin Belcher, Susan Crowley, Denise Esserman, Caroline Falker, Alicia Heapy, Svetlana Vassilieva, Samara Zuniga; Durham VA Health Care System Enrolling Site: Patrick Pun (PI), Wissam Kourany, James Lefler, Teresa Purdy, Jenika Hammond, Khristian Harris, Sara Hoffman, Jeanette Rutledge; Portland VA Medical Center Enrolling Site: Benjamin Morasco (PI), Christopher Blazes, Christopher Stauffer, Melissa Adams, Deza’Rae Collins, Richard Torres; Dallas VA Medical Center Enrolling Site: Jeffrey Penfield (PI), Monica Barbosa, Levi Beeks, Erik Guajardo, Sindi Sanchez; VA New York Harbor Health Care Enrolling Site: Mansi Mehta (PI), Adrian Cosmin, David Goldfarb, Sabrina Felson, Brian Sands, Frank Modersitzki; Scientific Data and Research Center: University of Pennsylvania: Laura Dember (PI), Martin Cheatle, John Farrar, Jesse Hsu, Steve Joffe, Kyle Kampman, Francis Keefe, J Richard Landis, Ted Barrell, Leah Bernardo, Natalie Kuzla; Patient Advisors: Dawn Edwards, Robert Grindstaff¹, Andre Hoover, Stephen Lerner, Roger Mims, Nina Quintana, Steven Sousa, Darlene Villareal, David White, Caroline Wilkie, Joel Williams; Pain Coping Skills Training Coaches: Veronica Dyer, New York University Langone Health; Eshika Kalam, Rogosin Institute; Blanca Contreras, University of Illinois Chicago; Carlyn Clark, University of Washington; Heather Howell and Kevin Payne, West Haven VA Medical Center; Pain Coping Skills Training Fidelity Monitors: Andrew Busch and Sarah Cameron, Hennepin Healthcare; Computer-Assisted Telephone Interviewing (CATI) Centers: University of New Mexico - Adamaris Arteaga Leon, Elena Ashley, Gabriela Chacon Palma, Katy Chalamidas, Oluoma Edeh, Lindsay Gear, Cameron Guy, Grace Kimura, Alexander Leon Cupe, Henry Luo, Katherine McDaniels, Sofia McLaren, Valeria Mejia, Claire Mullins, Uchechukwu Okereke, Gabriel Rudow, Giselle Rodriguez Sosa, Lily Sullivan, John Torres, Hugo Vilchis; University of Pittsburgh - University of Pittsburgh: Scott Beach, Amber Barton, Victoria Casilli, Lynda Connelly, Jane Dirks, Luke Farkas, Ana Geibel, Olivia Kirsch, Paula Kubrick, Patricia Lietz, Brandon Self, Susan Stringfellow, Olivia Wilson; National Institute of Diabetes and Digestive and Kidney Diseases: Kevin Abbott, Paul Kimmel, Jenna Norton, Tracy Rankin; Data and Safety Monitoring Board: Jane Atkinson, Bruce Barton, Melissa Bensouda, Roger B. Fillingim, Michael Freeman, Andrew Garland, Jennifer Gassman, Allen Nissenson, Keith Norris, Maile Robb¹, Friedhelm Sandbrink, Brigitte Schiller, Eric Storch, David Thomas, Roger Weiss.