This article has been
cited by other articles in ScienceCentral.
Abstract
Background
Pathologic examinations play an important role in medical services. Until recently, the overall status of pathologic examinations in Korea has not been identified. I conducted a nationwide survey of pathologic examination status using the insurance reimbursements (IRs) dataset from the Health Insurance Review and Assessment Service (HIRA). The aims of this study were to estimate current pathologic examination status in Korea and to provide information for future resource arrangement in the pathology area.
Methods
I asked HIRA to provide data on IR requests, including pathologic examinations from 2011 to 2015. Pathologic examination status was investigated according to the following categories: annual statistics, requesting department, type of medical institution, administrative district, and location at which pathologic examinations were performed.
Results
Histologic mapping, immunohistochemistry, and cervicovaginal examinations have increased in the last 5 years. Internal medicine, general surgery, obstetrics/gynecology, and urology were the most common medical departments requesting pathologic examinations. The majority of pathologic examinations were frequently performed in tertiary hospitals. About 60.3% of pathologic examinations were requested in medical institutions located in Seoul, Gyeonggi-do, and Busan. More than half of the biopsies and aspiration cytologic examinations were performed using outside services. The mean period between IR requests and 99 percentile IR request completion inspections was 6.2 months.
Conclusions
This survey was based on the HIRA dataset, which is one of the largest medical datasets in Korea. The trends of some pathologic examinations were reflected in the policies and needs for detailed diagnosis. The numbers and proportions of pathologic examinations were correlated with the population and medical institutions of the area, as well as patient preference. These data will be helpful for future resource arrangement in the pathology area.
Keywords: Insurance, Pathology, surgical, Cytological pathology, Reimbursement
Pathologic examinations play an important role in medical services. Until recently, overall pathologic examination status in Korea has not been investigated. Understanding the current pathologic examination status is important to establishing future plans for resource arrangement. Byeon and Kim [
1] reported the estimated pathologic examination status in Korea using the Health Insurance Review and Assessment Service (HIRA)–National Patient Sample (NPS). Because the HIRA-NPS data was statistically extracted (selection probability, 0.03), some amount of statistical error is inevitable [
2]. The current study is a follow-up study using the raw dataset from HIRA. The HIRA dataset did not contain any rejected inspections of insurance reimbursement (IR) requests or new diagnostic techniques that were not approved by various government organizations. However, this dataset is one of the largest medical datasets and it best reflects the current pathologic examination status in Korea. The aims of this study are to estimate the current pathologic examination status according to multiple parameters in Korea, to provide information for future resource arrangement in the pathology era, and to compare differences between NPS and the raw data.
MATERIALS AND METHODS
I requested HIRA to provide the IRs for various pathologic examinations (
Table 1) from 2011 to 2015 (data extraction was performed in May 2016). The HIRA provided the data after anonymizing the patient identification numbers and hospital identification numbers. Primary data processing was performed using the R Statistical Software (Foundation for Statistical Computing, Vienna, Austria) ver. 3.2.3 in a remote access system for HIRA. Secondary data processing was performed using the R Statistical Software ver. 3.3.1 using the t2.micro instance in the Amazon Elastic Compute Cloud (Amazon, Seattle, WA, USA). To evaluate the tendency of each pathologic examination number from 2011 to 2015, linear regression was applied. A p-value less than .05 was regarded as statistically significant.
I estimated the mean period between IR requests and completion of the IR inspections as follows. Since HIRA did not provide the exact date of completion for the IR inspections and provided their data on a monthly basisinstead, a table containing monthly pathologic examination status from January 2011 to December 2014 was formulated, as seen in
Table 1 (pathologic examinations in 2015 were excluded due to incomplete IR request inspections). Using this table as the reference table, I obtained sequentially cumulative data by adding up the monthly number, thereby marking the period pertaining to the 33, 66, and 99 percentile inspections of IR requests. Next, I made up the same table again as sequentially. I obtained the time more than 33, 66 and 99 percentile inspections of IR requests were done.
RESULTS
The abbreviations used in this study are as follows. AC, aspiration cytopathology; BFC, body fluid cytopathology; BONE, histopathologic examination for bone; CB, cell block; CERVIX, cervicovaginal cytopathology; EM, tissue electron microscopy; ENZYME, enzyme histochemistry; FS, emergency histopathologic examination during surgery; HER2 (FISH), HER2 gene fluorescence in situ hybridization; HER2 (SISH), HER2 gene silver in situ hybridization; HR, examination of hormone receptor in tissue; IF, tissue immunofluorescent microscopic examination; IHC, immunohisto(cyto)chemistry; L-, liquid based; LND, with lymph node dissection; LNDX, without lymph node dissection; MaligGro, resected specimen for malignant tumor requiring gross sectioning; MAPPING, histologic mapping of tumor; MSI, microsatellite instability; Non-maligGro, resected specimen requiring gross sectioning; NPB, number of paraffin blocks; OUTSIDE, outside slide interpretation; SPECIAL, special stain examinations.
A summary and the details for annual pathologic examination status are given in
Table 2 and
Supplementary Table S1, respectively. Among the main categories, it was found that the total numbers of MAPPING, IF, EM, IHC, CERVIX, and BFC increased during 2011–2015, while the total numbers of ENZYME and CB were decreased. In more detail, the following pathologic examinations increased in 2011–2015: non-maligGro with NPB ≤6, MAPPING LND, MAPPING LNDX, Masson’s trichrome stain, IF (IgG, IgA, IgM, C3, C4, fibrinogen, others), EM, IHC (interpretation by qualified doctor, HR, EGFR pharmDx kit), L-CERVIX, L-BFC, L-AC, CB after L-AC, and some of OUTSIDE (C5500009, C5912009, C5916009, C5917009). The following pathologic examinations decreased in 2011-2015: biopsy (more than 13 pieces), IF (IgE and hepatitis B surface antigen), ENZYME (acetylcholinesterase), CERVIX by smear, BFC (general), AC (conventional), and CB (AC). Note that no chloroacetate esterase examinations were performed during the last 5 years.
Most medical and dental departments requested various pathologic examinations (a summary is given in
Tables 3 and
4 and the details are given in
Supplementary Table S2), but the proportions were quite different. Internal medicine (13,917,799, 46.29%), general surgery (6,334,913, 21.07%), obstetrics/gynecology (3,444,796, 11.46%), and urology (1,769,651, 5.89%) were the most common medical departments (25,467,159, 84.70%) requesting pathologic examinations. These proportions were similar to those of the previous study [
1].
A summary and the details of the pathologic examination status according to type of medical institution are listed in
Table 5 and
Supplementary Table S3, respectively. The numbers of each type of medical institution were not included in this analysis. The majority of pathologic examinations were frequently performed in a tertiary hospital. Among the different pathologic examinations, biopsy and AC were frequently performed in clinics.
A summary and the details of the pathologic examination status according to the administrative district are listed in
Tables 6,
7, and
Supplementary Table S4, respectively. About 60.3% of pathologic examinations (20,787,770) were requested in medical institutions located in Seoul (12,249,590, 35.5%), Gyeonggi-do (5,517,990, 16.0%), and Busan (3,020,190, 8.8%).
A summary and the details of the pathologic examination status according to the location at which examinations were performed are listed in
Table 8 and
Supplementary Table S5, respectively. More than 90% of the resected specimens for malignant tumors requiring gross sectioning (MaligGro) LND, MaligGro LNDX, MAPPING, emergency histopathologic examination during surgery (FS), ENZYME, IHC, CB, and OUTSIDE were performed in their own hospitals. More than half of the biopsies and AC examinations were performed using outside services.
The mean periods between IR requests and the 33, 66, and 99 percentiles IR request inspection completions were 2.2, 2.8, and 6.2 months, respectively. The kernel density estimation plot for each percentile can be found in
Fig. 1. The 95 and 99 percentile values of the period for 99 percentile IR request inspection completion were 11 and 25 months, respectively.
DISCUSSION
In this study, I investigated various pathologic examination status according to the following categories: annual statistics, requesting department, type of medical institution, administrative district, and location at which pathologic examinations were performed in Korea, 2011–2015. In the last 5 years, the total numbers of MAPPING and IHC increased. These trends reflect the need for a more accurate pathologic diagnosis. Contrary to how the number of conventional cytopathologic examinations decreased, the number of variable liquid-based cytopathologic examinations increased. After FISH and SISH based HER2 gene examinations were approved by insurance companies in 2013, the number of HER2 gene examinations increased. These trends reflect both insurance and public health policies.
Throughout many medical institutions in Korea, pathologists will order a large amount of pathologic examinations for diagnosis. However, only 0.04% of pathologic examinations have been claimed by pathologic departments. These discrepancies come from the difference between actual and administrative claims. In tertiary and general hospitals, various medical examinations were requested based on the inpatient department. In clinics, medical examinations were usually requested based on the major disease code of the patient.
According to HIRA (
http://opendata.hira.or.kr/op/opc/olapMdclRcStatsInfo.do) and the Ministry of the Interior (
http://rcps.egov.go.kr:8081/jsp/stat/ppl_stat_jf.jsp), there were 53,252 medical institutions and 51,677,054 people in Korea in December, 2015 (see
Supplementary Table S6). The total number of pathologic examinations in each administrative district was positively correlated with both the number of medical institutions and people in the area (both p<.001). About 61.7% of OUTSIDE were performed in medical institutions located in Seoul. The proportion of OUTSIDE was higher than the population ratio of Seoul (10,022,181, 19.4%). This phenomenon reflects the patient preference for major medical institutions in Seoul.
Compared to previous studies using common pathologic examination codes in 2013 (approximately 5,440,288 pathologic examinations using the NPS data), there was a 3.2% increase in pathologic examinations (5,615,395 pathologic examinations in 2013). This difference was found to be in the acceptable range as determined by the pilot study. When using the raw HIRA data, implementation of basic algorithms using the NPS data is recommended.
In conclusion, even though this survey using the HIRA dataset did not reflect the exact current status, it is still quite accurate. I expect the present study to help with future operations for the Korean Society of Pathology in terms of understanding the current status and trends of pathologic examinations. I recommended that an in-depth analysis of the status of pathologic examinations considering the period between IR requests and inspection completions be made after at least 1 year.