INTRODUCTION
Spinal and spinal nervous diseases are some of the most frequently encountered problems in clinical medicine [
1]. Lower back pain affects up to 80% of the global population at some point in life, and 1% to 2% of the United States adult population is disabled due to lower back pain [
1,
2]. Back and neck pain are 2 of the most common reasons for visits to primary care physicians in the United States and cause considerable disability and financial burden [
3]. Similar to the existing trend in the Republic of Korea (ROK), the incidence of disc herniation was the highest globally in 2014 [
4]. Moreover, degenerative spinal diseases are also increasing in incidence as the average life span increases [
5]. Although most spinal diseases are not fatal, their associated morbidity exerts an enormous toll from both personal and societal perspectives [
3]. The substantial medical needs of affected patients, coupled with our poor understanding of the epidemiology of spinal diseases, have led to an ever-expanding array of medical costs and treatment options [
6,
7]. The estimated direct total medical expenditure in the United States for spinal care in 2006 was more than $85 billion, and the data suggest that the use and costs of spine care have been increasing at an alarming rate in recent years [
7].
Despite the vast amount of research devoted to spinal diseases, the epidemiology of this condition is not well understood, and the overall prevalence is unclear in many countries, even in the United States [
7,
8]. The ROK has a National Health Insurance Service (NHIS), which is a compulsory social insurance system that insures approximately 97% of the population [
9]. The majority of remaining 3% of population is covered by Medical Aid [
9]. All hospitals and clinics in the ROK submit medical records including diagnosis and operation codes of patients covered by the NHIS or Medical Aid to a Health Insurance Review & Assessment Service (HIRA) for review to be reimbursed for any healthcare services provided [
9]. The HIRA database nationally representative and contains data on multiple variables important to epidemiological research, such as diagnosis, treatment, procedures performed, surgical history, and treatment prescriptions.
The goals of this study were to describe the incidence and analyze trends related to spinal diseases based on a nationwide database in the ROK and to elucidate the health care burden for researchers, clinicians, patients, and families. Additionally, we aimed to investigate the distribution of medical expenses based on sex, age, and diagnostic code associated with spinal diseases.
MATERIALS AND METHODS
This study involved a retrospective analysis of data obtained from Healthcare Bigdata Hub, the Korean Statistical Information Service (KOSIS), and Open Data Portal from 2012 through 2016. The Healthcare Bigdata Hub, KOSIS, and Open Data Portal are maintained by the HIRA, Statistics Korea, and Ministry of the Interior and Safety, respectively. These databases include records of all patients who receive care at all Korean hospitals and clinics, and medical costs based on disease codes, age, and sex. The database included the beneficiaries of the NHIS and did not included population of Medical Aid. The data were acquired by disease codes according to the Korean Standard Classification of Diseases, seventh edition [
10] as same as the International Classification of Diseases, Tenth Revision (ICD-10) [
11].
1. Inclusion and Exclusion Criteria
We identified patients with spinal diseases and classified them according to the following ICD-10 codes for spinal diseases: deforming dorsopathy (M40–43), spondylopathy (M45–49), and other dorsopathy (M50–54) (
Table 1). Data for patients treated with these disease codes as a main diagnosis were included in this study. Patients diagnosed with these disease code as an accessory diagnosis were not included. Patients with spinal diseases due to congenital disease or trauma were excluded. The entire population data of the ROK were imported from the KOSIS census data to calculate the proportions of patients treated with spinal diseases relative to the entire population.
2. Data Variables
The total medical expenditure of beneficiaries of the NHIS in the ROK comprises paid medical expenses by insurers, copayments, unpaid medical expenses by insurers, and pharmacy preparation costs from pharmaceutical prescriptions. Copayments are paid by beneficiaries in addition to payments made by insurers, and made up 5%–20% of the total costs of items covered by the NHIS. Unpaid medical expenses by insurers are costs of devices or services which are not covered by the NHIS, and are paid by the patients themselves. Medication costs are shared between the insurer and the beneficiary, with the beneficiary paying 30%–50% of the medication cost.
Medical expenditure was calculated in Korean won (KRW, ), and converted to United States dollars (USD, $) using the average exchange rate from 2012 to 2016. We tried to demonstrate the costs in both USD and KRW. Otherwise, the standard unit of cost in this study was $, and the raw data for KRW are described in
Supplementary Table 1.
3. Analysis
The statistical analysis was mainly descriptive to determine the annual trends in the number and cost of spinal diseases as well as the differences in age, sex, and disease code distributions. Chi-square tests were used to identify any differences in the distribution of numbers and costs by age, sex, and disease code. A p-value of <0.05 was regarded to be significant. Counts, means, rates, ratios, proportions, and other relevant statistics were calculated using R 3.4.3 (R Foundation for Statistical Computing, Vienna, Austria) and/or Microsoft Excel 2016 (Microsoft Corporation, Redmond, DC, USA).
DISCUSSION
This study provides a comprehensive overview of the incidence and trends in spinal diseases in the ROK at the national level based on integrated data from the HIRA. The nationwide incidence of spinal diseases and medical costs increased continuously over the study period. While the incidence rate of spinal diseases was the highest among people in their 50s who mainly experienced disc herniation, the age-adjusted incidence rate was the highest among those in their 70s who mainly experienced spinal stenosis. M48 including spinal stenosis, which was frequent among older age groups, consumed more medical expenditure per patient.
The number of patients diagnosed with spinal diseases increased by 7.6% over the 5-year study period. As the age of population increased, the incidence of spinal diseases also increased and the peak incidence rate stratified by age was 42.6% in the 75–79 years age group. The reason for the increase in the number of patients may be related to the increase in the elderly population in the ROK because the age adjusted incidence rate of the population in ≥65 years old did not increase from 42.3% in 2012 to 40.9% in 2016, and proportion of the age group increased from 11.5% through 12.7% continuously as Korean society is aging. The population over 65 years of age in the ROK rapidly increased and made up 14.3% of the total population in 2018 [
12]. The World Health Organization and the United Nations define an “aging society” as one in which more than 7% of the population in ≥65 years old, an “aged society” as a society in which more than 14% of the population is ≥65 years old, and a “super-aged society” as a society in which more than 21% of the population is ≥65 years old [
13]. The ROK has become an aged society in 2018.
The number of claims increased by 3.1% over 5 years, with a 7.6% increase in the number of patients. The number of claims among male patients increased by 27.9% and 7.0% in both inpatient and outpatient clinics, and that of female patients increased by 26.3% and 0.5%, respectively. The reason for the mild increase in the number of claims may be due to the slight increase in claims among female outpatients who had the highest proportion of claims. They may have been hospitalized (as inpatients) and undergone surgical treatments instead of utilizing only outpatient services. Besides, the trend that male patients tend to have fewer hospital visits than female patients may be related with the result. Over the 5-year study period, the average cost per inpatient decreased by 9.1%, whereas the cost per outpatient increased by 18.3%. The reason for the increase in medical expenses was mainly due to +outpatient services. The reason for the increment in medical cost for outpatients may be related to pain interventions [
14]. Further prospective studies are needed on this topic to find out this relationship.
Although the ICD-10 code is used worldwide, the code system cannot clearly distinguish specific diseases of the spine and some diseases have multiple codes. Moreover, M54 including dorsalgia, coccygodynia, radiculopathy, and sciatica was a symptom related code, and was usually used as a preliminary diagnosis before diagnostic workup. Thus, M54 was hard to regard as a primary disease code and excluded from this analysis. The largest medical expenditure was on thoracic and lumbosacral intervertebral disc disorders (M51) followed by dorsalgia (M54), and other spondylosis including spinal stenosis (M48). The medical expenditure on M48 increased rapidly during the study period, which may be due to the increased number of patients and medical cost per claim. The majority of spinal diseases differed by age and sex. Among individuals younger than 19 years, the most common spinal disease was scoliosis (M41), and might have been adolescent idiopathic scoliosis. In that age group, deforming dorsopathy (M40–43) had the highest medical expenses, which may be related to corrective surgeries for deformities due to adolescent idiopathic scoliosis. In patients younger than 39 years, the most common disease was lumbar and thoracic disc herniation (M51) and male patients accounted for the higher proportion. Disc herniation among young men may be related with hard work, vigorous physical activity, and obligatory military service. Among patients older than 40 years, females accounted for the higher proportion. Among patients at least 60 years of age, other spondylosis including spinal stenosis (M48) accounted for the highest incidence and the largest proportion of total medical costs. Spinal stenosis among elderly women may be related with degenerative change due to long kitchen work and farm work [
15].
There are limitations that need to be acknowledged and addressed in the present study. The first limitation is with regards to the use of ICD-10 codes. ICD-10 codes are assigned for inpatient hospital diagnoses at discharge, and represent a patient’s diagnostic and procedural events [
11]. However, studies utilizing ICD-10 codes in databases should be interpreted with caution, as their original purpose was not intended for research. Unfortunately, there is neither a specific term nor a specific diagnostic ICD-10 code to describe degenerative spinal diseases such as cervical spondylotic myelopathy, ossification of the posterior longitudinal ligament, and ossification of the ligamentum flavum [
16]. This has resulted in the inconsistent use of diagnostic terms when referring to patients with these diseases, and subsequently, has given rise to ambiguity in critically exploring diagnoses, interventions, and outcomes for this prevalent and disabling set of conditions. A previous study has questioned the accuracy of administrative data for capturing clinical diagnoses [
17].
The second limitation is the included population of this study. This study included only beneficiaries of NHIS, which is 97% of the total population in the ROK. The remaining 3% of the population was covered with the Medical Aids, and were applied totally different system from the beneficiaries of NHIS. Because their incidence and medical expenses were not disclosed to public, we analyzed the data from the beneficiaries of NHIS. Although this study did not include the entire population of the ROK, we think it would be more important to analyze the population of 97%, which is subject to common regulations by NHIS, for health policy formulation.
The third limitation is related to the accuracy of the medical expenditures to which the findings can be generalized. Among total medical costs comprising paid medical expenses by insurers, copayments, unpaid medical expenses by insurers, and pharmacy preparation costs, the unpaid expenses and pharmacy preparation costs were not included in the medical expenditure. Although these were a relatively small proportion of the total medical costs, estimation of the exact cost of medical expenditures need to be made with caution.
Finally, retrospective surveys obtain information directly from affected individuals but may be subject to recall bias. The claimsbased data used in our study may have enabled us to avoid this limitation; they are not dependent on individual reporting and only detect subjects whose physicians coded for back pain associated with a given episode of care [
7]. Prospective methods of data collection for spinal diseases are superior for determining incidence and associations between patient characteristics, medical costs, and specific diseases.