Abstract
Objectives
Kawasaki disease (KD) is a common childhood vasculitis with increasing incidence in Canada. Acute KD hospitalizations are associated with high health care costs. However, there is minimal health care utilization data following initial hospitalization. Our objective was to determine rates of health care utilization and costs following KD diagnosis.
Methods
We used population-based health administrative databases to identify all children (0 to 18 years) hospitalized for KD in Ontario between 1995 and 2018. Each case was matched to 100 nonexposed comparators by age, sex, and index year. Follow-up continued until death or March 2019. Our primary outcomes were rates of hospitalization, emergency department (ED), and outpatient physician visits. Our secondary outcomes were sector-specific and total health care costs.
Results
We compared 4,597 KD cases to 459,700 matched comparators. KD cases had higher rates of hospitalization (adjusted rate ratio 2.07, 95%CI 2.00 to 2.15), outpatient visits (1.30, 95%CI 1.28 to 1.33), and ED visits (1.22, 95%CI 1.18 to 1.26) throughout follow-up. Within 1 year post-discharge, 717 (15.6%) KD cases were re-hospitalized, 4,587 (99.8%) had ≥1 outpatient physician visit and 1,695 (45.5%) had ≥1 ED visit. KD cases had higher composite health care costs post-discharge (e.g., median cost within 1 year: $2466 CAD [KD cases] versus $234 [comparators]). Total health care costs for KD cases, respectively, were $13.9 million within 1 year post-discharge and $54.8 million throughout follow-up (versus $2.2 million and $23.9 million for an equivalent number of comparators).
Conclusions
Following diagnosis, KD cases had higher rates of health care utilization and costs versus nonexposed children. The rising incidence and costs associated with KD could place a significant burden on health care systems.
Keywords: Children, Emergency department, Epidemiology, Hospitalization, Health care utilization, Health care cost, Kawasaki disease
Graphical Abstract
Graphical Abstract.
Kawasaki disease (KD) is a common childhood systemic vasculitis. We recently described that the incidence of KD has increased by nearly 50% in Ontario, Canada over the past two decades (1), and KD incidence in Ontario is similar to reported incidence in American studies (2–4). KD survivors are at risk of adverse short- and long-term cardiovascular and neurodevelopmental outcomes (1,5–11). In the USA, KD is associated with high acute hospitalization costs (median $5,000 to 30,000 USD) (2,3,12–15). The annual costs of KD hospitalizations in the USA is estimated to be $38 to 110 million USD (2,12). American children with KD have recently been reported to have higher short-term health care utilization within 90 days post-hospitalization, compared to their own pre-hospitalization baseline (16). To our knowledge, only one health economic study has been performed to date in the Canadian context, evaluating the cost-effectiveness of intravenous immunoglobulin (IVIG) treatment (17). Little is known about the long-term burden of illness and health care utilization among KD survivors, particularly in a universal health care system.
The aim of our study was to describe patterns of health care utilization and associated costs among KD survivors vs. matched controls in Ontario, Canada. Using linked population-based health administrative databases, we conducted a large retrospective cohort study to determine the incidence and costs of hospitalizations, emergency department (ED) visits, and outpatient physician visits over a 24-year study period. We hypothesized that KD survivors would have higher long-term health care utilization and costs.
METHODS
Setting and design
We conducted a population-based retrospective cohort study of all children newly diagnosed with KD between 1995 and 2018 in Ontario. The project was authorized under section 45 of Ontario’s Personal Health Information Protection Act and approved by Hamilton Integrated Research Ethics Board. The study was conducted and reported in accordance with STROBE (18) and RECORD (19) guidelines.
Study population
We included all children (0 to 18 years) that were hospitalized and diagnosed with KD between April 1, 1995 and March 31, 2018 (‘KD cohort’). KD was defined by the presence of an International Classification of Diseases, 9th Revision (ICD-9) [446.1] or ICD-10 [M30.3] code in the Canadian Institute for Health Information (CIHI) Discharge Abstract Database (20). This strategy has been recently validated in Ontario (positive predictive value: 93.5%) (20). Next, we identified the ‘nonexposed cohort’, by including all Ontario children (0 to 18 years) that were not diagnosed with KD. We excluded children with prior KD diagnoses between April 1, 1988 and March 31, 1995 (7-year lookback) and included only the first eligible hospitalization. We excluded children without Ontario Health Insurance Plan (OHIP) coverage (i.e., non-residents) and those with incomplete records. We matched each KD case with 100 non-exposed individuals based on index year, age, and sex.
Data sources
We utilized Ontario provincial health administrative databases stored at ICES (formerly known as Institute for Clinical Evaluative Sciences). ICES is an independent, non-profit research institute whose legal status under Ontario’s health information privacy law allows it to collect and analyze health care and demographic data, without consent, for health system evaluation and improvement. Records for all residents with OHIP coverage (>99% of the population) are captured through ICES. Annual emigration from the province is very low (~0.1%) and was the only reason for loss to follow up (21). We used the following databases: CIHI Discharge Abstract Database, CIHI Same Day Surgery, National Ambulatory Care Reporting System for emergency department visits after July 2000, OHIP for physician billings (diagnostic and fee codes), and the Registered Persons Database (a population registry of vital statistics). These datasets were linked through encrypted unique health insurance numbers and were analyzed at ICES. Study investigators had access to complete and uncleaned data for analysis. Descriptions of databases and administrative codes are included in Supplementary Appendix 1.
Demographics and co-morbidities
We determined patient demographics, pre-existing co-morbidities, and hospitalization characteristics for the KD and nonexposed cohorts. Demographic variables included were age, sex, rural status (community <10,000 persons) (22), neighborhood income quintile (by postal code) (23), and ethnicity (by surname-based algorithm for Chinese and South Asian individuals) (24). Hospitalization variables included were admission year, hospital, length of stay, PICU admission, and baseline coronary artery aneurysms (CAA). We looked back to birthdate (or Apr 1, 1991, if born prior) for pre-existing medical co-morbidities including prior cardiovascular events, congenital heart disease, juvenile idiopathic arthritis, systemic lupus erythematosus, other vasculitides, chronic kidney disease, end-stage renal disease, solid organ transplant, disorders of lipid metabolism, and hematological disorders (Supplementary Appendix 1). We determined each child’s Pediatric Medical Complexity Algorithm (PMCA) classification (complex chronic, non-complex chronic, or non-chronic disease) (25).
Outcomes
Our primary outcomes were rates of hospitalizations, ED visits, and outpatient physician visits (primary care, pediatrician, cardiologist, or rheumatologist). Our secondary outcomes were health care costs (in Canadian Dollars [CAD], adjusted for inflation to the 2018 year) associated with each event type, as well as composite health care cost, including hospitalizations, ED visits, and outpatient costs (i.e., lab services, physician billings, and capitation payments). Due to limited data availability in the National Ambulatory Care Reporting System, we could not evaluate ED visits prior to July 1, 2000. We therefore excluded these children from our ED visit analyses. For composite health care costs, we reported separate results for children with an index date prior to and after July 1, 2000, to account for ED visit cost data.
Statistical analysis and reporting
Baseline characteristics were reported as means (standard deviation [SD]) or medians (interquartile range [IQR]) for continuous variables and as numbers and percentages for categorical variables. These characteristics were compared using distribution-appropriate univariable comparison tests. A P-value of <0.05 was considered statistically significant. In accordance with ICES data privacy policies, cell sizes ≤5 individuals were not reported (NR).
Study enrollment date for the KD cohort was defined as discharge date +1 day following index hospitalization. Nonexposed children were assigned random enrollment dates based upon those of the KD cohort. Participants were then followed until death or March 31, 2019 (1 to 24-year follow-up). For hospitalizations, ED visits and outpatient physician visits, we determined the number of children with at least one event, total number of events and encounter rate per 1,000 person-years. This was stratified by time period: 0 to 1 year, 1 to 5 years, and >5 years following study enrollment. For outpatient physician visits, we evaluated encounters by physician type (i.e., primary care, pediatrician, cardiologist, or rheumatologist). We determined total costs, mean (SD), and median (IQR) costs associated with each event type, as well as composite health care costs over the above time periods. Finally, we created negative binomial regression models to determine the association between KD diagnosis and hospitalization, outpatient, and ED visit frequency, separately. These were reported as adjusted rate ratios, adjusting for the following covariates: PMCA classification and the number of composite health care visits (hospitalizations and outpatient visits) during the 1-year lookback prior to enrollment.
RESULTS
Baseline characteristics
We identified 4,597 children hospitalized for KD between 1995 and 2018 and 459,700 matched nonexposed comparators. Median follow-up duration was 11.1 years (IQR 5.5 to 17.0) for KD cases and 10.0 years (IQR 4.5 to 16.0) for nonexposed children. Fewer children in the KD cohort lived rurally and Chinese ethnicity was more frequent (Table 1). The KD cohort had higher rates of prior cardiovascular events, congenital heart disease, systemic connective tissue disorders, or coagulation/bleeding disorders, although these were all uncommon (each <5%). The KD cohort also had higher rates of health care utilization in the year preceding their index hospitalization. CAA were diagnosed in 119 (2.6%) KD cases during their index hospitalization.
Table 1.
Baseline characteristics
| Variable | KD cohort N=4,597 | Nonexposed cohort N=459,700 |
|---|---|---|
| Patient characteristics | ||
| Age at hospitalization, 0–4 years | 3,307 (71.9%) | 330,782 (72.0%) |
| Age at hospitalization, 5–9 years | 1,083 (23.6%) | 108,224 (23.5%) |
| Age at hospitalization, 10–18 years | 207 (4.5%) | 20,694 (4.5%) |
| Male gender | 2,763 (60.1%) | 276,300 (60.1%) |
| Rural status | 279** (6.1%) | 46,897 (10.3%) |
| Chinese ethnicity | 620** (13.5%) | 20,697 (4.5%) |
| South Asian ethnicity | 143 (3.1%) | 17,709 (3.9%) |
| Other ethnicity | 3,834** (83.4%) | 421,294 (91.6%) |
| Baseline health care utilization (in the preceding year) | ||
| Number of hospitalizations, mean ± SD | 0.22** ± 0.54 | 0.16 ± 0.48 |
| ≥ 1 hospitalization | 838** (18.2%) | 62,229 (13.5%) |
| Number of outpatient visits, mean ± SD | 11.56** ± 9.26 | 7.05 ± 9.63 |
| Number of outpatient visits, median (IQR) | 10** (6–15) | 5 (2–10) |
| No outpatient visits | 39** (0.8%) | 51,850 (11.3%) |
| 1-5 outpatient visits | 1,010** (22.0%) | 189,439 (41.2%) |
| >5 outpatient visits | 3,548** (77.2%) | 218,411 (47.5%) |
| Income quintile | ||
| 1 (low) | 947* (20.7%) | 98,908 (21.7%) |
| 2 | 855* (18.7%) | 88,056 (19.3%) |
| 3 | 885* (19.3%) | 90,252 (19.8%) |
| 4 | 978* (21.3%) | 93,723 (20.6%) |
| 5 (high) | 916* (20.0%) | 84,587 (18.6%) |
| PMCA classification† | ||
| Non-chronic disease | 4,250** (92.5%) | 441,799 (96.1%) |
| Non-complex chronic disease | 257** (5.6%) | 12,764 (2.8%) |
| Complex chronic disease | 90** (2.0%) | 5,137 (1.1%) |
| Past medical history | ||
| Any cardiovascular event | 172** (3.7%) | 11,236 (2.4%) |
| Juvenile idiopathic arthritis | NR | 23 (0.0%) |
| Systemic connective tissue disorder | 8** (0.2%) | 244 (0.1%) |
| Solid organ transplant | 0 (0.0%) | 48 (0.0%) |
| Chronic kidney disease | 11 (0.2%) | 660 (0.1%) |
| Congenital heart disease | 197** (4.3%) | 13,056 (2.8%) |
| Rheumatic heart disease | 14 (0.3%) | 888 (0.2%) |
| Disorder of lipid metabolism | 14 (0.3%) | 1,198 (0.3%) |
| Disorder of coagulation or bleeding | 16* (0.3%) | 863 (0.2%) |
| Index admission characteristics | ||
| Mean length of stay, in days ± SD | 3.55 ± 3.08 | N/A |
| Median length of stay, in days (IQR) | 3 (2–4) | N/A |
| PICU admission | 113 (2.5%) | N/A |
| Index hospital site | ||
| Toronto (Sick Kids) | 1,599 (34.8%) | N/A |
| Ottawa (CHEO) | 341 (7.4%) | N/A |
| McMaster (MCH) | 234 (5.1%) | N/A |
| London (LHSC) | 167 (3.6%) | N/A |
| Other | 2,256 (49.1%) | N/A |
HLH Hemophagocytic lymphohistiocytosis; MAS Macrophage activation syndrome.
* P-value < 0.05
**P-value < 0.001
†Pediatric Medical Complexity Algorithm (PMCA) is a classification system for identifying children with chronic and complex diseases using administrative health data. As is recommended for use with hospital discharge data, we looked back 3 years prior to hospital admission in the Discharge Abstract Database (DAD), using the “least conservative version” of the PMCA.
Health care utilization
Compared with nonexposed children, KD cases experienced higher rates of hospitalizations, outpatient, and ED visits at all time periods (0 to 1 year, 1 to 5 years, >5 years; Table 2). Within 1 year post-discharge, 717 (15.6%) of KD cases were re-hospitalized, 4,587 (99.8%) had ≥1 outpatient physician visit, and 1,695 (45.5%) had ≥1 ED visit. The rates of hospitalizations and ED visits among KD cases were highest in the first 3 months post-discharge (Supplementary Table 1). Differences in health care utilization between KD versus nonexposed cohorts decreased gradually over time (Table 2). We also evaluated the most prevalent diagnoses during hospitalizations (Table 3) and ED visits (Supplementary Table 2), comparing KD cases and nonexposed children. KD cases had a higher frequency of ED visits for nonspecific fever and rash within the first year post-discharge. Among KD cases within the first year, 9 to 18% of ED visits had a KD diagnosis. KD cases also had a higher frequency of hospitalizations for nonspecific fever and neutropenia. Among KD cases within the first year, 37 to 46% of admissions had a KD diagnosis and 2% had a diagnosis of CAA. These KD diagnoses during subsequent ED visits or hospitalizations may have represented recurrent/persistent symptoms, visits related to KD complications or coding of the history of KD.
Table 2.
Health care utilization following Kawasaki disease diagnosis
| Type of health care utilization | 0–1 year following diagnosis | 1–5 years following diagnosis | 5+ years following diagnosis | ||||
|---|---|---|---|---|---|---|---|
| KD Case | Comparator | KD Case | Comparator | KD Case | Comparator | ||
| Hospital admissions | N | N=4,597 | N=459,700 | N=4,576 | N=446,620 | N=3,571 | N=334,460 |
| Patients with ≥1 encounter | 717 (15.6%) | 12,140 (2.6%) | 441 (9.6%) | 22,418 (5.0%) | 463 (13.0%) | 33,496 (10.0%) | |
| Encounter rate (per 1000py) | 210.3 | 35.4 | 45.4 | 21.1 | 28.3 | 20.7 | |
| Adjusted rate ratio (95% CI) | 5.68 (5.57–5.78) | Ref | 1.64 (1.52–1.76) | Ref | 1.15 (1.04–1.26) | Ref | |
| Outpatient visits | N | N=4,597 | N=459,700 | N=4,576 | N=446,620 | N=3,571 | N=334,460 |
| Patients with ≥1 encounter | 4,587 (99.8%) | 393,683 (85.6%) | 4,447 (97.2%) | 404,126 (90.5%) | 3,389 (94.9%) | 296,710 (88.7%) | |
| Encounter rate (per 1000py) | 13,131.5 | 5,658.7 | 6,329.3 | 4,206.0 | 4,743.6 | 3,778.4 | |
| Adjusted rate ratio (95% CI) | 1.92 (1.89–1.94) | Ref | 1.18 (1.16–1.21) | Ref | 1.10 (1.06–1.13) | Ref | |
| ED visits | N | N=3,723 | N=372,141 | N=3,709 | N=363,349 | N=2,727 | N=257,503 |
| Patients with ≥1 encounter | 1,695 (45.5%) | 105,297 (28.3%) | 2,129 (57.4%) | 171,122 (47.1%) | 1,668 (61.2%) | 139,141 (54.0%) | |
| Encounter rate (per 1000py) | 877.0 | 486.6 | 492.1 | 349.7 | 358.0 | 314.3 | |
| Adjusted rate ratio (95% CI) | 1.52 (1.46–1.58) | Ref | 1.20 (1.16–1.25) | Ref | 1.07 (1.01–1.12) | Ref | |
Ref: reference group, for calculation of adjusted rate ratio.
Table 3.
Most prevalent diagnoses during hospitalizations* between 0-1 year after the index date
| Kawasaki disease cases | ICD-9 code ** | n (%) | ICD-10 code† | n (%) | ||
|---|---|---|---|---|---|---|
| 4461 | Kawasaki disease | 108 (36.9) | M303 | Kawasaki disease | 329 (46.4) | |
| 2880 | Neutropenia | 14 (4.8) | R509 | Fever, unspecified | 36 (5.1) | |
| 558, 5580 | Gastroenteritis or colitis | 10 (3.4) | I254 | Coronary artery aneurysm and dissection | 13 (1.8) | |
| V581 | Chemotherapy for neoplasm | 8 (2.7) | B349 | Viral infection, unspecified | 12 (1.7) | |
| 4939 | Asthma | 6 (2.1) | J4500 | Asthma | 12 (1.7) | |
| 7803 | Seizures | 6 (2.1) | K529 | Gastroenteritis or colitis | 11 (1.6) | |
| 7860 | Dyspnea and respiratory abnormalities | 6 (2.1) | J353 | Hypertrophy of tonsils with hypertrophy of adenoids | 10 (1.4) | |
| 7806 | Fever, unspecified | 6 (2.1) | J189 | Pneumonia | 8 (1.1) | |
| 486 | Pneumonia | NR | A084 | Viral intestinal infection | 7 (1.0) | |
| - | - | - | A080 | Rotaviral enteritis | 6 (0.9) | |
| Comparators | ICD-9 code ** | n (%) | ICD-10 code† | n (%) | ||
| 4939, 4930 | Asthma | 596 (11.4) | J4500 | Asthma | 802 (7.4) | |
| 486, 4860 | Pneumonia | 309 (5.9) | J189 | Pneumonia | 640 (5.9) | |
| 558, 5580 | Gastroenteritis or colitis | 306 (5.9) | K529, A099 | Gastroenteritis or colitis | 556 (5.1) | |
| 7803 | Seizures | 247 (4.7) | Z511 | Chemotherapy session for neoplasm | 283 (2.6) | |
| 2765 | Volume depletion | 205 (3.9) | J219 | Acute bronchiolitis | 251 (2.3) | |
| 4644 | Acute laryngitis (croup) | 181 (3.5) | A084 | Viral intestinal infection | 250 (2.3) | |
| 4661 | Acute bronchiolitis | 162 (3.1) | J050 | Acute laryngitis (croup) | 246 (2.3) | |
| - | - | - | J353 | Hypertrophy of tonsils with hypertrophy of adenoids | 212 (2.0) | |
| - | - | - | R509 | Fever, unspecified | 176 (1.6) | |
*Hospitalization diagnostic coding captured in the discharge abstract database (DAD).
**ICD-9 coding used from April 1, 1995 until March 31, 2002.
†ICD-10 coding used from April 1, 2002 until Mar 31, 2019.
NR: not reported due to small cell size.
Outpatient physician follow-up
Within 1 year post-discharge, 3,740 (81.4%) KD cases were seen by a primary care physician, 4,267 (92.8%) by a pediatrician, 2,202 (47.9%) by a cardiologist, and 153 (3.3%) by a rheumatologist (Supplementary Table 3). Over time, outpatient primary care physician visits became more common, while pediatrician and cardiologist follow-up declined. Between 1 and 5 years and >5 years post-discharge, 1,618 (35.4%) and 841 (23.6%) KD cases had ≥1 cardiologist visit, respectively. Primary care, pediatrician, cardiologist, and rheumatologist visits were all more common among KD cases versus nonexposed children; the largest differences were seen for pediatrician and cardiologist visits. In both cohorts, most children had ≥1 outpatient physician visit within 1 year post-enrollment (4,585 [99.7%] KD cases versus 389,955 (84.8%) nonexposed children, P < 0.001) (Supplementary Table 3).
Health care costs
KD survivors had higher composite health care costs at all time periods (0 to 1 year, 1 to 5 years, >5 years, and throughout follow-up; Figure 1 and Supplementary Table 4). For participants enrolled between 2000 and 2018 (i.e., with ED data available) the median health care cost (IQR) throughout follow-up was $9,174 (5,100 and 6,070) for KD cases vs. $3,148 (1,464 and 6,115) for nonexposed individuals. Total health care costs for KD cases in Ontario were $13.9 million (within 1 year post-discharge) and $54.8 million (throughout follow-up). For an equivalent number of comparator individuals, the total health care costs were $2.2 million and $23.9 million over the same time periods. Outpatient care was the largest cost driver for both KD (50% of total health care costs) and nonexposed cohorts (55%) (Supplementary Table 5). However, the proportion of costs attributable to hospitalizations was higher among KD cases (43% versus 36% nonexposed), due to higher hospitalization costs within the first year post-discharge.
Figure 1.
Composite health care costs among Kawasaki disease cases versus controls.
DISCUSSION
To our knowledge, this is the first study to evaluate long-term health care utilization and costs after KD diagnosis. Using Ontario health administrative databases, we report novel data on physician follow-up patterns, rates of hospital re-admissions, and ED visits, as well as relevant cost drivers. Compared to nonexposed children, KD survivors had significantly higher long-term rates of health care utilization and cost. Nearly all KD cases were seen by a pediatrician or primary care physician within 1 year post-discharge, while less than half were seen by a cardiologist.
Children diagnosed with KD typically require hospitalization for investigation and IVIG administration. American studies have estimated median costs associated with acute KD hospitalizations to be $5,652 to 32,170 USD (2,3,12–15). Across the USA, the cost of acute KD hospitalizations is estimated to be $38 to 110 million USD annually (2,12). Hospital costs for KD are higher than for other hospitalization causes in young children (e.g., bronchiolitis or gastroenteritis) (12), but less than all-cause pediatric hospitalizations (2). Cost drivers during KD hospitalizations include IVIG products, echocardiography, specialist consultations, and prolonged length of stay for fever monitoring and treatment. In Canada, IVIG products cost approximately $120 CAD per kilogram of body weight (per treatment) (26,27). Additionally, hospital costs are reported to be higher among weekend KD admissions, children with CAA, ‘refractory KD’ cases (requiring ≥2 IVIG doses) and children without health care insurance (13,14).
Recently, Johnson et al. described health care utilization among American KD hospitalizations between 2015 and 2017 using commercial insurance claim data (16). Comparing 90-day periods pre- and post-KD hospitalization, they found that mean health care costs were higher post-discharge ($3778 versus $2090 USD, P < 0.0001). Office visits and echocardiograms were the most significant health care cost drivers post-discharge. However, this study excluded children that received Medicaid or were uninsured; it may not be applicable to universal health care systems. Furthermore, health care utilization immediately prior to KD hospitalization may not be an appropriate comparator. The majority of their pre-KD hospitalization health care utilization occurred within 7 days prior to admission, which was likely related to the acute KD illness. Interestingly, we found that pre-KD hospitalization health care utilization was significantly higher among KD cases than nonexposed children. Health care visits related to either the acute KD illness (prior to diagnosis) or a triggering infectious event may contribute to this. However, most KD cases (77.2%) had >5 outpatient visits in the preceding year, which is unlikely to be explained by the KD illness alone. KD cases also had higher baseline frequencies of chronic disease, prior cardiovascular events, and congenital heart disease. Given the lack of specific diagnostic testing for KD, over-diagnosis is a possibility. Other authors have reported that KD demonstrates an over-diagnosis pattern, based on increasing KD incidence without a corresponding increase in the incidence of KD-associated CAA (4,20,28,29). Additional health care visits and surveillance among individuals with pre-existing comorbidities may have increased their likelihood of KD diagnosis.
To better understand post-discharge health care utilization and costs specifically attributable to KD diagnosis, we compared our KD cohort against a matched cohort of nonexposed children. Currently, there is minimal data on KD health care utilization in a universal health care system. In the Canadian context, Klassen et al. analyzed IVIG cost-effectiveness in the early 1990s (17). They found that children receiving IVIG (versus acetylsalicylic acid alone) had a shorter length of stay and lower health care costs. Among children receiving high-dose IVIG, the proportional cost-distribution was hospitalizations (26.9%), laboratory services (19.3%), parental costs (18.9%), physician services (16.1%), pharmacy (10.9%), and clinic visits (7.9%). In our study, hospitalization costs were a significant cost driver for KD cases within one year of discharge. Among KD cases, we found that there was a higher frequency of diagnoses for nonspecific symptoms including fever and rash during ED visits and hospitalizations within the first year post-discharge. This may reflect increased vigilance and anxiety toward symptoms associated with KD among caregivers and health care providers. Overall, outpatient care was the largest cost driver in both KD and nonexposed cohorts. The cost difference between KD and nonexposed cohorts was also greatest within the first year post-enrollment and decreased gradually over time.
Our results emphasize the need to evaluate inpatient care processes for children with KD, with an aim to standardize practice and identify potential cost savings. This may include policies to reduce inappropriate IVIG administration/wastage, improve out-of-hours echocardiography access, and facilitate earlier patient discharge. After discharge, KD follow-up care should be standardized and aligned with American Heart Association 2017 KD guidelines (5). In a universal health care system, follow-up for KD survivors without cardiac involvement (>95% of cases) (30,31) may be feasible in the primary care setting, to reduce unnecessary specialist visits. However, this must be supported by appropriate knowledge translation initiatives.
Our study has a number of strengths. A long follow-up period allowed us to assess differences in health care utilization and costs at various time intervals following KD illness. Ontario’s universal single-payer health care system and linked health administrative databases at ICES allowed us to perform detailed evaluation of health care events across the province, with minimal loss to follow up. Given the rising incidence of KD in Ontario, acute care and follow-up for KD survivors places a significant and growing burden on the provincial health care system.
Our study also has multiple limitations. Administrative coding for KD has been previously validated, with a high positive predictive value (20,32,33). However, we cannot fully characterize KD diagnoses (i.e., complete versus incomplete KD, missed KD or no hospitalization, treatment received). Administrative coding for our main outcomes (hospitalizations, ED visits, outpatient visits, and health care costs) are considered to be very reliable, given that accurate billing is a prerequisite for physician and facility reimbursement. Given the lack of validated administrative coding for CAA and the small proportion of KD cases with CAA present, we decided not to perform subgroup analysis for health care utilization or costs based on CAA status. Our composite health care cost outcome represents the majority of costs incurred by the provincial health care system. However, it does not include costs attributable to prescriptions, other out-of-pocket expenses for families such as private care or home care, or societal costs such as missed work by parents. Physicians that were practicing but not billing as specialists (e.g., not licensed as a specialist in Ontario; or a pediatric specialist billing as a general pediatrician) may have been misclassified. However, based upon physician billing patterns in Ontario, this would be uncommon. Finally, Chinese and South Asian individuals were identified using a surname-based algorithm (24). This database excludes surnames shared between ethnic groups such as South Asian Muslims or non-Han Chinese, and limits our conclusions regarding ethnic differences.
CONCLUSIONS
Using population-based health administrative databases in Ontario, we found that KD survivors had higher long-term rates of health care utilization and health care costs, compared to nonexposed children. The majority of KD survivors had outpatient physician follow-up with pediatricians and primary care physicians. The rising incidence and costs associated with KD place a significant burden on Ontario’s health care system. This justifies consideration of various cost-saving measures, aimed at rationalizing post-KD follow-up care.
Supplementary Material
ACKNOWLEDGEMENTS
This study was supported by ICES, which is funded by an annual grant from the Ontario Ministry of Health and Long-Term Care. This study received funding from institutional grants from Hamilton Health Sciences and McMaster University Department of Pediatrics. Parts of this material are based on data and/or information compiled and provided by CIHI. However, the analyses, conclusions, opinions, and statements expressed in the material are those of the authors, and not necessarily those of CIHI. We thank the ICES McMaster team, including Richard Perez for his dedicated work planning and supervising analysis for this study, as well as Erin O’Leary for project coordination.
Author Contributions: Cal Robinson, Rahul Chanchlani, and Michelle Batthish conceptualized and designed the study, coordinated and supervised data collection and analysis, drafted the initial manuscript, and reviewed and revised the manuscript. Hsien Seow and Elizabeth Darling conceptualized and designed the study, coordinated and supervised data collection and analysis, and reviewed and revised the manuscript. Ana Gayowsky conceptualized and designed the study, was primarily responsible for data collection and analysis, and reviewed and revised the manuscript. All authors have approved the final manuscript as submitted and agree to be accountable for all aspects of the work.
Contributor Information
Cal Robinson, Department of Pediatrics, McMaster University, Hamilton, Ontario, Canada; Division of Nephrology, Department of Pediatrics, University of Toronto, Toronto, Ontario, Canada.
Rahul Chanchlani, Division of Nephrology, Department of Pediatrics, McMaster University, Hamilton, Ontario, Canada; Department of Health Research Methods, Evidence and Impact, McMaster University, Hamilton, Ontario, Canada; ICES McMaster, Hamilton, Ontario, Canada.
Anastasia Gayowsky, ICES McMaster, Hamilton, Ontario, Canada.
Elizabeth Darling, McMaster Midwifery Research Centre, McMaster University, Hamilton, Ontario, Canada.
Hsien Seow, Department of Oncology, McMaster University, Hamilton, Ontario, Canada.
Michelle Batthish, Division of Rheumatology, Department of Pediatrics, McMaster University, Hamilton, Ontario, Canada.
ETHICAL APPROVAL
This study was approved by Hamilton Integrated Research Ethics Board (#7093).
FUNDING
Michelle Batthish was supported by a New Investigator Fund from Hamilton Health Sciences. Michelle Batthish was also supported by a Kawasaki Disease Canada Research Award. Cal Robinson was supported by Resident Research Award funding from the Department of Pediatrics, McMaster University.
CONFLICT OF INTEREST
All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
DATA AVAILABILITY
Data may be obtained from a third party and are not publicly available. A data request can be sent to ICES (formerly the Institute for Clinical Evaluative Sciences): https://www.ices.on.ca/About-ICES/ICES-Contacts-and-Locations/contact-form
REFERENCES
- 1. Robinson C, Chanchlani R, Gayowsky A, et al. Incidence and short-term outcomes of Kawasaki disease. Pediatr Res. 2021. March 30. Available from: http://www.nature.com/articles/s41390-021-01496-5. [cited April 11, 2021]. [DOI] [PubMed]
- 2. Holman RC, Belay ED, Christensen KY, Folkema AM, Steiner CA, Schonberger LB. Hospitalizations for Kawasaki syndrome among children in the United States, 1997-2007. Pediatr Infect Dis J 2010;29(6):483–8. [DOI] [PubMed] [Google Scholar]
- 3. Chang R-KR. Hospitalizations for Kawasaki disease among children in the United States, 1988-1997. Pediatrics. 2002;109(6):e87. [DOI] [PubMed] [Google Scholar]
- 4. Uehara R, Belay ED. Epidemiology of Kawasaki disease in Asia, Europe, and the United States. J Epidemiol 2012;22(2):79–85. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. McCrindle BW, Rowley AH, Newburger JW, et al. ; American Heart Association Rheumatic Fever, Endocarditis, and Kawasaki Disease Committee of the Council on Cardiovascular Disease in the Young; Council on Cardiovascular and Stroke Nursing; Council on Cardiovascular Surgery and Anesthesia; and Council on Epidemiology and Prevention . 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] [PubMed] [Google Scholar]
- 6. Friedman KG, Gauvreau K, Hamaoka-Okamoto A, et al. Coronary artery aneurysms in Kawasaki disease: Risk factors for progressive disease and adverse cardiac events in the US population. J Am Heart Assoc. 2016;15(9):5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Kato H, Sugimura T, Akagi T, et al. Long-term consequences of Kawasaki disease. A 10- to 21-year follow-up study of 594 patients. Circulation 1996;94(6):1379–85. [DOI] [PubMed] [Google Scholar]
- 8. Bang JS, Kim GB, Kwon BS, et al. Long-term prognosis for patients with Kawasaki disease complicated by large coronary aneurysm (diameter ≥6 mm). Korean Circ J 2017;47(4):516–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Holve TJ, Patel A, Chau Q, Marks AR, Meadows A, Zaroff JG. Long-term cardiovascular outcomes in survivors of Kawasaki disease. Pediatrics 2014;133(2):e305–11. [DOI] [PubMed] [Google Scholar]
- 10. Miura M, Kobayashi T, Kaneko T, et al. ; The Z-score Project 2nd Stage Study Group. Association of severity of coronary artery aneurysms in patients with Kawasaki disease and risk of later coronary events. JAMA Pediatr 2018;172(5):e180030. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Robinson C, Lao F, Chanchlani R, Gayowsky A, Darling E, Batthish M. Long-term hearing and neurodevelopmental outcomes following Kawasaki disease: A population-based cohort study. Brain Dev 2021;43(7):735–44. [DOI] [PubMed] [Google Scholar]
- 12. Belay ED, Holman RC, Maddox RA, Foster DA, Schonberger LB. Kawasaki syndrome hospitalizations and associated costs in the United States. Public Health Rep 2003;118(5):464–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. 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] [PubMed] [Google Scholar]
- 14. Lo JY, Minich LL, Tani LY, Wilkes J, Ding Q, Menon SC. Factors associated with resource utilization and coronary artery dilation in refractory Kawasaki disease (from the Pediatric Health Information System Database). Am J Cardiol 2016;118(11):1636–40. [DOI] [PubMed] [Google Scholar]
- 15. Ramphul K, Mejias SG, Joynauth J. Kawasaki disease among children in the United States. Reumatologia 2019;57(4):253–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Johnson SC, Chew M, Williams DC, Simpson A, Brinton D, Andrews AL. Assessing healthcare utilization and compliance in Kawasaki disease. J Pediatr. 2021;234:181–186.e1. doi:10.1016/j.jpeds.2021.03.022. [DOI] [PubMed] [Google Scholar]
- 17. Klassen TP, Rowe PC, Gafni A. Economic evaluation of intravenous immune globulin therapy for Kawasaki syndrome. J Pediatr 1993;122(4):538–42. [DOI] [PubMed] [Google Scholar]
- 18. von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP; STROBE Initiative . The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: Guidelines for reporting observational studies. Lancet 2007;370(9596):1453–7. [DOI] [PubMed] [Google Scholar]
- 19. Benchimol EI, Smeeth L, Guttmann A, et al. ; RECORD Working Committee . The REporting of studies Conducted using Observational Routinely-collected health Data (RECORD) statement. PLoS Med 2015;12(10):e1001885. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Manlhiot C, O’Shea S, Bernknopf B, et al. Epidemiology of Kawasaki disease in Canada 2004 to 2014: Comparison of surveillance using administrative data vs periodic medical record review. Can J Cardiol. 2018;34(3):303–9. [DOI] [PubMed] [Google Scholar]
- 21. Chanchlani R, Nash DM, McArthur E, et al. Secular trends in incidence, modality and mortality with dialysis receiving AKI in children in Ontario: A population-based cohort study. Clin J Am Soc Nephrol 2019;14(9):1288–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Bollman R. Rural Demography Update.Rural Ontario Institute; 2016. Available from: http://www.ruralontarioinstitute.ca/file.aspx?id=26acac18-6d6e-4fc5-8be6-c16d326305fe. [cited December 10, 2019]. [Google Scholar]
- 23. Canadian Institute for Health Information (CIHI). Health Indicators 2013 [Internet]. Canadian Institute for Health Information; 2013. May [cited December 18, 2019]. Available from: https://secure.cihi.ca/free_products/HI2013_EN.pdf [Google Scholar]
- 24. Shah BR, Chiu M, Amin S, Ramani M, Sadry S, Tu JV. Surname lists to identify South Asian and Chinese ethnicity from secondary data in Ontario, Canada: A validation study. BMC Med Res Methodol 2010;10:42. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Simon TD, Haaland W, Hawley K, Lambka K, Mangione-Smith R. Development and Validation of the Pediatric Medical Complexity Algorithm (PMCA) Version 3.0. Acad Pediatr 2018;18(5):577–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Blackhouse G, Gaebel K, Xie F, et al. Cost-utility of Intravenous Immunoglobulin (IVIG) compared with corticosteroids for the treatment of Chronic Inflammatory Demyelinating Polyneuropathy (CIDP) in Canada. Cost Eff Resour Alloc 2010;8:14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Strickland S. IVIG and SCIG Utilization in the Atlantic Provinces in FY 2016/17 [Internet]. Halifax: Nova Scotia Provincial Blood Coordinating Program; 2017. July [cited June 7, 2020]. Available from: file:///Users/California%20Love/Downloads/20180104-ivig-scig-utilization-atlantic-provinces-fy-2016-17.pdf [Google Scholar]
- 28. Coon ER, Wilkes J, Bratton SL, Srivastava R. Paediatric overdiagnosis modelled by coronary abnormality trends in Kawasaki disease. Arch Dis Child 2018;103(10):937–41. [DOI] [PubMed] [Google Scholar]
- 29. Lin YT, Manlhiot C, Ching JCY, et al. Repeated systematic surveillance of Kawasaki disease in Ontario from 1995 to 2006: Kawasaki disease in Ontario, Canada. Pediatrics International. 2010;52(5):699–706. [DOI] [PubMed] [Google Scholar]
- 30. Newburger JW, Takahashi M, Gerber MA, et al. ; Committee on Rheumatic Fever, Endocarditis, and Kawasaki Disease, Council on Cardiovascular Disease in the Young, American Heart Association . Diagnosis, treatment, and long-term management of Kawasaki disease: A statement for health professionals from the Committee on Rheumatic Fever, Endocarditis, and Kawasaki Disease, Council on Cardiovascular Disease in the Young, American Heart Association. Pediatrics 2004;114(6):1708–33. [DOI] [PubMed] [Google Scholar]
- 31. Oates-Whitehead RM, Baumer JH, Haines L, et al. Intravenous immunoglobulin for the treatment of Kawasaki disease in children. Cochrane Database Syst Rev. 2003;( 4):CD004000. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Bronstein DE, Besser RE, Burns JC. Passive surveillance for Kawasaki disease in San Diego County. Pediatr Infect Dis J 1997;16(11):1015–8. [DOI] [PubMed] [Google Scholar]
- 33. Kao AS, Getis A, Brodine S, Burns JC. Spatial and temporal clustering of Kawasaki syndrome cases. Pediatr Infect Dis J 2008;27(11):981–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Data may be obtained from a third party and are not publicly available. A data request can be sent to ICES (formerly the Institute for Clinical Evaluative Sciences): https://www.ices.on.ca/About-ICES/ICES-Contacts-and-Locations/contact-form