Introduction
Electronic medical record downtimes caused by system failures or ransomware attacks have exposed the reliance hospitals and healthcare systems have on the use of electronic medical records. Reference Uslu and Stausberg1,Reference Donnelly, Simpkin and Lillis2 More specifically, downtime events have also affected processes for cardiac registry data collection. Reference Uslu and Stausberg1 For many of today’s healthcare workers, electronic medical records are the only medical record system they have known, leaving them unprepared for extended paper charting in downtime situations. Reference Pageler, Longhurst and Wood3 In paediatric cardiology, where CHD patients require complex, multi-disciplinary care, the limitations of paper charting are particularly problematic, especially during an unexpected downtime. 4 As complex patient management has outgrown paper charting, downtime events have compromised the quality of medical record data for abstraction to cardiac registries and national reporting. 5 Registries, including Paediatric Cardiac Critical Care Consortium and the Society of Thoracic Surgeons, require high-accuracy data submission to support high-quality care through data-driven insights that reduce complications and improve outcomes. Reference Gaies, Donohue and Willis6
Hospital downtimes present unique challenges for data abstractors to maintain standards and processes for data abstraction and registry entry. These periods demand an even greater level of vigilance and adaptability to uphold standards. This article outlines key challenges faced by data abstractors in a paediatric cardiac programme during an extended downtime event and suggests strategies to ensure the continued accuracy of downtime charting to mirror the complexity of cardiac patients.
Materials and methods
This was a single-centre descriptive experience on how data integrity was maintained during a 27-day unexpected downtime event followed by a 10-day partial downtime where individual systems restoration and connection had not been completed. Throughout the downtime, our organisation adhered to Joint Commission, Centers for Medicare & Medicaid Services, and HIPPA-aligned contingency procedures, and no serious harm events occurred that were associated with the unavailability of the electronic medical record.
Before the downtime event, the heart centre’s team of data abstractors used a well-coordinated process for data capture and validation to ensure quality reporting across multiple registries, including but not limited to the Paediatric Cardiac Critical Care Consortium and Society of Thoracic Surgeons. These processes were fundamentally reliant on uninterrupted access to electronic medical records, which served as the primary infrastructure for nearly every phase of data abstraction. This was followed by a detailed review with the physician champion prior to registry submission. For each registry, the team utilises multiple electronic medical records and reporting tools shown in Figure 1.
Process for cardiac registry data collection and submission prior to downtime.
Figure 1 Long description
The flowchart illustrates the process for cardiac registry data collection and submission. It begins with automated reporting tools that gather patient lists, blood products, infections, and medications. Data from inpatient and outpatient electronic medical records (EMR) systems, including Cerner, Athena, and NextGen, are then fed into an internal database. This database, which includes REDCap, a centralized master folder, and a master patient list, supports rounding activities. After discharge, chart abstraction occurs, followed by the use of registry submission software such as Cardio Access and Arbornetrics. The data is then validated with an MD champion before being submitted to national registries like STS, PC4, PAC3, CNOC, IMPACT, NPC-OIC, and FON.
Utilising these tools and processes allowed for streamlined data collection while maintaining data integrity for the potential 1,700 data metrics reported to the Paediatric Cardiac Critical Care Consortium and Society of Thoracic Surgeons registries and served as the backbone of the data abstraction process. Every reported data element to all cardiac registries originated from discrete fields, clinician-entered notes, and time-stamped procedural records housed within the electronic medical record. In total, the combined potential data metrics reported across major registries (Paediatric Cardiac Critical Care Consortium, Society of Thoracic Surgeons, PAC3, and IMPACT) commonly used in cardiac programmes at paediatric hospitals encompass a potential 4,000 data metrics for each patient reflecting the comprehensive nature of data collection across cardiac registries. 7
When electronic medical records are fully operational, the data abstraction process is guided by a structured framework specific to each registry (although many elements are common between registries), ensuring consistency and traceability. The data abstraction team developed a standardised reference guide that outlined the order of preferred locations of where to pull electronic medical record data elements to promote consistency and ensure accurate data entry. In the event of an audit or question regarding data origin between data abstractors, this source reference provided a transparent guide to the initial data sources used for abstraction. This approach ensured that all metrics are consistently extracted from the same predefined locations across data abstractors, promoting uniformity and reproducibility in data collection and providing a high level of consistency and reliability. This pre-downtime method resulted in our centre receiving a 95–99.9% pass rate during all cardiac registry audits in the last 5 years.
When electronic medical records are unavailable, the integrated system of data capture and validation is significantly compromised. The absence of real-time access to documentation, lab values, and procedural data interrupts not only the abstraction process but also undermines data traceability, accountability, and audit readiness. The ability to instantly search records, validate timelines, trace data origins, and align documentation across disciplines disappears. Paper charting introduces fragmentation and inconsistency, leading to ambiguity. Data that would otherwise be in standardised fields must now be found from narrative notes, flowsheets, and handwritten margins (assuming they have been documented).Reference Lyons, Daly, Palmer, Horowitz and Cullinane 8
The downstream impact is significant: delays in identifying eligible patients, increased risk of missed or misinterpreted data, and compromised data quality for registry submission. In a high-stakes environment like congenital cardiology, where abstraction supports quality reporting and clinical benchmarking, this dependency on electronic medical records is not just a matter of efficiency but a matter of accuracy, accountability, and care continuity.
Results
During the downtime period, the data abstractor team manually managed data for 123 unique cardiac patient encounters totalling 762 patient days.
The downtime was initially assumed to be brief, so no formal alternative abstraction workflow was developed. However, as the outage extended, it became clear that the abstraction team could not rely on eventual electronic medical record restoration and needed a completely new method to:
-
• Identify all cardiac patients in real time.
-
• Track admissions, transfers, procedures, and complications.
-
• Ensure no patients were missed.
-
• Capture sufficient detail to support later registry abstraction.
One experienced abstractor led the development of a manual patient tracking system, coordinating with clinicians and administrators to gain access to unit census tools and provider tracking logs. Daily interdisciplinary rounds became critical for obtaining timely updates.
PHI protection, data governance, and document security during downtime
PHI was protected throughout the downtime process through strict data governance, controlled access, and secure storage across all platforms used for patient tracking and data collection. Documents containing patient information were maintained only within encrypted, institution-approved systems (REDCap, institutional email, and shared documents) with role-based access limited to designated data abstractors and essential clinical leaders. REDCap used prior to the downtime event continued to serve as a secure, parallel data repository for information abstracted from paper downtime records and direct data abstractor participation on rounds. Data collected were never stored on personal devices or unsecured media.
Although multiple redundant data sources were used to maintain continuity, the paper medical record remained the primary source of truth, and all secondary documentation was reconciled against it for accuracy. These practices ensured HIPAA compliance while allowing near real-time data capture, continuity of abstraction, and protection of patient information throughout the downtime period.
Six alternative data sources used to replace the electronic medical record
To reconstruct real-time patient tracking, the team relied on six data sources shown in Figure 2. Two sources included the OR Control board and REDCap, which were existing tools routinely used prior to the downtime event. The remaining sources were not part of the data abstraction team’s workflow but were quickly incorporated to compensate for the loss of the electronic medical record:
Process for cardiac registry data collection and submission during downtime.
Figure 2 Long description
The flowchart illustrates the process for cardiac registry data collection and submission during electronic medical record downtime. It begins with the EMR being down, leading to the unavailability of EMR and automated tools. The process then branches into several data sources: Daily Multidisciplinary Rounds, Cardiac Patient Census, Rounding Notes, OR Control Board, REDCap Patient Tracking, and Paper Chart. These sources feed into an Internal Database Collection or Master Patient List, where data is combined and cross-referenced. This internal database tracks admissions, transfers, discharges, procedures, complications, and demographics. The validated data is then submitted to national registries through registry submission software, with validation by an MD Champion.
-
1. CCU Census – a manual log maintained by charge nurses for admissions, discharges, and transfers.
-
2. Congenital Heart Surgery Rounding Reports – APP-maintained, password-secured, controlled-access, Excel document with patient location, procedures, and care updates.
-
3. OR Control Board – a third-party system allowing manual surgical case tracking and timestamp verification.
-
4. REDCap Patient Tracking System – a multidisciplinary shared database for surgical and clinical details.
-
5. Daily Multidisciplinary Rounds – real-time clinical updates.
-
6. Paper Chart – official record used for retrospective abstraction.
Creation of the master patient list
All information from all six sources was cross-referenced daily by the data abstraction team to create a master patient list. Admissions, procedures, complications, and care milestones were reconciled across sources to validate accuracy, identify discrepancies, and confirm timing of events.
This reconciliation followed a defined daily workflow in which a centralised master spreadsheet captured predefined data elements, including MRN, name, date of birth, admission unit and service line, procedures, and key hospitalisation milestones. These elements were compared across all six independent data sources, and discrepancies in identifiers, procedural timing, or clinical events were flagged and verified with the primary clinical team. Special attention was given to MRN changes that occurred when transitioning from downtime documentation to a live electronic medical record. This structured comparison and real-time discrepancy resolution ensured complete case capture despite the absence of electronic medical record functionality.
Transition to paper chart abstraction
Once patient tracking was reliable, the team transitioned to formal abstraction from paper charts. Because abstractors were physically present on site, they were able to access charts before they entered the months-long scanning process for electronic medical record re-entry.
To support this work, the team developed a Downtime Source Reference Guide that mapped where each required registry data element could be found in the paper record, including primary, secondary, and tertiary location options. This guide evolved as bedside staff adapted documentation practices, including creative solutions such as inserted documents for serial weight tracking.
Despite wide variability in paper documentation, the team’s deep knowledge of registry requirements allowed consistent identification of necessary data.
Impact on registry submission timeliness and data completeness
Participating registries require timely and complete data submission to preserve the value of real-time benchmarking and quality improvement feedback. The Paediatric Cardiac Critical Care Consortium requires case submission within 30 days of hospital discharge, while the Society of Thoracic Surgeons and other registries follow quarterly submission cycles with internal institutional expectations for near real-time abstraction to prevent backlog and data loss.
Prior to the downtime event, the data abstractor team met the Paediatric Cardiac Critical Care Consortium 30-day submission requirement 93% of the time, compared to an approximate collaborative average of 60%. 9 During the downtime and immediate recovery period, this rate declined substantially as the team shifted from electronic to paper-based abstraction while simultaneously developing a new patient tracking system. In the first months following electronic medical record restoration, on-time submission remained significantly below baseline as paper charts were still being processed centrally and scanning delays limited record availability. Two months post-downtime, on-time Paediatric Cardiac Critical Care Consortium submission reached its lowest point at 59%, reflecting the cumulative impact of delayed access to medical records rather than failure of patient or event identification. Following this nadir, submission rates began to improve as the downtime tracking system and paper source reference guide matured and record access stabilised.
As medical record access normalised and abstraction workflows matured, the backlog was reduced. From October through December of the same year, on-time Paediatric Cardiac Critical Care Consortium submission reached 96%, effectively restoring pre-downtime performance levels. Quarterly submission requirements for the Society of Thoracic Surgeons and other registries were met throughout the downtime period, although internal abstraction timelines were prolonged.
Importantly, no patients or cases were missed during this period. The redundancy of six data sources and direct data abstractor participation in daily rounds ensured complete patient identification and awareness of key clinical events in near real time, even when formal abstraction was delayed. This significantly reduced the time required to locate required data elements once charts became available and preserved the completeness of registry submissions.
Key takeaway
This experience demonstrated both the vulnerability of a fully electronic medical record-dependent abstraction workflow and the resilience of a highly trained data abstractor team capable of reconstructing a reliable tracking and abstraction system using nontraditional data sources. Primary intervention was not simply reliance on paper charts but the rapid development of a contingency system for identifying where data could be found when the electronic medical record was unavailable. The use of multiple redundant data sources allowed the team to maintain real-time awareness of all cardiac patients, which would not have been possible through retrospective paper chart review alone.
Direct participation of data abstractors in daily multidisciplinary rounds proved to be one of the most important lessons from this event. Abstractor presence led to immediate improvements in bedside paper documentation practices as clinicians became aware of which data elements were required for registry reporting. This resulted in adaptive documentation strategies that directly improved the quality and completeness of the paper record during downtime.
The downtime highlighted that effective abstraction is not solely dependent on access to the medical record but on deep knowledge of clinical workflow, patient movement, and the multiple locations where critical data are generated throughout hospitalisation.
Discussion
Downtime events expose the healthcare system and data abstractors’ reliance on electronic medical records, especially in high-data environments like paediatric cardiac care. As CHD patients increasingly live with complex, chronic conditions, accurate documentation is essential not only for care delivery but also for data abstraction to be able to submit to registries. Most modern hospitals are structured around electronic medical record workflows, and providers today often lack extensive formal training or experience in paper charting. When electronic medical records are unavailable, even routine data capture becomes fragmented and inefficient. Critical information such as daily weights, medication titrations, or transfer times often have missing or inconsistent documentation in paper charts, presenting a serious obstacle to accurate data abstraction for registry submission. Documentation varies by provider and is often delayed, which compromises data quality and reduces the precision of clinical benchmarking.
While many hospitals have robust clinical downtime protocols focused on continuity of care, there is little published guidance describing how to maintain data abstraction workflows during prolonged electronic medical record outages. In discussion with national registry representatives during this event, no formal recommendations or established practices for abstraction during extended downtime were identified. In January 2025, the Paediatric Cardiac Critical Care Consortium became the first cardiac registry to issue informal specific downtime guidance. The recommendations outlined two abstraction pathways: (1) deferring submission of affected cases or (2) implementing validated alternative abstraction strategies that preserve the Paediatric Cardiac Critical Care Consortium’s data quality standards. Supporting guidance emphasised the importance of prospective documentation practices, detailed data mapping through tools such as a source reference guide, and the designation of local champions to oversee abstraction efforts. Additional recommendations included establishing contingency plans for on-site abstraction support if primary teams are remote, implementing regular data quality reviews, and coordinating with IT and billing teams to ensure complete patient accrual through alternative census adjudication methods when electronic medical record reports are unavailable. 10
This informal guidance marks a significant advancement in registry operations. While organisations such as the Centers for Medicare & Medicaid Services and the National Healthcare Safety Network have issued policies for exceptional reporting circumstances, most clinical registries, including the Society of Thoracic Surgeons, Vermont Oxford Network, and the National Surgical Quality Improvement Programme, have not yet developed structured protocols for abstraction during downtime events. In this regard, the Paediatric Cardiac Critical Care Consortium is proactively addressing a critical gap in data governance and continuity.
Even short periods of compromised data quality can impact long-term benchmarking since missing or inconsistent entries may affect risk-adjusted outcomes and performance metrics. In cardiac registries that track longitudinal patient trajectories, the integrity of historical data becomes crucial. Institutions must be mindful that downtime gaps are not just operational setbacks but are also potentially long-lasting distortions in how their performance is evaluated and how quality improvement efforts are directed. Incomplete registry submissions can distort benchmarking, skew risk-adjusted outcomes, and compromise institutional rankings. As cardiac registries become increasingly central to quality reporting, reimbursement, and reputation, the need for accurate data regardless of electronic medical record status becomes more urgent.
This report reflects the experience of a single paediatric cardiac centre during an extended electronic medical record outage and may not be generalisable to institutions with different staffing models, documentation practices, or access to similar alternative data sources. The processes described were developed reactively during an operational crisis and are based on retrospective reconstruction rather than a prospectively designed contingency plan. Because formal efficiency and data completeness metrics were not collected in real time, assessment of effectiveness relies on surrogate measures such as registry submission timeliness and the absence of missed cases. Several tools used during downtime were unique to this institution’s clinical environment and may require adaptation in other settings. Despite these limitations, this report provides a detailed description of an abstraction contingency framework in an area where little published guidance currently exists.
Conclusion
The implications of electronic medical record downtime on data quality and registry submission are far-reaching, especially in high-acuity environments such as paediatric cardiac care. These events expose vulnerabilities in documentation and data systems that are often masked during standard operations. Data abstractors shoulder a heavy operational burden during these periods, reconstructing patient journeys from fragmented records, working extended hours, and safeguarding the integrity of institutional data. This critical work directly supports performance benchmarking, clinical analytics, and quality reporting efforts.
To ensure continuity and resilience, downtime data preparation must be embedded into standard institutional operations not as emergency improvisation but as deliberate, proactive planning. This includes aligning downtime documentation tools with the complexity of modern medicine. Electronic medical records have evolved to capture nuanced clinical data in ways that paper charting no longer mirrors. As medicine continues to advance, so too must paper-based workflows. Hospitals, registries, and clinical teams must collaborate to redesign downtime documentation systems that reflect today’s care standards, capturing the level of detail needed for accurate abstraction, communication, and performance evaluation. Only by rethinking these foundational practices can we protect data integrity, support clinical decision-making, and maintain high standards of care even in the face of digital disruption.
Acknowledgements
None.
Financial support
This research received no specific grant from any funding agency, commercial or not-for-profit sectors.
Competing interests
None.
