Behind the Scenes: Key Lessons Learned from the RELIEVE-AKI Clinical Trial

Abstract

Continuous kidney replacement therapy (CKRT) is commonly used to manage critically ill patients with severe acute kidney injury. While recent trials focused on the correct dosing and timing of CKRT, our understanding regarding the optimum dose of net ultrafiltration is limited to retrospective data. The Restrictive versus Liberal Rate of Extracorporeal Volume Removal Evaluation in Acute Kidney Injury (RELIEVE-AKI) trial has been conducted to assess the feasibility of a prospective randomized trial in determining the optimum net ultrafiltration rate. This paper outlines the relevant challenges and solutions in implementing this complex ICU-based trial. Several difficulties were encountered, starting with clinical issues related to conducting a trial on patients with rapidly changing hemodynamics, low patient recruitment rates, increased nursing workload, and the enormous volume of data generated by patients undergoing prolonged CKRT. Following several brainstorming sessions, several points were highlighted to be considered, including the need to streamline the intervention, add more flexibility in the trial protocols, ensure comprehensive a priori planning, particularly regarding nursing roles and their compensation, and enhance data management systems. These insights are critical for guiding future ICU-based dynamically titrated intervention trials, leading to more efficient trial management, improved data quality, and enhanced patient safety.

Introduction:

Acute kidney injury (AKI) is a frequent complication in critically ill patients, with incidence rates varying from 35% to 57% in the literature [1–4]. Approximately 23% of AKI patients require kidney replacement therapy (KRT), with continuous kidney replacement therapy being the predominant method in the intensive care units (ICU) [4]. Despite improvements in technology and contemporary knowledge regarding safer and more effective dialysis modalities, mortality rates among AKI patients remain high [5–7]. Several factors contribute to the high mortality in AKI patients requiring acute KRT, including advanced age, multiple comorbidities, hemodynamic instability, electrolyte disturbance, severity of their critical illness, and multiorgan failure [8, 9]. CKRT procedural-related factors may influence adverse event rates, such as catheter complications [10], inconsistencies of the utilized dialysis dosages [11, 12], the inappropriateness of CKRT initiation time [13–15], inappropriate replacement fluid composition for the individual patient’s metabolic needs [16], and finally, inadequacies of the net ultrafiltration rate [17].

Optimal fluid management is one of the essential pillars in the care of critically ill patients who have developed AKI and fluid overload. Several studies demonstrated that critically ill patients who experienced positive fluid balance and fluid overload have a significantly higher mortality rate [18–22]. Among critically ill patients with severe AKI and volume overload, CKRT is a cornerstone in fluid management. It provides net ultrafiltration (UF_NET), which aims to achieve euvolemic status. Previous studies demonstrated that patients with fluid overload who achieve higher UF_NET rates exhibit lower mortality rates [22–24]. In addition, recent findings indicate that UF_NET rates <1.01 or >1.75 mL/kg/h were associated with a lower survival rate [25], longer CKRT duration, and extended ICU stays [26]. While the association between UF_NET and mortality remains as a “J-shaped” curve [25], the optimal UF_NET rate in CKRT remains undetermined. No clinical trial has examined the effectiveness of UF_NET rates on long-term patient-centered clinical outcomes. Moreover, surveys indicate that over two-thirds of clinicians are willing to enroll patients in a clinical trial of protocol-based UF_NET [27]. Therefore, the REstrictive versus Liberal rate of Extracorporeal Volume removal Evaluation in Acute Kidney Injury (RELIEVE-AKI) study [28] was designed to assess the feasibility of implementing alternative UF_NET rates in adult critically ill patients with AKI who are undergoing CKRT.

The RELIEVE-AKI [28] is a pilot prospective, two-center, unblinded, parallel-group, comparative effectiveness, stepped-wedge cluster-randomized trial (ClinicalTrial.gov study number NCT05306964) to assess the feasibility of a large multicenter trial to compare two different UF_NET rates. The trial has been conducted across 10 ICUs at two hospital systems - 5 ICUs at the University of Pittsburgh Medical Center in Pittsburgh, Pennsylvania, and 5 ICUs at Mayo Clinic in Rochester, Minnesota. During each period of the study, ICUs are randomized to either a restrictive or liberal UF_NET strategy. For the first period, all ICUs were in the liberal arm. Subsequently, in each successive period, one ICU is randomized to the restrictive UF_NET arm. In both trial arms, the UF_NET rate is initiated at 0.5 mL/kg/h and gradually increased and maintained within the target range. In the liberal arm, the UF_NET rate is titrated to 2.0-5.0 mL/kg/h and maintained while patients have a clinical indication for fluid removal; in the restrictive arm, the UF_NET rate is maintained at 0.5-1.5 mL/kg/h (Figure 1).

Figure 1.

The study Intervention.

Our objective in this review is to assess the challenges and constraints encountered in conducting this type of study focusing on implementing CKRT and fluid management. We aim to comprehensively evaluate, and share learned insights on the challenges faced during the conduct of this complex trial, as well as potential solutions that could enhance the design and implementation of similar future studies (Figure 2).

Figure 2.

RELIEVE-AKI challenges and potential solutions.

1. Low recruitment rates during initial periods

One of the initial challenges faced in the RELIEVE-AKI trial was the low recruitment rate of eligible patients. In the first two months of the RELIEVE-AKI trial, both participating sites could enroll three patients in total. We implemented several interventions to change the recruitment rates.

Solution 1.1. Changing recruitment strategy

To recruit more efficiently, we employed the SEARRCH (screening electronic records, attending physician, registered nurses, rounds, consent, health information) approach, which involved changes from screening to intervention. Initially, we identified potential participants by meticulously Screening Electronic Health Records (EHR). Our next step involved approaching Attending physicians, including the primary intensivists and nephrologists, asking questions about the fluid removal plan, and obtaining the necessary permissions. For this purpose, we prepared a standard statement for the clinicians’ engagement, indicating: “Dear Clinicians, this patient qualifies for the RELIEVE-AKI trial. This trial is an NIH-sponsored feasibility study to assess differences between UF rates during CKRT. Both arms of the study will have UF rates considered standard of care but at the opposite side of the rate spectrum. This study follows the study protocol as long as patients can tolerate it. As the enrollment criteria require patients to be in equipoise, we ask if you would allow us to approach the patients or their legal representatives for potential enrollment. Please also share your thoughts on whether you plan to continue CKRT for 48 hours and answer the following questions: 1. Should emergent and rapid fluid removal occur now? 2. Should fluid removal be deferred now?” To ensure a sensitive and compassionate approach, we asked Registered nurses familiar with the emotional circumstances of the patients and their families if it was appropriate to discuss study enrollment. If the situation was conducive, our team participated in ICU Rounds with the critical care or nephrology teams before approaching patients or their families. Then, the research coordinator obtained Consent from patients or their legally authorized representatives (LARs), providing a detailed explanation of the trial risks and benefits in layperson’s terms and answering their questions. Finally, we documented consent forms through Health Information Management services, securing accurate and safe record-keeping. This structured plan guaranteed a thorough and ethical patient recruitment process.

Solution 1.2. Creating a brochure to show the trial design

The complexity of the CKRT intervention and the detailed information in the consent forms contributed to a lack of understanding among potential participants, adversely affecting their willingness to enroll in the trial. To address this issue, we created a simplified brochure (Supplemental Material 1). This two-page brochure effectively summarized the study, focusing on the critical elements of the trial, including the intervention and potential risks and benefits, in a straightforward format. This strategy aimed to facilitate a better comprehension of the study, particularly for participants in the ICU facing critical illnesses, thereby reducing patient and family stress and enhancing their willingness to participate.

Solution 1.3. Modifying eligibility criteria

Implementing inclusion and exclusion criteria is crucial in the recruitment process. Following a period of low enrollment rate at the beginning of the trial, we reassessed study eligibility for overly restrictive criteria. In addition to limiting enrollment, these excessively restrictive eligibility criteria could also impact the generalizability of the study results to a larger patient population. In reassessing the eligibility criteria, with the goal of enhancing the recruitment rates while ensuring a more accurate reflection of the real-world population, we implemented some changes in the study protocol. We revised our exclusion criteria by removing a body mass index (BMI) threshold, the presence of a Do Not Resuscitate/Do Not Intubate (DNR/DNI) order, and a history of intermittent hemodialysis during admission. Table 1 indicates the study inclusion and exclusion criteria, and Figure 3 shows the study flow diagram from screening to enrollment. These modifications led to an increase in the number of eligible patients. Notably, we observed excluding obese patients markedly limited the cohort of eligible patients.

Table 1.

Inlcusion and Exclusion Criteria of RELIEVE-AKI Clinical Trial

Inlcusions	Exclusions
1. Age 18 years or older	1. Respiratory distress due to pulmonary edema or fluid overload in unintubated patients
2. Stage 3 acute kidney injury according to the KDIGO criteria	2. Massive volume infusion (i.e., >200 mL/h for >6 hours of continuous infusion)
3. Started or intending to start CKRT for volume management	3. No intention to remove net fluid as determined by attending Intensivist or Nephrologist
4. Attending Intensivist or Nephrologist intending to remove net fluid using CKRT for at least 48 hours	4. Attending Intensivist or Nephrologist believes that the protocol will not be followed
	5. Continuous net fluid removal for >48 hours prior to study enrollment
	6. Patients on chronic outpatient hemodialysis
	7. Patients with history of, or current admission for kidney transplantation
	8. Patients on comfort measures only orders
	9. Moribund not expected to survive >24 hours
	10. Confirmed pregnancy
	11. Patients treated with ECMO, VAD, or IABP
	12. Organ donors with neurological determination of death (i.e., brain dead donors)
	13. Drug overdose requiring CKRT for drug clearance
	14. Enrollment in a concurrent interventional clinical trial with direct impact on fluid balance (e.g., >500 mL study drug administration

Abbreviations: KDIGO: Kidney Disease: Improving Global Outcomes, CKRT: Continuous Kidney Replacement Therapy, ECMO: Extracorporeal Membrane Oxygenation, VAD: Ventricular Assist Device, IABP: Intra-Aortic Balloon Pump

Figure 3.

The Study Flow Diagram from Screening to Enrollment

Solution 1.4. Expanding the participating ICU sites

One of the critical strategies to increase the enrollment rate was expanding the number of ICUs participating in the study. Initially, the focus was primarily on medical ICUs. We subsequently expanded our reach to include surgical and cardiovascular surgery ICUs at both sites, as they also had CKRT patients meeting all eligibility criteria. In addition, including surgical ICUs enhances the generalizability of findings to patients who develop post-operative AKI requiring CKRT. While this modification increased enrollment rates, it posed additional challenges in adherence to the study protocol within these new ICUs. To address this, we provided detailed information to each area and sought out feedback on the process.

Specifically, in cardiovascular surgery ICUs, where patients frequently required prolonged CKRT, coordinating continuous communication between research coordinators and nurses for the study calculator and data entry was challenging. Additionally, the high workload for nursing staff in these ICUs and the unique CKRT plans prescribed by surgical attending physicians due to the complexities of their patients and the need for various procedures, along with the potential use of Extracorporeal Membrane Oxygenation (ECMO) as an exclusion criterion, further complicated recruitment, and intervention. Assessing these challenges is an essential step before incorporating such ICUs in similar future trials. Development of strategies such as conducting preliminary surveys and engaging with ICU staff before in the participating ICUs could mitigate these challenges. This may enhance the successful implementation of the study and the reliability of its findings and ultimately contribute to improving patient safety. This is particularly significant as the study’s results are anticipated to shape future patient management protocols.

Our approach for all involved ICUs began with providing detailed information about the study’s goals and the importance of optimizing fluid removal rates in CKRT patients. This involved organizing webinars, attending nurse meetings, distributing newsletters and emails, creating a brief instructional video on using the study calculator, and providing one-on-one bedside training before the study intervention. This aspect aligns with the Awareness and Desire phases, the initial steps of the ADKAR model [29]. When facing resistance, which is common in the clinical trial process, we provided the team with training in CKRT fluid calculation and indications for rescue procedures to be used in emergent situations, ensuring the flexibility of this study as outlined in the study protocol. Moreover, we encouraged the nurses to share their views and implemented a feedback loop by sending them surveys following patient care. These initiatives are reflective of the Knowledge and Ability stages in the ADKAR model.

2. Intervention Challenges in the RELIEVE-AKI Study

Randomizing patients into a specific CKRT prescription in ICUs can be challenging. The complexity arises primarily because patients in the ICU often have rapidly changing clinical conditions that require constant monitoring and frequent adjustments in their treatment plans. Several unpredictable events (e.g., acute bleeding and recurrent sepsis) can affect the patients’ hemodynamic status, necessitating a careful reassessment of CKRT prescription parameters, including ultrafiltration rates. The ICU team, typically comprised of intensivists, nurses, and nephrologists, would have to collaborate to modify the CKRT prescription to adapt to the needs of critically ill patients and necessitates that physicians override the study protocols to adjust the UF_Net rate. Despite challenges in implementing these types of studies in critically ill patients with unpredictable and complex illnesses, conducting similar research is crucial for developing more personalized and effective patient care approaches in the ICU.

A web-based calculator was developed and provided to nurses to calculate the fluid removal rate as dictated by the study protocol [30]. To maintain patient data confidentiality, access to this tool was regulated by research coordinators at each site, necessitating frequent updates due to the regular turnover of nursing staff. Furthermore, it required dynamic collaboration with nurses, as several adjustments in their daily practices were necessary. These changes include considering IV infusions instead of the total intake for fluid removal calculations and refraining from adding bolus and intermittent medications administered <2 hours into the study CKRT calculator. These modifications increased the nursing staff’s workload, particularly in cardiac and cardiovascular surgery ICUs. Nurses are traditionally trained to incorporate the entire volume intake, i.e., oral and IV, into the EHR flowsheet and use their accustomed calculators. Transitioning to a new calculation method posed a significant challenge. Requiring nurses to adapt their practice to the study requirement requires time and multiple training sessions. Despite providing educational materials on using the study calculator, the time constraints and high-stress environment in ICU often prevented nurses from effectively engaging with these resources.

Furthermore, due to the fluctuating patients’ condition, frequent adjustments in the fluid removal plan were required during shifts. Although the study maintained standard UF rates in both arms, randomizing ICUs into restrictive or liberal arms and expecting nurses to comply with the protocol was a considerable challenge. Similar difficulties were observed in other studies [31], such as the OVATION pilot trial [32], which necessitated continuously titrated vasopressors to maintain patients within higher (75–80 mmHg) versus lower (60–65 mmHg) mean arterial pressure (MAP) targets during shock and is a complex task to adhere to upon admission [33]. Addressing this aspect for dynamic intervention and adherence is crucial before designing any future ICU-based fluid removal trials.

Solution 2.1. Using a Workflow-Aligned Intervention Tool

Given that this dynamic intervention necessitates new training and modifications based solely on IV fluid intake—which differs from routine nursing practices—using a study calculator that more closely aligns with nursing workflows is essential to ensure the intervention is pragmatic and consistent with the current practices. While this approach may minimally influence the accuracy of the UF net rate, it is expected to enhance collaboration.

Solution 2.2. Compensating ICU nurses

Since nurses play a crucial role in implementing intervention and managing their substantial workloads during this trial, compensating ICU nurses for their time and dedicated effort could be an additional solution to alleviate the increased workload associated with this study. At one site, investigators provided snacks as a token of appreciation. In contrast, at another site, they conveyed their gratitude via thank-you emails to the nurses, their supervisors, and ICU administrative directors. Although financial compensation could have been a motivating factor, its implementation was not feasible due to budgetary limitations within the scope of this trial and restrictions related to the health system and university policies.

Solution 2.3. Probable Advances in the Future by Adopting a Large Language Model

Artificial intelligence has already been effectively integrated into clinical practice in multiple settings and aids healthcare professionals in decision-making [34–36]. The increasing adoption of large language models (LLM) in medicine, such as Chat Generative Pre-Trained Transformer (ChatGPT), offers a promising solution to alleviate the nurses’ workload [37]. Integrating LLM technology into research and nursing workflows can assist nurses in diverse tasks, including documentation. This technology can simplify and automate calculations of UF_NET rates and provide real-time monitoring of patients’ data to adjust the fluid removal. This enables nurses to prioritize patient care over time-consuming data entry. Additionally, it can offer instructions and facilitate communication with nurses during dynamic study interventions, providing support for any questions when the main study team is not available at the bedside or online to respond [38]. Hence, this innovative solution can reduce the burden, mitigate burnout risk, and foster better collaboration for future research initiatives. However, further research in this area is necessary to fully understand and optimize the use of LLM technology in clinical research and practice.

3. Data Entry Challenges in the RELIEVE-AKI Study

The extensive data collection process, continuing up to the 28th day of admission following enrollment, encompasses numerous detailed aspects such as daily laboratory results, vital signs, medications administered, and hourly CKRT records. This heavy workload is highly costly and challenging, particularly for patients who require CKRT for extended periods.

Furthermore, uncertainties surrounding complex variables in data entry can lead to significant variability in data collection among the study coordinators. It is essential to simplify the process by providing a clear and concise explanation regarding variables in the protocol. The data coordinator team might not have specialized knowledge in nephrology, and therefore, understanding the complexity of the intake, output, and CKRT intervention calculations is one of the basics of accurate data collection. Addressing this issue is crucial in every trial to ensure accurate and reliable data collection [39].

Solution 3.1. Comprehensive training before trial initiation

We developed comprehensive standard operating procedures (SOPs) and a frequently asked questions (FAQ) document, along with robust definitions for various adverse events, and training before initiating the trial, which significantly reduced confusion and improved the accuracy of data collection, reporting of adverse events, and the overall analysis process.

Solution 3.2. Continuous training and regular meetings

Continuous training and regular meetings were organized to address any questions or uncertainties that emerged during the trial. These sessions played a crucial role in improving the team’s understanding of the data collection process, ensuring they remained informed and updated. This approach is especially vital in long-term trials, where there is a possibility of changes in coordinators or a need for refreshers on the initial training.

Solution 3.3. Consider adding a data entry supervisor

It is crucial to consider adding a data entry supervisor in every trial to enhance data collection accuracy. Given the principal investigators’ numerous project commitments, this designated individual would assume the responsibility of meticulously reviewing the precision and completeness of data entered after each enrollment, providing an invaluable extra layer of quality control and validation.

Solution 3.4. Requirement of an automated data abstraction system from medical record

Implementing an automated data abstraction system from medical records can significantly reduce the burden of manual data entry for research coordinators [40]. This technological solution not only facilitates the process but also minimizes the potential for human error, ensuring data accuracy and consistency [41]. In the validated electronic data abstraction forms, the collected data may be free of human errors. As long as the reliability of the abstracted data is monitored, it may add to the efficacy of similar projects.

By implementing these solutions, the challenges associated with extensive data collection can be mitigated, leading to enhanced data quality and reliability of trial findings.

Discussion:

Given that effective fluid management in critically ill patients who develop AKI continues to be a subject of debate, the need to conduct clinical trials is paramount. Yet, the preparation and execution of such complex trials within the busy ICU setting requires a multidisciplinary collaboration among participants, researchers, and the ICU team. Throughout this process, we faced numerous challenges but also identified and implemented solutions.

RELIEVE-AKI [28], as a pilot study, was designed to assess the feasibility of implementing a dynamic fluid management protocol for patients undergoing CKRT. Similarly, the Optimal Vasopressor Titration (OVATION) pilot trial investigated the effects of higher versus lower blood pressure targets for vasopressor treatment in shock by implementing a dynamic approach to continuously titrate vasopressors to maintain within specific mean arterial pressure (MAP) target, higher (75-80 mmHg) versus lower (60-65 mmHg) [32], and another pilot study by Panwar et al. [42] compared conservative versus liberal oxygenation targets for patients receiving mechanical ventilation. Both studies supported the feasibility of conducting larger-scale trials. Multicenter pilot studies are crucial for uncovering barriers within the ICU across various hospitals, suggesting potential solutions, and confirming the feasibility of conducting large trials. These insights are invaluable for improving future protocols, which will aid and facilitate future trials in similar contexts [31].

Enrollment difficulties are common in critical care settings. A recent systematic review of trials involving Acute Respiratory Distress Syndrome (ARDS) and sepsis revealed that the average enrollment rate was below one patient per site per month [43], which is likely an overestimation, given the review’s focus on common critical illness syndromes and publications in high-impact journals. Several key factors must be addressed to enhance enrollment rates per site. These factors include ensuring a sufficient pool of eligible patients, having an adequate number of research coordinators, and utilizing medical record resources effectively for screening.

Moreover, obtaining consent from patients or their surrogate decision-makers requires a sensitive approach by skilled and compassionate personnel [44]. Providing study information in a short, easily understandable [45] and less confusing content in the form of brochures and discussing this material during ICU rounds in the presence of their trusted clinicians, as practiced in this trial, can further improve the enrollment process. Additionally, expanding the number of sites proved effective in other trials. For instance, the OVATION trial [32] enrolled 118 patients across 11 academic hospitals in Canada and the USA over 15 months, and the HOT-ICU trial [46], which compared a lower-oxygenation group with a partial pressure of arterial oxygen (Pao2) target of 60 mmHg to a higher-oxygenation group with a Pao2 target of 90 mmHg over 90 days achieved enrollment of 2,928 patients by involving 35 ICUs across seven European countries over three years. However, site expansion requires significant efforts and additional funding. In this pilot study, we chose to increase the number of ICUs within each existing site due to budgetary constraints. Therefore, revising the initial protocol to streamline processes and increase flexibility by expanding involved ICUs and adjusting predefined exclusion criteria could enhance enrollment rates and better reflect real-world clinical practices.

Critical care trials often begin with an initial intervention followed by subsequent monitoring. Yet, the complexity of RELIEVE-AKI was amplified by the need for a dynamic intervention throughout the patient’s course on CKRT for up to 28 days, increasing the demand for significant effort and collaboration between the ICU nurses and research coordinators. Challenges such as an increased workload, adjustments in clinical practice (shifting from considering total fluid input to only intravenous fluids in the study calculator), and the lack of financial compensation for nurses could diminish their ongoing cooperation. Over time, these issues may lead to trial fatigue and compromise the precise delivery of the intervention [31]. Similarly, the OVATION [32] and HOT-ICU trials [46] involving complex dynamic interventions were designed to address the uncertainties surrounding the optimal therapeutic targets for blood pressure and oxygenation in critically ill patients, seeking to clarify the potential harms and benefits of varying treatment targets. Likewise, the RELIEVE-AKI trial employs the same approach for fluid removal targets, aiming to determine the best practices for managing fluid balance in this patient population.

Similar dynamic trials on fluid management have been conducted. For instance, the REVERSE-AKI pilot trial [47] compared a restrictive fluid management strategy to usual care among 100 adult critically ill patients with AKI across seven ICUs in Europe and Australia. They found that restrictive fluid management was feasible and resulted in a lower cumulative fluid balance and fewer adverse events compared to usual care. In contrast, the BALANCE trial [48] also compared conservative fluid management with usual care among adults with sepsis. This study was discontinued early following the enrollment of 30 patients, as it failed to meet the predefined threshold of 500 mL difference in mean daily fluid balance. Notably, the primary endpoint in the REVERSE-AKI trial was set at three days. In contrast, the BALANCE trial was extended to 14 days, suggesting that a longer dynamic intervention may impact adherence to the protocol. Critically ill ICU patients often experience dynamic changes in their conditions, including additional organ dysfunction or recovery, hemodynamic variabilities, hospital-acquired infections, and procedural requirements and their related complications. Such variations may require adjustments in management approaches, impacting the adherence to prespecified fluid management goals, particularly in trials with more extended intervention periods. Furthermore, other factors such as patient enrollment in emotionally and environmentally challenging ICU settings, the time of the intervention initiation, and continuous changes in standard-of-care practices may influence the ability to achieve the prespecified threshold for a distinct separation in fluid balance between the study groups. While previous dynamic trials on vasopressors [32], oxygenation [46], and fluid regimens [49] for critically ill patients showed no benefits on mortality, the heterogeneity of critically ill patients in ICUs and the variability in their management suggest that it would be more appropriate to adopt alternative endpoints for outcomes rather than focusing solely on mortality [50].

During dynamic interventions that require nurse collaboration, it is crucial to continuously monitor, train, and encourage the team. This process demands significant effort and time, especially given the constraints of time limitations, heavy ICU workloads, and the ongoing turnover of nurses. Adopting a method that aligns with nursing workflows and current practices might improve their collaboration. Another potential solution that requires more investigation is the adoption of large language models (LLMs) to assist nurses with documentation tasks [37] and communication with the research team, which could improve collaboration for future research initiatives. Despite the challenges mentioned in implementing clinical trials in the ICU, conducting further feasibility studies is essential to fully understand the uncertainties in managing critically ill patients.

Conclusion:

We encountered several challenges in patient recruitment, implementation of study intervention, and complexities in data collection during the RELIEVE-AKI clinical trial. To address these issues, several strategies were employed, such as streamlining enrollment, increasing flexibility in the eligibility criteria, providing extensive training before and during the trial, and ensuring comprehensive planning, particularly regarding nursing roles and their compensation and robust data management systems. The lessons learned from these experiences are instrumental in guiding future ICU-based trials, leading to more efficient trial management, improved data quality, and enhanced patient safety.

Highlights:

• The RELIEVE-AKI trial aims to determine the optimal net ultrafiltration rate for continuous kidney replacement therapy (CKRT) in critically ill patients.

• Challenges in this trial included low patient recruitment rates, the dynamic nature of the intervention, increased nursing workload, and complexities in data collection.

• The lessons learned included the need to streamline the intervention, enhance protocol flexibility, ensure comprehensive priori planning, and improve data management systems.

• These insights are critical for guiding future ICU-based dynamic trials, leading to efficient trial management, data quality, and patient safety.

Availability of data and materials:

Not applicable.