Patterns of care for glaucoma patients in Korea from 2002 to 2013 using the national health insurance service claims data.

Due to the rapidly increasing life-expectancy, the prevalence of glaucoma has increased steadily in recent years. We aimed to evaluate the patterns of care and primary treatment strategy patterns in Korea according to glaucoma subtypes to assess the quality of care for glaucoma patients.In this serial cross-sectional survey, the claims data from the Korean National Health Insurance Service was used to identify and group glaucoma patients into primary open-angle glaucoma (POAG), primary angle-closure glaucoma (PACG), other types of glaucoma, and ocular hypertension from 2002 to 2013. Records for outpatient visits, hospitalizations, drug prescriptions, admissions, and surgical interventions were used to determine the patterns of care and identify primary treatment strategies.Both the prevalence (0.11% in 2002 to 0.43% in 2013) and incidence rates (0.06% in 2003 to 0.11% in 2013) for glaucoma increased over time. The mean number of outpatient visits increased (4.9-6.0 visits per year), while the proportion of hospitalized patients (2.3-1.0% of patients) and duration of hospital stay (4.5-3.4 days among hospitalized patients) decreased between 2002 and 2013 for patients with POAG. The proportion of patients not being managed by medication or surgery decreased, with POAG and PACG patients receiving medications increasing from 70.9% and 59.2% in 2002 to 88.4% and 63.3% in 2013, respectively. Finally, while the proportion of trabeculectomy decreased (22.2% to 10.0% of surgical procedures in 2002 and 2013, respectively), more patients with PACG have received iridectomy (59.3% to 86.0% of surgical procedures in 2002 and 2013, respectively).Between 2002 and 2013, the pattern of care for both patients with POAG and PACG has shifted toward management by outpatient visits and intervention with anti-glaucoma medications in Korea.

Introduction

Glaucoma, a chronic eye disease resulting in optic nerve damage, leads to vision loss and blindness. Affecting more than 70 million people worldwide, glaucoma is the second leading cause of blindness in the world.[ Furthermore, glaucoma has previously been linked to a number of systemic illnesses affecting other organs such as the brain including Alzheimer disease.[ In Asia, the number of glaucoma patients is expected to reach 37 million by 2020, which would constitute approximately 47% of the world's population of glaucoma patients.[ Specifically in Korea, the prevalence of glaucoma patients increased from 0.79% to 1.05% from 2008 to 2013 among those aged 40 years or older, with an average annual increase rate of 9%.[ Due to increasing life-expectancy, the number of glaucoma patients is expected to increase further in Korea.

While a wide variety of medications and surgical methods aimed at treating glaucoma are available,[ relatively little is known about how actual care for glaucoma patients have changed over time, particularly according to glaucoma subtypes. Although studies investigating the prevalence of glaucoma[ and medication adherence[ in Korea have been conducted, there are limited studies on the patterns of care and primary treatment patterns of glaucoma according to each glaucoma subtype. Such studies are necessary to determine how management patterns have changed over time as well as to assess the quality of care of patients for a condition that is expected to continue to increase in prevalence. Furthermore, the results from such studies can be used to assist in clinical decision making for glaucoma patients as well as to determine the areas for improvement in a public health perspective.

We conducted a pattern of care study for glaucoma patients and newly diagnosed glaucoma patients using the Korean National Health Insurance Service (NHIS) database in this cross-sectional survey. Both prevalence and incidence rates of glaucoma were determined to assess the number of glaucoma patients from 2002 to 2013. Furthermore, the patterns of care and primary treatment strategies were determined for glaucoma patients according to glaucoma subtypes.

Methods

Study population

The study population was derived from the Korean National Health Insurance Service - National Sample Cohort (NHIS-NSC) database.[ Health insurance enrollment, provided by the Korean government became mandatory for all Korean citizens in 1989, resulting in an enrollment rate of 97%. Based on the claims data of the NHIS enrollees, the NHIS constructed a database using the proportional allocation method that contains information on sociodemographics, hospital outpatient department (OPD) visits, and admissions records. The NHIS database had been previously used multiple times for a wide variety of epidemiologic studies, and its validity has been described in detail elsewhere.[

In this serial cross-sectional survey, a total of approximately 900,000 participants, 803,791 in 2002 and 911,672 in 2013, were used to determine the prevalence of glaucoma for each year from 2002 to 2013. Furthermore, among glaucoma patients diagnosed for each year from 2003 to 2013, the proportion of those who were not diagnosed in previous years was used to determine the incidence rate of newly diagnosed glaucoma for each year. The Catholic University of Korea Institutional Review Board (IRB) approved with study (IRB number: VC16EISI0022) and waived the requirement for informed consent.

Key variables

Glaucoma patients were defined when a patient visited the ophthalmology outpatient or inpatient department at least twice for glaucoma according to the codes from the Tenth Revision of International Classification of Diseases (ICD-10) from the World Health Organization.[ Primary open-angle glaucoma (POAG, ICD-10 code: H40.1), primary angle-closure glaucoma (PACG, ICD-10 code: H40.2), and other types of glaucoma (ICD-10 codes: H40.3, H40.4, H40.5, H40.6) were included. We also determined ocular hypertension (ICD-10 code: H40.0). Newly diagnosed glaucoma patients were defined as those who visited the ophthalmology inpatient or OPD at least twice with a glaucoma ICD-10 code for each year among those who were not diagnosed with glaucoma in any of the previous years.

The average number of ophthalmologic department visits was calculated by determining the total number of ophthalmologic OPD visits per year for each glaucoma patient and determining the mean number of OPD visits for each glaucoma subtype during every year. The proportion of those hospitalized was determined by calculating the percentage of patients hospitalized in the ophthalmologic department for each year according to glaucoma subtype. Finally, the duration of hospital stay was determined by calculating the mean number of days hospitalized for hospitalized glaucoma patients in the ophthalmologic department every year.

Primary treatment strategy was divided into no treatment, medication, and surgery. Medication was defined as prescription of any of the following medications: prostaglandins, alpha agonists, beta blockers, carbonic anhydrase inhibitors, parasympathomimetics, and combination drugs, when a patient diagnosed with glaucoma did not undergo glaucoma-related surgical or laser procedures. Surgery was defined when a patient underwent iridectomy, trabeculectomy, glaucoma implant surgery, photocoagulation, trabeculotomy, ciliary body cryotherapy, nonpenetrating filtering surgery, sinusotomy, or other filtering surgery. Finally, no treatment was defined as a glaucoma patient who was not prescribed any glaucoma-related medications and did not undergo any glaucoma-related surgical procedures.

Statistical analysis

For each year, the trend of the crude and age-standardized prevalence of glaucoma patients, as well as the proportions for the descriptive characteristics of glaucoma patients including age (years, 0 to 19, 20 to 39, 40 to 59, 60 to 79, and 80 or older), sex, household income (lower half and upper half), and glaucoma subtypes were determined. Between 2003 and 2013, the linear trend of the crude and age-standardized incidence rates of newly diagnosed glaucoma patients, along with the proportions for the descriptive characteristics of newly diagnosed glaucoma patients, were determined. The age-standardized prevalence and incidence rates for glaucoma patients were determined using the age distribution of the study population in 2013. The proportions of glaucoma patients under each of the 4 primary treatment strategies were determined for each year according to each glaucoma subtype. The proportion of each surgical method for glaucoma patients who underwent surgery was calculated for each glaucoma subtype. Furthermore, the proportions of patients who were prescribed each prostaglandin or combination drug were determined according to glaucoma subtype. Finally, the proportions of each primary treatment strategy or household income according to subgroups of household income were determined for each glaucoma subtype.

All data collection and statistical analyses conducted in this study were done with SAS 9.4 (SAS Institute, Cary, NC) and STATA 13.0 (StataCorp LP, College Station, TX), respectively.

Results

Table 1 depicts the descriptive characteristics of glaucoma patients by year from 2002 to 2013. The prevalence of glaucoma patients increased from 0.11% to 0.43%. The proportion of glaucoma patients aged 80 or older increased from 4.3% to 10.9%. In 2013, 50.5% of glaucoma patients were female, while 62.6% of the patients were in the upper half of household income. Among glaucoma patients, the proportions of those with other types of glaucoma increased, while the proportion of patients with PACG decreased.

Table 1

Descriptive characteristics of glaucoma patients and newly diagnosed glaucoma patients by year.

The incidence of newly diagnosed glaucoma patients increased from 0.06% to 0.11%. The proportion of newly diagnosed patients aged 80 years or older increased from 3.7% to 9.6%. While the proportion of women was greater among those newly diagnosed with glaucoma in 2003, the proportion of women increased from 50.7% to 52.9% in 2013. While the proportion of newly diagnosed patients within the upper half of household income decreased, there were nonetheless more newly diagnosed glaucoma patients in the upper half of household income. While the proportion of other types of glaucoma increased, the proportion of PACG patients decreased.

The frequency of ophthalmologic department visits, proportion of hospitalizations, and average hospital stay duration for each glaucoma subtype are shown in Table 2. For both patients with POAG and PACG, the number of ophthalmologic OPD visits increased. For patients with POAG and PACG, the proportion of hospitalizations and the average duration of hospital stay decreased. A greater proportion of patients with PACG were hospitalized compared to that of patients with POAG.

Table 2

Frequency of ophthalmologic department outpatient follow-up visits and admissions by year according to glaucoma subtypes.

The proportions of each primary treatment strategy according the glaucoma subtypes are shown in Table 3. While 24.8% of patients with POAG were not treated and 70.9% managed with medications in 2002, 88.4% were managed with medications with only 9.3% of patients with POAG not treated in 2013. Furthermore, while 4.3% of patients with POAG underwent surgery in 2002, the proportion of those treated surgically decreased to 2.3% in 2013. For patients with PACG, the proportion of those not treated decreased. On the other hand, the proportions of those managed by medication increased. A greater proportion of patients with PACG were managed by surgery compared to that of patients with POAG.

Table 3

Proportion of each primary treatment strategy of glaucoma patients by year according to glaucoma subtypes.

Among those diagnosed with ocular hypertension, the proportion of those not treated decreased, while those managed by medication increased from 2002 to 2013. The proportion of ocular hypertension patients who underwent surgical procedures were minimal and remained largely unchanged. Among glaucoma patients diagnosed with other types of glaucoma, the proportion of those not treated decreased from 48.2% to 17.2%, while those managed with medications increased from 45.9% to 80.4%, respectively.

The proportions of surgical methods conducted among glaucoma patients who underwent glaucoma-related surgery for each glaucoma subtype between 2002 and 2013 are shown in Table 4. Trabeculectomy was the most common surgical method among patients with POAG. The second most common surgical method for patients with POAG, iridectomy, decreased in prevalence while glaucoma implant surgery also decreased in prevalence. The most common surgical method among patients with PACG, iridectomy, increased in prevalence while the prevalence of trabeculectomy decreased.

Table 4

Proportion of each surgical method by year according to glaucoma subtypes.

Table 5 shows the proportions of prostaglandins and combination drugs according to glaucoma subtypes from 2002 to 2013. For patients with POAG, the most frequently prescribed drug was latanoprost, with timolol and dorzolamide being the most common drug. Compared to 2002, patients with POAG were prescribed less latanprost and more travoprost, bimatoprost, tafluprost, and combination drugs in 2013. Among patients with PACG, latanprost was the most commonly prescribed drug in 2002 with 39.1%, while timolol and dorzolamide were the most common drug in 2013 with 52.0%.

Table 5

Proportion of each anti-glaucoma medication by year according to glaucoma subtypes.

The proportions of primary treatment strategy according to glaucoma subtypes within subgroups of household income are shown in Table 6. Among patients with POAG within the upper half of household income, the proportion of patients who were managed with medications increased from 68.3% to 87.4% between 2002 and 2013. Similarly, the proportion patients with POAG within the lower half of household income who were managed with medications increased from 72.0% to 89.0%. Among patients with PACG, the proportion of those managed by medication and surgery increased from 2002 to 2013 regardless of household income.

Table 6

Proportion of each surgical method by year according to glaucoma subtypes.

Table 7 depicts the proportions of household income among glaucoma patients by year according to glaucoma subtypes. In 2002, the proportions of patients with POAG and PACG within the lower half of household income were 28.7% and 32.2%, respectively. On the contrary, 71.3% and 67.8% of patients with POAG and PACG were within the upper half of household income in 2002. Similarly, the proportions of patients with POAG and PACG within the lower half of household income in 2013 were 37.5% and 38.6%, respectively. In contrast, 62.5% and 61.5% of patients with POAG and PACG were within the upper half of household income in 2013.

Table 7

Proportion of household income of glaucoma patients by year according to glaucoma subtypes.

Discussion

In this large-scale serial cross-sectional survey, we have determined the patterns of care and primary treatment strategies for glaucoma patients using a nationwide claims database. The NHIS-NSC database is an appropriate database to use for patterns of care studies and the results of our study is thus an accurate depiction of the practice patterns for glaucoma in Korea.

The increased prevalence of glaucoma is in accordance with previous studies that have reported a prevalence rate of 3.4%[ in rural areas of Korea among 671 participants aged 50 years or more evaluated by vision scores, slit lamp examination, angle width measurements, gonioscopy, and tonometry examinations. Two separate studies investigating the prevalence of glaucoma in a rural area of central Korea using ophthalmic examinations, optical coherence tomography, and scanning laser polarimetry reported prevalence rates of 3.5%[ for POAG and 4.5%[ for PACG among 1532 and 1426 participants, respectively. Furthermore, another study using the health claims data of 22 million Korean adults aged 40 years or older revealed a prevalence rate (2.8% in 2013) similar to our results, while also showing a 54% increase in glaucoma prevalence from 2008 to 2013.[ Among studies that predicted prevalence rates for glaucoma patients worldwide, Quigley and Broman estimated prevalence rates to be 1.96% worldwide and 1.18% in Southeast Asia among those aged 40 years or older.[ Furthermore, a systematic review of glaucoma prevalence revealed global prevalence rates of 3.54% in 2013 among those aged 40 years or older.[ Our results add to the findings of previous studies by lengthening the observation period to 12 years while also extending the study population to include those aged below 40 years.

In our study, the prevalence rates in 2013 for those aged older than 40 or 50 years are 0.80% and 1.62%, respectively (data not shown). The discrepancy between the prevalence rates from our study compared to those from previous studies conducted in Korea may be due to several reasons. First, due to the nature of the claims data, undiagnosed patients not examined by a physician could not be taken into account, possibly leading to underestimation of prevalence in our results. Particularly, previous studies have examined the entire study population for glaucoma regardless of symptoms, leading to the detection of previously undiagnosed patients and thus higher prevalence rates compared to those from our study.[ Second, we defined glaucoma patients as those who visited the ophthalmology department at least twice, which is a relatively strict definition compared to a previous study that defined glaucoma as those who visited the ophthalmology department at least once.[ The relatively strict criteria for glaucoma may have lead to an underestimation of the prevalence of glaucoma patients in our study. Nonetheless, the exact reasons for the discrepancy in prevalence rates between our study and previous reports are unclear and merit further investigation.

One possible explanation for the increase in prevalence is due to the increased life expectancy over the study period.[ This is partly supported by the fact that the proportion of glaucoma patients aged 60 years or older increased from 49.7% in 2002 to 65.7% in 2013. Furthermore, we have shown that the incidence rate of glaucoma also increased by 83.3% from 2003 to 2013, which carries further weight to the suggestion that improved life expectancy increases the number of people who are prone to the development of glaucoma, ultimately resulting in higher prevalence. Another possible contributing factor to the increased prevalence and incidence rates of glaucoma is the active promotion of awareness for glaucoma by ophthalmologic societies in Korea, thereby increasing awareness of the disease among the general population. However, the exact effect of increased awareness of glaucoma on the prevalence and incidence rates of glaucoma cannot be determined from our results and merit further research.

The majority of glaucoma patients were in the upper half of household income. This is in contrast with a previous study investigating the prevalence of glaucoma in Korea using the Korea National Health and Nutrition Examination Survey with a study population of 13,831 adults. In this study, patients with POAG were more likely to be in the lower half of household income compared to being in the upper half of household income (odds ratio 1.36, 95% confidence interval 1.12–1.65).[ Similarly, 62.5% of patients with POAG were within the upper half of household income in our study compared to 37.5% of patients with POAG within the lower half of household income in 2013. Furthermore, we have also shown that more PACG are within the upper half of household income (61.5% compared to 38.6% within lower half of household income). This tendency of more patients being within the upper half of household income may be explained by the fact that glaucoma lacks early symptoms,[ leading to low awareness of the disease,[ and is thus more likely to be diagnosed by coincidence upon routine ophthalmologic check-ups. Furthermore, diagnosis of glaucoma requires the use of diagnostic tools such as optical coherence tomography that are not usually covered by health insurance. Therefore, there may be a greater possibility of those with higher incomes being diagnosed with glaucoma compared to those with low incomes. Further research on the role of sociodemographic characteristics on the diagnostic rates of glaucoma are needed.

While the frequency of ophthalmologic outpatient visits increased for both patients with POAG and PACG, the number of hospitalizations decreased for patients with POAG but increased for patients with PACG. This suggests that while POAG is managed most commonly by medication, which would increase the need for hospital outpatient visits, PACG is managed by a combination of medication and surgery, leading to increased requirement of both outpatient visits and inpatient hospitalizations. This is supported by the fact that while more than 88% of patients with POAG were managed by medication in 2013, only 63.3% of patients with PACG were managed by medication, with 15.1% of patients with PACG having undergone surgery in 2013.

Advances in glaucoma medications over the past decade may have contributed to the increased frequency of ophthalmologic outpatient visits among glaucoma patients. For both patients with POAG and PACG, the average number of ophthalmologic department outpatient visits per year for each patient increased from 4.9 to 6.0 and 4.9 to 6.0 visits per year from 2002 to 2013, respectively. With the development of prostaglandin analogs, which have become first line drugs for glaucoma management,[ the control of intraocular pressure (IOP) has become more manageable compared to the previously used beta-blockers. Furthermore, the development of a number of combination drugs has made controlling IOP and the progression of glaucoma more manageable. Recently, with newly developed drugs targeting neuroprotection for glaucoma patients,[ the treatment of glaucoma has become more manageable by medication alone. Although this could lead to the reduction in the number of patients with POAG requiring surgery, such patients are more likely to be advanced cases, thus requiring closer follow-up and more frequent outpatient visits. This could have possibly contributed to the increased frequency of OPD visits while reducing the number and duration of admissions and hospital stay among patients with POAG. However, the exact reasons as to why the number of ophthalmologic department outpatient visits increased for glaucoma patients are unclear and merit further investigation.

Trabeculectomy was the primary surgical method of choice for patients with POAG undergoing surgery while iridectomy was the most common surgical procedure conducted for patients with PACG receiving surgery. The initial treatment strategy considered for patients with POAG is IOP-lowering topical medication. However, when medication alone is not sufficient in slowing the progress of POAG, IOP-lowering surgery may be considered. The fact that trabeculectomy is the most common surgical method of lowering IOP in patients with POAG appears to support the greatest proportion of trabeculectomy among patients with POAG undergoing surgery observed in our results.[ On the contrary, as PACG is caused by the obstruction of the angle by the apposition of the iris, the primary management of choice is laser iridotomy. However, when iridotomy is insufficient in lowering the IOP, surgical iridectomy is considered next, reflected in the highest proportion of iridectomy among patients with PACG undergoing surgery in our results.[

The most commonly prescribed prostaglandins and combination drugs were latanprost and timolol with dorzolamide, respectively. Although this remained true from 2002 to 2013, new drugs such as tafloprost, timolol with brimonidine, and timolol with brinzolamide have increased in prescription recently. This result is consistent with a previous study showing that among 1050 glaucoma patients in Korea, 457 (43.5%) patients were managed with prostaglandins.[ Furthermore, another study investigating the prescribing patterns of drugs among patients with POAG in India showed that timolol with brimonidine was the most common combination drug prescribed (30.8%).[ Previous studies, however, only observed management patterns over 1- to 2-year intervals. On the contrary, we have shown that while latanprost and timolol with dorzolamide remained the predominant prescribed drugs, new drugs such as tafloprost, timolol with brimonidine, and timolol with brinzolamide have recently increased in prescription frequency over a 12-year period.

Several limitations must be considered when interpreting the results of our study. First, as the claims database only accounts for diagnosed glaucoma patients, the presence of undiagnosed glaucoma patients may have resulted in the underestimation of prevalence and incidence of glaucoma. This may explain the slightly lower prevalence rate in our study compared to those in previous studies conducted in Korea.[ Second, as the claims data only collects records on medications or surgical methods that are insured, medications and surgical procedures that are not insured would have been neglected. This could potentially result in higher proportion of glaucoma patients who were not managed by medication or surgery, particularly in the earlier years of the observational period when insurance coverage was narrower. Third, as the data did not contain information on the outcome of glaucoma patients, we could not draw any conclusions as to how the patterns of care and primary treatment strategies differed in controlling the progress of glaucoma. Studies investigating how the change and difference in patterns of care affects the outcome of glaucoma patients are needed in the future. Finally, since this study was based on a large amount of routine data, future studies that use information from clinical charts are needed to further validate the findings from this study.

In conclusion, both prevalence and incidence rates for POAG and PACG have increased in Korea from 2002 to 2013. POAG is increasingly managed through OPD visits rather than hospital admissions. Such results may be useful in clinical decision making of patients with POAG or PACG when evaluating the primary treatment strategy. Furthermore, the increasing prevalence and incidence rates over time suggest glaucoma will become a more important disease to focus on in the future. Finally, the small but significant proportions of patients with POAG and PACG who are not management by medication or surgery highlights the importance of continued emphasis on managing glaucoma patients in a public health perspective. Future studies investigating the effect of quality of care on the outcome of glaucoma patients are needed.

Author contributions

Conceptualization: Sang Min Park, Donghyun Jee.

Data curation: Sang Min Park, Donghyun Jee.

Formal analysis: Seulggie Choi, Jin A Choi, Jin Woo Kwon, Donghyun Jee.

Funding acquisition: Donghyun Jee.

Investigation: Sang Min Park, Donghyun Jee.

Methodology: Seulggie Choi, Jin A Choi, Jin Woo Kwon, Sang Min Park, Donghyun Jee.

Resources: Donghyun Jee.

Supervision: Sang Min Park, Donghyun Jee.

Writing – original draft: Seulggie Choi, Donghyun Jee.

Writing – review & editing: Seulggie Choi, Jin A Choi, Jin Woo Kwon, Sang Min Park, Donghyun Jee.