INTRODUCTION
Obstructive sleep apnea (OSA) is a common sleep disorder characterized by recurrent episodes of partial or complete upper airway obstruction during sleep, resulting in intermittent hypoxia, sleep fragmentation, and daytime symptoms such as excessive sleepiness, fatigue, and impaired cognitive function [1,2]. OSA represents a major public health concern, with prevalence estimates ranging from 9.1% in men to 4.0% in women among middle-aged adults, and its incidence continues to increase with rising obesity rates worldwide [3,4]. Untreated OSA has been linked to multiple adverse health outcomes, including cardiovascular disease, metabolic disorders, and increased risk of motor vehicle accidents, underscoring the importance of timely diagnosis and treatment [5,6].
Polysomnography (PSG) remains the gold standard for diagnosing OSA, providing essential data on sleep architecture, respiratory events, and oxygen saturation [7]. However, accessibility has been limited, particularly when high costs made it unavailable to many patients [8]. Before June 2018 in South Korea, PSG was classified as a non-reimbursable diagnostic test, creating significant financial barriers. As a result, only patients with severe symptoms or strong clinical indications typically underwent PSG, leaving many mild to moderate cases undiagnosed and untreated [9,10].
In June 2018, South Korea’s national health insurance system introduced a pivotal reform, extending coverage to PSG and related treatments, including continuous positive airway pressure therapy [11,12]. This policy sought to improve access to OSA diagnosis and management, reduce socioeconomic disparities in healthcare access, and ultimately lessen the public health burden of untreated sleep disorders [13,14]. The immediate effect was a sharp rise in PSG prescriptions, reflecting increased diagnostic activity among previously underserved populations. Yet, the broader implications of this surge—particularly its influence on diagnostic outcomes and healthcare resource utilization—remain insufficiently understood. Prior research has highlighted possible consequences of expanding insurance coverage for diagnostic testing, including higher healthcare utilization and potential overdiagnosis among patients with low pre-test probabilities [15]. These trends may lead to unintended economic and social costs, including unnecessary treatments and pressure on healthcare resources. For PSG specifically, expanded insurance coverage may have increased the proportion of patients with normal or mild findings, raising questions about the policy’s efficiency and cost-effectiveness [13,14].
This study addressed these uncertainties by analyzing changes in PSG prescription patterns and diagnostic outcomes before and after the introduction of insurance coverage in South Korea. Using a multicenter retrospective design, we examined data from 17 institutions spanning the pre-coverage period (2015–2017) and post-coverage period (2019–2023). By comparing demographic characteristics, PSG utilization rates, and apnea-hypopnea index (AHI) distributions across these periods, we aim to clarify the impact of insurance coverage on OSA diagnosis and identify opportunities to optimize healthcare delivery.
The findings of this study hold significant implications for both policymakers and healthcare providers. By offering a comprehensive assessment of how insurance coverage shapes diagnostic practices and outcomes, this work contributes to the broader discussion on balancing accessibility with efficient resource allocation. Furthermore, the results may inform future policy adjustments to improve the effectiveness and equity of sleep disorder management in South Korea and in other healthcare systems facing similar challenges.
METHODS
This multicenter retrospective study was conducted under the Sleep Research Committee of the Korean Rhinologic Society. Invitations to participate were distributed to general and tertiary hospitals, and institutions that consented were included in the data collection process. Each participating center collected data from adult patients (aged ≥20 years) who first visited the outpatient clinic between January 1, 2015, and December 31, 2023, with a primary diagnosis of either OSA (G47.3) or snoring (R06.5). The collected variables included the date of first visit, age at presentation, sex, body mass index (BMI), eligibility for medical benefits, whether diagnostic PSG was performed, and, if performed, the corresponding AHI.
The primary aim of the study was to assess the impact of the 2018 policy reform that introduced insurance coverage for PSG. To achieve this, we compared data between two distinct periods: pre-coverage (2015–2017) and post-coverage (2019–2023). Key comparisons included the number of patient visits, PSG testing rates, and AHI results among patients who underwent PSG. AHI values were analyzed both as continuous variables and as categorical variables, with severity defined as follows: normal (<5), mild (5–14.9), moderate (15–29.9), and severe (≥30).
Prior to analysis, patients younger than 20 years were excluded. Extreme or implausible values in demographic or clinical variables were considered data entry errors and excluded by treating them as missing data. Continuous variables were compared using the Student t-test, while categorical variables were analyzed with the chi-square test. A p-value of <0.05 was considered statistically significant. Statistical analyses were conducted using SPSS software, version 23.0 (IBM Corp.).
Institutional Review Board (IRB) requirements were determined separately at each participating hospital. Where required, IRB approval was obtained in consultation with local investigators and documented accordingly. Sites that did not require formal review confirmed exemption in compliance with institutional policies.
RESULTS
Study population
A total of 20 hospitals participated in this multicenter study and submitted data. Three hospitals were excluded from the final analysis because they lacked pre-2018 data, which were essential for comparison before national insurance coverage for PSG was introduced. Thus, data from 17 hospitals were included in the final analysis. Of these, 11 hospitals were located in the Seoul metropolitan area (Seoul, Incheon, and Gyeonggi Province), and 12 were tertiary care centers as of 2023. The final study population consisted of 29,055 patients.
PSG utilization and AHI distributions
We compared patient characteristics between the pre-coverage period (2015–2017, n=7,800) and the post-coverage period (2018–2023, n=21,255). In the post-coverage period, the proportion of male patients decreased slightly (80.0% to 77.0%). The mean age increased from 45.8 to 49.8 years, and mean BMI rose modestly from 26.5 to 26.9 kg/m2. The proportion of patients receiving medical aid (Type 1 or 2) also increased from 2.6% to 3.2%. Following the implementation of insurance coverage, the proportion of patients undergoing PSG rose substantially, from 51.3% to 74.4% (p<0.001). The mean AHI also increased significantly, from 31.2±25.2 to 35.3±26.4 (p<0.001). When categorized by severity, the proportion of severe OSA cases increased from 42.8% to 48.8%, while the proportion of patients classified as normal (AHI<5) declined from 11.6% to 8.3% (Table 1).
These findings suggest that the policy change not only increased diagnostic utilization but also shifted the severity distribution, with a higher proportion of previously undiagnosed, high-risk cases being detected.
After insurance coverage began in 2018, the number of patients visiting clinics increased sharply in 2019. Although this number decreased slightly in subsequent years, it remained consistently higher than during the pre-coverage period. Likewise, the proportion of patients undergoing PSG increased significantly. Before coverage, fewer than 56% of patients received PSG, whereas from 2019 onward, the rate consistently exceeded 72%, a difference that was statistically significant (p<0.001) (Fig. 1 and Supplementary Table 1 in the online-only Data Supplement).
From 2015 to 2023, the proportion of patients with normal AHI decreased from 11.7% to 7.4%, while the proportion with severe OSA increased from 38.1% to 50.8%. This shift in severity distribution was statistically significant (p<0.001), indicating a trend toward diagnosing more severe OSA cases following the expansion of insurance coverage for PSG in 2018 (Fig. 2 and Supplementary Table 2 in the online-only Data Supplement).
DISCUSSION
Summary of key findings
This multicenter retrospective study provides a comprehensive analysis of changes in PSG utilization and diagnostic outcomes following the expansion of insurance coverage in South Korea. We observed a significant rise in PSG prescriptions, from 51.3% in the pre-coverage period (2015–2017) to 74.4% in the post-coverage period (2019–2023), reflecting improved access to diagnostic evaluation for suspected OSA.
Notably, the proportion of patients diagnosed with severe OSA increased from 42.8% to 48.8%, while the proportion of patients with normal AHI values declined from 11.6% to 8.3%. These trends suggest that the expansion of insurance coverage did not result in indiscriminate testing of low-risk individuals. Instead, it facilitated the identification of previously undiagnosed high-risk patients. This interpretation is further supported by the increase in mean AHI and the consistent rise in severe OSA detection across years following the policy change.
In addition, the demographic profile of patients undergoing PSG shifted during the study period. There was a modest but statistically significant increase in the proportion of female and older patients, indicating that the policy reform may have helped reduce disparities in access to sleep diagnostics. Taken together, these findings suggest that expanding insurance coverage enhanced diagnostic reach, particularly among populations with unmet clinical needs.
Comparison with previous studies
Our findings are consistent with prior research demonstrating increased healthcare utilization following insurance reforms that reduce financial barriers [16-18]. Unlike concerns about overdiagnosis, we found a shift toward more severe OSA diagnoses and fewer normal AHI results after coverage expansion. This contrasts with global evidence of diagnostic test overuse, where reviews have reported rates as high as 97.5% across different services [19]. The decline in low-risk findings in our cohort suggests appropriate targeting despite broader access.
In contrast to studies from the United States that reported greater low-value care following the Affordable Care Act [20,21], our findings indicate a more selective diagnostic pattern, likely influenced by Korea’s referral system and defined PSG coverage criteria. The 23.1% increase in PSG utilization is comparable to findings from Spain after the introduction of home polygraphy coverage [22]. However, our data uniquely demonstrate improved diagnostic yield. Given the high economic burden of untreated OSA [23], expanding insurance coverage for PSG may represent a cost-effective approach to identifying high-risk patients and preventing downstream complications.
Clinical implications
The demographic shifts observed after insurance coverage expansion provide important insights into unmet diagnostic needs. The increased proportion of female and older patients undergoing PSG suggests that these groups may have faced financial barriers that previously limited diagnostic evaluation. This is especially relevant since older adults and women are more likely to experience economic vulnerability, potentially restricting access to sleep testing. The rise in the proportion of patients receiving medical benefits (Medical Aid types 1 and 2), despite partial or full cost coverage even before 2018, further highlights that financial concerns were a major barrier to diagnostic access. These findings reinforce the importance of insurance coverage in reducing economic obstacles, particularly for socioeconomically disadvantaged groups.
Overall, PSG testing rates increased by approximately 23.1% after the policy change, consistent with the intended goal of improving diagnostic accessibility. Notably, contrary to concerns about moral hazard or overuse, the proportion of normal AHI cases decreased while the proportion of severe OSA cases rose. This pattern indicates that the policy facilitated access primarily for patients with high pretest probability who had previously remained undiagnosed due to cost constraints. Thus, the expansion of insurance coverage promoted more equitable and clinically appropriate use of PSG, helping to uncover high-risk patients who might otherwise have gone untreated.
Beyond cost-related improvements, our data also suggest that non-financial factors contributed to the observed patterns. The increased participation of medical aid recipients, despite minimal out-of-pocket burden before 2018, implies that improvements in patient awareness, referral pathways, and healthcare delivery may have played a role. Moreover, the demographic changes following coverage expansion reflect both reduced financial barriers and growing recognition of OSA risk in historically underdiagnosed populations, including women and older adults. These findings indicate that policy reforms can influence healthcare access through multiple mechanisms—including education, provider practices, and patient engagement—thereby extending the reach of diagnostic services to previously underserved groups.
Strengths and limitations of the study
This multicenter retrospective study offers several methodological strengths that enhance the validity and generalizability of its findings. By incorporating data from 17 institutions across South Korea, including both metropolitan and regional hospitals, it captures a wide spectrum of diagnostic environments. The large sample size of 29,055 adult patients provides robust statistical power, while the inclusion of both tertiary care centers and general hospitals ensures representativeness across different levels of healthcare delivery. In addition, the clearly defined comparison between pre-coverage (2015–2017) and post-coverage (2019–2023) periods, with the transitional year of 2018 excluded, strengthens causal inference by reducing confounding related to implementation variability [24]. The use of standardized diagnostic criteria such as AHI severity classification, consistent data collection protocols, and rigorous quality control measures—including exclusion of implausible values and appropriate handling of missing data—further supports the reliability of the results.
Nevertheless, several limitations should be considered when interpreting the findings. The retrospective design may introduce selection bias, as individuals who did not seek medical evaluation for sleep-related symptoms were not captured, potentially underestimating the population-level impact of insurance expansion. The absence of longitudinal follow-up also limits assessment of clinical outcomes, treatment adherence, and cost-effectiveness. Moreover, external influences such as heightened public awareness, evolving referral practices, or concurrent healthcare policy changes may have affected PSG utilization independently of the insurance reform. Finally, while demographic shifts such as increased representation of women and older adults were observed, these changes may reflect broader societal trends in health-seeking behavior rather than being attributable solely to improved financial access [25]. Future research should incorporate longitudinal follow-up, cost-effectiveness analysis, and qualitative approaches to more fully assess the broader impact of insurance coverage expansion.
Conclusion
The expansion of insurance coverage for PSG in South Korea significantly increased diagnostic testing and facilitated the detection of previously undiagnosed severe OSA. Rather than encouraging unnecessary testing, the policy improved access for high-risk and underserved populations, including women, older adults, and medical aid recipients. These findings underscore the value of reducing financial barriers to care and support continued efforts to ensure equitable access to sleep disorder diagnostics.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Supplement1
Print
