Approach to Quality Control Used for the HOPE Consortium Trial to Reduce Pain and Opioid Use in Hemodialysis

High-quality data are essential to ensure the validity and generalizability of findings generated from randomized clinical trials (RCTs). After decades of ranking among the lowest specialties for the number of RCTs conducted,1,2 nephrology has recently seen a remarkable growth in RCTs. One component essential to the conduct of RCTs is quality control, which helps ensure that the trial is executed as designed. The Trial to Reduce Pain and Opioid Use in Hemodialysis (HOPE Consortium Trial) offers insights into monitoring quality and addressing implementation challenges in a multicenter behavioral intervention trial for chronic pain in patients on hemodialysis. The HOPE Consortium Trial was a National Institutes of Health–funded multicenter trial designed to test whether a cognitive behavioral therapy (CBT)-based intervention known as pain coping skills training (PCST), compared with usual care, improved pain interference among individuals receiving maintenance hemodialysis.3 The PCST intervention consisted of 12 weekly, remotely delivered coach-led CBT sessions followed by an additional 12 weeks of interactive voice response (IVR) sessions delivered via daily automatic phone calls. Patient-reported outcomes (PROs) were ascertained using computer-assisted telephone interviewing administered by a centralized team.4 The trial demonstrated a greater reduction in pain interference (primary end point) in the PCST group. Favorable effects of the PCST intervention were also noted for quality of life, depression, and anxiety. To keep findings broadly generalizable, the trial had few exclusions and enrolled patients regardless of treatment adherence. Success, therefore, required a coordinated quality control approach that drew on expertise from nephrologists, pain specialists, behavioral scientists, biostatisticians, clinical trialists, and patients. To ensure data accuracy and adherence to best practices, we formed the HOPE Consortium Trial Quality Control (QC) Committee, which included multidisciplinary researchers, clinicians, and a patient advisor. Meeting monthly (or more often as needed), the Committee reviewed trial progress on enrollment, retention, death and nondeath withdrawals, intervention delivery, data collection, and staff workload and addressed challenges as they arose. As expected in a large, complex trial—especially during coronavirus disease 2019—several issues emerged. With an objective of informing quality control efforts in future clinical trials, we discuss three specific insights here. Completion of Questionnaires for PROs The trial used a centralized telephone-based approach for ascertainment of the PROs to ensure a standardized approach to the administration of the questionnaires by individuals masked to the participant's treatment assignment. Early in the trial, the QC Committee realized that completion rates for the PRO questionnaires, which included the primary outcome of the study, should be a focus of its efforts. Both the scorability of the questionnaires and the timeliness of completion were monitored. Timeliness of questionnaire completion was viewed by the Committee as important because of the hypothesized temporal relationship between the administration of the intervention and the outcomes. The rates of completion and the scorability of questionnaires were consistently high. Indeed, completion of questionnaires for the primary outcome was 100% among participants who had not withdrawn or died. However, it was soon evident that an increasing number of PRO ascertainment sessions were being completed outside the prespecified target windows. Root cause analysis was, therefore, conducted by the Committee in consultation with the PRO ascertainment team and the Data Coordinating Center. This effort identified the primary barrier to timely completion as difficulty scheduling telephone encounters among a medically complex population. Participants often reported that they were not aware that attempts were being made by the remote team to contact them by telephone. To address this issue, a comprehensive action plan was developed. The plan included incorporation of software to provide real-time data to the call schedulers, greater involvement of research coordinators in scheduling and reminding participants of their PRO ascertainment calls, keeping study phones at dialysis units, and expansion of the PRO ascertainment team. These efforts allowed the timely completion rates to stabilize at approximately 65%. Although using a centralized telephone-based approach has many conceptual and empirical strengths when collecting PROs on a large scale, timeliness of questionnaire completion—in contrast to completion itself or scorability (and therefore usefulness) of the selected outcome instruments—proved to be a significant challenge. Future trialists should be cognizant of this potential challenge. Target window periods for completion of assessments should be carefully planned to allow sufficient time for their completion, research coordinators should be prepared to expend effort in assisting with call scheduling, and adoption of call-scheduling software should be considered. Completion of Study Visits By Research Coordinators Although the intervention was delivered remotely by coaches trained in delivering CBT-based interventions, the study design called for ten visits (one visit every 4 weeks) to be conducted by research coordinators for each participant during the 36-week trial period. These visits were required to document medication changes, ascertain adverse events, and facilitate the scheduling of the initial PCST session and the PRO ascertainment sessions. Given that our trial enrolled participants on maintenance dialysis at 16 academic centers and 103 outpatient dialysis facilities, this was a significant undertaking. Missed study visits due to skipped dialysis treatments and hospitalizations were common particularly considering that the trial was conducted during the coronavirus disease 2019 pandemic but not necessarily modifiable by the research team. The QC Committee was, therefore, uncertain about the percentage of research coordinator study visits that should be expected to be completed. The Committee settled on a relatively high bar for the research coordinators of completion of 90% of the study visits. As the trial progressed, it became clear that achieving high visit completion required significant effort. If any site's completion rate dropped below 90%, the Committee reviewed site-specific data and worked with staff to identify and address modifiable issues, such as staffing challenges. This approach raised the overall visit completion rate to above 90% across sites. We can now highlight that completion of 90% of visits is a realistic goal for future studies that are conducted among individuals receiving hemodialysis, even in trials in which participants are not highly selected. However, reaching this goal requires attention and planning. The high hospitalization rate and high death rate among individuals receiving hemodialysis can significantly limit the rate of study visit completion. Retention rate in our trial was high (<5% of participants withdrew from a cause other than death), suggesting that study activities were not overly burdensome and, possibly, that the study participants valued the regular interactions with the research teams. By contrast, participant deaths were responsible for about three quarters of withdrawals, a sobering reminder to planners of trials in hemodialysis. PCST and IVR Completion Thresholds PCST had been used in nephrology trials before the HOPE Consortium Trial, providing a basis for reasonable delivery target for this intervention.5,6 The target dose of PCST needed for an intended effect in our trial, however, was not known during the planning phase. A mean of 10.8 (of 12) PCST sessions were delivered to the HOPE Consortium Trial participants who were randomized to the intervention; the 25th and 75th percentiles were 10 and 12, respectively. Considered in another way, >80% of participants in this arm received at least 80% of the planned PCST sessions. This was concordant with previous studies and can serve as a handy metric for planning and estimating the power of future trials.5,6 By contrast, less was known about how acceptable IVR would be to participants and how much of IVR would be needed to have the intended effect of helping participants maintain the skills learned during the coach-led PCST sessions. As an automated system, IVR is inherently less amenable to corrective action by trialists and research coordinators. At the end of the trial, 53% of IVR calls were completed, and >93% of participants participated in at least one IVR call. Given that IVR appeared to preserve, at week 24, the effect of the 12-week PCST intervention, these IVR thresholds, although modest, may form a basis for realistic expectations for future trials involving IVR. Moreover, regular review of completion rates by the QC Committee encouraged a warm handoff during the PCST phase, with participants getting a mock IVR call during their session time so that the coach could help navigate any technical issues. Implementing quality control principles during the planning and execution of clinical trials has important advantages.7,8 Our experiences, formed from the adoption of a comprehensive approach to trial quality which incorporated a multidisciplinary team and fostered regular communication, provide insights for future trials. We recognize that applying QC principles in clinical trials can be difficult when resources are limited. To address this, we relied on practical approaches such as holding committee meetings virtually instead of in person and using the expertise within our trial consortium rather than creating an additional external group. These strategies are especially relevant in resource-constrained settings. We also acknowledge that our emphasis on intervention fidelity and other quality control measures may limit the generalizability of the findings. Future real-world and implementation studies will be important to assess the intervention's effectiveness outside of a traditional clinical trial setting. In summary, quality control in trials strengthens scientific rigor, ensures responsible use of resources, builds public trust, and improves evaluation of an intervention's efficacy. In addition, quality control activities can provide insights into potential challenges to implementing the intervention in real-world practice.