#NephJC Chat
Tuesday, Jan 21st 2025, 9 pm Eastern on BlueSky
JAMA Intern Med. 2024 Dec 30. doi: 10.1001/jamainternmed.2024.7140. Online ahead of print.
Pain Coping Skills Training for Patients Receiving Hemodialysis: The HOPE Consortium Randomized Clinical Trial
Laura M Dember, Jesse Y Hsu, Rajnish Mehrotra, Kerri L Cavanaugh, Sahir Kalim, David M Charytan, Michael J Fischer, Manisha Jhamb, Kirsten L Johansen, William C Becker, Bethany Pellegrino, Nwamaka D Eneanya, Sarah J Schrauben, Patrick H Pun, Mark L Unruh, Benjamin J Morasco, Mansi Mehta, Nobuyuki Miyawaki, Jeffrey Penfield, Leah Bernardo, Carrie E Brintz, Martin D Cheatle, Ardith Z Doorenbos, Alicia A Heapy, Francis J Keefe, Erin E Krebs, Natalie Kuzla, Sagar U Nigwekar, Rebecca J Schmidt, Jennifer L Steel, James B Wetmore, David M White, Paul L Kimmel,Daniel Cukor
PMID: 39786400
Introduction
Many patients with end-stage renal disease (ESRD) prioritize how they feel and function over how long they live. Thus, there have been increased efforts to address the numerous symptoms that burden patients on dialysis. Pain is among the most common symptoms experienced by patients on dialysis, with approximately 60% reporting pain that is described as moderate or severe on most days (Davison S. Am J Kidney Dis, 2003). Pain is also one of the outcomes highlighted in the SONG-HD core outcomes in hemodialysis.
The perception of pain involves physiologic as well as psychological influences, and measurements of pain are notoriously subjective so that no two people experience pain in exactly the same way or intensity. Other psychological stresses, like anxiety and depression, can exacerbate pain and lead to feelings of helplessness. Among hemodialysis patients, long-term opioid use is associated with increased rates of falls, hip fractures, hospitalizations, dialysis withdrawal, and death (Kimmel et al. J Am Soc Nephrol, 2017). Patients treated with hemodialysis generally have low levels of physical activity and high rates of insomnia and depression, all of which can exacerbate chronic pain and complicate its management. Non-pharmacologic approaches, such as cognitive behavioral therapy, have shown benefits in non-dialysis patients with chronic pain, but have not been studied in patients with ESRD on dialysis. Given that all pharmacological options (see below) have dubious efficacy and worrisome toxicity, these options represent a different way to address pain.
The Study
Methods
HOPE (Helping People on Dialysis Manage Pain) was designed as a randomized clinical trial to evaluate approaches to reducing pain and opioid use among patients with chronic pain who were receiving maintenance hemodialysis for ESRD. The hypothesis was that pain coping skills training is effective at reducing pain and opioid use. Patients assigned to Pain Coping Skills Training (PCST) completed a structured, standardized, interactive program designed to reduce pain interference, opioid use, and the comorbid symptoms of depression, anxiety, and sleep disturbance. The program was adapted for the hemodialysis patient population and was delivered using live coaches via video telehealth and with interactive voice response via telephone. Patients in the usual care arm were given educational materials and resources about alternatives to opioid pain medications. These materials were also provided to participants in the PCST group.
This was a 24-week randomized control trial that enrolled patients in cognitive behavioral therapy versus usual care to treat fatigue, pain, and depression. The first 24-week randomized phase was followed by an exploratory, nonrandomized evaluation of buprenorphine as an alternative to full-agonist opioid medications for the subset of participants receiving opioid medication at 20 morphine milligram equivalents per day or higher. The total follow-up was 36 weeks. Buprenorphine is 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, but has limited data in ESRD.
Figure 1.1. Trial design. Dember et al. JAMA Int Med, 2024.
Patient-reported outcomes were obtained through various self-reported inventories: Brief Pain Inventory Severity and Interference, Single Item QOL, PHQ-9, GAD-7, PCS SF-6. The results of the scoring from these surveys were then compared between groups.
Inclusion Criteria
The trial was conducted at 16 centers and 103 outpatient dialysis facilities in the United States.
Participants were adults on in-center hemodialysis for >90 days with self-reported moderate to severe chronic pain (≥ 4 on a scale of 1-10 on most days over three months).
eTable 1. Inclusion and Exclusion criteria. Dember et al. JAMA Int Med, 2024.
Randomization
Patients were randomized 1:1 to pain coping skills training (PCST) or usual care, with stratification for opioid use. Baseline surveys were distributed before randomization. Participants and site research teams were not blinded to treatment assignments. Surveys were reviewed and scored by administrators who were blinded to patient treatment assignments.
Intervention
PCST patients received weekly 45-50 minute sessions led by a coach via video or telephone. There were also 12 weeks of daily automated interactive voice response (IVR) sessions. IVR sessions were conducted by telephone and required keypad input to interactively monitor patients with questions and skills refreshers. Modules were specific to address anxiety, stress, insomnia, and pain education. Coaches had a master’s degree in either social work or counseling with prior experience providing coaching. Coaches received additional training prior to trial initiation. Follow-up visits occurred every 4 weeks in person or by telephone. Adverse events, clinical outcomes, medication updates, and scheduling of surveys were accomplished during these visits.
Table 6.1. PCST telehealth sessions. Dember et al. JAMA Int Med, 2024. (MI: Motivational Interviewing, IVR: Interactive Voice Response, ACT: Acceptance and Commitment Therapy)
Funding
The trial was funded by the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) through the National Institutes of Health Helping to End Addiction Long-Term Initiative. A cool part of the trial design was the involvement of patient partners in the trial design and conduct throughout the period.
Figure: Set up of the trial conduct, from Dember et al, Contemp Clin Trial 2024
Primary Outcome
A Brief Pain Inventory (BPI) assesses pain-related interference of quality of life. The BPI is scored 0-10 on questions with a high score representing more pain interference.
Secondary Outcomes
Additional reports (outside of the BPI questions) of pain indicators and opioid use, as well as pain-related clinical events. The study also examined symptoms that accompany pain (i.e. anxiety, depression), and symptoms possibly related to pain treatments.
From the secondary outcomes, the following were prespecified to be included in the primary outcome: intensity, catastrophizing, quality of life, depression, anxiety, opioid use, and a composite of opioid use and pain interference.
Statistical Analysis
The primary outcome compared the change from baseline BPI survey scores to week 12 scores between PCST patients and usual care groups. An intention-to-treat approach included all data in a mixed-effects model. Data points included the study intervention, study week, opioid use, and enrollment site. The same modeling was used to compare BPI surveys from baseline to 24 weeks and then finally 36 weeks, to prove the durability of the intervention.
For secondary outcomes including death, falls and hospitalizations generalized estimating equation (GEE) models with Poisson distribution (which takes into account mean number of events in a given time period) were used to compare event rates. Sensitivity analyses were done to address missing data. One hundred datasets were imputed and pooled into a final estimate.
Results
The planned sample size was 640 dialysis patients. Assuming a standard deviation of 2-3, and an attrition rate of 20%, the trial was to have 90% power and be able to detect an inter-group difference of 0.62-1.17 points. The participant enrollment and follow-up distribution are seen below, with 643 patients undergoing initial randomization to CBT or usual care.
Figure 1. Flow diagram of participant enrollment. Dember et al. JAMA Int Med, 2024.
Baseline characteristics were similar between groups. The average age was 60 years for both groups, and nearly half of the participants were female. Although a diverse group of ethnicities were included, nearly 50% of patients self-identified as Black. Approximately 9% of participants reported Spanish as their preferred language. Most patients reported musculoskeletal (89%) or neuropathic (67%) as the underlying cause of pain. Nearly one-quarter of patients used opioids within the last 14 days, while one-fifth of the included patients used them during 3 of the last 6 months. The average morphine milligram equivalents/day (MME/day) was 15mg/day (the equivalent of almost 70 mg tramadol/day), while MME/d was higher in the pain coping skills training group. [Editorial note: These doses may seem high to people outside North America, but are lower than at the peak of the opioid marketing scandal see Van Zee AM J Public Health 2009. And Tramadol should be Tramadont].
Table 1. Baseline characteristics. Dember et al. JAMA Int Med, 2024.
Between 2021-2023, enrolled patients were randomized to cognitive behavioral therapy/pain coping skills training (PCST) versus usual care (319 vs 324). The median number of PCST sessions led by a trained coach was 12 (IQR 10-12) per participant. Video sessions were accomplished in 62% of PCST training. Additionally, approximately 63% of training sessions were done during hemodialysis treatment.
There were twenty-one thousand automated telephone calls during a 12-week interval, but patients only answered about 53% of those calls. A sample of those calls was reviewed, and 93% of answered calls met the criteria for successful completion. Withdrawal from the trial, other than death, occurred in 13 participants in the PCST group and 6 in the usual care group.
Primary Outcome
There was a greater disruption of pain symptoms in the PCST group compared with usual care.
The Brief Pain Inventory (BPI) interference score in PCST-treated patients was -0.49 (95% CI, -0.85 to -0.12; P= 0.009), statistically lower than scores from patients in the usual care group. The effect was preserved at 24 weeks but diminished by week 36. Results were similar in analyses that used imputed values for missing data points. A larger proportion of patients in the PCST group (50.9% versus 36.6%) had a decrease in BPI interference scoring of greater than 1-point at week 12 (OR 1.79, 95% CI 1.28-2.49) and at week 24 (OR 1.59, 95% CI 1.13-2.24), but not at week 36.
Table 2. Effect of PCST on BPI interference score. Dember et al. JAMA Int Med, 2024.
In a post hoc sensitivity analysis between groups, differences favored PCST at 12, 24, and 36 weeks. There was no evidence of an effect modification from demographic characteristics or baseline opioid use.
Figure 2. Changes in mean score of survey testing, PCST versus usual care. Dember et al. JAMA Int Med, 2024.
Secondary Outcomes
At week 24, changes from baseline in BPI severity, pain catastrophizing (tendency to exaggerate the threat of pain while feeling helpless), general anxiety, and depression surveys all favored PCST. However, at week 36 the effect appears to have persisted only for pain catastrophizing. The outcomes of falls, hospitalizations, and deaths were similar between groups. Serious adverse events (including adverse events of interest) rates did not differ between groups.
eTable 10. Clinical event outcomes. Dember et al. JAMA Int Med, 2024.
Discussion
We have previously discussed cognitive behavioral therapy for dialysis patients in the SLEEP-HD RCT (Mehrotra et al, Ann Int Med 2024┃NephJC Summary) for the management of insomnia. CBT was as effective as medications with fewer serious adverse events (remember “TRAZADONT”). We have to ask ourselves, why isn’t CBT being incorporated more into the care of patients on dialysis who are experiencing pain, anxiety, depression, and/or insomnia? The answers may seem obvious: cost, expertise, initiation effort, coordination, and time. However, we must seriously consider that when we embrace a team-based approach to the care of dialysis patients, why is psychological counseling and support absent from our comprehensive rounds? Perhaps as the evidence builds, we need to advocate more for a dramatic shift in this patient care paradigm.
The strengths of the trial is that this was a diverse and representative cohort, with high adherence despite a complex intervention. On the other side, it was an open label (unblinded) trial by design, with about ~10% loss from death (somewhat inevitable in a dialysis trial). No cost data were presented in this publication which is important as provision of this complex intervention will not be an easy task.
In this current RCT, the authors examined the use of CBT (in this instance pain coping skills training/PCST) to determine if there were any benefits for patients with ESRD in reducing pain-associated symptoms as quantified by self-reported patient surveys. The overall effect was modest, but significant to participants. The implementation of a program that includes both live coaching (telemedicine/phone calls) and automated survey calls to enforce training is something that could be implemented for patients performing dialysis at home and even at rural dialysis units. In this day and age when there are ever-increasing online psychology counseling apps and online services, there are fewer excuses to avoid integration of this essential element of care. Additionally, the fact that a majority of these interventions could be accomplished while the patient was on hemodialysis is promising, as some patients have difficulty scheduling and traveling for medical services during their non-dialysis days. It might be interesting to see in future studies if CBT on dialysis (a somewhat captive audience) is more effective than off dialysis.
Another argument for the integration of psychological services into monthly dialysis rounds is how quickly CBT improved pain and associated symptoms (12 weeks), while also being quickly extinguished (approximately 12 weeks after completion). Patients on dialysis are no different than the rest of us. There is a reason that “quitter’s day”, the day when people give up on their New Year’s resolutions, is just the second Friday in January. We all need constant encouragement to keep up good behaviors and turn them into habits. Future studies will need to focus on windows of opportunity and optimizing the number of interactions over fixed time that can consistently maintain gains in symptom amelioration. Honestly, even if CBT only causes small changes in behavior and perceptions of pain/anxiety/depression, the benefits appear to remarkably outweigh any risks.
Conclusion
In this RCT, cognitive behavioral therapy and pain coping skills training showed a modest but clinically significant improvement in pain symptoms for a majority of participants with a history of ESRD. Implementation of this important result remains a challenge to solve.
Summary Prepared by
Brian Rifkin
Hattiesburg Clinic, Hattiesburg MS
Reviewed by Cristina Popa, Swapnil Hiremath and Jade Teakell
Header Image created by AI, based on prompts by Evan Zeitler