Identifying Characteristics of Frequent and Highly Frequent Users of the Emergency Department: A Cross-Sectional Observational Study


 Introduction:

Since the characteristics of frequent emergency department (ED) users are heterogeneous, it is impossible to mitigate the overcrowding of the ED without the basic data of diagnoses and risk factors of frequent ED users. Our study will provide invaluable information that will help predict patient demand for medical resources while also providing important information that can be used to improve emergency medical services.
Methods

This is the cross-sectional observational study using records from The Korea Health Insurance Review Agency. Frequent ED users were defined as patients who visited an ED more than 7–17 times per calendar year and highly frequent ED users were defined as patients who visited an ED eighteen or more times during the same period. The diseases were ranked by prevalence in each of the three ED frequency groups (less frequent, frequent, and highly frequent ED user groups). Our study then developed two logistic regression models comparing frequent users with less frequent users and highly frequent users with frequent users. Standardized ß values were used to rank risk factor importance.
Results

Although less frequent ED users composed 98.98% of all patients, they only consisted of 92.27% of all ED visits. Compared with less frequent users, a greater proportion of frequent ED users were aged 65 years or older and were insured by Medicaid or Veterans Affair Health Care Program. Frequent ED users were also most strongly defined by wound dressing follow-up visits and liver diseases (standardized ß value of 3.29 and 2.31). However, this study did not show highly frequent ED users differed from frequent ED users in regard to the different disease categories.
Conclusion

The diagnoses and risk factors related to frequent ED visits in Korea identified in this study will be an important reference for future research aimed at reducing ED overcrowding. By further analyzing the risk factors associated with frequent ED use, non-emergency administrative systems or medical facilities can be utilized to reduce the overload on the ED.


Results
Although less frequent ED users composed 98.98% of all patients, they only consisted of 92.27% of all ED visits. Compared with less frequent users, a greater proportion of frequent ED users were aged 65 years or older and were insured by Medicaid or Veterans Affair Health Care Program. Frequent ED users were also most strongly de ned by wound dressing follow-up visits and liver diseases (standardized ß value of 3.29 and 2.31). However, this study did not show highly frequent ED users differed from frequent ED users in regard to the different disease categories.

Conclusion
The diagnoses and risk factors related to frequent ED visits in Korea identi ed in this study will be an important reference for future research aimed at reducing ED overcrowding. By further analyzing the risk factors associated with frequent ED use, non-emergency administrative systems or medical facilities can be utilized to reduce the overload on the ED.

Background
In the current South Korean emergency medical service system, there are 415 emergency departments(EDs) that are classi ed according to medical capacity and designated service area; as of Page 3/21 2018 there were 36 tertiary emergency centers, 121 secondary emergency centers and 258 primary emergency facilities. 16 Despite the su cient number of EDs, the hierarchical regionalism model has not been fully utilized in Korea's emergency medical system. As a result, there is more overcrowding in secondary emergency centers and tertiary emergency centers than in primary emergency facilities.
The national health expenditure in South Korea has increased by an average of $2.9 billion each year from 2014 and 2018. 1 The number of visitors to the emergency department also increased, and many of these visitors were frequent ED users. [2][3][4] In South Korea, as well as in other countries, there is a persistent problem of ED overcrowding. 8,9 The correlation between frequent users and ED overcrowding is a topic of interest not only to emergency physicians, but also to nancers, patients and policy makers. Hence, it is important to obtain a clear understanding of the link between frequent users and increasing volumes of ED patients. Further analysis of frequent ED users is necessary to understand the reason behind this increase in ED visits. [5][6][7] Purpose Although there have been several studies about frequent ED users, most studies only focused on a limited number of diseases or a patient population in a speci c region. [10][11][12][13] One recent study reported the characteristics of frequent ED users in Korea. However, previous studies did not analyze the frequent diseases found in the EDs in Korea. 14 These previous studies did not provide suitable information to make policies for an emergency medical service system. 10,12,17−25 To remedy these issues, our study conducted a systematic review on the causes of frequent ED visits. Instead of selecting for groups of diagnoses that were of interest, this study included all ED patients. Our study will provide invaluable information that will help predict patient demand for medical resources while also providing important information that can be used to improve emergency medical services.

Study setting and participants
A cross-sectional observational study was conducted using all ED visit records of the Korea Health Insurance Review Agency (HIRA). HIRA is a government organization that reviews and assesses medical claims for the National Health Insurance and maintains a database of medical records for the entire Korean population. HIRA provided identi able information that was speci c to each individual. These identi able information mbers anonymously corresponded to record les, which were utilized to provide a comprehensive picture of health care use and patient diseases. HIRA records showed that a total of 10,599,311 patients visited EDs in Korea from 2016 to 2017. It should be noted that HIRA data does not include patients who receive car insurance or worker's compensation. Due to this, the number of ED patients accounted for in this study was less than the actual number of ED patients.
De nition Of Less-frequent, Frequent, And Highly Frequent ED Users To measure ED visit frequency, each patient's most recent ED visit during the 2017 scal year was used as a reference point. In order to create each patient's total ED visit frequency, the ED visits that preceded the most recent visit were counted, beginning from one calendar year (365 days) before the reference point. "Frequent ED users" were de ned as patients who visited an ED more than 7-17 times per calendar year and "highly frequent ED users" were de ned as patients who visited an ED eighteen or more times during the same period. 15 Patients who were under 18 years old were excluded in this study, since their diagnoses contributed to less than 100 patients.

Selection Of Risk Factors
Risk factors associated with ED use were chosen according to a literature review and grouped by patient demographic, illness and disposition. Demographic factors (patient age, sex, insurance status) were also included and classi ed according to the most recent ED visit. Patient diagnoses were based on the 10th revision of International Statistical Classi cation of Disease (ICD-10) code algorithms. In the ICD-10, patient death was not classi ed as a separate group, so it was included in the discharge group.

Primary Data Analysis
To identify the risk factor of frequent and highly frequent ED visits, two separate logistic regression models were constructed. The rst model compared frequent ED users with less frequent ED users, while the second model compared highly frequent ED users with frequent ED users. For logistic regression analyses, patient-level demographic factors such as age, gender, type of insurance and disease categories such as chronic disease, pain-related disease, mental disorder, trauma-related disease, and minor cases were included. Standardized ß values were used to rank risk factor importance; this process can directly translate the size of these values by converting the unit ß values into standardized units.
Assuming equal variance, the standardized odds ratio is affected by the effect and sample size. The methods created by Agresti were used for the standardized odds ratios. All analyses were performed with SPSS (version 18.0; IBM SPSS Statistics, Chicago, IL).

Results
A total of 4,861,213 patients with 7,562,833 ED visits were analyzed in this study (Table 1). Less frequent ED users composed 98.98% of patients but only 92.27% of the total ED visits. Frequent ED users composed of only 0.62% of all patients but 5.94% of ED visits. Similar to this, highly frequent ED users composed of only 0.09% of all patients but 2.24% of ED visits. Each diagnosis was ranked from 1st to 30th in each group according to their ED visit frequency ( Table 2). A total of nineteen diagnoses increased its ranking from the less frequent group to the highly frequent group. Among those, eleven diagnoses -(M54) Radiculopathy, (M25) Other joint disorders, (R52) Pain, unspeci ed, (S33) Injuries of the lumbar spine, (F41) Anxiety disorders, (G44) Cluster headaches, (M48) Other spondylopathies, (K70) Alcoholic liver diseases, (M51) Intervertebral disc disorder, (L03) Cellulitis, and (F10) Alcohol related disorders -were not included in the top 30 in the less frequent group, although their ranks increased in frequent and highly frequent groups.  Compared with less frequent users, a greater proportion of frequent ED users were aged 65 years or older and insured by Medicaid or the Veterans Affair Health Care Program (VA program) ( Table 3). In addition to this, frequent ED users had many more instances of chronic diseases, such liver, lung and renal/endocrine disease, than less frequent ED users. Similarly, a greater proportion of frequent ED users had pain-related diseases or mental disorders when compared to less frequent ED users. Despite this, the less frequent ED users made up a greater percentage of the trauma-related injury group and minor care disease group. Although these disease groups occupied a high proportion of ED visits, the trauma-related and minor care disease groups were not presumed to exhibit a nature that would induce more frequent visits to the ED. In some instances, highly frequent ED users had characteristics that were similar to those of frequent users. Compared to the less frequent ED user group, the highly frequent ED user group had a higher proportion of male patients, a higher proportion of Medicaid or VA insured patients and a higher proportion of patients aged 45 to 74 years old. Compared to the highly frequent ED user group, the less frequent ED user group had a higher proportion of patients with standard health insurance and a higher proportion of patients aged 17 to 44 years old (Table 3). Highly frequent ED users were two or three times as high as frequent users in the pain-related disease group. In addition to this, highly frequent ED users showed a similar pattern when compared with frequent users in the chronic, trauma-related and minor care diseases groups.
The rate of patients hospitalized only once per year was higher for less frequent ED users (28.78%) than the rate for frequent ED users (13.56%) or highly frequent ED users (13.08%). On the other hand, the rate of patients hospitalized more than twice per year was higher in frequent ED users (48.36%) and highly frequent ED users (43.31%) than in less frequent users (6.78%). Overall, highly frequent ED users did not show higher rates of hospitalization than frequent users; however, it should be noted that highly frequent ED users have a relatively low rate of hospitalization.
From these descriptive ndings, the logistic regression was extended ( Table 4). The logistic regression models showed how various risk factors uniquely in uence each group and expressed the importance of these risk factors (standardized ß values). Compared with less frequent ED users, frequent ED users were characterized strongly by the wound dressing follow-up group and liver disease groups (standardized ß value of 3.29 and 2.31). The odds of being a frequent ED user were 26.86-fold greater for wound dressing follow-ups and 10.10-fold greater for patients with liver diseases, when compared to less frequent ED users. Additional factors of frequent ED use include lung disease (odds ratio = 4.40) and arthritis (odds ratio = 4.69). Demographics also showed that patients aged more than 65 years old and patients with Medicaid showed a greater odds ratio (2.793 and 3.45, respectively) of frequent ED users than less frequent ED users. Only two factors statistically signi cantly differentiated highly frequent ED users from frequent ED users ( Table 4). The odds of being a frequent ED user were 31 percent lower for males, and 49 percent lower for Medicaid patients. There were no speci c disease groups that caused the highly frequent ED user group to visit the emergency department more often than the frequent ED user group.

Discussion
This study was conducted using data from the Korea Health Insurance Review Agency. All patients who frequently visited the ED in Korea during one scal year were included in the analysis and the data for each patient was categorized by ICD-10 code. This is the rst study to include all ED users regardless of their diagnosis. Previous studies have selected disease groups or diagnoses of interest rst and then proceeded to analyze their relevance to ED use frequency. 19,20 In our study, frequent ED users were de ned as patients who visited the ED 7-17 times during the scal year. This is different from other studies, where frequent ED users were de ned as a patient who visited the 4 times or more during the scal year. 15 In addition to this, patients who visited the ED 18 times or more during the scal year were de ned as a highly frequent ED user -a patient who was more likely to unnecessarily visit the emergency department than frequent ED user. The top 30 diagnoses of each ED user frequency category (less frequent, frequent and highly frequent) was focused on (Table 2). By knowing the speci c types and characteristics of frequent diagnoses in the frequent and highly frequently ED user groups, this study hopes to help mitigate potential emergency department overuse.
Although there is a general perception that frequent users of the emergency department abuse and misuse emergency services, further investigation of the frequent ED user group provides more insight. 26,27 In South Korea, the government employs national health insurance based on the fee-forservice system. The current fee-schedule may incentivize patients to visit the emergency department as opposed to visiting non-emergency medical facilities. This may explain why there is a tendency for frequent ED users to be linked to minor medical conditions, such as wound dressing follow-up. Patients who were insured by Medicaid showed statistically signi cant differences among the three frequency groups. On the other hand, VA insured patients did not show statistically different differences among three frequency groups.
There is a subset of frequent and highly frequent ED users who had unmet social needs. The majority of these users had ongoing multi-dimensional medical, social, and mental health needs. For instance, patients with minor symptoms in these frequency groups may utilize emergency services as opposed to seeking primary care rst because of nancial reasons. Primary care requires settling co-pays when a patient receives prescription medications, however, emergency services do not charge co-pays if the patients are escorted via ambulance. Due to the varied and complex medical needs of the patients, it is not suitable to generalize about frequent ED users as a homogeneous group.
It is also noteworthy that individuals in the highly frequent ED user group had a higher number of visits to the emergency department than individuals in the frequent ED user group, despite exhibiting a lower rate of hospitalization than the frequent ED user group. This implied that individuals in the highly frequent ED user group visit the emergency department with milder symptoms than individuals in the frequent ED user group. The highly frequent ED user group was only affected signi cantly by gender and insurance status when compared to the frequent ED user group, implying that there were no speci c medical factors that caused the highly frequent ED user group to visit the emergency department more often than the frequent ED user group.
An interesting nding of this study showed that alcohol-related diagnoses were more noticeable in the frequent ED user group than in less frequent ED user groups. Although our study only included alcoholrelated disorders, it should be noted that liver disease and trauma-related diagnoses are often caused by alcohol. Future studies should further investigate different possible alcohol-induced diagnoses in order to analyze emergency department use resulting from alcohol abuse.
The objective of this study was to identify characteristics of frequent and highly frequent users of the emergency department. However, there were the following limitations to our study. The analyses of this study were conducted using data from National Health Insurance users in South Korea. National Health Insurance in South Korea does not cover medical conditions resulting from tra c accidents, worker's compensation, and suicide attempts. In addition to this, our data from NHI did not include patients under the age of 17. Due to these limitations, our ndings cannot be generalized to a more niche population, for instance pediatric patients or accident-related patient groups. In addition to this, it is possible that the diagnosis rate in this study may have been more conservative due to utilization of the main physician diagnostic code that was used to categorize each emergency department visit. This study also de ned the frequent ED user group as individuals who visited the emergency department 7 to 17 times during a scal year. It should be noted that since there is no standard for "frequent ED use", a sensitivity analysis can be conducted to better de ne the frequent ED user group. Overall, this study allowed for the comprehensive identi cation of diagnoses related to frequent ED visits in Korea. The ndings of this study can serve as an important reference for future research aimed at reducing the strain placed on emergency departments by extraneous usage. In order to improve the overload on the emergency departments, non-emergency medical facilities should be utilized for minor illnesses.

Conclusion
Future studies should focus on identifying factors speci c to the social and medical needs of patients, so that patients may be directed to the appropriate care facility. More speci c information on patient characteristics is important for many reasons. In countries that adopt the fee-for-service health insurance system, the ndings from this study would provide implications for future restructuring of healthcare management and policies. However, it should also be noted that the tendency to use emergency departments may differ from country to country, depending on the healthcare system that is employed. Consequently, the results of this study could also be compared to similar studies in other countries. This could potentially provide deeper insights into the relationships between different healthcare systems and health outcomes of different patient populations.