Changes in the Incidence of Cardiovascular Diseases during the COVID-19 Pandemic in Korea

Round 1

Reviewer 1 Report

The manuscript contain the descriptive study comparing the incidence of CVD before and during COVID-19 pandemic, there is a few improvement can be performed to improve the clarity for readers:

1. In the methods section: please clearly describe the data collection, how the authors collect data from NHI services, how to select which data enter into the analysis, is there any data omission and overlapping? is there any data cleaning ? how to ensure the validity of the raw data and data entry into the analysis?

2. In result section: please add the descriptive data of 6 CVD diseases in the NIH database, before dividing between before and during COVID-19.

3. In result section: please add the comparison of incidence of CVD based on the type of healthcare setting: primary clinic until tertiary hospital, and make an analysis based on the type of healthcare setting.   

Author Response

We sincerely thank the reviewer for the insightful and helpful comments on our manuscript and believe that these have helped greatly improve our manuscript. All of the authors have reviewed the manuscript and have attempted to address all of the reviewers’ comments. Please find our point-by-point responses to the reviewer comments presented below:

The manuscript contain the descriptive study comparing the incidence of CVD before and during COVID-19 pandemic, there is a few improvement can be performed to improve the clarity for readers:

1. In the methods section: please clearly describe the data collection, how the authors collect data from NHI services, how to select which data enter into the analysis, is there any data omission and overlapping? is there any data cleaning? how to ensure the validity of the raw data and data entry into the analysis?

We sincerely thank the reviewer for this valuable comment. According to the reviewer’s suggestion, we have further stated the specific information on NHI in the MATERIALS AND METHODS section in the revised manuscript as follows:

MATERIALS AND METHODS

Previous version: The entire Korean population is obligated to register with the National Health Insurance service. To collect data used for this study, we used records from the National Health Insurance service database which thus included all citizens of Korea (approximately 50 million people) without exception.

Revised version: The entire Korean population is obligated to register with the National Health Insurance service (NHI). To collect data used for this study, we used records from the National Health Insurance service database which thus included all citizens of Korea (approximately 50 million people) without exception. NHI contains all public and private information on patient demographics (age, sex at entry, income, and residency), medical use/transaction information, and claim database (diagnosis/prescriptions/consultation statements). Therefore, it can be used as a population-based, nationwide study for various diseases.

Previous version: Information on six CVDs, which are the most common in primary clinics, was extracted from the database. Each disease was identified using the diagnostic codes of the International Classification of Diseases-10 (ICD-10) as follows:

Revised version: Information on six CVDs, which are the most common in primary clinics, was extracted from the database to assess the monthly incidence of each CVD. Each disease was identified using the diagnostic codes of the International Classification of Diseases-10 (ICD-10) as follows:

2. In result section: please add the descriptive data of 6 CVD diseases in the NIH database, before dividing between before and during COVID-19.

We sincerely thank the reviewer for this observation. We have described the additional texts regarding the total numbers of every registered CVD in the RESULTS section in the revised manuscript as follows:

RESULTS

Revised version: From January 2018 to May 2021, cerebral hemorrhage and cerebral infarction were recorded in 1,327,131 and 6,795,513, respectively. 1,243,740 and 10,554,338 were diagnosed with myocardial infarction and ischemic heart disease. Also, 1,471,549 and 3,683,612 cases were registered with cardiac failure and arrhythmia.

3. In result section: please add the comparison of incidence of CVD based on the type of healthcare setting: primary clinic until tertiary hospital, and make an analysis based on the type of healthcare setting.

We sincerely thank the reviewer for the constructive and informative comments. As the reviewer has pointed out, it is vital to make an analysis based on the type of healthcare setting because it could influence the use of medical services. Unfortunately, we would apologize that we regret not being able to perform the additional analyses according to the healthcare setting due to current constraints on data access. Therefore, additional limitations of the study were included in the DISCUSSION section of the revised manuscript:

DISCUSSION

Revised version: Finally, we would not investigate the number of diagnoses based on the healthcare setting, such as whether CVD diagnoses were registered at a primary clinic or a tertiary hospital. Thus, we could not determine how the healthcare environment affected the medical visits for CVDs during the COVID-19 pandemic.

This manuscript has not been published or presented elsewhere in part or its entirety and is not under consideration by another journal. The study design was approved by the appropriate ethics review board, and informed consent was waived by the IRB, as this was a recording-based study with no patient contact. We have read and understood your journal’s policies, and we believe that neither the manuscript nor the study violates any of these. There are no conflicts of interest to declare. All the authors have approved the revised manuscript and agree with resubmission to your esteemed journal.

Reviewer 2 Report

In this manuscript, the authors report about a retrospective study on the difference in the incidence of several CVD before and during the COVID-19 pandemic in Korea, using a nationwide database. The authors allegedly found an heterogenous effect of COVID-19 on the incidence of several CVD during the outbreak.

The manuscript is interesting although not entirely  novel (previous studies have attempted at providing similar information). The findings are strengthened by the reliability of the underlying data (which are solid and nationwide). There are however several comments:

1) The numbers reported in tables are somewhat unclear. What "mean" and "standard deviation" stand for in this context? One should expect a measure of the incidence according to the research question. These numbers, which in the "statistics paragraph" are reported as "the mean number of diagnostic registrations before and during the COVID-19 pandemic" seems obscure and require further explainations (these are very difficult to interpret). Please report the incidence or explain more clearly what the numbers means.

- Regarding the codes of the diseases investigated, one should note that IHD and MI overlap for two defining codes (I21-I22) - there is a need for that? Perhaps the data regarding MI is more relevant. Also, while the codes are homogenous for most of the diseases, regarding "arrhythmia" the two codes actually refers to two very different disease (I48: AF and flutter; I49: other arrhythmia, which may comprise ventricular fibrillation etc.). I would suggest the authors to provide at least a sensitivity analysis using only the I48 code (this is important to compare these data with the studies that already studied the AF incidence during COVID-19).

- I appreciated the subgroups analysis stratified for sex and age - although an analysis using age as a continuous variable may have been more useful.

- I think that a more specific discussion of previous evidence is needed and should deal with three specific points: 1) how were these results potentially affected by the direct effect of COVID-19 (i.e., by the number of cases of CVD caused by COVID-19)? 2) Is it possible that these results were somewhat influenced by the burden of the pandemic in South Korea? In other words, is it possible that countries/regions with more cases may have experienced a surge in CVD which is not captured here? For example, a meta-analysis on the incidence of AF during COVID-19 (https://doi.org/10.3390/jcm10112490) showed how the incidence of AF is 8% during COVID-19 (and this may explain the slight increase observed here), but may be lower in Asian countries (as shown in meta-regression). These data should be discussed to give more context to the results presented. Finally, 3) there is a role for the underdiagnosis in explaining these findings? It is common knowledge that the burden of COVID-19 on the hospital systems have shifted resources from the prevention and diagnosis of the other treatments (including CVD) to the management of COVID-19 - is it possible that this have influenced the number of cases observed?

- I would suggest the authors, if feasible, to add an analysis regarding other specific types of CVD which get a lot of attention during COVID-19, and specifically myocarditis and pericarditis. These data would certainly increase the value and novelty of this study.

- A brief reflection on how vaccines (which were available for the last part of the observation period) may have influenced the results observed, reducing the risk of COVID-19-associated CVD, is perhaps useful.

Author Response

We sincerely thank you and the reviewers for your insightful and helpful comments on our manuscript and believe that these have helped greatly improve our manuscript. All of the authors have reviewed the manuscript and have attempted to address all of the reviewers’ comments. Please find our point-by-point responses to the reviewer comments presented below:

In this manuscript, the authors report about a retrospective study on the difference in the incidence of several CVD before and during the COVID-19 pandemic in Korea, using a nationwide database. The authors allegedly found an heterogenous effect of COVID-19 on the incidence of several CVD during the outbreak.

The manuscript is interesting although not entirely novel (previous studies have attempted at providing similar information). The findings are strengthened by the reliability of the underlying data (which are solid and nationwide). There are however several comments:

1) The numbers reported in tables are somewhat unclear. What "mean" and "standard deviation" stand for in this context? One should expect a measure of the incidence according to the research question. These numbers, which in the "statistics paragraph" are reported as "the mean number of diagnostic registrations before and during the COVID-19 pandemic" seems obscure and require further explanations (these are very difficult to interpret). Please report the incidence or explain more clearly what the numbers means.

We sincerely thank the reviewer for the constructive and informative comments. As the reviewer pointed out, the meaning of the mean number described in all tables had been hard to clearly understand. Therefore, we have modified the following paragraph in the MATERIALS AND METHODS and the titles of all tables in the RESULTS of the revised manuscript.

MATERIALS AND METHODS

Previous version: Information on six CVDs, which are the most common in primary clinics, was extracted from the database. Each disease was identified using the diagnostic codes of the International Classification of Diseases-10 (ICD-10) as follows:

Revised version: Information on six CVDs, which are the most common in primary clinics, was extracted from the database to estimate the monthly incidence of each CVD. Each disease was identified using the diagnostic codes of the International Classification of Diseases-10 (ICD-10) as follows:

RESULTS

Previous Table 1. Difference in the mean number and in the distribution of cardiovascular diseases before and during the COVID-19 pandemic.

Revised Table 1. Mean and standard deviation of the monthly incidence in six cardiovascular diseases before and during the COVID-19 pandemic.

RESULTS

Previous Table 2. Difference in the mean number and in the distribution of cardiovascular diseases before and during the COVID-19 pandemic stratified by sex in subgroup analysis.

Revised Table 2. Mean and standard deviation of the monthly incidence in six cardiovascular diseases before and during the COVID-19 pandemic stratified by sex in subgroup analysis.

RESULTS

Previous Table 3. Difference in the mean number and in the distribution of cardiovascular diseases before and during the COVID-19 pandemic stratified by sex in subgroup analysis.

Revised Table 3. Mean and standard deviation of the monthly incidence in six cardiovascular diseases before and during the COVID-19 pandemic stratified by age in subgroup analysis.

- Regarding the codes of the diseases investigated, one should note that IHD and MI overlap for two defining codes (I21-I22) - there is a need for that? Perhaps the data regarding MI is more relevant. Also, while the codes are homogenous for most of the diseases, regarding "arrhythmia" the two codes actually refers to two very different disease (I48: AF and flutter; I49: other arrhythmia, which may comprise ventricular fibrillation etc.). I would suggest the authors to provide at least a sensitivity analysis using only the I48 code (this is important to compare these data with the studies that already studied the AF incidence during COVID-19).

We sincerely appreciate your observation. We completely agree on the possibility that two diagnoses, ischemic heart disease, and myocardial infarction, coexist in one patient. What’s more, we realized that the disease properties of atrial fibrillation or flutter were highly different from other arrhythmias. However, we prioritized the common diseases in clinics when deciding on several cardiovascular diseases for this study. Accordingly, both ischemic heart disease and myocardial infarction were selected. Furthermore, we were unable to separately identify cases in which two or more different diagnostic names were registered in one patient because we counted the number of diagnostic registrations in the database. We would apologize that we cannot perform the sensitivity analysis the reviewer has pointed out due to the current restrictions on access to the database. Nonetheless, we agree with the reviewer’s comments, and we hope that further, larger investigations can be performed to provide firm evidence regarding this in the future.

- I appreciated the subgroups analysis stratified for sex and age - although an analysis using age as a continuous variable may have been more useful.

We completely agree with the reviewer’s comments. Unfortunately, we would apologize that we regret not being able to further subgroup analysis according to age as the continuous variable due to restrictions on data access.

- I think that a more specific discussion of previous evidence is needed and should deal with three specific points:

1) how were these results potentially affected by the direct effect of COVID-19 (i.e., by the number of cases of CVD caused by COVID-19)?

We sincerely appreciate this constructive advice. As the reviewer has pointed out, there is a limitation in that COVID-19 confirmed cases may be included in the cardiovascular disease group that we counted medical visits for this study. In particular, this assumption is highly probable because COVID-19 has been associated with multiple direct and indirect cardiovascular complications including acute myocardial injury, myocarditis, arrhythmias, and venous thromboembolism. However, in Korea, the first COVID-19 case was reported on January 20, 2020, and disease prevention and control began in earnest in Korea in March 2020. Up until May 9, 2020, the cumulative number of COVID-19 reported cases and deaths in Korea was 10,840 and 256, respectively. Therefore, it would be reasonable to regard the impact of coronavirus on the incidence of cardiovascular diseases during the investigation period of this study as indirect rather than direct.

2) Is it possible that these results were somewhat influenced by the burden of the pandemic in South Korea? In other words, is it possible that countries/regions with more cases may have experienced a surge in CVD which is not captured here? For example, a meta-analysis on the incidence of AF during COVID-19 (https://doi.org/10.3390/jcm10112490) showed how the incidence of AF is 8% during COVID-19 (and this may explain the slight increase observed here), but may be lower in Asian countries (as shown in meta-regression). These data should be discussed to give more context to the results presented.

We sincerely thank the reviewer for the constructive and informative comments. Since our study only investigated the trend of cardiovascular disease incidence regardless of the presence or absence of COVID-19 diagnosis, it is difficult to conclude the occurrence of each cardiovascular disease only in patients with COVID-19. Rather, during the period, Koreans were not seriously influenced by COVID-19, and the prevalence of COVID-19 was not that high, so it might be said that the incidence of cardiovascular disease in non-COVID-19 patients was identified.

Finally, 3) there is a role for the underdiagnosis in explaining these findings? It is common knowledge that the burden of COVID-19 on the hospital systems have shifted resources from the prevention and diagnosis of the other treatments (including CVD) to the management of COVID-19 - is it possible that this have influenced the number of cases observed?

We sincerely appreciate the reviewer’s comments. As the reviewer has indicated, the possibility that the incidence of cardiovascular diseases was diagnosed lower than the actual incidence during the period due to the burden of COVID-19 in the healthcare system cannot be excluded. However, as Korea once suffered from a lack of medical resources at the time of the outbreak of the Middle East respiratory syndrome (MERS), various strategies against the COVID-19 crisis seem to have been able to be carried out relatively quickly. For this reason, it is thought that the possibility of underdiagnosis is not high since Koreans were not heavily influenced by the COVID-19 outbreak. Nonetheless, we have described the following paragraph concerning the additional limitation in the DISCUSSION section of the revised manuscript:

DISCUSSION

Revised version: Third, the possibility that the incidence of CVDs was diagnosed lower than the actual incidence during the period cannot be completely excluded due to the prioritization of COVID-19 care and the resulting burden on the healthcare system.

- I would suggest the authors, if feasible, to add an analysis regarding other specific types of CVD which get a lot of attention during COVID-19, and specifically myocarditis and pericarditis. These data would certainly increase the value and novelty of this study.

We sincerely thank the reviewer for this valuable comment. Myocarditis and pericarditis are both imperative heart diseases. In particular, COVID-19 infection has been associated with multiple direct and indirect cardiovascular and heart complications including myocarditis. However, the purpose of this study focused on the changes in the incidence of not heart diseases but cardiovascular diseases related to the various influences of the COVID-19 pandemic. We would apologize for not being able to conduct the additional analysis for myocarditis and pericarditis because the data access currently was restricted. Even so, we have added the paragraph regarding the need to further analyses in the DISCUSSION section in the revised manuscript as follows:

DISCUSSION

Revised version: Moreover, further analyses for the COVID-19-related heart diseases, such as myocarditis and pericarditis are required to comprehend both direct and indirect impacts of COVID-19 on the various CVD as well.

- A brief reflection on how vaccines (which were available for the last part of the observation period) may have influenced the results observed, reducing the risk of COVID-19-associated CVD, is perhaps useful.

We sincerely appreciate the reviewer’s constructive advice. In Korea, COVID-19 vaccinations started on February 26, 2021, among health professionals who are directly involved in treating COVID-19 patients, employees, and residents in long-term care facilities, and first-line COVID-19 respondents [1]. Accordingly, the SARS-CoV2 vaccination rate in Korea was not high in May 2021, when our study was completed, so the effect of vaccination on the CVD incidence was not considered in this study.

[Reference 1]: Choi, WS; Cheong, HJ. COVID-19 vaccination for people with comorbidities. Infect. Chemother. 2021, 53, 155-158.

This manuscript has not been published or presented elsewhere in part or its entirety and is not under consideration by another journal. The study design was approved by the appropriate ethics review board, and informed consent was waived by the IRB, as this was a recording-based study with no patient contact. We have read and understood your journal’s policies, and we believe that neither the manuscript nor the study violates any of these. There are no conflicts of interest to declare. All the authors have approved the revised manuscript and agree with resubmission to your esteemed journal.

Round 2

Reviewer 2 Report

In this version of the manuscript, the authors have partly answered to previous comments. However, some of the issues raised seems to have not been taken into account appropriately.

Specifically:

- However, in Korea, the first COVID-19 case was reported on January 20, 2020, and disease prevention and control began in earnest in Korea in March 2020. Up until May 9, 2020, the cumulative number of COVID-19 reported cases and deaths in Korea was 10,840 and 256, respectively. Therefore, it would be reasonable to regard the impact of coronavirus on the incidence of cardiovascular diseases during the investigation period of this study as indirect rather than direct.

I understand the point of the authors - on the other side, their study span also till 2021, so one year after the outbreak of COVID-19. I think that a more thorough discussion of the aspects related to the impact of COVID-19 on the results (see previous review for further details) is still needed and it's not sufficient here.

- Regarding the ICD codes, it should be note that the issues raised in the previous round are severe limitations that should be at least mentioned appropriately in the manuscript. Also, I cannot understand what "We would apologize that we cannot perform the sensitivity analysis the reviewer has pointed out due to the current restrictions on access to the database." means. Please note that this is a very important point, since it is central in the study. This also apply for the other analyses suggested.

Author Response

We sincerely thank you and the reviewers for your insightful and helpful comments on our manuscript and believe that these have helped greatly improve our manuscript. All of the authors have reviewed the manuscript and have attempted to address all of the reviewers’ comments. Please find our point-by-point responses to each of the reviewers’ comments attached below:

The revision details are as follows:

Reviewer #2:

In this version of the manuscript, the authors have partly answered to previous comments. However, some of the issues raised seems to have not been taken into account appropriately.

Specifically:

- However, in Korea, the first COVID-19 case was reported on January 20, 2020, and disease prevention and control began in earnest in Korea in March 2020. Up until May 9, 2020, the cumulative number of COVID-19 reported cases and deaths in Korea was 10,840 and 256, respectively. Therefore, it would be reasonable to regard the impact of coronavirus on the incidence of cardiovascular diseases during the investigation period of this study as indirect rather than direct. I understand the point of the authors - on the other side, their study span also till 2021, so one year after the outbreak of COVID-19. I think that a more thorough discussion of the aspects related to the impact of COVID-19 on the results (see previous review for further details) is still needed and it's not sufficient here.

We sincerely appreciate the reviewer’s constructive advice. In addition, we apologize for not properly providing information on the cumulative number of COVID-19 reported cases in Korea in the study period. As the reviewer has pointed out, we conducted this study from January 2018 to May 2021. Therefore, the status of COVID-10 in Korea should have been investigated by May 2021. As a result, the cumulative number of COVID-19 reported cases in Korea was 140,337 Up until May 31, 2021.

The ultimate finding of this study was that the incidence of most cardiovascular diseases in Korea was maintained without meaningful changes before and after COVID-19. The reasons were discussed from various perspectives in the DISCUSSION section, but it was not possible separately to confirm the occurrence of cardiovascular complications in COVID-19 confirmed patients, thus this should be included in the limitation of this study. Therefore, we have modified the following paragraph of the DISCUSSION section in the revised manuscript:

DISCUSSION

Previous version: Second, the possibility that patients with CVDs included in this study had concomitant confirmed COVID-19 cannot be ruled out. This is important because cardiovascular comorbidities are common in patients with COVID-19 and COVID-19 patients can also show a variety of cardiac presentations [23], which can be a confounding factor in our results. However, as described above, the infection rate of COVID-19 was lower in Korea than that in other countries during the study period; thus, this possibility was probably not relatively high.

Revised version: Second, the possibility that patients with CVDs included in this study had concomitant confirmed COVID-19 cannot be ruled out. This is important because cardiovascular comorbidities are common in patients with COVID-19, and COVID-19 patients can also show a variety of cardiac presentations [23], which can be a confounding factor in our results. In fact, the cumulative number of COVID-19 identified cases in Korea was 140,337 by May 31, 2021, this possibility cannot be completely excluded. Subsequently, the effect of COVID-19 on the incidence of CVDs during the investigation period of this study is likely to have both direct and indirect.

- Regarding the ICD codes, it should be note that the issues raised in the previous round are severe limitations that should be at least mentioned appropriately in the manuscript. Also, I cannot understand what "We would apologize that we cannot perform the sensitivity analysis the reviewer has pointed out due to the current restrictions on access to the database." means. Please note that this is a very important point, since it is central in the study. This also apply for the other analyses suggested.

We sincerely thank the reviewer for the constructive and informative comments. As the reviewer has pointed out before, we also think that conducting a sensitivity analysis using only the I48 code is crucial for a better understanding our findings. However, the NHI database, the database we used for this study, is managed by the government, and if researchers want to use it for their research, they are given access to it for the indicated period after the data is applied for approved. Therefore, we could not conduct additional analysis due to time and access issues. Once again, we apologize for this.

This manuscript has not been published or presented elsewhere in part or its entirety and is not under consideration by another journal. The study design was approved by the appropriate ethics review board, and informed consent was waived by the IRB, as this was a recording-based study with no patient contact. We have read and understood your journal’s policies, and we believe that neither the manuscript nor the study violates any of these. There are no conflicts of interest to declare. All the authors have approved the revised manuscript and agree with resubmission to your esteemed journal.