Hospitalization deficit of in‐ and outpatient cases with cardiovascular diseases and utilization of cardiological interventions during the COVID‐19 pandemic: Insights from the German‐wide helios hospital network

Abstract Background Treatment numbers of various cardiovascular diseases were reduced throughout the early phase of the ongoing COVID‐19 pandemic. Aim of this study was to (a) expand previous study periods to examine the long‐term course of hospital admission numbers, (b) provide data for in‐ and outpatient care pathways, and (c) illustrate changes of numbers of cardiovascular procedures. Methods and Results Administrative data of patients with ICD‐10‐encoded primary diagnoses of cardiovascular diseases (heart failure, cardiac arrhythmias, ischemic heart disease, valvular heart disease, hypertension, peripheral vascular disease) and in‐ or outpatient treatment between March, 13th 2020 and September, 10th 2020 were analyzed and compared with 2019 data. Numbers of cardiovascular procedures were calculated using OPS‐codes. The cumulative hospital admission deficit (CumAD) was computed as the difference between expected and observed admissions for every week in 2020. In total, 80 hospitals contributed 294 361 patient cases to the database without relevant differences in baseline characteristics between the studied periods. There was a CumAD of −10% to −16% at the end of the study interval in 2020 for all disease groups driven to varying degrees by both reductions of in‐ and outpatient case numbers. The number of performed interventions was significantly reduced for all examined procedures (catheter ablations: −10%; cardiac electronic device implantations: −7%; percutaneous cardiovascular interventions: −9%; cardiovascular surgery: −15%). Conclusions This study provides data on the long‐term development of cardiovascular patient care during the COVID‐19 pandemic demonstrating a significant CumAD for several cardiovascular diseases and a concomitant performance deficit of cardiovascular interventions.

for several cardiovascular diseases and a concomitant performance deficit of cardiovascular interventions.

| INTRODUCTION
During the course of the COVID-19 pandemic, reduced hospitalization rates were described for multiple acute cardiovascular and noncardiovascular diseases. [1][2][3][4][5][6][7][8][9][10][11][12][13][14] Since most of the previous investigations focused on the early phase of the pandemic in spring 2020, data concerning the development of hospitalization rates following April 2020 are scarce. [15][16][17][18] Evidence of an increased case-severity respective mortality, especially in patients with cardiovascular diseases, led to concern that those reduced treatment numbers could negatively affect patients' long-term outcome. 9,15,19 Therefore, understanding patients' care pathways during the ongoing pandemic is of huge interest. Our group already introduced the cumulative hospitalization deficit as a metric to monitor cardiovascular hospitalizations across a multicenter hospital network in Germany. 20 This study expands these data with regard to the examined study period, provides data also on outpatient treatment and highlights the numbers of different cardiological interventions.  Tables 1-3

| Statistical analysis
Administrative data were extracted from QlikView (QlikTech, Radnor, PA). We calculated the total number of monthly and weekly admissions, with weeks defined so that the first day of the national protection phase (Friday, March 13th 2020) corresponds to the beginning of the investigated week. The cumulative hospital admission deficit (CumAD) was computed as the difference between the expected and observed cumulative admission number for every week in 2020, expressed as a percentage (95% confidence interval [CI]) of the cumulative expected number, which is defined as the weekly average across the time interval in 2019. The difference between the expected and observed cumulative admission number was assessed using a χ 2 test for the last week of the period. The p-values were adjusted for multiple comparisons using a Bonferroni correction. For all tests we apply a two-tailed 5% error criterion for significance.

| Study cohort
A total of 80 hospitals contributed 294 361 patient cases to the database (6 hospitals did not treat cardiovascular patients meeting the inclusion criteria) for the years 2019 and 2020 (140 658 in total and 92 082 during the study period in 2020; 153 703 in total and 106 544 during the control period in 2019). There were no differences in the distribution of baseline characteristics between the control period in 2019 and each of the studied months in 2020 with respect to gender, age groups, CCI or the treatment in areas with different COVID-19 case volumes (Table 1).

| Hospital admissions
There was a significant CumAD of −10% to −16% at the end of the study period for all investigated cardiovascular disease groups. In   Table 4). The cumulative number of performed interventions was significantly reduced for all examined procedures with −10% for catheter ablations (95%CI: −12; −8; p<0.001), −7% for cardiac implantable electronic device (CIED) operations (95% CI: −9; −6; p<0.001), −9% for percutaneous cardiovascular interventions (95%CI: −10; −8; p<0.001) and even −15% for cardiovascular surgery (95%CI: −16; −14; p<0.001). There has been no distinction made between in-and outpatient cases. Comparing weekly performance rates for each procedure group, a declining number of interventions was apparent from early (CIED implants) or mid-February (catheter ablations, cardiovascular surgery, percutaneous coronary interventions) to early or mid-June with a slight overcompensation period from end-June to early August within catheter ablations and CIED implants without reaching statistical significance for this temporal increase (Figure 3).

| DISCUSSION
The COVID-19 pandemic already led to profound restructuring processes in the affected health care systems. We investigated an administrative database with 294 361 in-and outpatient cases for temporal trend analysis in healthcare utilization related to several cardiovascular diseases. Extending the findings of a previous report of our working group, we showed a significant reduction of case numbers for all inves-  cardiological societies. 28 Our data did not provide information on the interventional treatment within specific disease groups and could therefore not show whether patients admitted with coronary artery diseases were treated equally compared to the pre-pandemic phase.
However, data from a Swedish registry indicates a similar interventional strategy in patients with acute coronary syndromes leading to the assumption that the pure reduction of admission numbers is responsible for our observations. 29  areas with low and even pronounced in those with high COVID-19 case numbers. 34 Once again, this is most likely due to a combination of the avoidance of patients entering the healthcare system or canceling their appointments, respectively, and healthcare providers postponing elective procedures according to the official recommendations. [35][36][37][38] Although there is no obvious explanation for the less pronounced effects shown in our data compared with some of the above mentioned studies, regional differences and the different COVID-19 case numbers within the investigated cohorts must be taken into account in the interpretation.
Data regarding the development of admission rates following April 2020 are scarce and limited to two studies investigating acute coronary syndromes and one manuscript describing the changes in hospitalization rates for acute heart failure. 15,16,18 Those three investigations showed a recovery phase in the later observational period similar to the increasing case numbers in the corresponding time interval seen in our analysis. However, comparability is limited as only patient cases up to May or mid-June were included in previous works, which has now been extended to mid-September in the present analysis.

DATA AVAILABILITY STATEMENT
The data underlying this article will be shared on reasonable request to the corresponding author.