Abstract
Objective. The purpose of the present study was to investigate the prevalence and incidence of juvenile systemic lupus erythematosus (JSLE) in Korea.
Methods. The data were collected from the National Health Insurance Claims Database of Korea. JSLE was identified using the diagnostic code M32 from the Korean Standard Classification of Diseases. Patients between 5 and 18 years old, who had at least 1 claim for JSLE from January 1, 2016, to December 31, 2017, as final diagnosis, were analyzed in the study. For prevalent cases, patients who used, at least 1 time, any type of medical services with a diagnostic code of M32 were selected. For incident cases, patients who did not use medical services with the M32 code 1 year prior and who were newly registered in 2017 were defined. Change-point analysis was used to find the age at which changes in prevalence and incidence occurred.
Results. The prevalence of JSLE was 6.92 per 100,000 persons and the incidence of JSLE was 2.76 per 100,000 person-years in patients between 5 and 18 years old. The prevalence and incidence of JSLE were higher in females than in males. According to the change-point analysis, we found that the incidence and prevalence of female patients increased rapidly at the ages of 14 and 15 years, respectively.
Conclusion. This Korean population-based epidemiological study of JSLE showed similar epidemiologic profiles to Asian population in other studies. The distribution of age, ethnicity, and pubertal status are important factors that influence population estimates of JSLE incidence and prevalence.
- change-point
- incidence
- juvenile systemic lupus erythematosus
- Korea
- prevalence
- systemic lupus erythematosus
The epidemiology of juvenile systemic lupus erythematosus (JSLE) varies between ethnic groups and countries. Previous nationwide population-based or cohort studies of JSLE demonstrated higher rates in non-White populations, Asian, and African American populations (Supplementary Data 1, available from the authors on request). They reported an incidence of JSLE from 0.36 to 2.5 per 100,000 persons and a prevalence of 1.89 to 25.7 per 100,000 persons1,2.
A few JSLE data were reported indirectly as a small fraction of the epidemiological data of adult SLE, categorized by 10-year age groups2,3,4. Moreover, the majority of studies in Korea have focused on those of adult SLE3,5.
This study aimed to investigate the prevalence and incidence of JSLE in the Korean population in 2017 based on national claims data.
MATERIALS AND METHODS
The data were collected from the exclusive National Health Insurance Claims Database, managed by the Health Insurance Review and Assessment (HIRA) service in South Korea. HIRA assesses the quality of healthcare services provided to patients and reviews the medical fees for reimbursement decisions in partnership with the National Health Insurance Service (NHIS). The Korean NHIS, developed in 1989, manages over 97.7% of the population, which is about 50 million6,7.
The study participants included all patients who had JSLE as a primary or secondary diagnosis in an outpatient visit or a hospital admission from January 1, 2016, to December 31, 2017. JSLE was identified using the diagnostic code M32 from the Korean Standard Classification of Diseases.
JSLE cases, which are registered in the copayment assistance policy for rare, incurable diseases by NHIS in South Korea7, were identified as true and accurate cases. To register the patient in this copayment assistance policy, a rheumatologist must confirm their diagnosis and apply on their behalf. The patients must meet at least 4 of the American College of Rheumatology (ACR) classification criteria revised in 1997 to satisfy the reimbursement policy of the NHIS8. Registration and copayment for a rare, incurable disease is strictly regulated by the Ministry of Health and Welfare, and the possibility of misclassification can be minimized by this process.
We collected data from patients of both sexes, from ages 0 to < 18 years. The patients had at least 1 claim for a JSLE diagnosis, and the first visit date for JSLE was obtained. This study was approved by the Institutional Review Board of Kyungpook National University Chilgok Hospital (KNUCH 2017-12-029).
The prevalence of JSLE in 2017 was calculated using the number of affected cases during the year divided by the total population, presented as cases per 100,000 persons. Data on the total South Korean population were estimated from the midyear resident registration population in 1-year age groups in 2017, available from the Korean Statistical Information Service (http://kosis.kr).
The incidence of JSLE was calculated using the number of new incident cases during the year divided by the total population of the year, presented as cases per 100,000 person-years (PY). Incident cases were defined as patients newly diagnosed with JSLE during 2017.
In both sexes, the annual age-specific prevalence and incidence were analyzed. We used a change-point simulation to find the age at which changes in prevalence and incidence occurred
Statistical analyses were performed using SAS 9.4 (SAS Institute Inc.) and IBM SPSS statistics 19.0 (IBM Corp.). A change-point analysis was conducted using R statistical software (www.r-project.org). A value P < 0.05 was considered statistically significant.
RESULTS
We analyzed the JSLE patients between 5 and 18 years old, as the claim data cannot identify a medical record of patients under 5 years (13 prevalent cases and 11 incident cases), which may be assumed as maternal SLE or neonatal SLE. There were 447 prevalent JSLE patients (93 males, 354 females) and 178 incident JSLE patients (45 males, 133 females).
The prevalence of JSLE was 6.92 (95% CI 6.90–6.94) per 100,000 persons (Table 1). The incidence of JSLE was 2.76 (95% CI 2.75–2.77) per 100,000 PY (Table 2).
Prevalence of JSLE.
Incidence of JSLE.
We estimated the sex- and age-specific prevalence and incidence of JSLE during the period; these were higher in females than in males. There was a marked increase in females during the period of adolescence. The difference between males and females was less prominent in very early–onset JSLE.
The prevalence of JSLE in males and females was 2.78 (95% CI 2.76–2.80) per 100,000 male population and 11.37 (95% CI 11.34–11.41) per 100,000 female population. The incidence of JSLE in males was 1.35 (95% CI 1.33–1.36) per 100,000 male population, and 4.27 (95% CI 4.25–4.30) per 100,000 female population.
According to the change-point analysis, we found that the incidence and prevalence of female patients increased rapidly at the ages of 14 and 15 years, respectively (Figure 1). The sensitivity analysis using different age intervals of 2–5 years showed that the inflection points were within the same age interval (Supplementary Data 2, available from the authors on request). There was no significant change point by age in male patients.
Change point of incidence and prevalence in female patients with JSLE. According to the change-point analysis, we found that the incidence (left) and prevalence (right) of female patients increased rapidly at the ages of 14 and 15 years, respectively.
DISCUSSION
Compared to population-based pediatric studies in other countries, the incidence of JSLE in this study was comparable with the United States (2.22–2.5 per 100,000) and higher than in the UK (0.73 per 100,000)2,9. The prevalence of JSLE was lower than in the US (9.73–12.6 per 100,000)2,9.
In terms of race, the incidence of JSLE among Asians in the UK (0.8–2.5 per 100,000), US (1.61–4.16 per 100,000), and New Zealand (1.17 per 100,000) was similar to that of this study9,10,11. The incidence of White patients in the UK (0.1 per 100,000) and US (0.5–1.33 per 100,000) were lower than in this study2,9,10,11,12,13. The prevalence of JSLE in Taiwan was 6.3 per 100,000 persons, similar to this study14. The prevalence of JSLE in White patients in the US (0.7–4.86 per 100,000 persons) was lower, and that of the Asia Pacific (18.3–25.7 per 100,000 persons) and African American (18.4 per 100,000 persons) populations in the US was higher than our results2,9,10,11,12,13,15. In light of the difference in ethnic proportions, we need to be more cautious in interpreting epidemiologic findings.
Through analysis of data categorized in 1-year age groups, a significant increase in the incidence and prevalence rates was seen in females with JSLE; this increase was not apparent in the males with JSLE. Change-point analysis of the incidence and prevalence of JSLE (Figure 1) showed that the change point occurred at 14 years for incidence and 15 years for prevalence. This result can be supported by a previous report that the mean age at menarche of Korean adolescents is 12.7 years, and overall 95.8% of the girls experienced menstruation before 15.0 years old16. A high incidence rate in patients under 3 years needs to be inspected more cautiously. There are possibilities of neonatal or monogenic causes of SLE17. However, we cannot identify a medical record of every single patient.
In the present study, the prevalence and incidence of JSLE were higher in females than in males, and there was a marked increase in females after adolescence. In many studies, there is clear sex predominance in SLE, affecting more females of childbearing age4,17,18,19. Many clinical and experimental studies have suggested that sex bias in autoimmune diseases such as JSLE may be influenced by sex hormones and sex chromosomes. Because sex bias in autoimmune diseases is stronger post puberty, hormones can initiate or exaggerate the autoimmune process. The female sex hormones estrogen and prolactin are considered to influence the immune response and modulate their coordinated response, such as allowing the survival of autoreactive B cells and skewing their maturation17,19. Sex chromosomes may affect autoimmune disease through differences in X chromosome gene expression, X chromosome gene dosage, and Y chromosome expression17,19.
A limitation of using National Health Insurance claim data is that there is no information on JSLE patients who have not visited a medical institution, and the diagnostic information may be inaccurate due to simple coding errors, classification errors due to lack of medical knowledge, and reimbursement system procedures. Further, our dataset does not include a rate of lupus nephritis or mortality cases that might be important prognostic information.
Despite these issues, national health insurance claims data covering almost the entire population can provide a meaningful, nationally representative study of small populations, such as children or those with rare diseases. Moreover, the use of physician-confirmed diagnoses of SLE according to the ACR criteria using the copayment system data might be a definite strength of this study. In addition, as the Korean and Taiwanese adult SLE epidemiological studies using health insurance data showed similar epidemiologic profiles3,20, our results can be compared to those reported in Taiwan using health insurance data.
In conclusion, there were 4 main findings from this study: The prevalence of JSLE was 6.92 per 100,000 persons; the incidence of JSLE was 2.76 per 100,000 PY; the prevalence and incidence of JSLE were higher in females than in males; and 14 and 15 years of age in females were identified as change point in incidence and prevalence, respectively. In epidemiologic studies in childhood, it might be more important to investigate a rate in smaller age interval rather than a 10-year interval, due to a small number of patients and various change points within the developmental stage.
Footnotes
↵Dr. S.G. Kwak and Dr. S.H. Park contributed equally as co–first authors.
This work was supported by the research fund of Rheumatology Research Foundation (RRF-2017-03).
- Accepted for publication April 9, 2020.
- Copyright © 2021 by the Journal of Rheumatology
