.png){kind=icon}
.png){kind=icon}
Return
Comparison of univariate and multivariate spatial scan statistics in detecting geographic clusters of croup cases in Alberta
.png){kind=icon}
OA
AI
S
B
R
DOI：10.1186/s12942-026-00468-4.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Croup is a common pediatric respiratory illness primarily affecting infants and toddlers. While understanding the geographic distribution and spatial clustering of croup healthcare encounters is crucial for public health planning and resource allocation, this area remains understudied. This study aimed to identify geographic clusters of croup healthcare encounters and areas of higher healthcare demand in Alberta, Canada, and compare univariate and multivariate spatial scan statistics using emergency department (ED) visits and physician claims data. We analyzed administrative health data for children aged $$\le$$2 years in 70 sub-regional health authorities (sRHAs) in Alberta from 2017/18 to 2022/23. Data were aggregated by sRHA and year. Kulldorff’s spatial scan statistics were used to identify croup clusters adjusted for sex distribution for each year separately. Analyses were conducted on each data source separately (univariate) and for both data sources simultaneously (multivariate). During the study period, there were 32,740 ED visits and 49,389 physician claims for croup, with males accounting for 63.5% and 62.2% of the encounters, respectively. Overall, 59.8% of patients with physician claims and 82.8% of ED patients (n=21,688) accessed both healthcare settings during the study period. Significant spatial clustering was consistently identified, with 1-4 clusters annually in univariate, and 2-5 clusters in multivariate analyses. Certain northern areas appeared most consistently in all years and methods. The COVID-19 pandemic year (2020/21) showed unique patterns with the highest relative risks. ED visits data demonstrated wider geographic coverage with consistent major metropolitan area involvement, while physician claims data showed frequent clustering in different metropolitan areas. Multivariate and univariate analyses showed overlapping but distinct findings, with multivariate analysis identifying two clusters where only physician claims data showed significantly higher case numbers than expected. Significant geographic clustering of croup healthcare encounters exists in Alberta, with northern regions most consistently identified as areas of higher healthcare utilization. Both univariate and multivariate spatial scan statistics detected significant clusters, with multivariate analysis providing additional insights by simultaneously analyzing ED visits and physician claims data. The distinct clustering patterns between data sources indicate different healthcare utilization behaviours and demonstrate the value of multivariate approaches for comprehensive spatial epidemiological analysis.
Keywords:
Multivariate spatial analysis
Univariate spatial analysis
Statistical cluster detection
Administrative health data
Spatial scan
Croup
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
3.2
Papers:
984
Citations:
2.8K
Organization
Cited Papers