, Introduction
The incidence of end-stage kidney disease (ESKD) has been increasing worldwide. Although the incidence of newly registered ESKD slightly decreased after the coronavirus disease outbreak, it is still one of the greatest socioeconomic burdens, causing excessive healthcare costs [1]. The incidence and prevalence of ESKD in Korea are high. This increasing incidence of ESKD is partly due to an aging society and an increase in morbidities such as diabetes and hypertension [2,3].
Patients with ESKD are likely to visit the emergency department (ED) more frequently and have poorer clinical outcomes after visiting the ED than those without chronic kidney disease (CKD) because of ESKD-related complications and underlying comorbidities [4]. Infection, electrolyte imbalance, and cardiovascular events are common causes of ED visits for patients with ESKD [5]. In a recent paper, we demonstrated that patients with ESKD have significantly higher rates of hospitalization and in-hospital mortality than do those without CKD [6].
Moreover, hemodialysis patients with low socioeconomic status have even poorer clinical outcomes. Our previous study using the Korean hemodialysis quality assessment database revealed that patients receiving Medical Aid (MA) have significantly greater risks of hospitalization and mortality than those receiving National Health Insurance (NHI) [7]. However, few studies have directly investigated the influence of health insurance status on acute healthcare outcomes in ESKD patients, and the existing evidence has been inconclusive [8,9]. Therefore, we compared the rates of hospitalization and in-hospital mortality after ED admission between MA recipients and NHI beneficiaries among patients with ESKD.
Methods
Study design
This retrospective observational study was conducted using ED visit data from regional and local emergency medical centers between January 1, 2018 and December 31, 2021. Data were obtained from the National Emergency Department Information System (NEDIS), which is operated by the National Emergency Medical Center in Korea. The NEDIS is a comprehensive database that captures nearly all ED visits across South Korea (covering >95% of emergency centers and hundreds of hospitals), recording detailed information on patient demographics, diagnoses, treatments, and outcomes for each visit [10].
This study was performed in accordance with the Declaration of Helsinki after approval by the Institutional Review Board of Ulsan University Hospital, and the requirement for informed consent was waived (No. UUH2024-01-017, UUH2023-10-034).
Study population
We included all adult patients aged 20 years or older who were diagnosed with ESKD and visited the ED during the study period. Patients with ESKD were identified via the International Classification of Diseases, 10th Revision (ICD-10) code N18.5 as either the primary or secondary diagnosis. ESKD was defined as CKD stage 5, regardless of whether the patient was undergoing hemodialysis or peritoneal dialysis. Additional exclusion criteria included missing values for key variables such as the Korean Triage and Acuity Scale (KTAS), ED outcomes (discharge, admission, or death), and ED length of stay. The patients were categorized according to their insurance type as either MA recipients or NHI beneficiaries before analysis.
Study variables and study outcome
The collected variables included age, sex, level of emergency medical center, insurance type, route of ED arrival, KTAS classification, length of ED stay, ED outcome, and cause of the ED visit. Ages were grouped into 10-year intervals. The levels of emergency medical institutions, as designated by the Ministry of Health and Welfare, were categorized into regional emergency medical centers, community emergency medical centers, and community emergency medical institutes. The routes of ED arrival included direct visits, transfer from another hospital, transfer from an outpatient clinic, and others. The KTAS classification, which is based on patient severity, is divided into five levels: resuscitation, emergent, urgent, less urgent, and nonurgent. The causes of ED visits were categorized into cardiovascular disease, cerebrovascular disease, infection, vascular access issues, electrolyte imbalance, and others. The primary study outcomes were the hospitalization rate and in-hospital mortality following ED visits according to health insurance type. Hospitalization following the ED visit was defined as admission to the general ward or intensive care unit. In-hospital mortality was defined as death occurring during hospitalization following the ED visit.
Statistical analysis
Categorical variables were compared via the chi-square test, whereas continuous variables were analyzed via t tests or analysis of variance. The results are presented as frequencies (percentages) for categorical variables and as the means with standard deviations for continuous variables.
To evaluate the impact of insurance type on hospitalization and in-hospital mortality among patients with ESKD, logistic regression analyses were conducted with adjustments for age, sex, and KTAS classification. All the statistical tests were two-sided, and a p-value <0.05 was considered significant. Analyses were performed via R software version 4.0.5 (R Foundation for Statistical Computing).
Results
Baseline characteristics of patients with end-stage kidney disease visiting the emergency department
A total of 26,930,142 emergency visits were recorded between January 2018 and December 2021 in the NEDIS database (Fig. 1). After excluding those with CKD stages 1 through 4 and those without data on the KTAS score, disposition, and length of stay, 21,706,470 ED visits were noted. Among them, 159,456 visits were ESKD-related. After excluding 1,662 ED visits without health insurance information, 157,794 ESKD-related ED visits were included in the analysis. Among them, 37,133 visits (23.5%) were from MA recipients, and 120,661 visits (76.5%) were from NHI beneficiaries.
The baseline characteristics of the patients according to health insurance status are presented in Table 1. MA recipients were mostly distributed in the 30- to 60-year-old age groups, whereas NHI beneficiaries were mostly older patients. Most patients visited the ED by direct visits (66.2%), while 25.3% were transferred from other hospitals. Compared with MA recipients, NHI beneficiaries had higher rates of KTAS classification, including resuscitation (3.8% vs. 3.3%), emergent (14.9% vs. 14.7%), and urgent (56.2% vs. 55.3%) (p < 0.001). The most common primary diagnoses for ED visits were infection (11.5%), vascular access problems (6.8%), and heart problems (6.4%).
Outcomes following emergency department visits according to health insurance status
Overall, 659 patients (0.4%) died in the ED, while 105,174 (66.7%) were hospitalized. MA recipients were more likely to be discharged (n = 12,931, 34.8%) compared to NHI beneficiaries (n = 36,097, 29.9%; p < 0.001) (Fig. 2). In contrast, NHI beneficiaries were more likely to be hospitalized (n = 81,773, 67.8%) than MA recipients were (n = 23,401, 63.0%; p < 0.001) (Fig. 3). Infection was the most common cause of hospitalization in MA recipients and NHI beneficiaries (Supplementary Table 1, available online).
Overall, 9,704 patients (9.2%) died during hospitalization. The proportion of in-hospital mortality was greater for NHI beneficiaries (7,926/120,661, 6.6%) than for MA recipients (1,778/37,133, 4.8%; p < 0.001) (Fig. 3). The most common cause of in-hospital death was infection (15.0%) (Supplementary Table 1, available online).
The risk of hospitalization and in-hospital mortality was lower for MA recipients before adjustment (Model 1 in Table 2) than for NHI beneficiaries. Even after adjusting for confounders (age and sex in Model 2, age, sex, and KTAS score, hospital service level, route of arrival, length of stay in Model 3), MA recipients demonstrated a lower risk of hospitalization (odds ratio [OR], 0.90 [95% CI, 0.88–0.92] in Model 2; OR, 0.89 [95% CI, 0.87–0.92] in Model 3) (p < 0.001) than did NHI beneficiaries. However, the risk of in-hospital mortality was not different between MA recipients and NHI beneficiaries after adjusting for confounders (OR, 0.97 [95% CI, 0.92–1.03] in Model 2; OR, 0.97 [95% CI, 0.92–1.03] in Model 3).
Factors associated with the risk of hospitalization via the emergency department according to health insurance status
The risk of hospitalization was 11% lower for MA recipients than for NHI beneficiaries after adjusting for age, sex, and KTAS score, hospital service level, route of arrival, and length of stay (Table 2). When we performed subgroup analysis according to age group, the risk of hospitalization was lower in MA recipients between 30 and 60 years of age (Fig. 4A). The risk of hospitalization did not differ between males and females and did not differ according to the KTAS score.
Factors associated with the risk of in-hospital mortality via the emergency department according to health insurance status
The risk of in-hospital mortality was not different between MA recipients and NHI beneficiaries after adjusting for age, sex, and KTAS score, hospital service level, route of arrival, and length of stay (Table 2). However, when we performed subgroup analysis according to age group, the risk of in-hospital mortality was greater for patients younger than 60 years. The risk of in-hospital mortality was 2.8 times higher in MA recipients than in NHI beneficiaries in their 30s and approximately 1.3 times higher than that of patients in their 40s and 50s (Fig. 4B).
Discussion
We compared the rates of hospitalization and in-hospital mortality after ED admission between MA recipients and NHI beneficiaries among patients with ESKD. Compared with NHI beneficiaries, MA recipients had a lower overall risk of hospitalization in the ED. Specifically, MA recipients between their 30s and 60s had significantly lower rates of hospitalization in the ED than did NHI beneficiaries. In contrast, the overall risk of in-hospital mortality did not differ according to health insurance status. However, MA recipients aged 60 years or younger demonstrated a greater risk of in-hospital mortality than did NHI beneficiaries.
In Korea, every citizen is enrolled in the NHI system if certain conditions are met. Therefore, approximately 97% of the citizens are NHI beneficiaries. Moreover, approximately 3% of the citizens are classified as MA recipients due to their low income. However, the proportion of MA recipients among patients with ESKD is greater (23.3%) than that among the overall patient population [7]. Although the government pays the medical expenses of MA recipients, their accessibility to hospitals is still low. There are several reasons for this. First, most MA recipients lose their jobs or change their jobs to part-time to earn income after they are diagnosed with serious illnesses such as ESKD. Therefore, the hospital utilization rate may decrease. In addition, MA patients may have limitations in terms of active laboratory testing or treatment due to items not covered by medical insurance. A previous study demonstrated that NHI beneficiaries with ESKD had greater out-of-pocket expenditures for admission and the outpatient department than MA recipients with ESKD did [11]. Compared with MA recipients, NHI beneficiaries are more likely to receive better medical care.
Although there has been an effort to improve overall care for MA recipients among patients with ESKD, MA recipients are still at risk of acute complications. Indeed, previous studies have demonstrated that patients with low socioeconomic status are likely to visit the ED more frequently than those with high socioeconomic status. One large-scale population-level study demonstrated that patients reporting low income, lack of community, low educational level, or social isolation have significantly greater risks of ED visits [12]. Another study using the National Center for Health Statistics database demonstrated that socioeconomic status and health insurance coverage are significant determinants of healthcare access [13]. A study with older patients with comorbidities also revealed that socioeconomic status is an independent risk factor for high ED visits after adjusting for confounders [14].
Our study is the first Korean study to classify the clinical outcomes of patients with ESKD visiting the ED according to health insurance status. Our previous study demonstrated that MA recipients have higher rates of hospitalization and patient mortality than do NHI beneficiaries after adjustment for confounders [7]. However, the hospitalization rate was lower for MA recipients after the ED visit than for NHI beneficiaries in the present study. A possible explanation could be that MA recipients who are economically disadvantaged often visit the ED but receive treatment as a temporary measure and then are discharged instead of being hospitalized. Since hospitalization makes it difficult for them to engage in economic activities, they try to receive treatment in the ED and are discharged. In a study conducted in the United States, hemodialysis patients who did not have health insurance were younger and had a lower rate of hospitalization [15]. In support of this information, our subgroup analysis revealed that patients in the most economically active young age group of 30–60 years had a lower rate of hospitalization (Fig. 4A). It is also possible that the patients who declined to be hospitalized from the ED room may have poorer clinical outcomes. In Fig. 2, MA recipients were more likely to be discharged after ED visit while NHI beneficiaries were more likely to be admitted to the hospital. MA recipients may be discharged because they received proper care from the ED room, but on the other hand, MA recipients may be discharged against medical advice due to their financial and personal problems. In contrast, there was no difference in in-hospital mortality between MA recipients and NHI beneficiaries. These results are in line with those of previous studies. One study demonstrated that, compared with patients with high socioeconomic status, older patients with low socioeconomic status have higher hospitalization rates but similar in-hospital mortality rates [16]. Another study also revealed that adult patients visiting the ED with sepsis do not have different in-hospital mortality rates according to health insurance status after propensity score matching [17]. Overall, there was no difference in mortality rates during hospitalization according to health insurance status in our study. However, attention should be given to the in-hospital mortality rates according to age groups. In our study, MA recipients aged 60 years or younger had a higher mortality rate than did NHI beneficiaries. Patients under age 60 years are usually economically active suggesting that they may be reluctant to visit the ED until they become seriously ill. Previous article by Meacock et al. [18] also suggested that higher in-hospital mortality among ED visitors during weekends may reflect a lower probability of admission during weekends. The reduced utilization of primary care and admission rate from ED among young MA recipients may result in serious illness and higher in-hospital mortality. Therefore, in cases where patients with ESKD with low socioeconomic status visiting the ED are hospitalized, the hospitalization rate may be low among younger patients, but the severity and short-term mortality can be greater than those in NHI beneficiaries.
This study has several limitations. First, a large amount of data may be missing because of the retrospective nature of this study. This study was performed via the NEDIS database, which relies on administrative data; therefore, important clinical data, including laboratory data, comorbidities, and duration of ESKD diagnosis, are missing. After adjusting for comorbidity risk, the health insurance status itself may not be an independent risk factor for hospitalization or in-hospital mortality. In addition, we could not classify the patients into those on hemodialysis, peritoneal dialysis, kidney transplant recipients, or predialysis populations since we only classified patients with ESKD using the ICD-10 code (N18.5). Therefore, the clinical outcomes of patients with ESKD include predialysis CKD stage 5, dialysis, and transplant recipients. We only examined in-hospital mortality among patients visiting the ED; therefore, we were unable to assess long-term patient mortality according to health insurance status.
In conclusion, this study highlights the different clinical outcomes following ED visits among patients with ESKD with different health insurance statuses. Compared with NHI beneficiaries, MA recipients were less likely to be hospitalized after an ED visit but had similar in-hospital mortality rates. However, MA recipients aged under 60 years showed higher in-hospital mortality, suggesting delayed care or more severe illness at presentation. These findings highlight the need to improve access and early intervention for socioeconomically disadvantaged ESKD patients.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Supplement table 1
Print
