How Can We Decrease Early Dialysis Initiation? An Interactive Quality Improvement Teaching Case for Health Care Providers and Narrative Review of Quality Improvement Methodology

Abstract

French

Purpose of Review:

Quality improvement (QI) initiatives use a team-based approach to problem-solving clinical and health system issues. All QI initiatives require the coordinated efforts of health care professionals and other stakeholders to encourage the provision of evidence-based clinical care. Most clinicians understand the principles of QI but may lack the training necessary to undertake individual projects.

Methods:

An educational, nephrology-oriented clinical case was created based on the IDEAL study on timing of dialysis initiation, a prioritized quality indicator in several provinces. The case illustrates how to utilize commonly employed QI methodology and to provide a pragmatic framework for both developing and running a QI project. Core concepts addressed in this review include how to perform a QI chart audit, identification of a quality-of-care problem, engaging stakeholders, and how to conduct a root cause analysis that leads to selection of QI measures and change solutions. Last, plan-do-study-act (PDSA) cycles and interpretation of data using run charts are highlighted.

Sources of Information:

PubMed and Google scholar were used as sources of published QI methodology.

Key Findings:

This nephrology-oriented QI case highlights how a core set of QI principles and tools can be used to improve clinical care. This review demonstrates that determining clear goals, utilizing evidence-based guidance to improve timing of dialysis initiation, engaging the appropriate stakeholders, identifying a feasible and measurable change, and tracking if that change leads to improvement are essential components of all QI initiatives. The above framework can be utilized in a variety of clinical areas both within and beyond nephrology-specific care.

Limitations:

Considerations regarding QI-specific data analysis were not addressed as they were beyond the scope of this review.

Keywords

quality improvement quality assurance dialysis start education advanced chronic kidney disease

Introduction

Quality improvement (QI) initiatives use a team-based approach to problem-solving clinical and health system issues and require knowledge of the best available clinical evidence, but also knowledge of how best to prioritize ideas and track the changes that are tested along the way. All QI initiatives require coordination between health care professionals and other stakeholders. These projects should align with patients’ and health care professionals’ priorities, and determining how to organize, document, and plan each step is critical for successful and sustainable change. As the majority of Canadian nephrologists who participate in or lead QI initiatives may not have formal training in this area,¹ the purpose of this educational, nephrology-oriented clinical case and review of QI methodology is to provide a framework for clinicians to develop a QI project, to highlight core QI concepts, and to guide clinicians in successfully implementing and troubleshooting a QI project.

Methods

A PubMed and Google Scholar search was used to include relevant QI methodology. We then created a fictional clinical case to illustrate how commonly employed QI tools can be used to improve clinical care at a local level.

Review

The Clinical Case

A 70-year-old woman is followed in your local multidisciplinary kidney clinic. She feels well and has no symptoms suggestive of uremia. Her clinical examination shows mild ankle edema and metabolic parameters are normal. Her eGFR is 11 mL/min/1.73 m² from 12 mL/min/1.73 m² the month prior. She asks the team, “. . . when should I start dialysis?”

Question 1 : Based on available evidence, when should we initiate dialysis?
a) Start now, do not wait until she has a clinical or metabolic indication
b) Defer dialysis until there is an indication (eg, hyperkalemia, volume overload, symptoms suggestive of uremia) or eGFR drops to 6 mL/min/1.73 m², whichever comes first.
c) Uncertain

(see the following text for answers)

The nephrology team ask themselves, “What does the evidence tell us about when to start dialysis?” and “What QI strategies will ensure we follow evidence-based care?”

eGFR and Dialysis Initiation

The suggested timing of dialysis initiation comes from the IDEAL Study, a multicenter randomized control trial that studied patients with advanced chronic kidney disease (category G5) randomized to early dialysis initiation (eGFR, 10-14 mL/min/1.73 m²) or late (eGFR 5-7 mL/min/1.73 m²).² Over 3.6 years of follow-up, there was no difference in survival or other clinical outcomes between the 2 groups. The conclusion from the IDEAL study was that the intent-to-defer dialysis strategy does not negatively impact clinical outcomes including survival and likely results in substantial cost savings, due to higher dialysis and transportation costs in the early-start group. A limitation of this study was their use of the Cockcroft-Gault equation rather than MDRD. The assay used to develop the Cockcroft-Gault was likely 10% to 20% higher than MDRD, and at the time of dialysis initiation, the eGFR using the Cockroft-Gault formula versus the MDRD formula was 12 versus 9.8 and 9 versus 7.2 mL/min/1.73 m² for the early versus late start groups, respectively.³ An intent-to-defer strategy is accepted by the Canadian Society of Nephrology (CSN), who released clinical practice guidelines on this topic in 2014⁴; therefore, the answer to the question is b.

Patient engagement in developing QI initiatives is integral to any project, as priorities may differ between those of health care providers, including dialysis-free time and the ability to travel versus lab values.⁵ A model of shared decision-making between the patient and health care team should focus on addressing the various priorities, and discussion and decisions surrounding kidney care should focus on the expected changes of advancing kidney disease.^6,7

Approaching eGFR and Dialysis Initiation From a Quality Improvement Perspective

The first steps in conducting a QI project involve understanding local program practice. Reviewing local data may reveal unanticipated findings. For example, between 2001 and 2007, the mean estimated glomerular filtration rate at initiation of dialysis in Canada increased from 9.3 to 10.2 mL/min per 1.73 m², and the proportion of early starts rose from 28% to 36%.⁸

Chart Audit—how many charts should be reviewed in your kidney clinic to identify this quality-of-care problem?

A common misperception among care providers is that we need an extremely large sample size to obtain an answer.

Question 2 : If it is expected that 50% of patients in your clinic are starting dialysis with an eGFR >9.5 mL/min/1.7 3m², and your desired performance is 30%, how many charts must you review to be convinced that a quality-of-care problem exists?
a) 250
b) 100
c) 40
d) 12

In Table 1, Etchells et al demonstrate the approximate sample size required to reject the null hypothesis (using a conventional 2-tailed P value of .05) that observed performance (from an audited sample) is consistent with the desired system performance. This information can help determine how many charts require auditing. Clinicians may be surprised to see how few charts need to be audited to determine a quality-of-care problem.

Table 1.

Minimum Sample Sizes Required for QI Projects Based on Observed and Desired System Performance.

Observed system performance (%)	Desired system performance
Observed system performance (%)	80%	90%
95	26	140
90	70	Not applicable
85	260	180
80	Not applicable	50
75	280	28
70	80	20
66	45	15
60	25	10
50	12	6
40	10	5
20	5	5

Source. Adapted from Etchells et al.

(Answer to question 2 is d)

When are patients starting dialysis in your practice?

The team reviews eGFR at dialysis initiation in the kidney clinic over the preceding 6 months, and they find that 50% of patients start dialysis with an eGFR of >9.5 mL/min/1.73 m². A literature review of this topic can also be used to determine baseline performance and root causes.

Creating a Problem and Aim Statement

Having identified a quality-of-care problem, the next step is to develop a “problem statement.” A problem statement is a framework upon which we then create an “aim statement”—the focus of our quality improvement initiative.

Example of a problem statement:
Patients are often started on dialysis with an eGFR>9.5 mL/min/1.73 m².
This practice does not improve patient outcomes.

Effective aim statements are S.M.A.R.T: They are specific about what they target to achieve. They are measurable and should include a specific outcome or process measure (eg, no. of patients doing x). They should be achievable, where the desired level of improvement should be feasible. Determining the level of improvement that is feasible is often challenging and may be a moving target. This depends on many factors, including the resources available to establish QI interventions. They should be relevant, where the aim should be meaningful to stakeholders. Last, the aim statement should be time-bound, where a realistic target date should be set.

Question 3. Based on the clinical case and your literature review , choose the most appropriate aim statement for this project:
a) We aim to decrease the proportion of patients starting dialysis early.
b) We aim to decrease the proportion of patients starting dialysis early
c) (eGFR >9.5 mL/min/1.73 m²).
d) We aim to decrease the proportion of patients starting dialysis early
e) (eGFR >9.5 mL/min/1.73 m²) to <30% by the end of the next calendar year.

The answer to question 3 is c (a is incorrect because it is neither specific nor time bound. b is incorrect because it is not time bound. c is correct—it defines an early start, provides a target and a date. It is achievable, as some programs are below 20%)

The next steps are to create a root cause analysis team. This includes stakeholders invested in the outcome with knowledge of the issue. Creating a root cause analysis team is critical. Those affected by the proposed changes should be those helping create and guide the process.

Who are our stakeholders?

Stakeholders in this QI initiative can include, for example, the nephrology division head, nephrologists, nurse practitioners, nursing staff, dieticians, social workers, clinical managers, nephrology trainees, patients, caregivers, and information technology support staff.

To help identify and categorize stakeholders based on their influence and interest in a project, a power vs interest grid can be an effective visual tool (Figure 1).^9,10 This grid can help develop strategies to engage and manage all involved stakeholders, to minimize resistance against the project, and to maximize support.

Figure 1.

Power vs interest grid.

Patient Engagement

Engaging a patient champion while choosing the members of your team is an essential way of incorporating the voices of patients and caregivers into decisions affecting clinical care. Patient engagement improves communication between patients and health care providers, leading to increased patient and provider satisfaction, which may lead to a better embrace of potential changes. This empowers patients to become involved in their own health care and can result in meaningful changes to health care services. Any potential barriers to patient engagement within an organization need to be addressed.¹¹

Question 4. Which one of the following is not a root cause analysis tool?
a) The 5 Why’s
b) Pareto chart
c) Fishbone / Ishikawa diagram
d) Driver diagram

The answer to question 4 is d. The 5 why’s, Pareto chart, and fishbone / Ishikawa diagram are all root cause analysis tools. A driver diagram is a tool designed to help prioritize change ideas.

Identifying a cause before developing potential solutions is a key tenet of QI work. To figure out why patients may be starting dialysis at higher eGFRs, the QI team may use several root cause analysis tools including the 5 Why’s, a fishbone / Ishikawa diagram, a failure mode and effect analysis, and a Pareto diagram. Patient champions will play a big part in identifying potential areas for improvement. This step may include consulting with the patient champions who are part of your team or reaching out to a larger number of patients through surveys, interviews, or focus groups.

1. The 5 Why’s: a brainstorming exercise where team members consider the root cause behind patients starting dialysis at higher eGFRs (Figure 2).¹²

Figure 2.

The 5 Why’s—Determining the root cause of a clinical problem.

2. Fishbone / Ishikawa diagram: A fishbone diagram (Figure 3) is a cause-and-effect diagram with 6 process categories. This diagram helps identify possible causes of a problem and helps brainstorm causes relating to specific categories. Each “fishbone” branch is labeled with a category. The 4Ss, 5Ps, 6Ms is an example of different categories that can be used (4S: surroundings, suppliers, systems, skills; 5P: patients, providers, policies, procedures/processes, place/equipment; 6M: m achine, m ethod, m aterials, m easurement, hu( m an), and environment ( m other nature).

Figure 3.

Example of a Fishbone / Ishikawa diagram considering the reasons why people followed in the multidisciplinary kidney clinic may be starting on dialysis early.

3. Failure mode and effect analysis (FMEA): This tool helps evaluate processes to identify how and where they might fail and to assess the impact of various failures. The FMEA tool helps us to proactively change processes to prevent future failures instead of reacting once failures have occurred (see Table 2).¹²

Question 5. What is the Pareto principle?
a) That only 50% of root causes should be prioritized
b) That root causes must be identified to solve a QI problem
c) That 80% of outcomes come from ~ 20% of the causes

Table 2.

Example of Failure Mode and Effect Analysis (FMEA) Tool.

Steps	Failure mode	Failure causes	Failure effects	Likelihood of occurrence (1-10)	Likelihood of detection (1-10)	Severity (1-10)	Risk profile number (RPN)	Actions to reduce occurrence of failure
	What could go wrong?	Why would the failure happen?	What would be the consequences of the failure?	10 being the most likely	10 being the most likely NOT to be detected	What is the likelihood that the failure mode will cause severe harm?	Multiply the 3 scores. The lowest score will be 1, the highest possible score will be 1000.	List the possible actions to improve safety systems, especially for failure modes with the highest RPNs.
1	Patient symptom checklist completed.	Lack of time by allied health.	Patient may be experiencing symptoms that are not picked up by the team.	5	7	7	245	Reminder in template patient note to address symptom checklist.
Etc.

The answer to question 5 is c—the Pareto principle states that 80% of the outcomes or consequences can be traced back to roughly 20% of the causes. This is otherwise known as the 80/20 rule.

4. The Pareto diagram: A Pareto diagram is a bar chart depicting how often different categories of events occur. The most commonly occurring event may be the root cause and this helps maintain focus on causes that will have the greatest impact if solved (Figure 4).¹²

Figure 4.

Example of a Pareto diagram.

Figure 5.

Example of an affinity diagram where change ideas are developed and categorized.

Based on the root cause analysis, the QI team decide the greatest potential impact will stem from addressing

- Difficulties reviewing patients with low eGFR regularly in clinic

- A lack of expertise with symptom management

Next steps: brainstorming solutions and change ideas

Several tools have been developed to encourage brainstorming and organizing change ideas that address the root causes identified above. An affinity diagram (Table 3) allows teams to be creative but also to include and categorize change ideas that are both measurable and achievable. These change ideas should be directed at solving the root causes (eg, difficulty reviewing patients with low eGFR on a regular basis in the clinic and lack of expertise with symptom management).

Table 3.

Hierarchy of effectiveness.

A visual representation of a hierarchy of effectiveness is another brainstorming tool that helps organize the following change ideas by increasing effectiveness (see Table 4).¹³

Table 4.

Example of different QI measures and their definitions.

QI Measure	Definition	Examples based on Clinical Case
Outcome measure	- A measure of what we are trying to achieve by capturing system performance.- Usually you have only one.	- The proportion of patients starting dialysis with an eGFR <9.5 mL/min/1.73 m².
Process measure	- Measures the things you are doing to achieve your aim capturing the changes your QI efforts make to the inputs or steps that contribute to system outcomes.- These can be different for each PDSA cycle.- There are often multiple process measures.	- The proportion of patients receiving weekly telephone calls to evaluate uremic symptoms.- The proportion of visits where a symptom management checklist is completed.
Balancing measures	- Evaluate system performance and potential changes.- Measures unintended consequences or trade-offs.	- The additional time spent by nursing staff per patient per week.- Additional time spent calling patients to review symptoms.

Question 6. Which type of change idea may be most effective based on the hierarchy of effectiveness that will help reduce the proportion of patients starting dialysis early? a) Forcing function and constraints (eg, if no metabolic indication, wait 1 week before b) starting the patient on dialysis) c) Training and education (eg, infographics for patients and providers on symptom control) d) Simplification and standardization (eg, introduce a symptom control nurse)

The answer to question 6 is a—design-oriented features are sometimes the most effective at creating change as they rely less on human factors and will create less variability between health care providers.

How do we decide which change idea to test first?

A “Possible / Implement / Challenge / Kill” (PICK) chart helps categorize and identify whether a change idea will be easy or difficult to implement.¹⁴ Items in the left upper corner will have a high payoff and are relatively easy to implement. Items in the bottom right corner will have a lower payoff—they may be challenging to implement, and potentially reasonable to defer. Remember that change ideas should address root causes no matter where they fall on the PICK chart.

Once a change idea has been selected for further testing, it is important to track whether our changes will lead to improvement. The Institute for Healthcare Improvement (IHI) Model for Improvement is a model which focuses on setting aims and developing measures to see whether a change results in improvement. At the core of this model is the Plan-Do-Study-Act cycle, where small scale interventions are implemented to test for change, and where subsequent cycles are modified to learn from these tests.^15,16 In the “Plan” phase, we describe the objective and the change being tested. Predictions of what will happen are formulated, and action steps are created, with a plan for collecting data. The “Do” step is where the change ideas are tested, data are collected, and we describe what happens. We then “Study” and analyze the collected data, compare outcomes with what was originally predicted, and summarize what has been learned. Last, we “Act” and decide the next best steps, considering any changes for the next Plan-Do-Study-Act (PDSA) cycle.

Quality improvement initiatives collect data on many measures. Outcome measures reflect what we are trying to achieve, with one outcome measure typically analyzed for a given change idea. The proportion of patients starting dialysis with an eGFR <9.5 mL/min/1.73 m² is an example of an outcome measure. Process measures quantify what we are doing to try to reach a particular outcome.

Question 7. Which one of the following is a process measure that we can use to measure the changes we implement in our multidisciplinary kidney clinic? a) Proportion of patients with completed symptom management checklist b) Proportion of patients starting dialysis with an eGFR >9.5 mL/min/1.73 m² c) Time (minutes) spent per week calling patients by nursing staff

The answer to question 7 is a—the proportion of patients with a completed symptom management checklist. B is an outcome measure, and c is a balancing measure (or measure of trade-offs, or unintended consequences)

Process measures can be different for each PDSA cycle, with multiple process measures being often evaluated (eg, proportion of patients receiving weekly calls to evaluate symptoms, or the proportion of visits where a symptom management checklist is completed). Last, balancing measures assess unintended consequences or trade-offs. The additional time spent per patient by nursing staff per week or additional time spent calling patients to review symptoms is an example of balancing measures.

Figure 6.

Example of a design and person-oriented change ideas organized by their effectiveness.

Figure 7.

Example of a PICK chart (possible, implement, challenge, kill).

Creating a run chart

A run chart is a graphical display of data, where data from each PDSA can be highlighted (Figure 8).¹⁷ This visual representation highlights whether the intervention affects the outcome or key process measure, and indicates whether a QI change is showing a nonrandom signal.¹⁷ The y axis is the QI indicator (eg, 0%-100%), and x axis represents time. Two lines are placed on the chart: (1) median—based on current data, and (2) your chosen target. A run chart highlights if changes lead to improvement, and if that improvement is then sustained. Astronomical data points are points which obviously differ, where all studying the chart would agree the point is unusual.¹⁷ Figure 9 illustrates a run chart displaying how different PDSA cycles affect the proportion of patients starting dialysis with an eGFR > 9.5 mL/min/1.73 m². A trend represents 5 consecutive points going either up or down (eg, consecutive months where the percentage of patients starting dialysis early decreases), and a shift represents 6 consecutive points that are consistently above or below the median (eg, 6 consecutive months where the percentage of patients starting dialysis early is <35%) (Figure 10).¹⁷

Figure 8.

Example of a run chart template.

Figure 9.

Example of a run chart identifying change in practice over time. Outliers, or astronomical data points are highlighted. A run chart can be interpreted by counting the number of times the line connecting the data points crosses the median which indicates change in practice over time.

Figure 10.

Run chart showing the percentage of patients starting dialysis early (eGFR >9.5 mL/min/1.73 m)².

Clinical Case—Conclusions

The nursing team starts calling the patient weekly to evaluate for symptoms suggestive of uremia, and a symptom management checklist is introduced. The patient ultimately initiates dialysis at an eGFR of 8 mL/min/1.7 3m² with onset of generalized pruritus and worsening fatigue, in line with current CSN recommendations regarding dialysis initiation.⁴

Summary

Health care systems have a commitment to continuously improve the quality of care: care that is safe, effective, timely, efficient, patient-centered, and equitable. This nephrology-oriented QI case highlights how a core set of QI principles can be used to develop a QI project. Identifying the quality issue and determining clear goals, engaging the appropriate stakeholders to understand the problem from a range of perspectives, identifying a feasible and measurable change, and using data to track the impact of change on improvement are essential components of all QI initiatives. This case used some of the common methods and principles in quality improvement. This may help spread improvement interventions through growing a culture of improvement and its methodology in renal and the general medical practice. This article will be available in the form of an educational interactive case on the Canadian Society Nephrology (CSN) website at the following address https://bit.ly/41U4wXo. Additional QI-specific tools and resources can also be found at https://www.csnscn.ca/csn-committees/csn-quis-committee/

Footnotes

Acknowledgements

The authors value the support and guidance from the Canadian Society of Nephrology and the Canadian Society of Nephrology Quality Improvement and Implementation Science (CSN-QUIS) Committee.

Declaration of Conflicting Interest

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: K.L. has no conflicts of interest relevant to this study. S.A.S. has received consulting fees from AstraZeneca and honoraria from Otsuka and Novo Nordisk. T.G. has no conflicts of interest relevant to this study. E.S.-V. has no conflicts of interest relevant to the study. All authors approved the final version of the submitted manuscript. We certify that this manuscript nor one with substantially similar content has been published or is being considered for publication elsewhere.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Ethics Approval and Consent to Participate

Formal research ethics board review was not required by the University of Ottawa based on the Tri-Council Policy Statement.

Consent for Publication

All authors approved the final version of this manuscript.

ORCID iDs

Khaled Lotfy

Epsita Shome-Vasanthan

Samuel A. Silver

Tamara Glavinovic

Data Sharing

Please contact the corresponding author for any reasonable requests to share this data.

References

Iskander

McQuillan

Nesrallah

Rabbat

Mendelssohn

DC.

Attitudes and opinions of Canadian nephrologists toward continuous quality improvement options. Can J Kidney Health Dis. 2017;4:2054358117725295. doi:10.1177/2054358117725295.

Cooper

Branley

Bulfone

, et al. A randomized, controlled trial of early versus late initiation of dialysis. N Engl J Med. 2010;363:609-619. doi:10.1056/NEJMoa1000552.

Pollock

Cooper

Harris

DC.

When should we commence dialysis? The story of a lingering problem and today’s scene after the IDEAL study. Nephrol Dial Transplant. 2012;27(6):2162-2166. doi:10.1093/ndt/gfs125.

Nesrallah

Mustafa

Clark

, et al. Canadian Society of Nephrology 2014 clinical practice guideline for timing the initiation of chronic dialysis. CMAJ. 2014;186:112-117. doi:10.1503/cmaj.130363.

Song

Lin

Gilet

Arnold

Bridgman

Ward

SE.

Patient perspectives on informed decision-making surrounding dialysis initiation. Nephrol Dial Transplant. 2013;28(11):2815-2823.

Cabrera

Hansson

Kliger

, et al. Symptom management of the patient with CKD: the role of dialysis. Clin J Am Soc Nephrol. 2017;12:687-693.

Evangelidis

Tong

Manns

, et al. Developing a set of core outcomes for trials in hemodialysis: an international Delphi survey. Am J Kidney Dis. 2017;70(4):464-475.

Clark

Rosansky

, et al. Association between estimated glomerular filtration rate at initiation of dialysis and mortality. CMAJ. 2011;183:47-53. doi:10.1503/cmaj.100349.

Eden

Ackermann

Making Strategy: The Journey of Strategic Management. Newcastle upon Tyne, UK: SAGE; 1998.

10.

Bryson

JM.

What to do when stakeholders matter: stakeholder identification and analysis techniques. Public Manag Rev. 2004;6:21-53.

11.

Ontario

. Engaging with patients and caregivers about quality improvement: a guide for health care providers. Published 2016. Accessed December 12, 2024. http://www.hqontario.ca/portals/0/documents/qi/qip/patient-engagement-guide-1611-en.pdf

12.

Patient Safety Essentials Toolkit. Published 2023. Accessed November 29, 2023. https://www.ihi.org/resources/tools

13.

Parties

Canadian Incident Analysis Framework. Edmonton: Canada Patient Safety Institute; 2012.

14.

Stamatis

DH.

Six Sigma Fundamentals: A Complete Introduction to the System, Methods, and Tools. Boca Raton, FL: CRC Press; 2019.

15.

Ogrinc

GS.

Fundamentals of Health Care Improvement: A Guide to Improving Your Patient’s Care. Oakbrook Terrace, IL: Joint Commission Resources; 2012.

16.

Langley

Moen

Nolan

, et al. The Improvement Guide: A Practical Approach to Enhancing Organizational Performance. Hoboken, NJ: John Wiley & Sons; 2009.

17.

Perla

Provost

Murray

SK.

The run chart: a simple analytical tool for learning from variation in healthcare processes. BMJ Qual Saf. 2011;20(1):46-51. doi:10.1136/bmjqs.2009.037895.