Impact of Nationwide COVID-19 Lockdowns on the Implantation Rate of Cardiac Implantable Electronic Devices

Aim The COVID-19 pandemic resulted in a significant decrease in the number of hospital admissions for severe emergent cardiovascular diseases during lockdowns worldwide. This study aimed to determine the impact of both the first and the second Danish nationwide lockdown on the implantation rate of cardiac implantable electronic devices (CIEDs). Methods We retrospectively analysed the number of CIED implantations performed in Denmark and stratified them into 3-week intervals. Results The total number of de novo CIED implantations decreased during the first lockdown by 15.5% and during the second by 5.1%. Comparing each 3-week interval using rate ratios, a significant decrease in the daily rates of the total number of de novo and replacement CIEDs (0.82, 95% CI [0.70, 0.96]), de novo CIEDs only (0.82, 95% CI [0.69, 0.98]), and non-acute pacemaker implantations (0.80, 95% CI [0.63, 0.99]) was observed during the first interval of the first lockdown. During the second lockdown (third interval), a significant decrease was seen in the daily rates of de novo CIEDs (0.73, 95% CI [0.55, 0.97]), and of pacemakers in total during both the second (0.78, 95% CI [0.62, 0.97]) and the third (0.60, 95% CI [0.42, 0.85]) intervals. Additionally, the daily rates of acute pacemaker implantation decreased during the second interval (0.47, 95% CI [0.27, 0.79]) and of non-acute implantation during the third interval (0.57, 95% CI [0.38, 0.84]). A significant increase was observed in the number of replacement procedures during the first interval of the second lockdown (1.70, 95% CI [1.04, 2.85]). Conclusions Our study found only modest changes in CIED implantations in Denmark during two national lockdowns.


Introduction
The COVID-19 pandemic has put most health care systems worldwide under enormous pressure. A number of societal lockdowns have been put in place since the start of 2020. On 13 March 2020, Denmark was one of the first European countries to impose a nationwide lockdown, which meant, for example, the reprioritisation and relocation of health care workers, the postponement of scheduled (elective) treatments, the closing of cultural and educational institutions, and non-essential government employees being required to work from home. This reprioritisation and postponement of treatment was controversial, and may have resulted in a lower risk of COVID-19-related death, but a higher chance of cardiovascular death during the pandemic [1].
Studies have shown that, in Italy, which was severely affected by COVID-19, a significant decrease in the number of hospital admissions for severe emergent cardiovascular diseases during the pandemic was observed [2]. Other studies have shown that this was also the case in Denmark, even though it has been less severely affected by COVID-19 than Italy [3][4][5]. In addition, countries including Greece, Peru, England, and Germany experienced a reduced number of implantations of cardiac implantable electronic devices (CIEDs) during the pandemic [6][7][8][9].
The purpose of this study was to investigate if there was a significant reduction in the number of CIED implantations during both COVID-19 lockdowns in Denmark, and whether non-acute surgery was cancelled or postponed, as recommended by health authorities.

Study Population
The data presented in this study were collected from the Danish National Patient Registry (DNPR), which is a population-based administrative registry. The registry collects data on diagnoses, treatments, and examinations related to all in-hospital and outpatient clinic visits, as well as administrative data. DNPR data are updated continuously [10]. The registry is linked to the Danish Civil Registration System (CRS) on an individual level. The CRS contains data on, among other things, the name, sex, and date of birth of all people living in Denmark. Furthermore, indications for device implantation were collected from the Danish Pacemaker and Implantable Cardioverter Defibrillator (ICD) register. The register is an official clinical quality database and records details of all cardiac device implants and explants in Denmark [11]. The Danish health care system is funded by taxes and is available to the Danish population at no additional cost. The present study included all men and women aged 18 years with either an implantable cardioverter defibrillator (ICD; NOMESCO-code: BFCB00-02, BFCB04), pacemaker (NOMESCO codes: BFCA01, BFCA03, and BFCA07), cardiac resynchronisation therapy (CRT; NOMESCO codes: BFCB03 and BFCA04-06), or implantable loop recorder (ILR; NOMESCO code: BFCA50) implanted, in any Danish hospital. Pacemakers and ICDs were not further differentiated on subtype. In a secondary analysis, all replacement procedures across all three device types were included (NOMESCO codes: BFCA1, BFCA10-12, and BFCB5, BFCB50-53).
Our study complied with the principles of the Declaration of Helsinki.

Stratifications
The data on de novo implantations were stratified in two ways: by implantation type and indication; and by implementation type only. The indications were categorised into two groups: non-acute and acute. For pacemakers, acute indications were defined as third-degree atrioventricular (AV) block, or syncope 2 weeks before implantation, and non-acute indications as all other indications in relation to pacemaker implantation. For ICDs, acute indications were defined as cardiac arrest or ventricular arrhythmia with no diagnosis of ischaemic heart disease, myocardial infarction, or bypass 6 months prior to implantation. Non-acute ICDs were defined as all other implantation-related indications. Both hospitalised and ambulatory patients were included. In the case of one person having several indications, the most recent was used. If a person had more than one indication on the same date, the indications were ranked and used as follows: (1) ventricular arrhythmia/cardiac arrest; (2) thirddegree AV block; and (3) all other indications. The numbers of CRTs were too few to stratify in this way. The replacement procedures and ILRs were not stratified further.

Approval
Register-based studies do not need ethical approval in Denmark. Furthermore, this study was approved by the Capital Region of Denmark (approval P-2019-191) in accordance with General Data Protection Regulation.

Study Period
The Danish government imposed the first nationwide lockdown on 13 March 2020 and began lifting restrictions on 15 April 2020. The first study periods were then defined as weeks 11-16 of 2019 and 2020. The second lockdown was imposed on 17 December 2020 and lasted until 28 February 2021. Our data covered up to 29 January 2021. Therefore, the second study periods were defined as weeks 51 of 2019 and 2020 up to and including week 5 of 2020 and week 4 of 2021.
As our second study period spanned 1 January, we analysed the data using a custom year, which started on 29 January. In addition, each lockdown was further divided into 3-week intervals. Consequently, our first study periods were defined as the third and fourth 3-week interval, and our second periods as the 16th, 17th, and 18th 3-week interval.

Analysis
We retrospectively analysed the number of implantations during weeks 11-16 of 2020 to data from the same weeks in 2019 and weeks 51-04 of 2020/21, to data from weeks 51-05 Impact of the Nationwide COVID-19 Lockdowns on the Rate of CIED Implantations of 2019/20. We compared the total number of de novo implantations (ICDs, pacemakers, CRT, and ILRs), shown as a graphical illustration of the calculated daily rates. We also compared the number of implantations of ICDs, pacemakers, CRTs, and ILRs. This was done by calculating the rate ratios between the years for each device by using an exact Poisson test, using 2019 as the reference for the first study period, and 2019 spanning into 2020 for the second. The exact Poisson test performed an exact test of our null hypothesis that the ratio between two rate parameters was 1.0. The rate ratios were shown graphically as forest plots. Furthermore, the number of pacemaker and ICD implantations were stratified by indication. The indications were divided into two groups: non-acute and acute. The proportions between these numbers were shown graphically. Stratification by indication was not possible for CRTs and ILRs, as the numbers would be too few to maintain the patients' anonymity in this study, as was the case for rate ratios of the number of ICDs implanted with an acute indication during the second and third 3-week interval. For replacement procedures both daily rates and rate ratios were calculated and were shown in the same way as de novo implantations.

Results
The total number of de novo CIED implantations decreased during the first lockdown by 15.5%, from 524 in 2019 to 443 in 2020. Divided by device type there was a change of -16.  (Table 1). Regarding pacemakers, 310 were implanted with a non-acute indication and 70 with an acute indication in 2019. In 2020, 256 were implanted with a nonacute indication (-17.4% vs 2019), while 60 were implanted with an acute indication (-14.3%; Table 1). Regarding ICDs, 66 were implanted with a non-acute indication and 17 with an acute indication in 2019. In 2020, 49 were implanted with a non-acute indication (-25.8% vs 2019), while 19 were implanted with an acute indication (111.8%; Table 1).
During the second lockdown, the total number of de novo CIED implantations decreased by 5.1% from 623 in 2019/20 to 591 in 2020/21. Divided by device type, there was a change of  Table 2). In 2019/20, regarding pacemakers, 343 were implanted with a non-acute indication and 107 with an acute indication. In 2020/ 21, 325 were implanted with a non-acute indication (-5.2%), while 85 were implanted with an acute indication (-20.6%; Table 2). The ICDs implanted in 2019/20 were too few to be further stratified. In 2020/21, 71 were implanted with a nonacute indication, while 24 were implanted with an acute indication (Table 2). Daily rates of the total number of de novo implantations of all three types of CIED appeared to be lower during both lockdowns compared to the year before ( Figure 1). Table 1 Numbers and daily implantation rates of cardiac implantable electronic devices in the first COVID-19 lockdown.   The first six rows show the number of implantations done in the first, second, and third 3-week intervals for both 2019 and 2020/21 in total, stratified by de novo or replacement, de novo by device type, and pacemakers and implantable cardioverter defibrillators (ICDs) by non-acute or acute indications. The next two rows show the corresponding numbers in total for the reference period and during lockdown, respectively. The final six rows show the corresponding daily rates again divided between the first, second, and third 3-week intervals. 2020 was the year of the second lockdown, although it spanned into 2021. a The number of acute ICD implantations could have been calculated from the total and non-acute ICDs and would have been too few, and have therefore been left out for 2019. Abbreviations: CRT, cardiac resynchronisation therapy; PM, pacemaker; ILR, implantable loop recorder; ICD, implantable cardioverter defibrillator.
In order to show whether the differences were statistically significant, we used rate ratios of the daily rates during the first and second lockdown, respectively, using the year before as a reference.
During the first lockdown both the total number of de novo and replacement procedures, as well as the number of de novo procedures, decreased significantly during the first 3week interval (Figure 2). In relation, the number of pacemakers implanted with a non-acute indication decreased significantly during the first 3-week interval (Figure 3).
During the second lockdown, the daily rates of de novo implantations across all device types decreased significantly during the third 3-week interval, while the daily rates of replacement procedures increased significantly during the first 3-week interval ( Figure 2). Additionally, the total number of pacemaker implantations decreased during both the second and third 3-week intervals (Figure 2). In relation, the daily rates of acute pacemaker implantations decreased significantly during the second 3-week interval, whereas the daily rates of non-acute pacemaker implantations decreased during the third 3-week interval (Figure 3).
In the time between the two lockdowns all other daily rates were comparable to the year before, except that there was a higher rate of ILRs during 3-week interval 14 and a lower rate of non-acute ICDs during interval 15 (Supplementary Table 1).

Discussion
The present study is the first study to investigate the consequences of the two Danish nationwide lockdowns during the COVID-19 pandemic, imposed on 13 March and 17 December 2020, respectively, on the number of CIED implantations. The total number of de novo CIEDs was reduced in both the first and the second lockdown. Overall, we observed mostly lower daily rates of the total number of CIEDs implanted. When divided into acute and non-acute indications for CIED, a significant decrease was seen in the daily rates of non-acute pacemaker implantations during the first interval of the first lockdown. In the second lockdown, both acute and non-acute pacemaker implantations decreased significantly during the second and third 3-week intervals, respectively.
Several national studies have been published providing national rates of CIED procedures during COVID-19 lockdowns. In northern Greece, a decrease of 48% was seen in CIED implantations during their first lockdown, while the number of implantations during their second lockdown was comparable to the year before [6]. Catalonia has reported an overall decrease of 56.5% in pacemaker and ICD implantations during the first wave of COVID-19, affecting both elective and urgent implantations, while a study from England also showed decreases in pacemaker, CRT, and ICD implantations for both elective and urgent indications during the first lockdown [8,12]. Peru reported a decrease of 73% in de novo pacemaker implantations for all aetiologies after the beginning of their first social isolation period, and in Germany a decrease of 18% in pacemaker implantations was recorded during their first lockdown [7,9]. Our study also revealed lower daily rates; however, when comparing the changes using rate ratios, we found more significant changes during the second lockdown compared with the first. Keeping in mind that most other studies have been done on the first lockdown/COVID-19 wave only, it appears that several other countries have experienced a more extensive and significant impact of COVID-19 on CIED implantation compared with Denmark. In line with our study, another Danish study that investigated cardiovascular disease during the COVID-19 lockdown found a significant drop in admission rates for new-onset ischaemic stroke and ischaemic heart disease during week 4 of the second lockdown [13], which corresponds to Figure 2 Rate ratios of the daily rates of cardiac implantable electronic device (CIED) implantations defined by both device type and 3-week interval during the first and second COVID-19 lockdowns in Denmark with 2019 as the reference. The first interval corresponds to the first 3 weeks of lockdown, the second interval to the next 3 weeks of lockdown, and the third interval to the last 3 weeks, where applicable. Implantable loop recorders (ILRs) during the second interval of the first lockdown were too few to be calculated as rate ratios, and are therefore not included in the graph. Abbreviations: CI, confidence interval; CRT, cardiac resynchronisation therapy; ICD, implantable cardioverter defibrillator; PM, pacemaker; ILR, implantable loop recorder.
Impact of the Nationwide COVID-19 Lockdowns on the Rate of CIED Implantations the drop in acute pacemaker implantations during the second 3-week interval found in our study. The Danish study found even larger reductions during the first lockdown, which was not reflected in our study [4].
In order to evaluate if non-acute electronic device procedures were postponed, as recommended by the Danish government, we stratified our data accordingly. During the first interval of the first lockdown, compliance with recommendations from the Figure 3 Rate ratios of the daily rates of pacemaker and implantable cardioverter defibrillator (ICD) implantations defined by an non-acute or acute indication and 3-week interval during the first and second COVID-19 lockdown in Denmark, with the previous year as the reference. The first interval corresponds to the first 3 weeks of lockdown, the second interval to the next 3 weeks of lockdown, and the third interval to the last 3 weeks, where applicable. The ICDs implanted according to an acute indication were too few to be calculated as rate ratios for the second and third 3-week interval, and are therefore not included in the graph. Abbreviations: CIED, cardiac implantable electronic device; CI, confidence interval; PM, pacemaker.
Danish government were demonstrated, to some extent, as only non-acute pacemaker implantation rates decreased significantly. This was only the case during the first 3-week interval, which could suggest a different prioritisation of patients further into the first lockdown. During the second lockdown, however, we observed an increase in replacement procedures during the first interval, a decrease in acute pacemaker implantations during the second interval, and a decrease of the number of nonacute pacemaker implantations during the third 3-week interval. The increase in replacements could be due to an excess left over from the first lockdown, and the following decrease in acute and non-acute pacemakers could be down to patients staying away from hospitals for fear of contracting COVID-19. Other possible explanations could be the inadequate redistribution of health care resources and prioritisation of patients, or a larger reduction of available staff due to COVID-19-related absenteeism during the second lockdown vs the first, even though instructions issued by the Danish health authorities on how to redistribute health care resources during the lockdowns tried to account for the latter [14]. However, these reasons are speculative, and the only information that can be concluded is that our results showed a greater discrepancy between recommendations from the Danish government and the surgeries performed during the second lockdown. It should be noted that guidelines from the Danish health authorities mainly provided recommendations on how to manage the large number of hospitalised patients with COVID-19 [14]. Furthermore, it was highlighted in the guidelines that surgical procedures should only be postponed on the basis of the clinical assessment of each patient. No definite list of indications that should result in surgery being postponed or not was announced, only a few examples were given: in cardiology, elective pacemaker implantations were given as an example of a procedure to be postponed. However, to our knowledge, most Danish invasive cardiologists agreed on the prioritisation of indications, which was also the basis for how they were stratified in this study, and the general impression is that they were followed, although our study cannot rule out the opposite.
On the contrary, patients in need of a non-acute ICD seem to have not had their procedure postponed during either lockdown. Only some data on acute ICDs were available, and these indicate sufficient treatment, at least during the first interval of both lockdowns. The number of CRTs and ILRs were too few to differentiate by indication, and therefore too few to determine the types of patients affected by any changes.
Overall, a decrease in acute life-saving procedures is problematic and should be avoided. Knowledge obtained from the pandemic and lockdowns is important, and attention to the effects of subsequent lockdowns during this and future pandemics is essential to maintaining the balance of health care resources and the right level of care for cardiovascular patients.
The prioritisation of acute and life-threatening conditions, and patients in need of a CIED during future and subsequent lockdowns or pandemics could prevent a later increase in the need for more extensive care and possibly even in mortality.

Strengths and Limitations
This was a nationwide study and the inclusion and selection biases were therefore minimal. In comparison to most other studies published on CIED implantation during lockdown, this study compared implantation rates with the year before using rate ratios. This is a strength of this study; however, these stricter analytics result in higher requirements, making it more difficult to show significant changes. A limitation is that Denmark is a relatively small country with a population of 5.8 million people, and the number of implantations performed is therefore equally small. This means, as mentioned earlier, that we could not stratify our data on CRTs and ILRs based on indication, as the numbers would have been too small. In addition, this was an observational study, and therefore analyses on changes in implantation rates and the COVID-19 lockdowns are only observational and do not necessarily prove a causal relationship.

Conclusion
Our study found a significant decrease in both the total daily rates of de novo and replacement CIEDs, de novo CIEDs only, and in the daily rates of non-acute pacemaker implantations during the first Danish nationwide lockdown. During the second lockdown, a significant decrease in the total daily rates of pacemaker implantations was observed, divided between both non-acute and acute indications, while an increase in the number of replacements was seen. The reasons behind these findings cannot be concluded, but our results reveal that CIED implantation rates were only modestly affected in Denmark during two national lockdowns.