Current growth in health care costs is unsustainable. The ability to define real-world “effectiveness” of health care delivery is therefore of vital importance. To bend the cost curve and ultimately achieve sustainability in health care, medical providers and surgical treatments of the highest quality and effectiveness must be preferentially used and purchased. As the current US health care environment continues to evolve, it will be essential for all spine surgeons to understand and be facile with the principles of evidence-based health care reform, value-based care, and how registries may power these efforts.
The Institute of Medicine defines health care quality as follows: “The degree to which health care services for individuals and populations increase the likelihood of desired health outcomes.”1 Safety and effectiveness (quality), patient-centered outcomes, as well as cost of care are each central components of the health care value equation (benefit/cost). Both sides of the value equation are of critical importance. Health care reform may simply reward cheaper care (cost-based purchasing) rather than more valuable care (value-based purchasing), if we as providers do not demonstrate “efficacy.” Value-based purchasing, in which providers are held accountable for both the quality and cost of health care services, will soon dominate most incentive models in an increasingly resource-restricted medical economy.
The Institute of Medicine, Agency for Healthcare Research and Quality, and the Patient Protection and Affordable Care Act call for prospective, well-designed registries for the measurement of real-world effectiveness to facilitate evidence-based health care reform. Prospective registries, if structured correctly, have the potential to provide the necessary statistical power and real-world health delivery setting required for true value measurement in both individuals and populations. Furthermore, measuring and understanding patient-centered effectiveness (facilitated by registry formats) is of vital importance for comparative effectiveness research, and is a central pillar of practice-based learning and quality improvement.
The aims of this article are to (1) describe methodological considerations of spine surgery registries, (2) critically examine the trade-offs of the wide variety of registry designs, (3) review existing spine surgery registries, and (4) briefly describe the Vanderbilt Prospective Spine Registry (VSR) as a case example of a registry effort that the authors feel optimizes patient centeredness, nonconfounded methodology, quality control, and clinical data accuracy.
METHODOLOGICAL CONSIDERATIONS OF SPINE SURGERY REGISTRIES
It is often assumed that group-mean efficacy observed in tightly focused randomized clinical trials (RCTs) reflect real-world effectiveness. Although high internal validity of RCTs ensures data quality, strict inclusion/exclusion criteria limit external validity and therefore the ability to generalize the conclusions of these analyses. Results from RCTs are often (inappropriately) applied to patients not represented by research cohorts, in health care settings not represented by research settings, and by surgeons not represented by study surgeons. Moreover, there are significant costs and complexity in ongoing RCTs, as well as, ethical issues related to randomization. In contrast, patient registries are cost-effective, easily scaled, document care in real-world environments, and are increasingly supported by numerous health care stakeholders. Some observers have gone so far to state that registries and registry trials may represent “the next disruptive technology” in clinical research.2
Well-constructed prospective registries with appropriate a priori design can produce level 1 prognostic evidence (predictive analysis), and level 2 evidence on effectiveness of care. Characteristics of quality registries are outlined in Table 1. A high-quality registry possesses many of the same characteristics as an RCT, including: prospective data collection (as opposed to retrospectively retrieving data from administrative data sets), clinical (nonclaims) data point entry, patient-reported outcomes (and not process measures), high follow-up rates, quality control and data validation, and study designs that control for confounding biases.
The term “registry” is currently used to describe a wide variety of centralized data-collection efforts centered on health care treatments, safety, and outcomes. The vast majority of these represent efforts to track hospital or payer/purchaser claims and billing records. The advantage of these data sets is the feasibility and low cost of collection. Claims-based systems can provide large-scale population assessments as well as more granular cost information. However, claims are notoriously inaccurate when serving as proxies for clinical end points of safety or effectiveness of care. Another common large repository of data is centered on hospital-based quality programs that track compliance with a set of mandated standards. Although such process measures can be helpful at incentivizing hospitals and providers to adopt best care approaches, they do not provide high-level evidence in the analysis of registry data. Hence, the Institute of Medicine, Agency for Healthcare Research and Quality, and Patient-Centered Outcomes Research Institute have called for evolution of well-designed, prospective, clinical-variable, patient-powered longitudinal outcome registries to generate evidence that will guide health care reform. There are many processes, methodological, and end point measurement factors that are important to consider when critically examining evidence generated from these emerging spine surgery registries.
First, the enrollment process needs to be transparent and unbiased. Total care enrollment, or randomized enrollment, results in the least confounding; however, costs and resources can preclude scalability. Representative sampling as used by National Surgical Quality Improvement Program or the National Neurosurgery Quality and Outcomes Database (N2QOD) is a practical solution and can minimize selection bias. The need for a written consent to participate in a registry has come under debate given the information is used for quality improvement.3 A verbal consent has been recognized as adequate for this purpose with submission of an institutional review board approval for specific hypothesis driven questions. First, once a patient has provided verbal consent, a clear definition of enrollment eligibility and appropriate grouping of patients by pathology and treatment, are necessary to prevent inaccuracies. Second, the incidence of missing covariate data and follow-up rates of major end points should be transparent and minimized. Missing data for baseline variables of disease, risk factors, or treatments reduces the reliability of registry analysis. The greatest challenge to registries is long-term follow-up. Given these challenges the N2QOD and Swedish Spine are currently the only multicenter clinical spine surgery registries collecting prospective, longitudinal patient-reported outcomes with a 12-month follow-up rate more than 75%. Finally, publicly transparent quality control mechanisms are of vital importance. This can be accomplished through weekly missing data feedback reports, site visits on data accuracy validation, and site education of standardized operating procedures or case definitions.
A significant hurdle that must be overcome in nationwide implementation of a high-quality registry is the significant costs incurred to maintain data-collection infrastructure and perform routine quality assurance. The N2QOD registry, established by the American Association of Neurological Surgeons, requires a full-time employee data extractor and an annual subscription fee. For the Vanderbilt Prospective Spine Registry, the cost per patient is approximately $160 in total and is paid for in combination through institutional and grant funding. National Surgical Quality Improvement Program subscription fees and patient enrollment create a cost that is slightly higher. To provide perspective, the cost per enrolled patient in RCTs can exceed $6000.4 Registries like N2QOD are proving that high-quality patient-reported outcomes data can be collected on a large scale at a fraction of the cost of traditional RCTs. Nevertheless, hospital systems and surgeon groups with limited resources, may initially face challenges with participating in high-quality longitudinal patient-reported outcome registries.
The VSR is highlighted below as a registry that adheres to total case enrollment, avoids the use of process measure or claims-data proxies, is prospective with an a priori design for prognostic analysis, uses internal case auditing and missing data surveillance, measures patient outcome end points, and spans a full 12-month period from index procedure. It also addresses the perspective of multiple stakeholders by measuring return-to-work time, direct and indirect health care costs, and patient quality of life and satisfaction.
Summary of Existing Spine Surgery Registries
The following review of available prospective clinical spine surgery registries (nonclaims based), is meant to give the reader a reference of registry resources, but is not meant as a comprehensive critical review. Furthermore, we did not focus on administrative claims data sets, nonspine registries, or other retrospective initiatives. We were able to obtain detailed information on 13 existing spine surgery registries through various internet-based resources. Details of this analysis are presented in Tables 2 and 3. All the 13 registries prospectively collect patient information (several supplement this with retrospective data). All registries were continuing to accumulate data as of January 2014. Of the 13, 3 registries had start dates before 2000, 3 between 2001 and 2005, 5 starting in 2006, and 3 were indeterminate. Registry populations were in the range from 300 patients (Scolisoft Registry) to as large as 75,000 patients (SweSpine Registry). We were unable to identify the total number of patients for 3 of the registries. Patient follow-up periods were in the range from 12 months (N2QOD) to 5 years (Rick Hansen and SweSpine Registry) and follow-up rates were in the range from 22% to 79%. Longer follow-up times consistently produced lower follow-up rates.
Most registries focused on regional or national data collection; however, 3 collected patient information internationally. We were unable to identify the data sources for 3 of the registries. Most of the registries reviewed required the attending physician or hospital assistants to input the patient information; however, several allowed patients to submit their own information and some used a combination of both methods. Seven of the registries reported baseline patient demographic information. Of the 13 registries, relevant outcome measures included Oswestry Disability Index, pain scores, EuroQol-5D, complications, and 2 registries reported radiographical data. Only 3 registries reported some level of quality assurance to ensure data quality and integrity (N2QOD, Vanderbilt, and Kaiser Permanente Registries). The N2QOD and Swedish Spine are currently the only multicenter clinical spine surgery registries collecting prospective, longitudinal patient-reported outcomes with a 12-month follow-up rate more than 75%.
The Vanderbilt Prospective Spine Registry (Case Example)
The authors highlight the VSR solely as a case example of meaningful data use. This prospective registry is on the very granular end of the data-collection spectrum, which may not be feasible with limited resources. Providers from the departments of neurosurgery, orthopedic surgery, physical medicine and rehabilitation, and pain management comprise the multidisciplinary spine care center of the VSR.
All patients undergoing elective spine surgery at the Vanderbilt Spine Center are enrolled into a prospective, longitudinal registry. Outcome assessments are administered to all patients without a change in standard of care or procedures delivered. As such, the Vanderbilt University Institutional Review Board deemed this effort as quality improvement with recognition that aggregate data could power research. Because the registry data process and storage was found to be of minimal risk to patients, institutional review board granted a waiver of written informed consent for participation in the registry, but required verbal consent via a standardized phone script during baseline patient interviews. Once a surgical case is scheduled, it is also posted in the centralized Microsoft Outlook calendar of the Spine Registry. Automated e-mail notifications are subsequently sent to registry associates at the appropriate times for data collection (preoperatively, as well as at 3 and 12 mo postoperatively). All data are collected and managed using REDCap (Research Electronic Data Capture) web-based data capture tools hosted at Vanderbilt University.5 Prospectively collected variables include patient demographics, comorbidities, outcome metrics, surgical variables, and morbidity data. Variables collected are listed in Table 4.6–8 A subset of patients who were considered surgical candidates but chose to undergo continued medical management, were also enrolled into the registry. In addition, all patient-reported health care resource utilization during the follow-up period is collected to estimate direct health care costs. At each follow-up, productivity losses due to spine-related problem (i.e., missed work days for those employed outside of the home and missed homemaking days for those who reported housekeeping as their primary activity) are recorded. Unpaid caregiver opportunity costs are also included. Indirect costs are estimated using the standard human capital approach by multiplying the change in hours worked by the gross-of-tax wage rate based on self-reported wages.9 The VSR requires 3 full-time employees, the use of a nearly free web-based module and data storage system, office space, computer use, and phone expenses. The VSR total cost is roughly $168,000 per year at a cost per patient enrolled of $160. These costs are covered through a combination of institutional and grant support.
Measuring Effectiveness and Value of Care
The Vanderbilt Prospective Spine Registry has enabled us to demonstrate the effectiveness of surgical care provided to our patient population, as demonstrated through validated patient-reported outcome metrics, Figure 1. This information has allowed us to discuss novel contracting strategies with “second opinion” programs and gain preferred provider status with local health plans. Detailed data on the effectiveness of care rendered, especially return-to-work capacity and times, have been helpful in contracting spine services with self-insured employers and third-party payers, Figure 2. In addition to revealing effectiveness of care, validated patient-centered outcomes measurements can be used to demonstrate the value of the spine care being delivered.
Along with the aforementioned effectiveness metrics, collection of utility data for all patients enrolled into the Vanderbilt Prospective Spine Registry has allowed for assessment of the relative value for each of the various spine procedures performed at our institution. Utility scores are obtained from preference-based health state measures such as the EuroQol-5D or Short Form-6D.10,11 The utility scores calculated from these metrics (quality adjusted life years) reflect how a certain health state is valued by a society and our patients. Routine collection of this data allows cost of care to be assessed in the context of patient-centered effectiveness, which is the foundation of value-driven versus cost-driven care. By dividing the quality adjusted life years gained following a given procedure by its 1-year cost (or 2-yr cost), we are able to calculate and compare an absolute value measure (cost/quality adjusted life years gained) for each of the spine treatments provided, (Table 5). Value-based decisions on “best care” are thus facilitated. For example, in patients with structural spine pathology who are surgical candidates (lumbar stenosis, spondylolisthesis, and disc herniation that fails to respond to initial medical management), 1-year direct and indirect costs from medical management are significantly lower than surgery. However, surgical care is a much better “value” compared with medical management because the increased costs of surgical care are more than offset by its increased effectiveness.
Quality Improvement and Practice-Based Learning
A final example of enhanced practice-based learning is the ability to identify patient subgroups least likely to respond favorably to surgical spine care. In our practice, we have seen that increasing preoperative scores on the Zung Self-Rating Depression Scale (suggests greater levels of depression) correlate significantly with decreased surgical effectiveness, Figure 3A, decreased satisfaction with surgery, Figure 3B, and attenuated return-to-work capacity, Figure 3C. As depicted in Figure 4, large patient-reported outcomes registries provide sufficient statistical power to more meaningfully display real-world effectiveness at the individual patient level. Such data provide the substrate to build predictive models to identify best care paradigms and to predict, before care is initiated, which patients have the best or least chance of successful surgical outcomes. Our scatter plots suggest that care may not be optimized in nearly 20% of cases despite evidence-based diagnosis and treatment approaches, Figure 4. This is an example of how registries that are able to measure ineffective care from the patients' perspective can identify health care waste (high cost/no benefit), and empower initiatives to reduce this waste and improve overall value.
Prospective, longitudinal, patient-reported outcomes registries are powerful tools that allow measurement of cost, safety, effectiveness, and health care value across clinically meaningful episodes of care. Registries entirely based on claims or billing data, safety measures alone, process measures or other proxies of outcome, offer valuable insights, but do not provide comprehensive data to drive patient-centered value-based reform. The Patient Protection And Affordable Care Act, along with various value-based health care reform initiatives, have established prospective, patient-centered, clinical data registries as a preferred method to power large-scale evidence-based reform at the disease and patient level. Prospective registries should include the following characteristics: clear definition of eligibility for enrollment and careful grouping of patients by disease and treatment, longitudinal collection of patient-reported outcomes, missing covariate data and follow-up rates of major end points should be transparent and minimized, and quality control measures must be in place to ensure data validity. Although the cost of high-quality, patient-centered clinical registry data remain significantly less than RCT data, the infrastructural costs remain a barrier for universal adoption among health care systems. To our knowledge, the N2QOD and Swedish Spine are currently the only multicenter clinical spine surgery registries collecting prospective, longitudinal patient-reported outcomes with a 12-month follow-up rate more than 75%. As more spine-focused registries emerge and their integration into the US and international health care delivery evolve, the evidence to power value-based reform will be enabled.
- To bend the cost curve and ultimately achieve sustainability in health care, medical providers and surgical treatments of the highest quality and effectiveness must be preferentially used and purchased.
- Value-based purchasing, in which providers are held accountable for both the quality and cost of health care services, will soon dominate most incentive models in an increasingly resource-restricted medical economy.
- The vast majority of “high level” medical evidence (i.e., tightly focused randomized clinical trials) has not accounted for important variations in individual patient characteristics, disease states, or health care delivery settings.
- Prospective, longitudinal, patient-reported outcomes registries are powerful tools that allow measurement of cost, safety, effectiveness, and health care value across clinically meaningful episodes of care.
1. Lohr KN, Schroeder SA. A strategy for quality assurance in Medicare. N Engl J Med 1990;322:707–12.
2. Lauer MS, D'Agostino RB Sr. The randomized registry
trial—the next disruptive technology in clinical research? N Engl J Med 2013;369:1579–81.
3. McGirt MJ, Speroff T, Dittus RS, et al. The National Neurosurgery Quality and Outcomes Database (N2
QOD): general overview and pilot-year project description. Neurosurg Focus 2013;34:E6.
4. Emanuel EJ, Schnipper LE, Kamin DY, et al. The costs of conducting clinical research. J Clin Oncol 2003;21:4145–50.
5. Harris PA, Taylor R, Thielke R, et al. Research electronic data capture (REDCap)—a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform 2009;42:377–81.
6. Badia X, Diaz-Prieto A, Gorriz MT, et al. Using the EuroQol-5D to measure changes in quality of life 12 months after discharge from an intensive care unit. Intensive Care Med 2001;27:1901–7.
7. Jansson KA, Nemeth G, Granath F, et al. Health-related quality of life (EQ-5D) before and one year after surgery for lumbar spinal stenosis. J Bone Jt Surg Br 2009;91:210–6.
8. Johnson JA, Coons SJ, Ergo A, et al. Valuation of EuroQOL (EQ-5D) health states in an adult US sample. Pharmacoeconomics 1998;13:421–33.
9. Hodgson TA, Meiners MR. Cost-of-illness methodology: a guide to current practices and procedures. Milbank Mem Fund Q Health Soc 1982;60:429–62.
10. Brazier J, Usherwood T, Harper R, et al. Deriving a preference-based single index from the UK SF-36 Health Survey. J Clin Epidemiol 1998;51:1115–28.
11. EuroQol Group. EuroQol—a new facility for the measurement of health-related quality of life. Health Policy 1990;16:199–208.
12. McGirt MJ, Speroff T, Godil SS, et al. Outcome science in practice: an overview and initial experience at the Vanderbilt Spine
Center. Neurosurg Focus 2013;34:E7.