Obstructive sleep apnea (OSA) is a disease in which hypopnea or apnea is repeated for more than 10 seconds during sleep due to stenosis or obstruction of the upper airway and is accompanied by frequent arousal and decreased blood oxygen saturation [1]. Its prevalence in middle-aged men and women are considerably high, at 9.1% and 4.0%, respectively, and are gradually increasing with the rising obese population [1]. OSA causes symptoms such as fatigue, daytime sleepiness, and headache, and increases not only diseases such as coronary artery diseases, heart failure, stroke, depression, and diabetes, but also vehicle and workplace-related accidents, and ultimately increases mortality [1].

Treatment methods for OSA include body weight loss, lifestyle improvement, surgery, an mandibular advancement device, and positive airway pressure (PAP) [1]. PAP was developed for the first time by Dr. Sullivan of Australia in 1981 [1]. It opens the closed upper airway by injecting air into the nose using an air pump [1]. Although PAP is the most effective treatment for OSA, its use is limited because its adherence rate is low and the patient’s cost burden is high. However, its use rapidly increased as it was started to be covered by National Health Insurance in 2018 [2]. On contrary to the increase in the use of PAP, the proportion of surgical treatment is gradually decreasing. In the case of uvulopalatopharyngoplasty (UPPP) which is the most representative procedure, 3,958 cases were performed nationwide in 2013, but it has gradually decreased to 2,644 cases in 2021 [3]. Although complications such as severe pain and persistent foreign body sensation were also a problem with UPPP, the low treatment success rate was a bigger problem. According to a meta-analysis published in 1996 by Sher et al. [4], the treatment success rate was only about 40%. However, subsequent study reported that the success rate of UPPP was over 60%, and considering the low adherence rate and irregular use time of PAP, UPPP was a more effective treatment [5].

Recently, cost-effectiveness analysis has become a very important concern across all fields of medicine. If a new treatment method is a little more effective than the existing treatment but its cost is too high, it is socially unacceptable as the cost-effectiveness is consequentially low [6]. On the contrary, if the effect is much superior to the existing treatment method and the cost is increased only a little or if the effect is similar but the cost is reduced, the cost-effectiveness is increased [6]. Although PAP is certainly beneficial to an OSA patient as it could reduce complications [1], a study argued that its costeffectiveness is not satisfactory considering treatment cost [7]. While the treatment costs of UPPP could be measured, it was impossible to carry out cost-effectiveness analysis as there were no clear data related to the preventive effect of complications from UPPP. However, as data from the National Health Insurance were released for research purposes in Korea, it was possible to perform the cost-effectiveness analysis of UPPP [8,9]. Although PAP is mainly used for the treatment of OSA patients recently in Korea, UPPP may be a more cost-effective treatment method. If so, this study can help patients choose treatment options and influence the establishment of insurance policy.

The purpose of the present study is to perform a cost-effectiveness analysis of PAP and UPPP in South Korea.

Most cost-effectiveness analyses of PAP have been conducted in Western countries [17-20]. The results are very different for each study, but to enjoy one unit of quality-adjusted life year, which means one year of perfect health, it costs between KRW 4 million and KWR 60 million [6,7,17-20]. However, when analyses are conducted from a social point of view, the cost is significantly reduced [6,17-20], because PAP is a more cost-effective treatment when non-medical expenses and productivity-loss cost resulting from diseases or accidents are taken into account. Cost-effectiveness analyses are greatly influenced by the medical environment of the relevant country or society [6]. It is because medical service costs for the same treatment are different, and non-medical expenses and productivity-loss cost are also calculated differently. In Western countries, since the treatment costs of complications or accidents resulting from OSA were overwhelmingly higher than the PAP costs, the cost-effectiveness of PAP could have been evaluated to be high in general. On the other hand, in Korea, since medical expenses are very low, there may be no significant benefit in terms of cost-effectiveness because the treatment costs of complications or accidents are low and the treatment costs of PAP are relatively high [7].

According to our result, KRW 925,095 has been used for 10 years in the case of UPPP to reduce KRW 297,870, the treatment costs of complications and accidents. On the other hand, KRW 10,246,948 of diagnosis and treatment costs were used to reduce complications and accident treatment costs by KRW 562,683 won for 10 years when receiving PAP treatment. Therefore, UPPP is overwhelmingly more cost-effective than PAP. When the study of Català et al. [18] is applied to our study, in the case of PAP treatment, KRW 10,977,403 is required to be spent to enjoy one unit of quality-adjusted life year [7]. It is difficult to make a direct comparison because there is no data of quality-adjusted life year for UPPP. However, based on the reduction in the treatment costs of complications and accidents calculated in this study, as the difference in quality-adjusted life year between the Surgery group and the No Treatment group is estimated to be 0.53 times the difference between the PAP Treatment group and No Treatment group, KRW 1,219,006 is required to be spent to enjoy one unit of quality-adjusted life year in the case of UPPP [18]. UPPP is overwhelmingly more cost-effective than PAP treatment from a quality-adjusted life year perspective.

One of the most astonishing results of this study was that PAP or UPPP had very little effect on reducing the expected treatment costs of complications or accidents of OSA. It is because the incidence rate of complications or accidents in OSA patients is not high [10]. As shown in Table 1, in the No Treatment group of patients with moderate to severe OSA, the probability of having a stroke was 0.91% but, when using PAP, it decreased to 0.64%. In other words, it can be said that PAP treatment reduces the incidence of stroke by about 30%, which may be thought to be a great effect, but the absolute value of reduction is as small as 0.27%. In addition, the effect of OSA on the incidence of such a complication is also insignificant. The concept that reflects the contribution of a certain risk factor to the incidence of a specific disease is a population-attributable fraction [21], and the population-attributable fractions of OSA for complications are very small showing a value of 4.8% for coronary heart disease, 1.5% for heart failure, 4.8% for stroke, 3.6% for depression, 1.7% for diabetes, 3.8% for vehicle accidents, and 1.3% for work-related accidents [7,22]. Accordingly, even when OSA is actively treated, the benefit that can be obtained by reducing actual complications or accidents may be not very large. Although various other complications are known to occur, as their population-attributable fractions are lower than those of the representative complications included in this study, their effects on the reduction of the treatment costs are presumed to be not large. In addition, as the treatment costs are very low even when a complication or an accident occurs because the medical expenses in Korea are very low, the preventive effects of PAP or UPPP also decrease. When the expected treatment cost for 1 year is obtained by multiplying the probability of the incidence of complications or accidents and the average treatment cost, it is KRW 108,209 in the case of the PAP Treatment group and KRW 178,369 in the case of the No Treatment group, and the difference is only KRW 70,160. This means that PAP treatment only has the effect of reducing medical expenses by KRW 70,160 for a year. If medical expenses were 10 times higher than the current expenses, the savings would have increased 10 times to KRW 701,600.

Another noteworthy finding is that UPPP has an effect of reducing the incidence of complications [8,9]. Although it was presumed that UPPP would reduce the incidence of complications also in the past, direct evidence was not sufficient. This was because it was difficult to conduct long-term monitoring of the patients who had undergone UPPP. However, as the National Health Insurance data were released for research purposes in Korea, the long-term result of UPPP could be observed, and the preventive effect on some complications could be proven, at least indirectly [8,9]. Although it was difficult to directly compare it with PAP, it was confirmed that the treatment costs of complications and accidents were reduced from those of the No Treatment group. However, the reduction was smaller than that of the PAP Treatment group and, as this also corresponds to the result expected, various assumptions of the present study are thought to be appropriate. In addition, it was far more cost-effective than PAP, and this was because the surgical expense is very low although the effect of UPPP was not superior to PAP.

In a cost analysis, the viewpoint of the cost analysis is very important [6]. The present study analyzed costs from the insurer’s point of view, and only official medical expenses, meaning only the expenses that fall under insurance benefits and non-insurance benefits were taken into account. In the case of healthcare system viewpoint, non-medical expenses such as transportation expenses, time cost, and nursing expenses are included and the social viewpoint is the most comprehensive concept that includes even the productivity-loss cost [6,7]. However, since data were not sufficient in the case of the healthcare system viewpoint or social viewpoint, calculation could not be made in the present study. If the harmful effects of OSA are investigated from a wider point of view as in the cases of other countries, PAP treatment will be found to be more cost-effective than in the result presented by this researcher as the expenses required for the treatment of complications or accidents will increase and productivity loss will be included.

This cost-effectiveness analysis has limitations [7]. First, the analysis was conducted using a simple model, not the diagnosis and treatment records or billing data of actual OSA patients. However, it is difficult to take into account very diverse diagnosis and treatment realities and a reasonable model may help in more clearly understanding the real situation. Second, as there were no data related to the incidence probabilities of complications or accidents in Korea, foreign country data were used. Third, PAP treatment had a high dropout rate, which was not taken into account for simple analysis. Fourth, although it is required to take into account not only the expenses but also the quality of life, it could not be performed due to the lack of data. Fifth, as the effects of UPPP were analyzed under somewhat unreasonable assumptions for the items that did not have clear data, the result may be different from the real situation. In addition, as no sufficient matching was achieved for the two groups, the accurate incidence rate of complications may be different. Lastly, it should be also taken into account that the treatment effects of PAP and UPPP may vary depending on the patients group.

In conclusion, it is presumed that UPPP is far more cost-effective for the treatment of OSA patients in the Korean medical environment.