High-quality healthcare is that which increases “the likelihood of desired health outcomes and is consistent with standards of health care.”1 Inherent in such a definition is the critical role of measurement of patient outcomes and the ‘standards of healthcare,’ such as those reflected in evidence-based clinical practice guidelines.2 Unfortunately, when the quality of healthcare in the US is measured, significant ‘gaps’ in care, unexplained variation in care delivery and outcomes, and medical errors are often found.3 As such, a major focus of the current national healthcare debate is how to improve the quality of care in order to fill these gaps, reduce inappropriate variation, and eliminate medical errors.
Measurement is central to quality improvement (see Figure 1). Only with the collection of data can clinicians or hospitals compare, or benchmark, their care and outcomes to peers and national standards. Benchmarking helps identify targets for performance improvement, and the feedback of data with benchmarking, or ‘audit and feedback,’ can lead to improvements in care.4 Major strides in quality improvement require a process of benchmarking of data coupled with changes in systems of care and engagement of local clinical leaders and administrative support. The process must be iterative, to assess whether the changes made worked. Therefore, quality improvement is a process that includes being able to answer the questions ‘How am I doing compared with others?’ and ‘Am I getting better?’ and is most successful in a culture that supports performance improvement efforts. Unfortunately, the tools and culture for effective quality improvement are often lacking in healthcare settings.
The National Cardiovascular Data Registry
With a mission centered on the collection and use of data to support quality improvement, the American College of Cardiology (ACC) launched the National Cardiovascular Data Registry (NCDR™) (www.ncdr.com) more than a decade ago. There were several key reasons for the development of the NCDR: first, the cardiovascular profession wanted to demonstrate leadership in defining and improving quality of care; second, individual hospitals or practices can collect their own performance data, but rarely have the ability to accomplish regional or national benchmarking on their own; third, data collection needs to be standardized for comparisons to have validity; and finally, the data elements collected should support measurements of care that reflect current evidence-based clinical practice guidelines and quality metrics. The ACC, in collaboration with the American Heart Association (AHA) and other professional organizations, develops the data standards, guidelines, and quality measures for cardiovascular disease.2 The ACC was therefore well positioned to establish national clinical registries to promote quality improvement.
Currently, the NCDR includes six registry programs (see Table 1). Over 2,400 hospitals across the US participate in at least one NCDR program. The CathPCI registry alone includes more than 10 million patient records. The IC3 program, even in its pilot phase, includes over 175 practices, representing 623 office locations in 48 US states and two US territories, and has over 150,000 patient records.
It is important to emphasize that the NCDR registry programs are ‘multi-organizational,’ whereby the ACC partners with other professional organizations in specific areas where appropriate. For example, the ACC and AHA partner in the ACTION Registry®—Get With The Guidelines (GWTG™), just as they partner in the development of data standards, guidelines, and quality measures for acute myocardial infarction (MI). The CathPCI registry naturally links the ACC with the Society for Cardiovascular Angiography and Intervention (SCAI) and, similarly, the Implantable Cardioverter Defibrillator (ICD) Registry naturally links the ACC and the Heart Rhythm Society (HRS). The CARE Registry reflects the multistakeholder nature of carotid revascularization through its seven partnering organizations.
The data collection goal for each NCDR program is to assemble a parsimonious data set based on existing data standards that will enable assessment and feedback of guideline concordance of care and quality metrics, including both process of care and outcome measures. Examples of quality metrics for the CathPCI Registry include door to balloon time for patients with ST-segment elevation MI, vascular complications, need for urgent coronary artery bypass graft surgery, and risk-adjusted in-hospital mortality based on the validated, National Quality Forum-endorsed NCDR risk model. Examples of ICD Registry metrics include peri-procedural complications, guideline-indicated medication use rates, and utilization measures such as type of ICD used. The ACTION Registry— GWTG measures key processes of care for patients with MI ranging from time to reperfusion therapy (for both direct arrival and transfer patients) and delivery of evidence-based medications throughout hospitalization to risk-adjusted in-hospital mortality.
The CARE Registry, which includes assessment of both carotid stenting and carotid endarterectomy, collects data necessary to evaluate procedural indications, peri-procedural complications, and patient outcomes, including 30-day outcomes. Of note, the neurological outcomes captured in CARE include an independent neurological examination using the National Institutes of Health (NIH) stroke assessment tool, a key feature for accurate assessment of patient outcomes following carotid revascularization.
Finally, the new office-based IC3 program collects the data necessary to evaluate the ambulatory care performance measures for patients with coronary artery disease, heart failure, hypertension, diabetes, and/or atrial fibrillation.5
For all NCDR programs, the goal is to have data collection as embedded as possible into the routine clinical workflow. For example, vendors of the hemodynamic recording systems used in many cardiac catheterization laboratories are certified by NCDR, such that data entered at the point of care are automatically captured for subsequent transmission to the CathPCI registry. With the IC3 program, the NCDR works with individual practices regarding their preferred method of data collection and transmission. Increasingly, data can be extracted from electronic medical records so data are truly collected as part of routine care documentation.
A core quality of care function of the NCDR programs is to provide reports back to participating hospitals and practices regarding their care. These reports include both regional benchmarking (to ‘like’ hospitals/practices) and national benchmarking. A sample section of such a benchmark report is contained in Figure 2. It is intended that this feedback with benchmarking will be used by hospitals and practices to target areas for improvement. For example, a hospital participating in the CathPCI Registry may find that they have a high vascular complication rate compared with their region or the nation. This should lead to an examination of cases with such complications—a root cause analysis—to determine whether process changes can be made to reduce complications. Moreover, the NCDR and broader ACC and SCAI communities provide educational resources and sharing of best practices to support quality improvement for participants.
Recognition and Linked Incentives
The NCDR now has a recognition program for participants achieving a high quality of care. This program started in the ACTION Registry— GWTG, naturally building from the successful AHA Get With The Guidelines Recognition Program, and will be expanded throughout the NCDR programs. The NCDR Rewards and Recognition program will annually recognize hospitals and practices based on criteria linked to established quality metrics, quality improvement, and leadership in the field of quality of care.
In addition to local use of NCDR reports to support quality improvement, data from the CathPCI registry and other NCDR programs are increasingly used by hospitals and practices for external reporting to states, health systems, and payers. For example, five states currently require participation in the CathPCI registry, as do United Healthcare, Blue Cross/Blue Shield, and Wellpoint for their cardiovascular quality programs. The Centers for Medicare and Medicaid Services (CMS) mandate participation in the ICD Registry for data on primary prevention ICDs for Medicare patients, and the IC3 program is certified as a reporting solution for the CMS Physician Quality Reporting Initiative.
There are several important reasons why data from NCDR programs are increasingly desired by states and payers for quality assessment. First, the data are clinical rather than administrative, or claims, data. While claims data may be sufficient for some quality assessment purposes, claims data generally do not include key variables such as disease severity (e.g. coronary anatomy) or comorbidities (e.g. degree of renal insufficiency) that can directly affect clinical decisions and risk for adverse outcomes. Moreover, claims data generally cannot differentiate comorbidities from complications and usually lack indications for procedures, medications given, and exclusion criteria or intolerance of medications for a given patient. Furthermore, data restricted to that available to single states or payers generally do not allow direct benchmarking to national data and/or do not allow benchmarking of quality of care based on all of the patients cared for by a given hospital or practice.
For quality improvement to be effective, data must also be trusted as being valid and accurate by clinicians. In contrast to claims data or other sources, NCDR participants submit their own clinical data, which are then evaluated for internal consistency and completeness by the NCDR. The NCDR maintains a robust data quality program, including use of data standards, training, and data checks for completeness and consistency, and an on-site auditing program to assess the accuracy of data entry by facilities. Thus, the data in NCDR are clinical data reflective of the ‘denominator’ of care at a hospital or within a practice and are transparent to participants with regard to the validity of the data and use of the data to evaluate their quality of care. All of these factors are important to empower quality improvement.
Quality Initiatives and Tools
In addition to benchmarking and feedback of clinical data and quality metrics, NCDR promotes quality improvement through allied quality initiatives and tools. For example, NCDR actively promoted the Door-To-Balloon Alliance (www.d2balliance.org) among the NCDR community of hospitals, and was a major source of measurement of the impact of the initiative. As another example, the ACTION Registry-GWTG will serve as the primary measurement source to evaluate the impact of Mission Lifeline (www.americanheart.org/missionlifeline).
With regard to quality improvement tools, the NCDR developed CathKIT, to guide local evaluation, and ‘Plan–Do–Study–Act’ interventions, to improve the quality of care in the cardiac catheterization laboratory.6 In addition, the NCDR now partners with leading regional percutaneous coronary intervention (PCI) quality programs, including the Blue Cross Blue Shield of Michigan Cardiovascular Consortium (BMC2) and Northern New England Cardiovascular Study Group (NNE). BMC2 and NNE both implement and evaluate quality improvement interventions (e.g. an intervention to reduce vascular complications). Where successful, these interventions can then be disseminated nationally via the NCDR.
The NCDR also promotes quality of care through education in several ways. First, the NCDR Annual Meeting is a conference of NCDR participants with an agenda focused on data quality and quality improvement. Experts in data collection and quality improvement lead interactive workshops and presentations such as ‘How to use your data for QI.’ Second, the American Board of Internal Medicine offers Maintenance of Certification (MOC) credit linked to participation in NCDR registries, as well as participation in the D2B alliance, as part of its performance improvement modules (www.abim.org/pims/choose/specialty/cardiovascular-disease.aspx). With these modules, cardiovascular specialists can use the NCDR to measure their quality of care (e.g. their CathPCI metrics), plan and execute a local performance improvement intervention (e.g. to improve door to balloon time), and document results of the intervention as reflected in their hospital’s NCDR quality metrics (e.g. whether or not their NCDR door to balloon time measure improved). Finally, the ACC as well as other NCDR partners offers a wide range of educational resources, such as those available at www.cardiosource.com for cardiovascular professionals and www.cardiosmart.com for patients.
A key aspect of the NCDR mission is to leverage the clinical registry data for health outcomes research. The Institute of Medicine (IOM) has proposed six domains of high-quality healthcare, namely care that is effective, equitable, efficient, safe, timely, and patient-centered.3 Health outcomes research is focused on the study of these six domains of quality in order to improve healthcare delivery and patient outcomes.7
The NCDR thereby views outcomes research as an essential tool to provide new knowledge on quality of care and the determinants of patient outcomes, learning from the collective practice of NCDR hospitals and practices. Examples of outcomes research from the NCDR best illustrate this point:
- In a study utilizing CathPCI registry data, Kutcher et al. compared hospitals performing PCI with and without on-site cardiothoracic surgery, a topic of considerable debate with regard to quality of care and patient outcomes, yet lacking in evidence.8 This study found that PCI centers with off-site cardiac surgery had similar procedure success, complications, emergency cardiac surgery rates, and risk-adjusted mortality. These data support the safety of PCI at facilities without cardiac surgery but only at facilities where “rigorous clinical, operator, and institutional criteria are in place and where data are submitted and reviewed in a comprehensive multicenter registry such as NCDR.” This study therefore relates to the effectiveness and safety domains of quality as described by the IOM.
- In a study utilizing ICD Registry data, Curtis et al. compared the outcomes of patients who received an ICD implantation by an electrophysiologist or non-electrophysiologist.9 They found that non-electrophysiologists currently implant about 30% of ICDs in the US; however, implantation by a non-electrophysiologist was associated with a significantly higher risk for peri-procedural complications. Moreover, these patients were less likely to receive cardiac resynchronization therapy when it was indicated. It is very unlikely that a question such as this would be evaluated in a randomized clinical trial, and thus these data provide critical evidence regarding the importance of sub-specialty certification and the IOM quality domains of effectiveness and safety with regard to ICD implantation.
- In a study utilizing data from the ACTION Registry, Diercks et al. evaluated the proportion of patients with ST-segment elevation MI who receive a pre-hospital electrocardiogram (ECG), as recommended in a Scientific Statement from the AHA.10 They found that just over one-quarter of patients transported by emergency medical services receive a pre-hospital ECG, yet patients who did receive a pre-hospital ECG were more likely to receive reperfusion therapy and had shorter door to reperfusion times, and there was a trend toward lower in-hospital mortality. These findings shed light on the timeliness and effectiveness of a key, modifiable process of care in relation to patient outcomes.
- In a study funded by the US Food and Drug Administration (FDA), a subset of hospitals participating in the CathPCI registry participated in a detailed investigation of bleeding complications associated with various groin closure devices.11 This study revealed that one device, VasoSeal, was associated with a significantly higher rate of vascular complications. The results of this study led to the removal of the device from the market. Thus, this study had a direct impact on patient safety and also addresses the IOM domain of equity, as the complication risk was found to be particularly high in women.
Looking ahead, the role of NCDR in quality improvement will significantly expand, paralleling advances in health information technology and the rapidly changing healthcare environment in the US (see Table 2). First, point-of-care reporting is currently being implemented, whereby participants can generate instant local reports based on their NCDR data as a complement to their quarterly outcomes reports that include formal regional and national benchmarking. Such reporting capability enhances local quality assessment and improvement efforts. Moreover, NCDR is moving to offer clinical decision support tools in several registry programs. An example is a pre-catheterization risk estimator for vascular complications, based on the validated NCDR bleeding risk model, which will also suggest strategies to reduce bleeding risk for patients at elevated risk.
As electronic health records become more widespread, data collection for NCDR will become more and more ‘transactional,’ whereby clinical information documented in electronic records will automatically populate clinical registry programs. Thus, as clinicians document their care, they will have the necessary quality assessment and improvement tools based on their own ‘data collection.’
NCDR will be increasingly linked to performance improvement measurement: for example, continuing medical education is increasingly focused on the measurement of performance improvement (so-called ‘PI-CME’). The ACC is currently initiating PI-CME projects, whereby measurement of performance improvement is accomplished via NCDR data. Cardiovascular practitioners will increasingly be able to utilize participation in NCDR to document performance improvement for PI-CME and maintenance of board certification.
NCDR will increasingly serve as a ‘broker’ of quality of care data for cardiovascular professionals: for example, it is anticipated that hospitals and practices will increasingly want to leverage NCDR data to satisfy regulatory and other reporting requirements, such as for ‘pay for quality’ programs. It is hoped that the ACC and its partners in the NCDR can effectively work with entities such as government and payers to reinforce the professional data standards, guidelines, and quality metrics for cardiovascular practice that should be used to evaluate and reward quality healthcare.
It is expected that partnerships will continue to grow across NCDR programs to include additional professional organizations, consumer groups, business groups, government agencies, and/or other entities. This will make NCDR even more ‘multistakeholder,’ thus reinforcing a unified mission of the use of clinical registries to measure and improve quality of care, serving patients, the cardiovascular profession, and society.
NCDR is moving to expand the frontiers of research that can directly inform improved care: for example, the NCDR is embarking on collaborative research with the Society of Thoracic Surgeons (STS) to support comparative effectiveness studies in relation to modes of coronary revascularization. As another example, NCDR is working in partnership with the Duke Clinical Research Institute to build the infrastructure necessary to conduct clinical trials among interested NCDR participant sites. It is hoped that such clinical trials will be more efficient than de novo clinical trials because of the ability to recruit sites and utilize the clinical registry data already being collected, and because post-trial surveillance can occur naturally via the existing clinical registry.
The NCDR is also becoming international: the NCDR now has active collaboration with hospitals, clinicians, and societies from more than 15 countries. Hospitals outside the US are joining NCDR, with a key goal being the ability for them to benchmark their care against US care. Moreover, many international societies want to adopt the NCDR data elements and definitions for their clinical registries to support movement toward international standardization and the ability to collaborate on improving global quality of care.
Finally, NCDR registries will increasingly capture longitudinal care and patient outcomes, moving toward a model of true ‘patient-centered’ registries. It is critical that quality of care and patient outcomes be evaluated over time, rather than just for single episodes of care or procedures. Thus, with full attention to data security and patient privacy, NCDR is moving to capture longitudinal care and patient outcomes. This will also support longitudinal surveillance (device and medication) to enhance patient safety. The future may also include the integration of data from patient health records into NCDR registry programs, which could then allow the realization of a truly patient-centric health registry, integrating data from both episodes of care (procedures, clinic visits, hospitalizations) and the critical health information from patients between episodes of care (e.g. symptom and functional status, medication use, and measurements such as home blood pressures).
The NCDR has undergone tremendous growth from its inception just 10 years ago to the current array of national registry programs. The mission of NCDR is to improve the quality of cardiovascular patient care by providing information, knowledge, and tools, implementing quality initiatives, and supporting research that improves patient care and outcomes. Such coupling—using clinical data for benchmarking against current professional standards, linking measurement to performance improvement (and, increasingly, performance reporting in the current healthcare environment), and backed by the resources of NCDR and its partnering organizations—has significant potential to fill gaps in care quality, reduce inappropriate variation, and reduce medical errors. This is achieved as a partnership with clinicians and hospitals, whereby they are engaged and effective agents in promoting quality improvement, rather than simply being ‘judged’ by external entities. As such, the NCDR has the potential to help the US cross the ‘quality chasm’ and achieve healthcare that is effective, equitable, efficient, safe, timely, and patient-centered.