Article Text
Abstract
Introduction Continuous renal replacement therapy (CRRT) is a continuous form of dialysis used to support critically ill patients with acute kidney injury. The ideal delivery of CRRT requires ongoing monitoring and reporting to adjust practice and deliver optimal therapy. However, this practice occurs variably.
Methods QUALITY CRRT is a multicentre, prospective, stepped-wedged, interrupted time series (ITS) evaluation of the effectiveness, safety and cost of implementing a multifaceted CRRT quality assurance and improvement programme across an entire healthcare system. This study will focus on the standardisation of CRRT programmes with similar structure, process and outcome metrics by the reporting of CRRT key performance indicators (KPIs). The primary outcome will be the quarterly performance of CRRT KPIs. Secondary outcomes will include patient-centred outcomes and economic outcomes. Analysis will compare pre-implementation and post-implementation groups as well as for the performance of KPIs using an ITS methodology. The health economic evaluation will include a within-study analysis and a longer-term model-based analysis.
Discussion The effective delivery of CRRT to critically ill patients ideally requires a standardised approach of best practice assessment and ongoing audit and feedback of standardised performance measures. QUALITY CRRT will test the application of this strategy stakeholder engagement and stepped-wedged implementation across an entire healthcare system.
Ethics and dissemination This study has received ethics approval. We will plan to publish the results in a peer-reviewed journal.
Trial registration number NCT04221932.
Protocol version 1.0 (15 June 2020).
- intensive & critical care
- dialysis
- health services administration & management
- quality in health care
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
Strengths and limitations of this study
Quality continuous renal replacement therapy (CRRT) involves the implementation of CRRT key performance indicators (KPIs) across an entire healthcare system.
Study includes pilot programme followed by broader stepped-wedged roll out of CRRT KPIs across all Intensive Care Units (ICUs) performing CRRT.
Included CRRT KPIs informed from current evidence-base as well as stakeholder surveys.
Study is limited to Intermittent Renal Replacement Therapy (CRRT) and does not include IRRT.
Introduction
Continuous renal replacement therapy (CRRT) is a continuous method of blood purification that provides slow uninterrupted clearance of uremic toxins and enables acid–base, electrolyte and volume homeostasis while preserving haemodynamic stability.1 2
CRRT is the most common initial form of dialysis in ICU settings
The recent epidemiological study, AKI-EPI, revealed that CRRT was the most common form of initial acute RRT for patients with severe AKI.3 These patients have greater illness severity, are more likely to die and have significantly increased healthcare utilisation when compared with their non-CRRT critically ill counterparts.2 As our population ages, becomes more medically complex, and presents with greater severity of illness, the utilisation of CRRT is likely to increase and become an increasingly vital component of life-sustaining therapy.3
CRRT is expensive but there are substantial opportunities to improve costs
CRRT is a costly and labour-intensive resource.4 In the setting of increasingly constrained healthcare resources, intervention is needed, which may identify and eliminate inefficiencies, improve performance and decrease waste while improving provider satisfaction and achieving better patient outcomes.5 6 Currently, performance indicators for CRRT are not routinely measured, and as such, we are not in a position to understand or identify the inefficiencies or gaps in the quality of care of CRRT delivered to our sickest patients.6
Current CRRT practices are not standardised
In our healthcare system, CRRT is delivered as per individual unit protocols and practice patterns and is not consistently monitored (ie, initiation strategies, anticoagulation techniques, dose delivered, ultrafiltration, etc). Discrepancies from best practices and lack of standardisation of CRRT delivery can result in unplanned CRRT interruptions, decreased treatment time, inadequate dose delivery and impaired clearance of toxic metabolites, which can lead to worsened patient outcomes.7 8
Such suboptimal practice variation may relate to the lack of well-developed key performance indicators (KPIs) for CRRT delivery and performance, and the associated audit and feedback function such KPIs can facilitate. KPIs are measures that can be used to monitor the performance of healthcare delivery.9 They are necessary and can improve reliability of care, standardise complex interventions and provide a platform to measure and monitor performance and the impact of practice changes.10 11
Recently, previous phases of work have identified and prioritised KPIs for CRRT care.12 13 Implementing these CRRT KPIs may change practice to provide effective, validated and standardised CRRT.12 13 Though several previous programmes of work have looked to implement these CRRT KPIs into clinical practice, but no programme has rigorously tested the implementation of this structure and monitoring across an entire healthcare system.14–16
Objectives and research questions
Primary objective
The primary objective is to improve the quality of care delivered to critically ill patients receiving CRRT in Alberta, as measured by CRRT KPI development, monitoring and performance.
Secondary objectives
These will include patient-centred outcomes (ie, Intensive Care Units (ICU) mortality and length of stay, duration of CRRT therapy and 90-day renal recovery) and cost of health services, including unit-specific CRRT costs.
Research hypotheses
Can we improve the performance of CRRT programmes through the implementation of evidence-based clinical practice guidelines and provision of targeted multifaceted CRRT audit, feedback and education sessions?
Will the implementation of standardised CRRT programmes our healthcare system’s ICUs result in decreased healthcare systems costs?
What is the impact of a multifaceted quality assurance and improvement programme on the efficacy and safety of care in critically ill patients requiring CRRT across our healthcare system?
Methods
Trial design
The QUALITY CRRT trial is a pragmatic, multicentre, population-level, stepped-wedged, ITS evaluation of the implementation of an evidence-based CRRT quality assurance and improvement programme to standardise the delivery of CRRT in the 15 adult general and cardiac ICUs and 3 paediatric ICUs in our healthcare system that provide CRRT (table 1). It conforms with the Standard Protocol Items: Recommendations for Interventional Trials Checklist for study protocols (see online supplemental appendix 1).
Supplemental material
Trial oversight
QUALITY CRRT will be led by a small but specialised steering committee, whose members bring extensive experience with CRRT programmes and clinical leadership, implementation science and healthcare systems research. This pan-provincial team will be based at the University of Alberta Hospital and will include representation from the Critical Care Strategic Network of Alberta Health Services (the provincial body which provides provincial liaison, networking and coordination of adult and paediatric critical care in Alberta).17 The steering committee will be responsible for programme management, development and implementation of minimum standards for CRRT programmes, KPI reporting, targeted education and overall trial management.
Patient and public involvement
While this study currently does not directly include patients in its design, the Critical Care Strategic Clinical Network includes patient representatives on its core committee and is represented on the study team. The study objectives and research hypotheses have been developed along with these members. Finally, the results of this study will be disseminated to patients and families leveraging the strengths of the Critical Care Strategic Clinical Network. This will be conducted through online resources, publications and public engagement events (ie, Café Scientifiques).
Population and eligibility
This study will be conducted at all ICUs in Alberta capable of providing CRRT. All subjects in this study will be critically ill patients (ie, paediatric and adult) receiving CRRT as part of their care. There will be no exclusion criteria. The inclusion criteria are purposely broad in scope to capture a system-level sample of critically ill patients. This will be done so that these new KPI monitoring processes may be developed and implemented as policy, and outcomes measured on a population level.
All new ICU admissions receiving CRRT in the 15 adult and 3 paediatric ICUs in Alberta who provide this therapy will be included in this project. In 2019, there were 12 132 adult and 1592 paediatric admissions per year with 5.6% and 1.4% of these patients (ie, 680 adult and 22 paediatric patients) receiving CRRT. As this study will be conducted over a 4-year period, thus data on approximately 3000 adult and paediatric (ie, 2900 adult and 100 paediatric) patients will be included in this project.
Interventions, duration and frequency of follow-up
The project consists of a 24-month baseline phase to measure current CRRT practice and a 24-month intervention phase to implement a standardised CRRT programme targeting ICUs-based CRRT KPIs and monitor performance and compliance of participating sites. Data from the 24-month intervention phase will be used to model long-term health economic outcomes.
Baseline phase
Baseline data collection
Baseline clinical and resource utilisation data will be collected on all patients having received receiving CRRT between 1 November 2017 and 31 October 2019.
Stakeholder survey
A healthcare system-wide survey of care providers and stakeholders at participating ICUs will be conducted to identify and establish agreement on the most appropriate KPIs to measure at their ICU during the intervention phase. The survey will be administered through Survey Monkey (www.surveymonkey.com).
Intervention phase
KPI benchmark reporting
The primary study intervention will be the implementation of audit and feedback on CRRT KPI benchmarks identified by the individual ICU teams in the baseline survey. We will implement a minimal bundle of potential CRRT KPIs with evidence to measure will include CRRT programme structure, filter life, downtime, delivered dose, ultrafiltration achieved, alarms, adverse events, ICU mortality and renal recovery (table 2).6 12 13 Reports will be implemented and reviewed with ICU stakeholders ad hoc and at quarterly intervals.
Prior to implementation of the reports, each ICU will receive multifaceted education strategies tailored to their site and informed by local CRRT leaders, champions and stakeholders (table 3). Education strategies will include, (1) interprofessional grand rounds, seminars and webinars supported by a web-based information repository and (2) identification of site champions to provide onsite advocacy and education. The intervention will be multidisciplinary, targeting CRRT prescribers, nurses, unit operational leaders and educators. After the intervention is implemented, quarterly audit and feedback reports and quarterly tele/videoconference and/or in-person visits will be conducted to support the ICUs. The content of this feedback and methods will be individualised to individual ICU needs and preferences.
While the initial education strategy will contain similar themes across all sites, each site will be encouraged to facilitate and participate with our working group in their own audit and educational activities to address unit-specific shortcomings in their CRRT KPI performance. A central website repository of troubleshooting tools that will be hosted by the Critical Care Strategic Network of Alberta Health Services will be available for sites which are not achieving KPI benchmarks.
The CRRT KPI reporting programme will be implemented in a stepped fashion with a pilot occurring at the General System ICU (GSICU) at the UAH over a 3-month period to ensure feasibility, proper reporting and compliance. This will lead to optimisation of the tools prior to more generalised use. The pilot will be followed by a stepped-wedge roll out at centres across Alberta over the subsequent 12 months.
Intervention data collection
At the end of the intervention phase, clinical and resource utilisation data will be collected on all patients receiving CRRT during the 24-month intervention period (table 4).
Outcomes
Primary outcome
The primary endpoint measures are quarterly changes in the performance of the CRRT process KPIs:
Average filter lifespan, measured in hours.
Downtime, as percentage of prescribed time.
Delivered dose, as a percentage of prescribed dose.
Ultrafiltration achieved, as a percentage of prescribed ultrafiltration.
Alarms as recorded per machine, per day.
Secondary outcomes
Patient-centred:
Mortality: ICU, hospital, 90-day post discharge.
Length of stay: ICU and hospital.
Duration of CRRT treatment in hours.
Renal recovery 90 days post ICU discharge.
Health economic:
Supply costs: dialysis filters, fluids and dialysis catheters.
Medication costs: anticoagulation and renal-specific replacement medications (eg, erythropoietin analogues, calcium binders, etc).
Healthcare worker costs: physician billing and nursing (hours).
ICU and hospital stay costs (length of stay).
Progression to end stage renal disease: projected chronic dialysis costs.
Quality-adjusted life years (QALYs).
Health-related quality of life (HRQoL).
Total healthcare costs.
Data management
Data elements will include patient-centred variables: (ie, demographics and type of admission (medical, surgical and trauma)), clinical characteristics (ie, comorbid diseases and primary diagnosis), illness severity (ie, Acute Physiology and Chronic Health Evaluation II (APACHE II), Sequential Organ Failure Assessment and Clinical Frailty Score), treatment intensity (ie, duration of renal replacement therapy, mechanical ventilation and vasoactive therapy), ICU and hospital lengths of stay, and outcomes (ie, renal recovery, mortality and HRQoL), and CRRT-associated cost data (ie, filter use, prescription/dose, machine alarms/downtime, coagulation, adverse events, re-hospitalisations and progression of renal disease). A schedule of data variables to be captured is summarised in online supplemental appendix 2.
Supplemental material
Data sources will include TRACER and Enterprise data repository, AHS Data Integration, Management and Reporting administrative databases, the Nephrology Information System, the Patient-based Renal Information System and Baxter Healthcare.18
All study documents will be kept in a locked filling cabinet in a locked office, and computer files will be encrypted and stored on a secure network for 5 years following completion of the study.
Co-enrolment
QUALITY CRRT is a pragmatic, real-world, quality improvement and assurance programme. Due to the healthcare systems scope of the programme, there are no patient-level interventions. Accordingly, there will be no limitations to co-enrolment or specific patient or clinician practices.
Statistical analyses
Analysis will be conducted between the pre-implementation and post-implementation groups. Analyses of the primary and secondary outcomes will involve summary measures obtained by aggregating the endpoints. Analyses will be performed using SAS Enterprise Guide V.7.1 (Cary, North Carolina, USA). Baseline comparisons will be performed using χ2 test for equal proportions with results to be reported as frequencies with percentages. Continuous normally distributed variables will be compared using t-tests and reported as means with SD, while non-parametrically distributed will be compared using Wilcoxon rank sum tests and reported as medians and IQRs. In case of small sample size, Fisher’s exact test will be used.
ITS analyses using autoregressive integrated moving average models will be employed for important risk factors to account for temporal trends and to determine whether there were changes in the clinic outcomes at the intervention period (compared with the baseline period) and associated with implementation of the evidence-based acute RRT pathway.
Cost-effectiveness or net-benefit (investment–return) analysis using a decision tree will be adopted to compare return (or benefit, B) and investment (or cost, C) of the evidence-based RRT pathway. Reduction of healthcare systems costs, including inpatient services (length of stay of primary admission, number of readmissions and readmission LOS), outpatient services (emergency room visits and clinic visits), physician services (specialist visits and general practitioner visits) and ongoing new end-stage renal disease, will be estimated based on generalised linear models. Cost effectiveness will be analysed by estimating incremental cost and effectiveness based on QALYs gained. QALYs will be calculated based on HRQoL as measured by the EQ-5D-5L in adults and the PedsQL in children. Patients will be sent letters with study team contact information in order for them to contact our team in order to complete these questionnaires.
Performance of CRRT KPIs
Our primary outcome will be the iterative performance of selected CRRT KPIs. Based on prior work, KPIs might include filter life (measured in hours), delivered dose (measured in mL/kg/hour), downtime (measured in percentage of time), ultrafiltration realised (measured in percentage of prescribed) and access alarms (measured in total number per day). We will aim to both compare the performance of these KPIs to historical controls, as well as prospectively through an ITS analysis. The ITS analysis will allow us to follow variable changes over time, will allow for assessment of gradual change and is consistent with traditional quality improvement initiatives.
Patient-centred outcomes analysis
The patient-centred outcome analysis will include ICU, hospital and 90-day mortalities, ICU and hospital lengths of stay, duration of CRRT treatment and renal recovery measured at 90-day months. While this study is not designed to evaluate the effect that the implementation of the reporting of CRRT KPIs will have on mortality, lengths of treatment and stay or renal recovery, these are important patient-centred outcomes that will need to be considered as balancing measures for CRRT KPI reporting and implementation of our multifaceted knowledge translation intervention.
Health economic evaluation
The economic evaluation will comprise two parts: (1) a within-study analysis and (2) a longer-term, model-based analysis.
The within-study analysis will focus on costs and outcomes collected during the study period. It will include total quarterly unit-specific CRRT-associated costs following the implementation of the CRRT KPI reporting programme. This endpoint will be determined from our provincial CIS and Alberta Blue Cross databases. Specifically, we will evaluate and compare the (1) costs of supplying CRRT filters, (2) costs of CRRT fluids, (3) cost of CRRT anticoagulation and (4) costs and utilisation of dialysis catheters. Costs will be calculated in part using CRRT process measures captured by our CRRT KPIs (ie, filter life and number of filters used, anticoagulation modality, dose delivered, effluent used, etc). CRRT-associated costs were selected as an important secondary outcome as these will be most immediately affected with the implementation of the CRRT KPI quality assurance programme across unit.
We will also determine healthcare systems costs to include total ICU and hospital stay associated costs, ongoing new end-stage renal disease (ie, chronic RRT) costs, total healthcare costs and outcomes (mortality and QALYs). Modelling analysis will provide cost estimates from both a healthcare system and societal perspective (capturing costs to the health service, social care providers and patients). Results will be reported as the incremental net benefit and incremental cost-effectiveness ratios. Uncertainty will be captured in the analyses through probabilistic sensitivity analysis and reported using cost-effectiveness acceptability curves, showing the likelihood the intervention will be cost effective over a range of values of willingness-to-pay for specific outcomes.
Planned subgroup analyses
Pre-specified subgroup analysis will include ICU patients to (1) adult versus paediatric, (2) female versus male, (3) academic versus community ICUs, (4) cardiovascular ICUs versus medical/surgical ICUs, (5) high volume versus low volume centres (ie, as per quartiles) and (6) patients requiring acute RRT versus those on chronic dialysis. Adult, paediatric, female and male patients are fundamentally different patient populations and deserve specific study.
Cardiovascular ICU patients differ from general medical/surgical patients as often these patients are immediately postoperative, have a specific timing of insult (ie, cardiac surgery) and hence have different pathophysiology related to their critical illness. It is important to delineate academic versus community ICUs as, for mechanically ventilated patients (ie, another form of critical life-sustaining therapy) with acute respiratory distress syndrome (ARDS), mortality rates differ significantly.19 Finally, higher ARDS hospital case volume has also been associated with lower ARDS hospital mortality and it will be important to determine if this association is present in CRRT.20 We will perform the above analyses for health economic evaluations, patient and process of case measures to include our prespecified primary and secondary outcomes for each subgroup. Each analysis will be accompanied by a test for interaction between treatment and subgroup to ascertain whether effects differ significantly between subgroups.
Ethics approval and consent to participate
This project is an evaluation of impact of a multifaceted CRRT quality assurance and improvement programme on patient outcomes and healthcare resource utilisation in Alberta ICUs delivering CRRT. All diagnostic and management strategies are within standard of care and all data with relevance to the project are already routinely captured as part of standard patient care by means of machine-specific data cards or clinical charting. No added trial-specific investigations or clinical documentation is required.
This evaluation was reviewed by the University of Alberta’s Health Research Ethics Board (study ID: Pro00075274; 22 January 2020) and a waiver of consent was granted based on the premise this project represents health services implementation and evaluation compatible with a quality assurance and improvement initiative (see online supplemental appendix 3).
Supplemental material
Any protocol modifications will be submitted to the appropriate relevant parties.
Dissemination
The findings of QUALITY CRRT will directly inform and guide policy on establishing evidence-based best-practices guidelines for delivering CRRT in Alberta ICUs. In addition, establishing evidence-based benchmarks across the entire healthcare system will enable systematic evaluation of CRRT performance. These outcomes will help create a framework for the standardisation of CRRT programmes across Alberta and other jurisdictions providing CRRT (table 2).
Alberta’s comprehensive ICU clinical information and analytics infrastructure (Connect Care, eClinical TRACER) will be leveraged to implement a CRRT Quality Dashboard, accessible to all Alberta ICU practitioners. The dashboard will contain statistics on KPI benchmarks to provide real-time feedback on individual ICUs performance in delivering CRRT.
A central website containing a summary of CRRT guidelines and best practices and a repository of troubleshooting tools on attaining KPI benchmarks will be developed and made available to all Alberta CRRT practitioners.
We are proposing to publish the study results. Furthermore, this work will be presented at local, provincial and national critical care and nephrology meetings. Finally, QUALITY CRRT will serve as the basis for a broader programme of work, dialysing wisely, which will aim to transform the fashion in which acute dialysis is conducted in Alberta.
Discussion
The importance of the quality and management for critically ill patients with acute kidney injury requiring CRRT has been previously recognised.5 6 Previous studies have focused on single unit or individual hospital-level quality improvement and assurance interventions (table 5).14–16 Griffin et al, first conducted such a quality improvement study at the University of Colorado Hospital, where they assessed the magnitude in variability in CRRT dosing. They followed specific implementation that included optimising their electronic medical record to calculate CRRT dosing in real time to then comment on dosing and provide guidance and education in order to better adhere to national guidelines. This led to the doubling of the rate of appropriate CRRT dosing and reduction in variability.14 Mottes et al, at the University of Cincinnati Children’s Hospital, created a ‘CRRT Dashboard’ which tracked important KPIs such as ‘filter life’, ‘mean prescription dose’ and ‘fluid balance’, and found that this platform provided a significant means for measuring adherence to robust standards on the delivery of CRRT, specifically in the process of care.15 Finally, most recently a group from the University of Kentucky Medical Centre reported the development, implementation and subsequent outcomes associated with a quality assurance system to support the provision of CRRT in the ICU.16 This was the largest programme to date, numbering 1185 adult patients on CRRT over a 34-month period. Using the monitoring of evidence-based KPIs and targeted education, they doubled the appropriate use of citrate-based anticoagulation, improved the appropriateness of CRRT dosing, increased filter life while decreasing machine alarms and maintaining similar CRRT duration and patient mortality while reducing CRRT costs. While these programmes demonstrate that the implementation of evidence-derived KPI-based CRRT quality assurance programmes are effective in improving the efficiency and quality of CRRT, none of these programmes have sought to do this on an entire healthcare systems level. QUALITY CRRT will build on the experience of these programmes in order to scale such a quality improvement and assurance initiative across a provincial health system of ICUs which provide CRRT.
Strengths and limitations
While QUALITY CRRT focuses on standardising CRRT programmes across an entire provincial healthcare system by ensuring a robust framework is in place and the monitoring of CRRT performance and delivery occurs, this is limited to only continuous RRT. Intermittent RRT can also occur in the acute setting for critically ill patients in the ICU. Accordingly, the experience and infrastructure realised in QUALITY CRRT will pave the work for additional critical care nephrology programmes aimed at improving all forms acute RRT (ie, continuous and intermittent) in the ICU.
Ethics statements
Patient consent for publication
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Contributors SMB and OR were responsible for the conception, design and planning of this study. ER assisted in the development of continuous renal replacement therapy key performance indicators. VL and XW have assisted in creating the analysis plan and will work with interpretation the data. DO, NF and DZ assisted with manuscript preparation. All authors approved the final drafting of this manuscript.
Funding QUALITY continuous renal replacement therapy was supported by University of Alberta Hospital's Fund (grant number: RES0040497) and Baxter Healthcare's investigator initiated research (grant number: RES0044818).
Competing interests SMB and OR have received honoraria from Baxter Healthcare. The study sponsors had no role in protocol development, trial management or data analysis and reporting.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.