Abstract
Objectives
Kawasaki Disease is the most common cause of paediatric acquired heart disease and one of the most resource-intensive acute paediatric conditions to treat. Hospitalization for Kawasaki Disease is a stressful event for children and caregivers, leading to increased use of emergency resources after resolution of disease. A quality improvement clinical pathway was introduced to alleviate strain on inpatient resources and reduce length of stay by 33% over 12 months.
Methods
All patients diagnosed and treated for Kawasaki Disease at this institution were included. The primary outcome measure was length of stay. Patients with complicated Kawasaki Disease (i.e., Admission to intensive care) were excluded from outcome measurement, however process and balancing measures were collected for all patients. Balancing measures included rate of urgent care access post-discharge and patient satisfaction.
Results
Forty-nine patients were included in the study, with length of stay data collected for 19. The average length of stay was 46 hours (17% reduction from a pre-intervention baseline of 55 hours). Process measures showed >75% uptake in change ideas following two plan, do, study, act (PDSA) cycles. Patient satisfaction was high with no safety concerns after discharge. Approximately $25,402 was saved over 49 admissions through reductions in length of stay and laboratory investigations.
Conclusion
While the primary objective was not met, a sustained 17% reduction in length of stay was achieved without compromising patient safety or satisfaction. This initiative successfully implemented a standardized clinical pathway through strong collaborator engagement and patient-driven change ideas.
Keywords: Mucocutaneous Lymph Node Syndrome, Quality Improvement, Patient Care Management, Clinical Pathways
Background knowledge
Hospitalization for Kawasaki disease (KD), an acute vasculitis of childhood and the leading cause of acquired paediatric cardiac disease globally, is a sudden and distressing experience for families. Without treatment, 20% of patients develop coronary artery aneurysms, increasing risk for severe complications such as thrombosis and myocardial infarction (1). Following hospitalization, families report high levels of anxiety driven by distressing hospital interventions and a lack of KD-specific educational resources (2). KD is also one of the most expensive acute childhood conditions to treat (3). These issues highlight the need to optimize KD care to streamline hospital resource utilization while maintaining patient safety and satisfaction.
Standard treatment for uncomplicated KD includes IV immunoglobulin (IVIG) and echocardiography monitoring for coronary artery aneurysms (4). For IVIG-resistance and complicated KD, guidelines from the American Heart Association (AHA) also describe treatment with corticosteroids both as upfront intensification and as an option for IVIG-resistance while recognizing that institution treatment varies widely for the above situations (5). Factors that can complicate KD include: age <6 months, presence of coronary artery aneurysms on the first echocardiogram, cardiogenic shock and IVIG-resistance (fever recurrence 24 to 36 hours after IVIG completion) (5). For 6 to 8 weeks at minimum after discharge, patients are monitored with serial echocardiograms for cardiac sequelae and KD recurrence (4).
Patient impact
Although patients generally have excellent outcomes after KD treatment (4), caregivers report high levels of anxiety from the sudden deterioration in their child's health, frequent blood draws and lack of information about expectations after diagnosis (2). Families highlighted a need for improved KD-specific educational resources and peer support following diagnosis (2).
Resource use
The average admission for KD in America is about 3 days, and costs approximately $5000 to $30,000, with IVIG echocardiography and subspecialty consultation as major inpatient expenditures (3,6,7). In a 10-year Canadian health administrative review, KD cases incurred $13.9 million in cost within the first year of diagnosis compared with $2.2 million for age-matched controls (3). Acute hospitalization accounts for approximately one-third of costs within the first year of diagnosis, with prolonged fever monitoring and pending investigations further increasing costs of hospitalization (3,6).
Local problem
At our Canadian tertiary paediatric centre which manages about 100 cases of KD annually, these challenges with KD care are amplified. Bed occupancy is consistently over capacity (averaging 107% in 2023) on paediatric medicine units, creating pressure on hospital resources and personnel. All patients with KD admitted to paediatric medicine receive subspecialty consultation with follow-up within 2 weeks of discharge in a dedicated KD clinic.
The emergence of multisystem inflammatory syndrome induced by COVID-19 (MIS-C), a condition similar to KD (8), spurred the development of KD/MIS-C clinical care pathways. The clinical working group partnered with KD Canada—a national organization supporting families affected by KD, to champion this quality improvement (QI) initiative.
Intended improvement
Leveraging institutional KD expertise and follow-up infrastructure, the project aimed to reduce strain on hospital resources. Although similar QI initiatives in KD have not previously been reported, strategies for safe LOS reduction in other acute paediatric inpatient settings have included standardized clinical care pathways (9–11).
Study
Given the impact of hospitalization on patient experience and the high cost of prolonged LOS, this project aimed to reduce average LOS by 33% for uncomplicated KD cases by December 2024.
Methods
Ethical issues
QI approval (QIP-2023-08-31T14-42-41) was obtained from Quality Management. Patient data was de-identified. Satisfaction survey responses were anonymized and collected on REDCap (12).
Setting
The core project team includes four physicians and trainees in Paediatric Medicine and Rheumatology. Collaborators included trainees, allied health and staff from Paediatric Medicine, Rheumatology, Pharmacy, Cardiology, Laboratory Medicine and patient representatives from KD Canada.
The hospital electronic medical record (EMR) is EPIC, which allows for electronic charting, orders and communications. EPIC employs Smartphrases—a customizable standard text that can be shared by multiple users via text shortcuts to create documentation (i.e., consult notes). The after visit summary (AVS) is another customizable document which allows providers to communicate post-discharge instructions to patients.
Planning the intervention
All patients diagnosed with and treated for KD at our institution were included in the study. LOS, the primary outcome, was collected only on uncomplicated KD cases. Complex KD cases with medically necessary reasons to prolong LOS were excluded from primary outcome measurement but included in process and balancing measures as the planned intervention could affect all KD admissions. Complex KD cases included: shock, IV antibiotics >24 hours in duration, IVIG-resistance or coronary artery lesions on echocardiogram. LOS was defined as time from admission to the ward until discharge. A baseline chart review of 38 uncomplicated KD cases from January to December 2023 revealed a stable baseline average LOS of 55 hours.
The current state was process mapped with chart review data from six baseline charts and presented to collaborators (Figure 1A). As activities between post-IVIG bloodwork and discharge were not consistently documented, we surveyed collaborators for a list of necessary activities preventing discharge to create the Pareto diagram (Supplementary Figure S1). Thirty collaborators then voted on the top root causes of prolonged admission: delayed echocardiograms, rheumatology discharge approval, fever monitoring, awaiting post-IVIG labs and unclear follow-up plans. These root causes were addressed through the development of a new clinical pathway for KD (Figure 1B). Sedated echocardiograms, a highly ranked root cause, were not a target of change because stakeholders deemed the current process necessary for maintaining quality of care.
Figure 1.
Process map of uncomplicated KD hospitalization before and after implementation of the New Clinical Pathway. The swim lane (A) illustrates the previous process of inpatient care for patients with uncomplicated KD. Boxes in white are not measured as part of LOS but provide context for the workflow that occurs prior to patient transfer to the ward. (B) Illustrates the new process of inpatient flow after the clinical pathway was implemented. MOH = Ministry of Health; NVA = no value added step and NVN = no value added; VA = value added step, but necessary step. Average time required for each step extracted from chart review of 6 charts (average LOS 51 hours), denoted in red with arrows. If not noted, the time required for the step was not extractable from the chart.
Interventions, process and balancing measures
Five interventions were incorporated into the new clinical pathway (Supplementary Figure S2):
Remove echocardiography approval: During the pandemic, KD echocardiogram approvals were completed by a specific cardiologist due to increased demand. As patient volumes decreased, echocardiography approval was no longer required. This change automates echocardiography scheduling with placement of the order—eliminating an unnecessary step. Time to echocardiography from admission was measured as a secondary outcome measure (Table 1).
Develop discharge checklist: AHA currently recommends 36 hours of monitoring after IVIG administration for IVIG-resistance (5). The discharge checklist (Supplementary Figure S3) establishes a minimum fever-free period post-IVIG of 24 hours. With the 8-hour IVIG infusion, this timing approximates the 36 hours of monitoring. Furthermore, review of the baseline charts demonstrated that patients with IVIG-resistance developed fever well before 24 hours. Rate of untreated fever post-discharge was measured to detect unintended consequences of this change. The discharge checklist was incorporated into a standardized consult note Smartphrase owned by Paediatric Medicine and Rheumatology providers which eliminated the need for Rheumatology approval prior to discharge. Rate of consult Smartphrase use was measured to evaluate the uptake of this intervention.
Eliminate unnecessary bloodwork: In 2023, Laboratory Medicine requested that clinicians reduce unnecessary direct antiglobulin tests (DATs). DATs were being drawn pre/post-IVIG to identify IVIG-mediated hemolysis occurring primarily when patients received a second dose of IVIG as treatment for IVIG-refractory KD. Each DAT is expensive and increases the workload for laboratory technicians by 2 hours. In 2019, as corticosteroids were increasingly used for upfront treatment of MIS-C, this institution standardized corticosteroid use in KD both for upfront intensification and for IVIG-resistance given the risk of hemolysis and need for ongoing IV access with a second dose of IVIG. By transitioning to corticosteroids for IVIG-resistance, there was no longer a need for pre/post-IVIG DAT. As a forcing function, DAT and post-IVIG labs were removed from the order set, requiring providers to order tests manually if clinically required. The rate of pre- and post-IVIG DAT orders were used to measure the success of this intervention.
Standardize discharge instructions: To standardize workflow, instructions for referral to the KD clinic and patient instructions were outlined in the consult and AVS Smartphrases respectively (Supplementary Figure S4). Rate of emergency department presentations post-discharge and urgent calls to rheumatology were measured to identify safety issues with unclear instructions regarding reasons to seek care.
Improve patient educational resources: Interviews with patient partners revealed that there was a lack of information provided on symptoms and consequences of KD, prognosis following treatment, expectations for follow-up and symptoms to monitor post-discharge. Through multiple developmental PDSAs, the institution-based patient education and KD Canada websites were updated. A new AVS Smartphrase was developed to disseminate the resources at discharge, and the rate of AVS use was examined as a process measure (Supplementary Figure S4). This intervention aims to educate families on important reasons to return to care (i.e., fever) and reduce unnecessary emergency services access after discharge.
Table 1.
Family of measures for the reduction of LOS in uncomplicated KD project.
| Type of measure | Measure |
|---|---|
| Outcome* | Length of stay (hours) |
| Process | % of DATs drawn before IVIG (Intervention C) % of DATs drawn after IVIG (Intervention C) % of patient charts containing the consult Smartphrase in the subspecialty consult note (Intervention D) % of patient charts containing the AVS Smartphrase (Intervention E) |
| Balancing | Patient satisfaction Between hospital discharge and the patient's first KD clinic visit:
|
| Additional measures | Time to echocardiogram (hours) Cost savings (Canadian Dollars) |
*The outcome measures were only measured for the 19 uncomplicated KD cases. All other measures were collected for all 49 KD patients.
AVS = After Visit Summary; DAT = Direct Antiglobulin Test; IVIG = IV immunoglobulin.
Method of evaluation
Charts from the KD clinic were reviewed for the family of measures (Table 1) on a weekly basis by two reviewers. The primary outcome measure, LOS, was calculated for patients with uncomplicated KD. Time to echocardiogram and cost savings were secondary outcome measures collected for all KD patients. As the change ideas were broadly applicable to all admissions, process and balancing measures (Table 1) were collected for all KD patients after their first outpatient clinic to assess the fidelity of change ideas and for unintended consequences of the intervention.
Time to echocardiogram from admission was measured as an indirect method of assessing the effect of removing echocardiography approval. Cost savings were calculated in Canadian dollars by adding savings from reduced lab tests and decreased LOS from each admission. Maximum lab test savings were $282 per admission (sum of: complete blood count and reticulocytes $13.25; chemistry $10.75; up to 2 DATs per admission $129). LOS savings were calculated by multiplying the number of hours below the median LOS (55 hours) achieved in the admission by the estimated daily admission cost of $1736 per 24 hours (rate obtained from administrative data) (13).
Patient satisfaction (balancing measure) was measured by an anonymized REDCap survey mirroring the existing general inpatient satisfaction survey completed by families admitted to Paediatric Medicine. Surveys were distributed at the first follow-up appointment.
Two PDSA cycles addressing process measure uptake of <75% were completed after implementation of the new clinical pathway in December 2023.
Analysis
The outcome measure was assessed with statistical process control charts (SPC), specifically an X-mR chart. The baseline was determined to be stable as data were within 2 standard deviations of the average. The chart was analyzed for significant shifts, trends and astronomical points. Process measures were plotted approximately quarterly. Patient satisfaction survey responses were compared with the general inpatient satisfaction survey responses using the Fisher Exact Test.
Results
Outcomes
The new clinical pathway was disseminated in December 2023 to the paediatric medicine, cardiology and rheumatology groups via shared online drives, physical posters, e-newsletters and in-person presentations.
Over 13 months, 49 patients were admitted and treated for KD. 19 of those patients (39%) had uncomplicated KD and met inclusion criteria for primary outcome measurement (LOS). The average LOS was reduced from 55.2 to 45.7 hours—a 17% reduction with a downward shift from February to December 2024 (Figure 2). The shift was sustained through the summer despite personnel changeover both within the project and in the hospital with new trainees starting in July 2024. Time from admission to echocardiography was 27 hours at baseline with a nonsignificant reduction to an average of 24 hours post-intervention for all patients with KD. An estimated $25,402 Canadian Dollars were saved over 49 KD admissions by reducing LOS ($13,922) and unnecessary bloodwork ($11,480).
Figure 2.
Length of stay for patients with KD. Uncomplicated KD LOS from January 2023 to 2025. The clinical pathway was implemented in December 2023. The median LOS at baseline is 55.2 hours. Circled is a shift from the baseline LOS to 45.7 hours that occurred February to December 2024.
KD admission characteristics
The 49 patients were aged 3 months to 15 years with 10 patients (20%) aged <12 months who received upfront steroid intensification with IVIG. Nineteen patients (39%) had coronary artery lesions on their first echocardiogram. No patients with normal echocardiograms at baseline developed new coronary lesions after discharge, and no coronary artery lesions worsened after discharge. All patients with IVIG-resistance received corticosteroids as second line treatment, and no patients required transfusion for IVIG-mediated hemolysis.
Plan, do, study act cycles
In February 2024, the first PDSA cycle following implementation of the clinical pathway was conducted in response to high rates of DAT ordered on admission (Figure 3). An email reminder was sent via the Paediatric departmental newsletter with a subsequent reduction in DAT testing.
Figure 3.
Process measures (A–D). Rate of process measure uptake measured quarterly. Arrows denotethe desired direction of change. Target value of process measures: A, 0%; B, 0%; C, 100%; D, 100%. PDSA cycles occurred in December 2023, March 2024 and May 2024. AVS = After Visit Summary; DAT = Direct Antiglobulin Test; IVIG = IV immunoglobulin.
In May 2024, rates of pre-IVIG DAT rose from 25% to 67%. Upon chart review, an old consult Smartphrase that contained instructions to draw DAT was still in use. The Smartphrase was subsequently deleted with good process measure response.
Balancing measures
Over the course of the intervention, only 3 out of 49 KD patients returned to the ED between discharge and the first follow-up in KD clinic (Table 2). These patients had uncomplicated KD and came to the ED appropriately for fever, which was diagnosed as infection. No patients had untreated KD-related fever after discharge, which suggests good patient education regarding symptoms of recurrent KD.
Table 2.
Pre- and post-implementation balancing measures: rate of events occurring between discharge and first follow-up in the KD clinic for all patients with KD.
| Pre-intervention (N = 38) | Post-intervention (N = 49) | |
|---|---|---|
| ED visits | 8% (3) | 6% (3) |
| Rheumatology phone calls | 26% (10) | 6% (3) |
| Untreated fever | 3% (1) | 0 |
Patient satisfaction survey
There were 18 out of 49 responses to the patient satisfaction survey (37% response rate). Feedback was positive, with families feeling thankful for the medical information provided about KD. When compared with the inpatient paediatric medicine satisfaction survey, there were no significant differences (Supplementary Table S1). All patients responded that: they were treated with courtesy and respect by their care providers (Q.5), a healthcare provider discussed medication administration post-discharge (Q.8) and they received follow-up instructions for the KD clinic. At 83%, Q.10 regarding counselling on return to normal activities, was the lowest performing question and could be addressed in future PDSA cycles.
Discussion
Summary
The KD inpatient clinical pathway introduced in December 2023 resulted in a 17% decrease in LOS for uncomplicated KD from baseline sustained over 10 months. The clinical pathway was likely responsible for the reduction in LOS given the correlation with sustained uptake of change ideas. While the target of a 33% reduction in LOS was not met, the clinical pathway has improved hospitalization by addressing priorities in patient satisfaction, workplace efficiency and resource stewardship without compromising patient safety. Bold approaches to standardizing inpatient KD care were also safely established through this clinical pathway such as the shift in practice to corticosteroid use for IVIG-resistance, establishing a minimum fever-free period prior to discharge and shifting bloodwork monitoring to the outpatient setting.
Limitations and future directions
Low patient numbers limited the ability to conduct multiple rapid PDSA cycles. Although 49 KD patients presented over 13 months, only 19 uncomplicated cases (39%) were eligible for outcome measurement given the medical complexity of the other 30 patients. This may be attributable to the transfer of stable KD patients to community hospitals for treatment.
Another limitation of the project is that educational materials were available only in English and French. Patient ethnicity and preferred language were not collected. The ethnicity of patients with KD in Canada is highly varied, and therefore how ethnicity affected the perception of care delivery is a knowledge gap of the project.
Lastly, the setting of the project (consult services, EMR and infrastructure) may not be applicable to all hospitals.
Future work could incorporate diversity and language in the development of educational resources and exploration of further shifts in KD care to the outpatient setting.
Conclusion
The design of a KD clinical pathway has been a novel project that improved workflow efficiency and waste reduction. The lessons learned from this project fill a knowledge gap in existing literature that may help other institutions to optimize care for patients with KD.
Supplementary Material
Acknowledgements
The authors wish to thank the patients and families from Kawasaki Disease Canada who have helped provide invaluable insight into this work, everyone who participated in this project at the Hospital for Sick Children for contributing to making the hospital a learning health system, and the instructors and tutors of the University of Toronto Masters of Science in Quality Improvement and Patient Safety for their guidance during this project.
Contributor Information
Audrea Chen, Division of Paediatric Rheumatology, Department of Paediatrics, Stollery Children's Hospital, University of Alberta, Edmonton, Alberta, Canada; Division of Paediatric Rheumatology, Department of Paediatrics, Hospital for Sick Children, University of Toronto, Toronto, Ontario, Canada.
Beth D Gamulka, Division of Paediatric Medicine, Department of Paediatrics, Hospital for Sick Children, University of Toronto, Toronto, Ontario, Canada.
Hanof Alkhdher, Division of Paediatric Rheumatology, Department of Paediatrics, Hospital for Sick Children, University of Toronto, Toronto, Ontario, Canada.
Macarena Palomer, Division of Paediatric Medicine, Department of Paediatrics, Hospital for Sick Children, University of Toronto, Toronto, Ontario, Canada.
Rae S M Yeung, Division of Paediatric Rheumatology, Department of Paediatrics, Hospital for Sick Children, University of Toronto, Toronto, Ontario, Canada.
Supplementary data
Supplementary data are available at Paediatrics & Child Health online.
Funding
The study did not receive funding.
Potential conflict of interest
The authors reported no conflicts of interest. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest.
References
- 1. Kato H, Ichinose E, Kawasaki T. Myocardial infarction in Kawasaki disease: Clinical analyses in 195 cases. J Pediatr 1986;108(6):923–27. doi: 10.1016/s0022-3476(86)80928-3 [DOI] [PubMed] [Google Scholar]
- 2. Chahal N, Clarizia NA, McCrindle BW, et al. Parental anxiety associated with Kawasaki disease in previously healthy children. J Pediatr Health Care 2010;24(4):250–57. doi: 10.1016/j.pedhc.2009.07.002 [DOI] [PubMed] [Google Scholar]
- 3. Robinson C, Chanchlani R, Gayowsky A, et al. Health care utilization and costs following Kawasaki disease. Paediatr Child Health 2022;27(3):160–68. doi: 10.1093/pch/pxab092 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. McCrindle BW, Rowley AH, Newburger JW, et al. Diagnosis, treatment, and long-term management of Kawasaki disease: A scientific statement for health professionals from the American Heart Association. Circulation 2017;135(17):e927–99. doi: 10.1161/CIR.0000000000000484 [DOI] [PubMed] [Google Scholar]
- 5. Jone PN, Tremoulet A, Choueiter N, et al. Update on diagnosis and management of Kawasaki disease: A scientific statement from the American Heart Association. Circulation 2024;150(23):e481–e500. doi: 10.1161/CIR.0000000000001295 [DOI] [PubMed] [Google Scholar]
- 6. Ghimire LV, Chou FS, Mahotra NB, Sharma SP. An update on the epidemiology, length of stay, and cost of Kawasaki disease hospitalisation in the United States. Cardiol Young 2019;29(6):828–32. doi: 10.1017/S1047951119000982 [DOI] [PubMed] [Google Scholar]
- 7. Money NM, Hall M, Quinonez RA, et al. Association of models of care for Kawasaki disease with utilization and cardiac outcomes. Hosp Pediatr 2022;12(3):266–81. doi: 10.1542/hpeds.2021-006364 [DOI] [PubMed] [Google Scholar]
- 8. Tsoukas P, Yeung RSM. Kawasaki disease and MIS-C share a host immune response. Nat Rev Rheumatol 2022;18(10):555–56. doi: 10.1038/s41584-022-00820-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Ogdon TL, Loomba RS, Penk JS. Reduced length of stay after implementation of a clinical pathway following repair of ventricular septal defect. Cardiol Young 2024;34(1):101–104. doi: 10.1017/S1047951123001245 [DOI] [PubMed] [Google Scholar]
- 10. Madsen NL, Porter A, Cable R, et al. Improving discharge efficiency and charge containment on a pediatric acute care cardiology unit. Pediatrics 2021;148(3):e2020004663. doi: 10.1542/peds.2020-004663 [DOI] [PubMed] [Google Scholar]
- 11. Gellert GA, Davenport CM, Minard CG, et al. Reducing pediatric asthma hospital length of stay through evidence-based quality improvement and deployment of computerized provider order entry. J Asthma 2020;57(2):123–35. doi: 10.1080/02770903.2018.1553053 [DOI] [PubMed] [Google Scholar]
- 12. Harris PA, Taylor R, Thielke R, et al. Research electronic data capture (REDCap)—A metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform 2009;42(2):377–81. doi: 10.1016/j.jbi.2008.08.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Indicator Results for Hospital for Sick Children | CIHI. <https://www.cihi.ca/en/indicators/cost-of-a-standard-hospital-stay> (Accessed September 30, 2023).
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.