System-wide CRRT quality program targets variation in dialysis delivery for the critically ill
Continuous renal replacement therapy CRRT occupies a difficult place in critical care: it is lifesaving support for patients with acute kidney injury, yet its delivery depends on a chain of decisions, machine performance, nursing expertise, anticoagulation management, and timely troubleshooting. When any link in that chain slips, dose delivery can drift away from the prescription, filter life can shorten, and patients may receive less consistent renal support than intended. The QUALITY CRRT protocol is built around that reality. It proposes that a structured quality assurance and improvement program, applied across an entire healthcare system, can standardize practice and make performance visible enough to drive change.
The study is designed as a multicentre, prospective, stepped-wedged, interrupted time series evaluation. In practical terms, that means participating sites will not all receive the intervention at the same moment, allowing investigators to compare performance before and after implementation while also accounting for changes over time. The implementation package centers on the reporting of CRRT key performance indicators and on aligning programs around common structure, process, and outcome metrics. The protocol is not testing a new dialysis membrane or a novel anticoagulant. It is testing whether better measurement, feedback, and shared standards can improve how CRRT is actually delivered at the bedside.
KPI monitoring is the intervention, not an administrative afterthought
In CRRT, small operational differences can have outsized clinical consequences. Therapy interruptions, suboptimal effluent delivery, inconsistent documentation, and variable escalation pathways can all erode the intended dose of treatment in patients who have little physiologic reserve. A quality program built around recurring KPI review is meant to expose those failures in real time. Standardized reporting also makes it possible to compare units using the same language, which matters in a therapy that is often embedded in different local workflows and staffing models.
The primary outcome is the quarterly performance of CRRT KPIs. Secondary outcomes extend beyond process alone and include patient-centered and economic measures. That combination matters because an intervention can improve a dashboard while leaving the patient experience unchanged, or it can consume substantial resources without producing measurable clinical benefit. By linking process measures to downstream outcomes, the protocol asks whether a more disciplined delivery system can translate into better care rather than cleaner documentation.
The interrupted time series framework is well suited to this question. CRRT quality improvement rarely produces an instantaneous step change; instead, practice tends to shift gradually as teams adopt new habits, supervisors reinforce standards, and clinicians begin to trust the reporting process. An ITS design can capture whether KPIs improve over time after implementation and whether those changes persist rather than fading after the initial enthusiasm of rollout. The stepped-wedge structure adds another layer of realism, since a health-system-wide intervention often has to be introduced in phases rather than all at once.
Why standardization matters in acute kidney injury support
CRRT is not simply “dialysis in the ICU.” It is a complex therapy for critically ill patients with acute kidney injury, often delivered in the setting of hemodynamic instability, multiorgan failure, vasopressor use, and rapidly changing fluid goals. In that environment, variation in practice can reflect both patient complexity and system inconsistency. A patient may need therapy adjustments within hours, and the margin for error is narrow. Standardization therefore becomes a patient safety strategy, not just a management preference.
By organizing CRRT programs around similar structure, process, and outcome metrics, QUALITY CRRT attempts to create a common operational baseline across sites. That baseline can help identify where variation is clinically meaningful and where it is simply noise. It also creates a platform for audit and feedback, one of the few intervention types that can influence frontline behavior without requiring a wholesale redesign of intensive care delivery. In theory, repeated feedback can improve adherence to prescribed therapy and reduce avoidable interruptions, both of which are central to effective extracorporeal support.
The protocol also includes a health economic evaluation with both a within-study analysis and a longer-term model-based analysis. That component is important because system-wide quality programs carry administrative and implementation costs. A program that improves care but is unsustainable financially may struggle to survive beyond the study period. Economic evaluation will help determine whether the added infrastructure needed for KPI reporting and feedback is justified by downstream gains in efficiency, resource use, or patient outcomes.
Health-system implementation will test whether bedside practice can be changed at scale
The trial’s scope extends across an entire healthcare system, which makes the implementation question as important as the clinical one. Many improvement projects work well in a single unit where local champions are deeply involved, then weaken when expanded to multiple hospitals with different cultures and workflows. QUALITY CRRT is specifically structured to examine whether stakeholder engagement plus stepped implementation can overcome that barrier. The intervention is therefore part quality science, part systems engineering.
This design also acknowledges that CRRT quality is not determined solely by the nephrology or intensive care teams. It depends on nurses, perfusion-like technical support, pharmacists, physicians, and local leaders who can sustain compliance over time. A KPI framework can only influence behavior if the data are credible, timely, and actionable. Without that loop, measurement becomes paperwork. With it, measurement becomes a mechanism for practice change.
The protocol has received ethics approval and is registered under NCT04221932. Version 1.0 was dated 15 June 2020. Because this is a protocol paper, no outcome data are presented yet, and no claims about effectiveness, safety, or cost-effectiveness can be made from the publication itself. The most important near-term question is whether a structured, system-wide quality program can move CRRT from variable delivery toward a more consistent and auditable standard of care. The answer will depend on whether process gains show up in the quarterly KPI data and whether those gains survive once the implementation phase is complete.