The Danish Pacemaker and ICD Register: Review of Clinical and Research Potential

Introduction

Clinical registries are essential for surveillance and quality assurance in cardiovascular medicine. Established in 1982 as a hardware-oriented cardiac device registry and clinical quality database, the Danish Pacemaker and ICD Register (DPIR) has accumulated over 40 years of patient-, hardware-, and procedural data on cardiac implantable electronic device (CIED) therapy in Denmark.¹ It is among the largest, most comprehensive, and longest-running registries dedicated to CIED therapy worldwide. DPIR supports daily care, monitors safety, and offers valuable insights into long-term, real-world clinical practice. It has also been a valuable resource for epidemiological research. This review evaluates the content and data quality of DPIR, explores its research applications, and assesses its role in improving care for patients with CIEDs.

Organization

DPIR was established in 1982 by physicians from all Danish CIED implanting centers as a hardware-oriented cardiac device registry. Its objectives were later expanded to supporting daily care, documenting the quality of care associated with CIEDs, defining minimum standards of care, and identifying critical domains demanding careful attention.² More recently, DPIR was approved as an electronic medical record system. Implantable cardioverter defibrillator (ICD) implantations were included in 1989, and cardiac resynchronization therapy (CRT) devices in 1997. DPIR is located at Odense University Hospital and nested within the Arrhythmia Working Group of the Danish Society of Cardiology. The registry is funded by CIED manufacturers through a fee imposed on every device and lead sold in Denmark. In turn, the companies gain access to aggregated data from the registry.

In Denmark, pacemaker implantations are centralized to 10 public cardiology centers and one private hospital. Together, these centers serve the entire Danish population of approximately 6.0 million inhabitants (Figure 1). Five centers also perform ICD implantations, and four offer CRT. The distribution of centers across Denmark is determined by the Danish Health Authority. All centers report data to DPIR.

Figure 1 Overview of CIED centers: Four highly specialized university hospitals perform implantation of all CIED types as well as CIED extractions. One university center (Roskilde) performs pacemaker- and ICD implantations, while six centers implant pacemakers only. Pediatric CIED procedures are centralized at a single center in Copenhagen. *Also performs pediatric CIED implantations and extractions. **University Center. Abbreviations: CIED, cardiac implantable electronic device; CRT, cardiac resynchronization therapy; ICD, implantable cardioverter defibrillator.

The DPIR Steering Committee is comprised of one physician from every public CIED implantation center (including the Chair) (10 individuals) and a Data Manager (11 members in total). Daily management is overseen by the Chair, the Data Manager, and one nurse technician. The nurse technician provides technical support and is not a member of the DPIR Steering Committee. Key clinical quality indicators and standards from DPIR are defined by the Danish Health Authority and presented in an annual report.¹ Data from DPIR can be provided for research purposes, but all research emanating from the registry must be pre-approved by the Steering Committee.

Setting

Healthcare in Denmark is universally tax-financed and freely available to all Danish residents.³ This includes CIED therapy. The Danish Civil Registration System (CRS) comprises a complete and ongoing cohort of Danish residents dating back to 1968.⁴ Registration occurs at birth or upon immigration by assignment of a unique ten-digit personal identifier, the Civil Personal Registration (CPR) number. The CPR number serves as a prerequisite for individual-level linkage across the vast network of Danish registries.^3,4 The CRS is continuously updated with information on kinship, residency, migration, and vital status, enabling epidemiological research on complete populations with the potential for lifelong follow-up.⁴

Data Entry

All patients who undergo a surgical CIED procedure in Denmark are entered into DPIR using the CPR number, which enables individual-level linkage to other Danish registries. Entry occurs at first implantation and is completed, as a minimum, for all procedures involving hardware changes. The implanting physician or dedicated staff members perform the registration immediately after the procedure.

In 2007, DPIR became fully web based. This standardized the data collection procedure and enhanced its applicability as a clinical tool. Registered users can access the system, which enables physicians to view device-specific patient data across regions and over time, independently of electronic medical record systems. After implantation, in accordance with European Legislation for medical devices, patients are equipped with a patient ID (implant) card with information about the implanted device and leads, pacing mode, hardware model and serial numbers, and the date of implantation.

Data Content

DPIR records administrative patient data, clinical indication for CIED therapy, perioperative management, implanted hardware, procedure data, and implant-related complications. Indication for CIED therapy is provided only at first implantation. This includes information about cardiovascular diagnosis, prior percutaneous interventions or coronary artery bypass graft surgery, indication for CIED therapy, primary presenting symptom, New York Heart Association (NYHA) functional class, atrioventricular (AV) conduction status, QRS duration, and left ventricular ejection fraction (LVEF) (Table 1). A complete list of variables and their coding in DPIR per December 2025 is provided in Supplementary Table 1.

Table 1 Completeness of Procedural Data Variables in the Danish Pacemaker and ICD Register

Procedure data is provided for every intervention (Tables 1 and 2) and includes pre- and perioperative management and status (anesthetics, antibiotic prophylaxis, priority, body temperature, C-reactive protein, and leucocyte count, anticoagulant or antiplatelet therapy), procedure-specific information (procedure type, device type, procedure duration, fluoroscopy time and amount), and information about the implanted device (site of implantation, leads and locations, and hardware model). Many of these indicators were added to the registry in 1997.² For each case of CIED or lead removal, or lead abandonment, the reason for the explant procedure is documented. This is crucial for surveillance of serious device-related complications including CIED infections and lead-related complications. CIED infections typically warrant complete device extraction, whereas lead dysfunction often necessitates lead removal or abandonment. A separate sheet was added in 2007 to include complications for which hardware exchange or explant procedures are not necessarily indicated: superficial surgical site infections, pocket hematomas, central venous complications, deep venous thrombosis, pulmonary embolism, pneumothorax, hemothorax, pericardial effusion or tamponade, hardware complications, and stroke. Figure 2 presents the timeline of DPIR indicators, and Figure 3 shows the DPIR data structure.

Table 2 Completeness of Lead- and Device-Specific Data

Figure 2 Timeline over important structural (red) and content (black) changes in the Danish Pacemaker and ICD Register (DPIR). Bold text highlights significant time points (bold black) and major structural changes (bold red) to the registry. Abbreviations: CRP, C-reactive protein; CRT, cardiac resynchronization therapy; ICD, implantable cardioverter defibrillator; LBB, left bundle branch; LV, left ventricular.

Figure 3 Data structure in the Danish Pacemaker and ICD Register (DPIR).

Quality Indicators

A set of quality indicators and standards for CIED therapy in Denmark was defined in accordance with recommendations from the European Society of Cardiology (ESC) Working Group for Cardiac Pacing Quality Indicators and the European Heart Rhythm Association (EHRA).⁵ This data is shared with the Danish Clinical Quality Program under The Danish Healthcare Quality Institute,⁶ and made publicly available in an annual report for transparency and accountability.¹ Key quality indicators in DPIR are surgical site infections resulting in CIED removal within 365 days of implantation (maximum 2%), early implant-related complications with major clinical impact or resulting in a CIED reoperation within 120 days of implantation (maximum 5%),⁷ venous access site (minimum 50% should apply the cephalic cut-down technique),⁸ and the annual number of procedures per operator (minimum 50 procedures per operator per year).^7,9 Failure to meet the national standards elicits targeted efforts to identify, adjust, and remedy potential issues. This may occur locally or at a national level.

Completeness and Validity

Materials and Methods

For this paper, we evaluated variable completeness (all variables) and the validity of 29 key variables in the DPIR (Table 3). Completeness was assessed as the proportion of available data in the registry relative to the total number of registered procedures (for procedural variables) or patients (for baseline demographics). Both true missing entries and entries recorded as “unknown” were classified as missing. Owing to major structural revisions of the DPIR over time, completeness was evaluated across three distinct periods: 1982–2001, 2002–2010, and 2011–2023. The time periods were chosen based on data availability. Only variables with full coverage within each period were included in the analysis. Medical records served as the reference standard.

Table 3 Positive Predictive Values (PPVs) for Procedural Data Variables in the Danish Pacemaker and ICD Register (DPIR). Data Availability Refers to the Records with Information Present in the Database for a Given Variable, Reflecting Its Completeness

Validity was assessed by three independent assessors (MHJPF, FNJ, FM) through medical record review in accordance with the definitions provided in Supplementary Table 1. The validation sample was divided among three reviewers. In cases of uncertainty during data extraction or classification, the case was first discussed among the reviewers to reach consensus and, if needed, subsequently with an external reviewer (JBJ or JCO). Medical record review was performed for a combined total of 700 randomly selected procedures; pacemaker (PM), cardiac resynchronization therapy (CRT), and implantable cardioverter-defibrillator (ICD) from three implanting centers (two tertiary and one regional). We validated 300 PM, 200 CRT-P/D, and 200 single- and dual-chamber ICD procedures (100 of each device type per center) including a random selection of first implants and reintervention procedures. Another 200 lead extractions were separately validated focusing on reason for reintervention and mode of extraction only. Numeric variables were validated as integers, as fractional values were not uniformly reported. Lead type and position (n=500), and procedure duration and fluoroscopy amount (n=200) were validated only for a subset of patients. During follow-up, one center (Odense University Hospital) transitioned to a new electronic medical record system, resulting in reduced availability of certain variables for validation: height, weight, biochemistry, ECG characteristics, procedure duration and fluoroscopy amount. Accordingly, validation of these variables was limited to implantations performed after the transition. Data on left ventricular (LV) lead positioning and pacing thresholds were inconsistently reported across centers and available for only a very small subset of patients; consequently, these variables were not included in the validation analysis.

Positive predictive values (PPVs) were calculated as the proportion of DPIR data confirmed by medical record review with 95% confidence intervals (CIs) estimated using the Wilson score method. Missing data were excluded from the primary PPV analyses; however, sensitivity analyses were conducted using best-case (all missing considered confirmed) and worst-case (all missing considered unconfirmed) scenarios. For selected variables, analyses were further stratified by acute versus elective procedures to account for potential misclassification during off-hours.

All statistical analyses were performed in Stata 19 SE (StataCorp, USA). The study was approved by the DPIR Steering Committee. Access to patient medical records was granted for each participating center (Viborg, Aarhus, and Odense) and limited to entries within five years of approval in accordance with institutional guidelines (January 1, 2020, to December 31, 2025). In accordance with Danish law, informed consent and institutional review board approval was not required for this type of study. All data accessed for this study were handled in accordance with relevant Danish data protection and privacy regulations.

Completeness

Completeness results are presented in Table 1 and Table 2, and in Supplementary Figure 1. Most variables exhibited high completeness (>98%), whereas NYHA functional class (46–78%), left ventricular ejection fraction (LVEF) (50–88%), and baseline atrioventricular (AV) conduction at implant (26%) demonstrated lower completeness. Completeness was generally higher in more recent years, likely reflecting improvements following full digitalization of the registry. NYHA class was primarily reported for patients with clinically manifest heart failure, while AV conduction status was predominantly available for patients with an indication for bradycardia pacing.

Validity

All procedures were eligible for validation through medical record review. Positive predictive values (PPVs) are presented in Table 3 and in Supplementary Table 2, stratified by procedure priority. PPVs ranged from 86 to 100%. In total, 23 of 29 variables demonstrated PPVs ≥95%, including indication for cardiac implantable electronic device (CIED) therapy (97%, 95% CI 95–98%), lead type (100%, 95% CI 99–100%), and venous access site (98%, 95% CI 97–99%) -variables that have frequently been employed in research. Reason for reintervention, which is used to identify complications requiring a reintervention procedure, had a PPV of 91% (95% CI 86–94%). PPVs were consistent between acute and elective procedures (Supplementary Table 2). Overall, the PPVs remained high in both the best- and worst-case analyses (Supplementary Table 3), indicating robust validity under varying assumptions.

Limitations

Although the overall validity was excellent, results should be interpreted with caution for variables with small sample sizes due to a high proportion of missing data or low completeness, or variables with multiple non–mutually exclusive response options (eg., indication for CIED therapy, AV conduction, QRS duration, and reason for lead or generator reintervention; Supplementary Table 1). Furthermore, it was not the objective of this study to assess the causes of generator or lead removal or negative predictive values. Although the validation was conducted in only three centers across two regions, we have no reason to suspect that the findings are not generalizable to other Danish regions.¹⁰ In contrast, it is less certain whether our findings can be extrapolated to earlier time periods than 2020–2025. However, Consistency with earlier validation of defibrillator leads,¹¹ supports the extrapolation of the observed high data quality to earlier periods.

Although complications are recorded in the DPIR, they were not assessed in this study due to underreporting of complications that do not require a CIED reintervention such as postoperative pneumothorax, superficial infections, or conservatively managed hematomas.⁷ An audit of complication registration in DPIR performed in 2009 revealed a sensitivity for complications of 43% (95% CI 39–47).¹² This limitation arises because DPIR is primarily a hardware-oriented registry. Primary data entry is performed immediately after implantation by the implanting physician or designated staff, whereas later or minor complications are often treated by non-electrophysiologists without access to the registry, leading to incomplete reporting. Exceptions are complications requiring device reinterventions, most notably definite CIED-related infections, in which case complete hardware removal is recommended,^12–14 or lead-related reinterventions.^12,15 These events are systematically recorded under reason for lead or device removal. However, cases of definite CIED infection in which device removal is deferred (eg., in frail or high-risk patients) remain subject to underreporting in the registry.

Temporal Trends in Implant Activity

DPIR has captured the (near) totality of CIED therapy in Denmark for over 40 years, accumulating more than 200,000 CIED procedures across more than 133,000 individuals. Per February 27, 2024, 52,039 persons were living with a CIED in Denmark. As shown in Figures 4 and 5, the annual number of CIED procedures has been steadily increasing as has the number of high-complexity CIED device implantations and reoperation procedures (battery changes, upgrades, downgrades, lead-related interventions and device removals), which are associated with higher complication rates.^7,16

Figure 4 Implant activity 1982–2023 per device type excluding lead interventions only. Colors indicate device type (pacemaker, ICD, CRT or other device type). CRT-P/D = cardiac resynchronization therapy pacemaker/defibrillator. Abbreviations: ICD, implantable cardioverter defibrillator; PM, pacemaker.

Figure 5 Total number of procedures performed in 1982–2023. Colors indicate registration type (first, second, third, fourth or more than fifth registration).

Research Potential

The near-optimal epidemiological framework served as the basis for a wide range of research using DPIR data. Below are examples of how research emanating from DPIR influenced clinical care for patients with CIEDs.

Prevalence and Prevention of CIED-Related Complications

CIED implantation carries a risk of serious complications, including infection, lead dysfunction, cardiac perforation, tamponade, and pneumothorax. Randomized controlled trials are often underpowered to adequately assess such events, and few prospective CIED registries match DPIR in completeness, accuracy, and long-term follow-up.

To understand the risks and predictors of CIED-related adverse events, a previous study identified all patients undergoing a CIED procedure between May 2010 and April 2011 in Denmark using DPIR (n=5,918).⁷ Detailed information about minor and major complications occurring within six months of the procedure were collected from medical chart review in all patients.⁷ Early implant-related complications occurred in approximately 9.5% of patients (higher than previously assumed) while major complications (defined by severe clinical impact or reoperation) were observed in 5.6%. Complication risks were higher among women, underweight patients, those receiving complex devices, and in procedures performed at low-volume centers (<750 procedures/year) or by low-volume operators (<50 procedures/year).

Subsequent analyses of 28,860 Danish patients examined specific risks for pneumothorax¹⁷ and lead-related complications.⁸ Procedural factors were the strongest independent predictors, and subclavian venous access was associated with a markedly increased risk of pneumothorax (odds ratio (OR) 7.8, 95% confidence interval (CI) 3.0–10.8), discouraging the use of intrathoracic venous puncture.^16,18

Data from DPIR enabled researchers to identify complete populations of CIED recipients and thereby determine complication rates at the population level. These studies established contemporary benchmark rates for future reference and, through risk-factor assessments, helped define quality indicators for CIED therapy. For example, annual operator and center volume, and systematic reporting of CIED-related adverse events, were included as recommended indicators for the quality and outcomes of cardiac pacing by EHRA,⁵ and they informed the EHRA consensus document on optimal implantation technique for pacemakers and ICDs.¹⁹ Standardized registration is essential for identifying and addressing undesirable patterns of care.

CIED Infections

Given the clinical and economic burden of CIED infections,²⁰ they have been studied extensively using data from DPIR. Complete CIED removal is recommended in case of definite CIED infection,^13,14 and because DPIR is highly hardware oriented, infections are well-documented in the registry. Two population-based cohort studies originating from DPIR examined the incidence of CIED infections and their associated risk factors; first in 2011 in 46,299 pacemaker patients (1982–2007),²¹ and later in 2019 in 97,750 patients with pacemakers, ICDs or CRTs (1982–2018).²²

The overall incidence of CIED infections after first pacemaker implantations was 2.04/1000 device years (DY) for pacemakers implanted between 1982 and 2018.²² Infection rates were highest for higher complexity devices, and considerably higher after CIED reoperation procedures; as high as 18.94/1000 DY for CRT-D reoperations. These findings underscored the significance of device and procedure complexity for the risk of infection.

A key strength of DPIR is its ability to track detailed device histories for each patient across centers and administrative regions.³ When linked with other Danish national health registries, DPIR has facilitated identification of high-risk patient subgroups,^23–25 enabling the development of targeted preventive strategies. For example, pocket infections were shown to occur more frequently after CRT reoperations,²³ and subsequent analyses using DPIR demonstrated that prophylaxis with an antibiotic-eluting envelope was associated with a reduced infection risk in this population.^20,26 As healthcare systems face increasing financial constraints, systematic data collection is essential to guide the targeted use of preventive strategies in patients at highest risk.²⁷

Lead Performance

In December 2011, the U.S. Food and Drug Administration (FDA) issued a Class I recall for the 7- and 8-French Riata defibrillator leads (St. Jude Medical) after multiple reports of conductor externalization and lead failure.²⁸ Hardware recalls pose serious risks to patients and must be addressed promptly through correction, monitoring, or device removal.²⁹

DPIR maintains accurate hardware registrations, which enabled rapid identification of all 299 patients implanted with the recalled Riata leads. This facilitated retrospective chart review, organization of an examination program to assess the prevalence of externalized leads, and long-term follow-up of patient outcomes.^30,31 DPIR thus served both as a clinical instrument for addressing urgent patient safety issues, and as a provider of patient data for investigating the mechanisms, prevalence, and risk factors for conductor externalization and electrical failure. The resulting studies demonstrated that the prevalence of lead externalization in the Danish cohort was comparable to international findings, that the degree of externalization increased over time, that its anatomical distribution differed between single- and dual-coil leads, and that externalized leads were associated with an elevated long-term risk of electrical abnormalities. Subsequent studies have used the registry to examine lead-related outcomes in other subpopulations of patients with CIEDs to identify causes of failure and prevent future adverse events.^15,32,33

More recently, several new pacing and ICD modalities have been introduced, including conduction system pacing (CSP), leadless pacemakers, and subcutaneous and extravascular ICDs. The long-term durability and extractability of these systems remain uncertain, particularly for CSP, an emerging modality that has been widely adopted beyond controlled trial settings.^34,35 DPIR began recording His-bundle lead placements in 2009 and added left bundle branch area (LBBA) lead placements in 2023. These updates will enable monitoring and longitudinal assessment of CSP lead performance and outcomes within the Danish healthcare setting.

Perspectives

Clinical registries are essential for independent monitoring of clinical performance and quality of care, and for documenting the prevalence and impact of cardiovascular disease on outcomes of CIED therapy. With a population now exceeding 133,000 individuals with CIEDs, combined with the ability to link individual-level data to numerous national databases in Denmark,³ the research potential of DPIR is extensive. Maintaining a registry with high completeness and validity, however, requires a strong organizational structure, perseverance, and adaptability from all participating centers.

Beyond fostering numerous registry-based studies, the association of Danish CIED implanting centers has enabled several investigator-initiated national randomized controlled trials in clinical electrophysiology and device therapy, including the DANISH trial,³⁶ the DANPACE³⁷ and DANPACE II³⁸ trials, and the ongoing DANISH-CRT trial.³⁹ As demonstrated by the Western Denmark Heart Registry⁴⁰ and the SWEDEHEART Registry,^41,42 clinical registries may also serve as platforms for RCTs. Registry-based RCTs leverage the existing infrastructure for randomization, data collection, and follow-up to provide pragmatic answers to key clinical questions, while substantially reducing the costs and time compared with conventional RCTs.

Conclusion

Maintaining clinical registries is labor-intensive but essential to preserve both scientific integrity and economic sustainability in cardiac electrophysiology and device therapy. DPIR comprises a complete, continuously updated cohort of patients with CIEDs, including detailed data on patients, hardware, and procedures. For more than 40 years, it has supported CIED therapy in Denmark, making it an invaluable tool both in clinical practice and for epidemiological research. Completeness is high, and most variables demonstrate excellent validity for research purposes. Through a wide range of population-based investigations, the registry has played a key role in shaping clinical decision-making in Denmark and beyond.