ABSTRACT
Introduction High-quality clinical trials are needed to establish the safety, efficacy, and real-world use of potential therapies for acute kidney injury (AKI) prevention. In this consensus workshop, we identified patient and caregiver priorities for recruitment, intervention delivery, and outcomes of a clinical trial of cilastatin to prevent nephrotoxic AKI.
Methods We included adults with lived experience of AKI, chronic kidney disease, or risk factors for AKI (e.g., critical care hospitalization), and their caregivers. Using a modified nominal group technique approach, we conducted a series of hybrid in-person/virtual discussions covering 3 clinical trial topic areas: (1) consent and recruitment; (2) intervention delivery; and (3) trial outcomes. Participants voted on their top preferences in each topic area, and discussion transcripts were analyzed inductively using conventional content analysis.
Results Thirteen individuals (11 patients, 2 caregivers) participated in the workshop. For consent and recruitment, participants prioritized technology enabled pre-screening and involvement of family members in the consent process. For intervention delivery, participants prioritized measures to facilitate intervention administration and return visits. For trial outcomes, participants identified kidney-related and other clinical outcomes (e.g., AKI, chronic kidney disease, cardiovascular events) as top priorities. Analysis of transcripts provided insight into care team and family involvement in trial-related decisions, implications of allocation to a placebo arm, and impact of participants’ experiences of AKI and critical illness.
Conclusion Findings from our workshop will directly inform development of a clinical trial protocol of cilastatin for nephrotoxic AKI prevention and can assist others in patient-centered approaches to AKI trial design.
INTRODUCTION
People living with chronic diseases frequently require hospitalization, where they are exposed to a variety of medications.1–4 Some of these medications, such as chemotherapeutics for treatment of cancer, immunosuppression for organ transplantation or autoimmune conditions, antibiotics for infections, and contrast dyes for imaging procedures, can cause acute kidney injury (AKI) in up to 25% of patients, especially when used in combination.5–7 The consequences of AKI include poor patient outcomes such as prolonged hospitalization, chronic kidney disease (CKD) and kidney failure, cardiovascular events, and death,6, 8, 9 as well as health system burden related to high acute and long-term chronic disease care needs.10 However, beyond avoidance of potentially implicated medications, which is often neither safe nor feasible for hospitalized patients, no pharmacologic therapies are currently available for prevention of nephrotoxic AKI.5
Uptake of nephrotoxins in the proximal tubules of the kidneys is a major contributor to the pathogenesis of AKI.11–14 A small molecule called cilastatin can prevent tubular drug uptake and kidney injury through its inhibitory action on two proteins (dipeptidase-1 and megalin).15–17 While cilastatin prevents kidney injury in cell culture and animal models of nephrotoxic AKI, it has been only indirectly tested in human trials using a combined formulation approved for clinical use – i.e., imipenem-cilastatin, where cilastatin prevents tubular degradation and extends the action of the antibiotic, imipenem.18 In a recent systematic review, pooled results from 10 studies showed lower risks of AKI and better kidney function among patients treated with imipenem-cilastatin compared to inactive or active controls.19 Given the lack of alternatives and evidence suggesting fewer adverse events with cilastatin alone than with imipenem-cilastatin formulations already approved for clinical use,19 a well-designed and adequately powered trial is needed to establish the efficacy of cilastatin alone for this indication.
Over the past two decades, international initiatives have increasingly integrated the perspectives of patients, or people with lived experience of a health condition, into research activities.20 Though patients can meaningfully contribute their expertise as users of the health system at any stage of research, their engagement from the beginning to develop protocols can enhance research feasibility and relevance.21 Evidence for involving patients in the co-design of clinical trials includes several benefits, ranging from more patient-centered recruitment practices, informational materials, and outcome selection, to improved experiences for research participants and greater adherence to the trial intervention.22–24 Given the individual and health system burden of nephrotoxic AKI and demand for novel therapeutic agents for AKI prevention, integrating patient preferences into the design of an interventional AKI clinical trial is critical for supporting rigorous, pragmatic, and patient-centered research. Therefore, we undertook a consensus-based workshop with people with lived experience of AKI or risk factors for AKI to identify preferences and priorities related to recruitment, intervention delivery, and outcomes for a clinical trial of cilastatin to prevent nephrotoxic AKI.
METHODS
Study design and setting
We held a half-day hybrid in-person/virtual workshop in December 2023 at the University of Calgary. Two thirds of participants attended the workshop in person with the remaining participants attending virtually using the Zoom™ platform. We used a modified nominal group technique (NGT),25 an accepted consensus building approach, to generate and prioritize preferences related to the design of a clinical trial of cilastatin for nephrotoxic AKI prevention among people with lived experience of AKI or risk factors for AKI. During the workshop, three vignettes (i.e., clinical scenarios involving fictional patients and caregivers) were used to help focus and guide discussions related to three key aspects of clinical trial design: recruitment and consent, intervention delivery, and outcomes (Supplementary Table S1). This study was approved by the Conjoint Health Research Ethics Board at the University of Calgary (REB23-1564).
Participants and Recruitment
We recruited 13 adult participants who were comfortable communicating in English and who had either experienced or cared for someone with AKI, CKD, and/or risk factors for nephrotoxic AKI. We purposively sampled participants from nephrology and critical care patient advisory groups in Alberta, Canada and among participants from related research expressing interest in being contacted about future studies. Research team members distributed email invitations to potential participants and responded to those indicating interest with additional information about the workshop.
Participants were provided with packages by email one week before the workshop. These packages included a summary of the topic area, workshop agenda, three vignettes (Supplementary Table S1), a consent form, and instructions for parking, lunch, and use of the virtual platform, if required. In the topic summary, definitions and examples of areas for discussion (i.e., recruitment/consent, intervention delivery, and outcomes) were provided. We asked participants to review the vignettes in advance of the workshop and reflect on how a clinical trial in AKI could be designed with the example patients’ and caregivers’ needs in mind. Research team members were available by phone or email to answer questions prior to the workshop and to troubleshoot issues in real time for virtual attendees. All participants provided written informed consent prior to workshop commencement.
Data collection
An overview of the workshop phases and flow is provided in Figure 1. First, one facilitator (MJ) welcomed participants, explained the purpose of the workshop, and provided a program overview. One virtual and two in-person groups were established, each with 4-5 participants as well as a facilitator (MJ, MJE, KF) and note-taker (EB, SL, BR) with advanced training in qualitative and workshop methodology. Each group participated in three separate small-group discussions for each topic area of trial design – 1) consent and recruitment, 2) intervention delivery, and 3) outcomes. Experienced facilitators led the small-group discussions using a topic guide (Supplementary Table S2) and one of the vignettes to guide the conversation (Supplementary Table S1). Facilitators ensured participants had equal opportunity to contribute by directly inviting them to share their thoughts and views, redirecting the flow of the conversation, and refocusing the discussion around the vignette when required. Following each small group, a facilitator or group participant summarized key points from their group’s discussion for the larger group. Before the final prioritization exercise, research team members met to consolidate and categorize preferences within each topic area.
Using cumulative dot voting,26, 27 participants were asked to vote on three preferences that they considered most important under each of the consent and recruitment, intervention delivery, and outcomes topic areas (for a total of 9 dots per participant). Participants voted by either placing a physical dot beside their choice (in person) or by selecting their preferred options using the polling feature on Zoom™ (virtual). All discussions were audio-recorded using a handheld recording device and transcribed by an experienced transcriptionist. One week after the workshop, participants were invited by email to provide their feedback on the workshop format and processes in an evaluation survey (Supplementary Table S3).
Data analysis
We summarized demographic data provided by participants descriptively. Preferences were ranked in each topic area by tallying the total number of votes and ranking results as high (≤7 votes), medium (3-6 votes), or low (<3 votes) priority. Priority categories were determined based on the number of workshop participants, available selections within each topic area, and results of other similar consensus-based exercises.28, 29 Results from the post-workshop evaluation were summarized descriptively (Supplementary Figure S1).
Transcripts from the small and large group discussions were reviewed and inductively analyzed to elaborate on the prioritization results and additional insights raised during group discussions. Using conventional content analysis,30 three research team members (MJE, SL, BR) reviewed the transcripts independently and discussed them as a group to generate a list of relevant codes representing distinct ideas. These codes were then sorted into categories, or key concepts, within each of the three topic areas. The key concepts were refined further through discussion among the broader research team and review of handwritten field notes taken during the workshop that captured additional non-verbal cues and group dynamics. We ensured methodological rigour through our transparent and reflexive approach to data collection and analysis, systematic application of consensus-based methods with experienced facilitators, researcher and data triangulation, and provision of rich descriptions to support our findings.31
Patient Engagement
Two patient partners (DB, ML) with lived experience of AKI and/or CKD were part of the core research team supporting development of the cilastatin trial protocol. Both collaborated on the design, conduct, interpretation, and reporting of this project and participated in the workshop. Another patient partner (HD) was the co-lead of the Nephrology Research Group’s Patient and Community Partnership at the University of Calgary and helped develop and coordinate the consensus workshop. Patient partners reviewed final outputs and contributed to manuscript preparation. We shared a graphical summary of the prioritization results and thematic findings with all workshop participants one month after the workshop and invited them to provide feedback, offer alternative interpretations, and request clarification. We have reported this work in accordance with the Guidance for Reporting Involvement of Patients and Public (GRIPP2; Supplementary Table S4).32
RESULTS
Thirteen people participated in the workshop, including four with prior AKI, seven with CKD, six with conditions putting them at risk of AKI (e.g., cardiovascular disease, diabetes, nephrotoxic medication exposure), and two with experience of caregiving for a person with AKI or CKD (Table 1). Seven participants (54%) identified as women, seven (54%) were greater than 65 years of age, and 7 (54%) were retired. Most participants (69%) resided in an urban location. Reduced kidney function (i.e., eGFR of 30-60 mL/min/1.73m2) at the time of the workshop was reported by seven (54%) participants, either among themselves or the corresponding patient for caregiver participants; two (15%) patients had received a prior kidney transplant. In the following sections, we summarize results from the prioritization exercise and key concepts arising from small- and large-group discussions in relation to identified priorities (Tables 2 and 3).
Preferences for trial recruitment and consenting processes
Within the recruitment and consent topic area, participants highly prioritized healthcare team members’ access to electronic health records for identifying and recruiting eligible patients to the trial (11 votes). This included acceptance of a waiver of consent for access to health data to screen for eligibility based on potentially nephrotoxic medication exposures (i.e., technology enabled pre-screening). Participants also prioritized the ability of family members to provide informed consent for trial participation on behalf of patients who are critically ill and/or unable to provide consent themselves (8 votes).
During group discussions, participants emphasized the importance of rapport and “trusted relationships” with personnel approaching patients about trial participation, regardless of their role. Participants discussed wanting assurances from their care team that the intervention would be in their best interests and clear communication of the expectations of trial participation. Participants favoured the use of electronic health records for eligibility screening as a way of streamlining recruitment and avoiding delays in initiating a potentially life-saving therapy. While participants acknowledged privacy concerns, they suggested the benefits of prompt participant identification and recruitment through access to limited and necessary electronic health data outweighed these risks. Participants indicated that consent for trial participation should be provided by patients themselves with involvement of family members where possible, but that informed consent from family members or trusted physicians, who are responsible for “life and death decisions for you”, would be acceptable if patients’ condition precludes active care participation (e.g., sedated, critically ill, ventilated).
Preferences for intervention delivery
In the prioritization exercise, most participants indicated that placement of a new intravenous access would be acceptable if needed to deliver the trial intervention (10 votes). Participants also identified the provision of support and reimbursement for return trial visits, whether for further intervention doses or for monitoring, as a high priority (9 votes).
Participants from all three small groups raised concerns about the acceptability of the trial’s placebo arm without prompting. Given the high stakes of AKI and potential benefits of the intervention, participants indicated it would be important to communicate to patients upfront that they have a 50% chance of receiving cilastatin and why a trial designed in this way is needed to establish its safety and efficacy. Participants prioritized their participation in a trial with the potential to avert AKI over details regarding precisely how and when cilastatin would be administered. However, they did indicate that use of an existing intravenous (IV) cannula would be preferrable to placing a new one, particularly for patients with difficult peripheral venous access or needle phobia, and that trial medication dosing should be coordinated with other routine care activities. Small group discussions also covered the convenience of intervention dosing and study monitoring as an inpatient, where patients are “continually monitored” and “hooked up to an IV”, and preferences for receiving the intervention drug only while hospitalized or, if required after discharge, through home visits. This related to participants’ expressed concerns about the safety of long-term intravenous cannulation and logistical challenges of having to return to hospital for ongoing treatments. They also relayed anticipated challenges to surveillance after hospital discharge, such as travel, parking costs, and lost wages, for which they expected the study team would provide support, as well as a preference to minimize facility-based study visits.
Preferences for trial outcomes
Top priorities for trial outcomes included short- and long-term measures of kidney function (e.g., AKI, need for dialysis; 10 votes) and other clinical events (e.g., cardiovascular events, death; 8 votes).
Discussions about trial outcomes centered on averting adverse renal outcomes, specifically preventing AKI, AKI progression, and need for dialysis and leveraging routinely collected clinical and laboratory data for outcome ascertainment. Kidney and other clinical endpoints were largely discussed in relation to the complexity of hospitalized patients with AKI and the anticipated negative impact on quality of life and mental and physical health. Although quality of life outcomes were not prioritized during the voting exercise, participants discussed attaining one’s previous level of functional and mental wellbeing as an important long-term outcome.
They also identified a need to measure quality of life objectively using validated tools (e.g., patient-reported outcome measures) and in a way that considers the impact on both patients and caregivers. While they did not express a preference for specific instruments or tests to ascertain outcomes, participants valued trends in kidney function, quality of life, and functional status over time. Participants also raised concerns about the long-term safety of cilastatin and trial participation. They suggested that defining “expectations going forward” for monitoring over the course of the trial, including timing, responsible care team members, and long-term safety, would reassure participating patients and families and those considering enrolling in the trial.
Evaluation survey
All participants (n=13) completed a post-workshop evaluation survey (Supplementary Table S3 and Figure S1) and indicated the workshop goals were communicated clearly and the materials were presented in an organized and well-paced way. Eleven participants stated that their opinions were captured in the large-group report-back summaries, and all but one participant felt the final voting results reflected the opinions and preferences discussed during the workshop. Three participants did not feel the vignettes added value to the workshop, with one suggesting they may have detracted from discussion about the experiences of the individual patient and caregiver participants.
DISCUSSION
This consensus workshop was undertaken to explore the preferences and priorities of people with lived experience of AKI or risk factors for AKI, and to integrate these perspectives into the design of an upcoming trial for nephrotoxic AKI prevention. Over a series of group discussions and final voting exercise, participants identified priorities across aspects of trial design that included, most notably, the use of technology (i.e., access to electronic health records) and involvement of trusted individuals in trial recruitment, logistics of intervention administration (i.e., IV access, during hospitalization), support for study follow up, and emphasis on kidney-related and quality of life outcomes. While this workshop centered on the proposed trial intervention, cilastatin, our findings can also help other trials for AKI develop patient-centred approaches to recruitment and consent processes, intervention delivery, and outcome selection.
Low accrual rates and delays in identifying eligible patients put the viability of clinical trials at risk, and both are common when research staff must manually find participants for trials. Pre-screening potential participants for clinical trials can increase the efficiency of trial recruitment by quickly identifying people who may qualify for a study prior to approaching them and proceeding with informed consent. Technology enabled pre-screening is an increasingly accessible way to ensure systematic and timely identification of potential participants for modern-day trials as the availability of digital clinical information systems expands.33 Although privacy legislation governs how health data can be accessed in each jurisdiction, patient perspectives on use of such digital approaches are important to consider, as they may affect the willingness of patients to participate in trials. Technology enabled pre-screening is particularly relevant to trials seeking to enroll people with or at risk of AKI, as this population is widely distributed across hospital units and clinical services, making traditional manual approaches inefficient.
Participants identified the use of technology enabled pre-screening as a priority and considered access to this information by health care providers or members of the research team as acceptable. In fact, the use of algorithms within computer clinical information systems to identify people who would meet the eligibility criteria for a trial was the top ranked option for trial recruitment by our participants. Our findings are generally consistent with other research describing patient perspectives on the use of digital health tools in health research.34–36 In a recent literature review, Kassam et al. reported that most studies exploring this topic found that participants were willing to share personal health information digitally for clinical research provided there was clarity about who can access the information, for what purpose, and how privacy will be ensured.34 In a survey of the general population in the United States, Kim et al. found that participants were more willing to have their electronic health data shared when participants believed it would improve research quality and when they valued research benefit over privacy.35 A recent national survey exploring consent in the digital health ecosystem in Canada found that most respondents preferred data sources to be accessible by health care providers and delegates as the default option.36 These reports, in conjunction with our findings, imply a general patient acceptance and support for the use of technology enabled pre-screening approaches for clinical trials in AKI when the value of access to the information for the success of the trial is clear, and privacy and security concerns are appropriately addressed.
Participants in our workshop recognized the potential to avert nephrotoxic AKI through trial participation, which meant most, if not all, were willing to comply with processes for administering the intervention if recommended by research and care teams. However, they made suggestions for integrating intervention delivery with other routine care activities while in hospital to minimize burdens of clinical trial participation, such as need for additional testing and/or return visits, financial challenges, and concerns about safety.37, 38 Embedding trial processes within clinical workflows, such as timing cilastatin administration with other medications or coordinating laboratory tests with routine inpatient bloodwork, could further reduce the burden for healthcare team members and help to promote trial feasibility.39 Important concerns raised by participants about allocation to a placebo arm when the cilastatin intervention might help prevent poor AKI-associated outcomes align with patients’ perspectives on rare disease trial design from a qualitative study by Gaasterland et al.40 As patients view a novel intervention as a source of hope, the possibility of not receiving a potentially beneficial intervention may compromise this hope, the perceived benefits of trial participation, and ultimately patients’ willingness to enrol in trials of this design.40 In acknowledging these perspectives early in the design phase, recruitment materials and communication strategies can be co-developed with patients to clearly articulate trial processes and justification for a well-designed, placebo-controlled trial to establish intervention efficacy and safety.
Several narrative reviews have been published on the selection of outcomes for AKI, although these papers have focused on methodological aspects of outcome measures, limitations with the clinical criteria for ascertaining AKI, and requirements for regulatory drug approval.41–46 These existing publications have been exclusively written from the perspectives of clinical researchers, and although reports have called for greater patient participation in the study design of trials for AKI,42 few studies have explored patient priorities for AKI trial outcomes. The Standardised Outcomes in Nephrology (SONG) initiative is an international project that aims to establish core outcome measures across the spectrum of kidney disease for trials and other forms of research based on the shared priorities of patients, caregivers, clinicians, researchers, and policy makers.47 Although the SONG initiative has not addressed outcomes for AKI trials, findings from a focus group study conducted with patients and care providers of people with CKD may be relevant to trials for AKI, including the high priority assigned to outcomes of kidney function, mortality, fatigue, life participation, and mental health.48 We similarly found that patients prioritized measures of kidney function, both short and long term, as the top ranked trial outcome, followed by clinical outcomes including survival, cardiovascular events, and kidney failure for AKI trials. Notably, our findings from patients align with the most recent recommendations from AKI trialists, which highlight the occurrence of AKI as a key endpoint for phase 2B prevention or attenuation trials, and major adverse kidney events (including death, dialysis, or a sustained reduction in kidney function) for phase 3 AKI prevention, attenuation, or treatment trials.41–43
Our study is strengthened by the involvement of people with lived experience in workshop organization and the capture of diverse experiences and perspectives related to AKI. However, we acknowledge some limitations. First, the time allotted for small-group discussions may have been insufficient for participants to reflect and elaborate fully on important experiences, and some participants may have felt uncomfortable sharing their perspectives in this forum. We used skilled facilitators encouraged respectful interactions and ensure all participants had the opportunity to contribute. Second, while use of vignettes helped to focus small-group discussions on each of the topic areas, they may have unintentionally underemphasized participants’ own experiences. Third, the priorities brought forward to the voting exercise were compiled in real-time following small-group discussions, which means that preferences expressed by one participant or not discussed at length may not have been captured among voting options. However, results from the workshop evaluation survey suggest that participants felt the outcome reflected the content of group discussions. Fourth, the views and priorities of participants, who were largely white, cisgender, and highly educated, may not reflect those of underrepresented groups who are at risk of nephrotoxic AKI. Although priority categories are broad and discussions did cover aspects of socioeconomic disadvantage and health inequity, this area warrants future dedicated study. Lastly, the hybrid workshop format may have influenced the quality of interactions differentially between in-person and virtual attendees, although this drawback is outweighed by the inclusivity and diversity of participation enabled by the hybrid approach.
CONCLUSION
In our consensus workshop, patients and caregivers prioritized technology enabled pre-screening and integration of trial processes and intervention delivery with routine care activities to streamline participation in a clinical trial of cilastatin for preventing nephrotoxic AKI. Participants’ prioritization of kidney-related and other clinical endpoints related in large part to their desire to avoid sequelae of AKI, such as dialysis dependence, and restore physical and mental wellbeing following hospitalization. The perspectives shared by patients and caregivers will uniquely inform development of our clinical trial protocol and can also help others to develop patient-centered approaches for recruitment and consent, intervention delivery, and outcome selection for AKI trials.
FUNDING
The study was funded by the Canadian Institutes of Health Research (CIHR) Team Grant: Intervention Trial in Inflammation for Chronic Conditions - Evidence to Impact; Funding Reference Number LI3 189373.
DATA SHARING
All data produced in the present study are available upon reasonable request to the authors.
DISCLOSURE
A pre-print version of this manuscript is available at MedRxivs: https://www.medrxiv.org/content/10.1101/2024.03.06.24303823v1
CONFLICTS OF INTEREST
The authors report no conflicts of interest.
Data Availability
All data produced in the present study are available upon reasonable request to the authors.
Footnotes
Minor changes to text and manuscript format only.