Psychosis, including schizophrenia and similar psychotic disorders, is representative of severe mental illness and is not uncommon. It places a huge economic burden on both patients and their families. According to a study from 2005, the overall cost of schizophrenia was estimated to exceed 3,000 million USD in the South Korea [1]. Specifically, unemployment was identified as the largest component of the overall cost, while the direct healthcare cost was estimated to exceed 400 million USD.

To properly handle the issue of severe mental illness, it is necessary to accurately understand the situation. Many studies investigating the epidemiology of schizophrenia have been conducted worldwide and have reported a lifetime prevalence of 0.3-0.7% [2-4], with an estimated 23.6 million people being affected globally [5]. In Korea, a series of nation-representing psychiatric epidemiologic studies (Korean Epidemiologic Catchment Area Study) were also performed at 5-year intervals. In a third survey in 2011, the lifetime prevalence of psychotic disorders and schizophrenia were 0.6% and 0.1%, whereas the 12-month prevalence were 0.4% and 0.1%, respectively [6]. The follow up study in 2016 reported a 0.5% lifetime prevalence and 0.2% 12-month prevalence of schizophrenia spectrum disorders (including schizophrenia, schizophreniform disorder, schizoaffective disorder, delusional disorder and brief psychotic disorder) [7]. These community studies have limitation that while they are well designed, their sample size of 5,102-6,022 were not sufficient for accurate estimation of rarer disorders like schizophrenia.

Therefore, we analyzed the South Korean database of Health Insurance Review and Assessment Service (HIRA), which contain the health utilization data of all Korean nationals to estimate the incidence and prevalence of schizophrenia. Two previous studies have reported the prevalence of schizophrenia using this Korean national database; however, they comprised either 1 year of the database1 or 2 years, with a focus on first onset psychosis [8].

Throughout the 5 years of observation, 598,951 patients were diagnosed with schizophrenia or similar psychotic disorders in South Korea. Our investigation is the first to take into consideration multiple years of the HIRA database containing data of almost all Korean nationals to calculate both the prevalence and the incidence of SSP in South Korea. However, it should be noted that the prevalence and incidence mentioned here are from “treated” cases, meaning that the “true” prevalence or incidence must be higher as there would be some cases not under treatment. However, both the 12-month prevalence and the IR are much beyond the expectations of the common clinical wisdom and of the estimates from community studies. Therefore, we suppose the “true” incidence and prevalence rates not to be much different from the numbers reported in this study.

The prevalence of schizophrenia or similar psychotic disorders found in our study was in line with a previous study using the HIRA database from year 2005 [1], or similar studies from other nations [4,3,11]. However, this rate is much higher than that reported in the national-representing KECA series, which reported only 0.1-0.2% 12 month prevalence of schizophrenia [6,7,12,13]. The low prevalence seen in the KECA series might be due to the characteristics of the community samples. Psychosis-like symptoms tend to be underreported, in part due to the stigma associated with the disorder and in part due to the high specificity but low sensitivity of community-surveying instruments of the Composite International Diagnostic Interview (CIDI) [14]. Moreover, considering that schizophrenia is a relatively rare disorder, the sample size was too small to estimate accurate prevalence in CIDI studies. Indeed, with a few exceptions [15], substantial CIDI studies including samples from US, Europe and worldwide [16-18] did not report the prevalence of psychosis in peer-reviewed journals. In another study, Cho et al. [19] investigated the prevalence rates of schizophrenia in the combined populations hospitals, psychiatric nursing facilities and homeless asylums and reported the lifetime and one year prevalences of schizophrenia and other psychotic disorders as 1.19% to 1.32% and 0.51% to 0.61%, respectively.

With regard to the incidence of psychotic disorders, a recent study conducted in 2017 showed the number of newly diagnosed psychotic patients in the Republic of Korea in 2006 and 2007 [8], which can be translated into 29.8 per 100,000 PY, a much lower rate than the numbers found in our study. Since the above-mentioned study used the same HIRA dataset, at different time periods, we can assume that the incidence of treated schizophrenia or similar disorders, if not the true incidence, is increasing in Korea.

Several explanations related to the observed increase in incidence are possible. First, the constant increase of psychiatric beds, and somewhat lessened stigma to the psychiatric treatment in Korea in recent years might explain the increase [20], since this estimate is of treated cases and not true incidence in itself. If this is the case, this result showing higher treatment incidence rate in older age groups than in younger age groups might indicate severe treatment delays for schizophrenia in Korea.

However, this increase might represent real changes in the true incidence rate of psychosis. Some studies from other countries suggested use of illegal drugs and possible cohort effects as reason of increased incidence [20,21]. In this study, the higher incidence found in the older age group might indicate cohort effects. The generation was born before 1950 and has spent their young age either during World War II or the Korean War, experiencing rapid transitions towards an industrialized economy during their working years, and unprecedented rapid societal changes during their whole adulthood. Therefore, these socio-economic factors may have left them more vulnerable to major psychotic disorders, although this hypothesis needs to be explored further to fully determine what the risk had been.

It is also of note that, although several studies from the 1990s and early 2000s described a decreasing incidence of schizophrenia [3,22-24], more recent studies tend to be reporting increasing incidences [20,21,25]. Nevertheless, the incidence rate of schizophrenia in this study is very high compared to that reported for other countries. There were similar studies based on national/universal health insurance data from Taiwan and Quebec, and these studies reported incidence of schizophrenia as 45-94/100,000 PY [3,26]. Study using GP database of Netherland indicated incidence of 22/100,000 PY over 10 year-period, and a study using the Danish Psychiatric Central Registry reported the incidence of schizophrenia to be 31-40/100,000 PY [25,27]. For a meta-analysis result, a study from England showed a pooled incidence of 31.7/100,000 PY for all psychosis and of 15.2 for schizophrenia [28], and another meta-analysis reported a 23.7/100,000 PY [11].

This high incidence in Korea may be in part due to the methodological limitations of this study. Because the dataset used is from 2010, the cases identified as new cases in 2011 or 2012 only had either 1 or 2 years of clearance ahead of them; therefore, they may include relapse cases. Indeed, the study from Quebec showed that compared to the 10-year clearance period, more than 6 years of clearance is required to achieve an excellent Kappa coefficient value [26]. However, the incidence rate of year 2015 would be reasonable estimate, since data from 2015 contains 5 years of clearance ahead of them, and especially estimate for the younger age groups (15-24 years old) are likely to be accurate. There might be also diagnostic validity issues resulting from the diagnostic leniency required to use medications off-label in South Korea, especially in the older age group, since dataset was from claims data of health insurance. However, we only included cases with primary diagnosis of schizophrenia, it would be less likely. Nevertheless, diagnostic validity for this study would be improved if additional information like prescribed medications, specialty of physicians who made relevant diagnosis and frequency of outpatient visits was available.

Regardless of these limitations, the present study is one of the first to report a reliable incidence rate of schizophrenia in South Korea and to draw attention to the urge of service requirements to address mental health in public. We found the incidence of schizophrenia or similar disorders to be very high, with the possibility of prevailed delayed treatment until the older age. Exploring the possible cohort effects that render the older people more vulnerable and adapting a more rigorous early detection (for younger adults) and treatment programs for psychosis are of extreme importance nation-wide.