The South Korean healthcare system is governed by the Ministry of Health and Welfare and covers approximately 97% of the overall Korean population. The claims data of Health Insurance Review and Assessment Service (HIRA) contains comprehensive information such as patients’ diagnosis, procedures, surgeries, treatment, and prescription history. Recently, the Ministry of Health and Welfare and HIRA initiated a global research collaboration “OpenData4Covid19” project (https://hira-covid19.net/). This project provides, via their website, insurance claims data for those who tested for SARS-CoV-2 infection, along with all the claim records up to 3 years before the individual’s date of SARS-CoV-2 test. As of May 27, 2020, data up to May 15, 2020 was released, and Korea Centers for Disease Control & Prevention (KCDC) data was merged to HIRA claims data with respect to SARS-CoV-2 confirmation and mortality.

The dataset included data of 234,427 subjects who had been tested for SARS-CoV-2 infection. Among 215,184 patients older than 19 years, 64,243 patients with hypertension were included in the study. All SARS-CoV-2 infection cases were confirmed via reverse transcriptase polymerase chain reaction. This study was approved by the Institutional Review Board of Seoul National University Bundang Hospital (IRB No. X-2005-613-902) and conformed to the tenets of the Declaration of Helsinki. The requirement for informed consent was waived off as all study participants were anonymized.

SARS-CoV-2 positive cases were predefined in the provided dataset, based on both COVID-19 case data from the KCDC. Claims data for the prescription of ACE inhibitors or ARBs were screened. Subjects were confirmed to be on ACE inhibitors or ARBs if their prescription had covered more than 4 weeks of an 8-week duration before the SARS-CoV-2 test (medication possession ratio > 50%). Prescription data on other classes of antihypertensive medication (beta-blocker, calcium-channel blocker, and diuretics) were also screened in the same manner.

Baseline comorbidities were also defined using ICD-10 codes as follows: hypertension, I10–13, I15; heart failure, I50; history of myocardial infarction, I21, 22; and diabetes mellitus, E11–14. Chronic obstructive pulmonary disease (COPD) and asthma were defined using a combination of ICD-10 codes (J43.1–9, J44 for COPD; J45, J46 for asthma) and prescription history of relevant drugs at least twice in 2019. The detailed claim codes used for the definitions are provided as a supplementary file.

This study has two main outcomes: [1] positive SARS-CoV-2 test among subjects that had been tested, and [2] mortality among SARS-CoV-2 (+) patients. The secondary outcome measure was severe clinical course among SARS-CoV-2 (+) patients, defined as a composite of requiring oxygen therapy, admission to the intensive care unit (ICU), and mortality. A composite of ICU admission and mortality were also evaluated. The flowchart of the scheme of the study is presented in Fig. 1.

Pearson’s chi-square test or Fisher’s exact test were used for categorical variables, and the Student’s t test was used for continuous variables. Outcomes from logistic regression were presented as odds ratios (ORs) with 95% confidence intervals (CIs). Comparison between RAS blocker user and non-user were performed in patients with hypertension. To address potential sources of bias, propensity score matching was additionally performed. The baseline clinical characteristics were incorporated into a non-parsimonious logistic regression model to compute the propensity score for the use of RAS blocker. The included covariates were age, sex, diabetes mellitus, heart failure, history of myocardial infarction, asthma or COPD, and use of other classes of antihypertensive medications. Then, propensity score-matching iteration from the eighth digit to the third digit in the SARS-CoV-2 tested population, and to the first digit in the SARS-CoV-2 (+) population were performed. In total, 20,178 matched pairs regarding the use of RAS blocker were selected from the overall SARS-CoV-2 tested population (c-statistics = 0.671) and 535 pairs from those that tested positive for SARS-CoV-2 (c-statistics = 0.666). The baseline characteristics and outcomes were compared again between the patients who were not on RAS blockers (control group) and the patients on RAS blockers (RAS blocker group) at the time of SARS-CoV-2 testing. All statistical tests were performed using SAS version 9.4 (SAS Institute, Cary, NC, USA), and p < 0.05 was considered statistically significant.

In total, 38,231 (59.5%) patients with hypertension were on ACE inhibitors or ARBs at the time of testing. The comparison of baseline characteristics between the RAS blocker user and non-user in the overall hypertensive and matched population are presented in Table 1. In the overall hypertensive population, the RAS blocker group patients were younger than those of the control group. Male sex and diabetes mellitus were more prevalent, and the proportion of heart failure and COPD/asthma were lower in the RAS blocker group. Most patients (96.3%) in the RAS blocker group used ARBs. After matching by propensity score, these parameters became comparable between the groups.

The ORs for positive SARS-CoV-2 tests are presented in Table 2. The positive rate of SARS-CoV-2 tests was significantly higher in the RAS blockers group than in the control group before (OR, 1.25; 95% CI, 1.13 to 1.39; p < 0.001) and after adjustment for age, sex, diabetes, heart failure, history of myocardial infarction, airway diseases (COPD or asthma), and other classes of antihypertensive medications (OR, 1.22; 95% CI, 1.10 to 1.36; p < 0.001). Other than RAS blocker usage, younger age and female sex were associated with higher SARS-CoV-2 positivity rates in the multivariate analysis. Diabetes mellitus, cardiovascular and airway diseases were correlated with lower SARS-CoV-2 positive tests. Analysis in the propensity score-matched population also showed significantly higher SARS-CoV-2 positive rate in the RAS blocker group than that in the control group (OR, 1.16; 95% CI, 1.03 to 1.32; p = 0.017).

In total, 7,083 subjects tested positive for SARS-CoV-2 infection, and among them, 1,609 patients had history of hypertension. The clinical outcomes were compared according to the use of RAS blockers in the SARS-CoV-2 (+) patients with hypertension. The baseline characteristics of the overall SARS-CoV-2 (+) with hypertension and the matched population are presented in Table 3.

The overall mortality rate in this dataset was 3.2% (227/7,083). Among patients with hypertension, the mortality rate was 10.3% (165/1,609), and the predictors for mortality are shown in Table 4. Use of beta-blocker, older age, diabetes mellitus, cardiovascular disease, and airway diseases were significantly associated with a higher mortality in the univariate logistic regression analysis. In multivariate analysis, age (per 10 years, adjusted OR, 3.11; 95% CI, 2.56 to 3.79; p < 0.001), diabetes mellitus (adjusted OR, 2.24; 95% CI, 1.56 to 3.22; p < 0.011), and airway diseases (adjusted OR, 1.95; 95% CI, 1.21 to 3.13; p = 0.006) were significantly associated with higher mortality. Patients on RAS blockers showed lower mortality in the univariate analysis (OR, 0.59; 95% CI, 0.43 to 0.82; p = 0.001), but there was no significant difference after adjustments for comorbidities and other classes of medications (adjusted OR, 0.81; 95% CI, 0.56 to 1.17; p = 0.265) or matching by propensity score (OR, 0.88; 95% CI, 0.59 to 1.33; p = 0.545). With respect to a composite of ICU admission and mortality, there was no significant difference between the RAS blocker and the control groups after adjustment or propensity score matching (adjusted OR, 0.95; 95% CI, 0.73 to 1.22; p = 0.669) (adjusted OR, 1.05; 95% CI, 0.79 to 1.40; p = 0.720). Analyses on severe clinical course (a composite of oxygen therapy, ICU admission, and mortality) showed similar trends (Fig. 2).

Baseline characteristics after matching by propensity score for each class of antihypertensive medications in the SARS-CoV-2 tested and the SARS-CoV-2 (+) population with hypertension are provided as Supplementary Tables 1 and 2. Other classes of antihypertensive medications were not associated with the rate of SARS-CoV-2 positive tests (Supplementary Table 3). Clinical outcomes of SARS-CoV-2 infection were also not associated with the use of other classes of antihypertensive medications after adjustments for comorbidities and other classes of medications or matching by propensity score (Supplementary Table 4).

ACE2 is the functional receptor for SARS-CoV-2 cell entry [2], and although the results are inconsistent, select preclinical animal studies showed higher ACE2 expression after the use of RAS blockers [12–14]. In contrast, data from human studies showed that RAS blockers have little to no effect on urinary or plasma ACE2 levels [15]. It is also uncertain if changes in urinary or plasma ACE2 levels reflect changes in tissue ACE2. Despite insufficient evidence, there are still concerns about the use of RAS blockers, particularly that they may alter ACE2 and play a role in COVID-19. In this study, the use of RAS blockers was associated with a significantly higher rate of SARS-CoV-2 infection after adjustment or matching by propensity score, indicating that more caution may be needed while managing patients on RAS blockers.

In contrast, a recent studies from Italy and US reported that the use of RAS blockers did not affect the risk of developing COVID-19 [9,16]. The systematic review from these studies also showed similar outcomes [17]. The disparity in results may be explained by the differences in the study population. The Korean government has conducted extensive testing and tracing measures, including the use of advanced information technology [18] to control the spread of the virus. This enabled the identification of a large group of individuals who are suspected to have been exposed to, but not infected with, SARS-CoV-2 as the control group. In contrast, in the previously mentioned Italian study [16], the subjects not tested for SARS-CoV-2 infection were designated as the control group and those infected with SARS-CoV-2 may have been unintentionally included. The Italian Ministry of Health had issued stringent testing policies, and testing was limited to those with mild symptoms or did not require hospitalization [19]. Tests for SARS-CoV-2 were also limited in the US and contact tracing was not widely available [20]. In the study by Reynolds et al. [9], 46.8% of the patients tested for COVID-19 were positive. In contrast, the positive rate was 3.3% in our study population, suggesting that this dataset is well designed to assess the risk of SARS-CoV-2 infection. It was suggested in the systematic review that previous studies were mostly from symptomatic patients and therefore these results may not apply to patients with mild or no symptoms [17].

Another study from Korea reported that use of RAS blockers was not associated with the increased risk of COVID-19 [21]. The previous study included individuals without hypertension and those that were not tested for SARS-CoV-2 infection. Moreover, the underlying comorbidities of the population were not assessed and there were no adjustments for comorbidities or other antihypertensive medications.

Other than the use of RAS blockers, age, sex, diabetes, cardiovascular and airway diseases were also associated with higher risk of SARS-CoV-2 infection. Comorbidities suggested as risk factors for severe COVID-19 were associated with lower SARS-CoV-2 infection rate in our population. There is no clear biological explanation for this. One possible explanation is that changes in behavioral pattern may have contributed to these results. Subjects at higher risk for severe outcomes might had been more cautious than healthy subjects with respect to personal protection and hygiene during this pandemic. A recent report on the association between risk perception of COVID-19 and self-protective behaviors may support this explanation [22].

In addition to the risk of infection, the use of RAS blockers was also suggested to affect the clinical course of SARS-CoV-2 infection [4,5], and some media sources even recommended discontinuation of RAS blockers in patients with known or suspected COVID-19. Despite these theoretical concerns and uncertainties, there is evidence of potential benefit, rather than harm, of RAS blockers in the treatment of COVID-19. In preclinical studies, RAS blockade attenuated SARS-CoV-1-induced acute lung injury [23] and other viral-induced lung injuries [24]. Previous studies have suggested that the use of RAS blockers did not have an adverse effect on the clinical course of COVID-19 [8–10,16]. There were even suggestions that RAS blockers were associated with a lower mortality rate [7]. This study also showed that the pre-existing use of RAS blockers did not increase the risk of mortality, admission to the ICU, or oxygen therapy. These findings support the recommendations from relevant professional societies that advocated for the continued use of ACE inhibitors/ARBs unless clinically indicated [25].

It is noteworthy that the general trend of SARS-CoV-2 data in South Korea is different from those in China, Italy, Spain, or United States. The proportion of positive test results and case fatality rate were relatively lower, and the overall population was younger [19]. South Korea had experienced an exponential spread of SARS-CoV-2 infection earlier than most countries. By late February 2020, the number of COVID-19 infections was the second highest after China. Using the extensive testing and tracing measures, the spread of virus was controlled without overwhelming the capacity of the healthcare system, which may have translated into the relatively lower fatality rate than that in other countries. Therefore, our data may represent the effects of the RAS blockers on the clinical course of COVID-19 in the setting of adequate healthcare management.

This study had some limitations. This is an observational study, thus there may be additional unknown confounders or bias affecting our results despite propensity score matching. Since most of the study population was Asian, it is not known if the results could be applied to population of other ethnicities. Also, the effects of RAS blockers on SARS-CoV-2 infection may be different across the strains of COVID-19. The risk of infection may be influenced by the preventive measures taken by each individual. As this study was based on claims data, additional beneficiaries’ information such as severity of conditions, laboratory results, and smoking status was not available. In addition, the use of RAS blockers was determined by prescription and may not coincide with the actual drug use. Dosage or adherence to drugs were not evaluated in this study. Further, less than 5% of subjects in the RAS blocker group were prescribed ACE inhibitors, and thus it was not feasible to assess the effect of ACE inhibitors and ARBs separately.

In this Korean nationwide claims dataset, the use of RAS blockers was associated with a higher risk of SARS-CoV-2 infection but not with higher mortality or other severe clinical courses. Possible precautions should be taken in patients on RAS blockers, in order to reduce their exposure to SARS-CoV-2, and help control the spread of COVID-19.