Health-care costs in the United States have increased substantially over time. From 1980 to 2007, the percentage of gross national product spent on health care has increased from 8.8% to 16%1. Total joint replacement is one of the most costly diagnosis-related groups, with >600,000 procedures performed each year in the United States2,3. The demand for total joint replacement is expected to increase in the United States as a result of advances in medical technology, an increased prevalence of obesity, and an increasing aging population2,4. By 2030, annual volumes are projected to increase by 673% for primary total knee arthroplasty and by 174% for primary total hip arthroplasty4. The demand for revision total knee and total hip arthroplasty is also projected to increase by 601% and 137%, respectively. Total knee and total hip arthroplasty costs are also expected to increase dramatically, with annual hospital charges estimated to reach $40.8 billion for primary total knee arthroplasty and $17.4 billion for primary total knee arthroplasty by 20155. Similar increases in cost are projected for revision total knee and total hip arthroplasty. Wilson et al. estimated that by 2030, total knee arthroplasty and total hip arthroplasty will cost Medicare over $50 billion6.
In addition to increases in demand and cost, recent concerns about metal-on-metal bearing surfaces7 and recent implant recalls and advisories have emphasized the need to monitor total joint outcomes nationwide. The identification of procedures and implants associated with higher revision rates could prevent revision procedures, improving care and addressing the increased cost and demand associated with this procedure. Joint registries provide one potential solution for reducing total joint replacement implant variation and revision rates. The Swedish Hip Register8-11 has demonstrated the effectiveness of national registries in changing clinical practice. Feedback from the registry on cement techniques resulted in changes in practices and a reduction in revision rates. Sweden's revision burden (percent revision of total cases) is half that of the estimated revision rate in the United States. The Swedish registry not only has reduced total hip arthroplasty revision rates but also has reduced implant variation, with six implants being used for 70% of total hip arthroplasty procedures8,9.
While other countries such as Sweden8, Australia12, Denmark13, Germany14, Canada15, and Norway16 have established national registries to track and monitor devices, the United States currently does not have a mechanism for post-market surveillance of orthopaedic devices. Within the United States, registries have been limited to institutional or regional geographical efforts such as the Mayo Clinic17 or the HealthEast registry18. Recently, the American Academy of Orthopaedic Surgeons (AAOS), Agency for Healthcare Research and Quality (AHRQ), and United States Food and Drug Administration (FDA) have explored mechanisms for the development of a national United States total joint replacement registry. A national United States total joint replacement registry could potentially reduce total joint replacement revision rates through the early identification of implant failures and by providing feedback to surgeons on techniques and implants associated with higher revision rates.
The Total Joint Replacement Registry
Recognizing the benefits of implant registries for our 8 million patients in eight geographical regions, the surgeons in our integrated health-care delivery system developed a total joint replacement registry. The total joint replacement registry was developed in 2001 and was modeled after the Swedish registry8,11, with a focus on a minimal dataset and revision surgery as the end point. The goals of the total joint replacement registry include (1) monitoring revisions, reoperations, and complications such as surgical site infection, pulmonary embolism, and deep-vein thrombosis, (2) immediately identifying and notifying patients and surgeons during implant recalls or advisories, (3) identifying patients at risk for revisions and complications, (4) assessing the clinical effectiveness of implants and techniques, and (5) providing the framework for more in-depth clinical research studies19-25. The total joint replacement registry has been implemented at fifty hospitals in the six geographical regions in which we own our own hospitals, with 90% voluntary participation from >350 orthopaedic surgeons, 80% of whom are fellowship-trained. As of March 2010, the total joint replacement registry had >100,000 total joint replacement cases registered, with 17,000 cases added each year.
Anterior Cruciate Ligament Reconstruction Registry
On the basis of the success of the total joint replacement registry, our surgeons also implemented an anterior cruciate ligament reconstruction registry in 2005, which tracks surgical procedures, techniques, graft types, fixation types, and implants. The goals of this registry include (1) identifying risk factors that lead to degenerative joint disease, graft failure, meniscal failure, and failure to return to sports, (2) evaluating outcomes associated with different graft types and fixation techniques, (3) describing the epidemiology of anterior cruciate ligament reconstruction procedures and patients, (4) monitoring complications, (5) assessing and comparing procedure incidence rates, and (6) constructing a framework for future studies tracking anterior cruciate ligament reconstruction outcomes. Similar to the total joint replacement registry, the anterior cruciate ligament reconstruction registry is implemented in six regions, with thirty-four sites and 208 different surgeons participating. Although the anterior cruciate ligament reconstruction uses a similar methodology, this registry is modular, with collection of patient-reported outcomes at specific centers. As of March 2010, 9200 cases were registered, with 2500 new cases added each year.
The purposes of the present article are (1) to present total joint replacement demographics, survival, and revision risk factors from a large community-based practice; (2) to highlight the impact of the total joint replacement registry on clinical practice, patient safety, cost effectiveness, and research within our integrated system; and (3) to present anterior cruciate ligament reconstruction demographics, survival, and reoperation and complication rates from the same community-based practice.
Data Source
The present study includes total joint replacement registry data from April 2001 to March 2008 and anterior cruciate ligament reconstruction data from February 2005 to June 2008. Registry data are collected prospectively through standardized documentation at the point of care (i.e., during preoperative, intraoperative, and all postoperative encounters). The standardized forms collect information on demographic characteristics, implant characteristics, surgical techniques, and outcomes (e.g., revisions, reoperations, infections, deep-vein thrombosis, and pulmonary embolism). Registry forms are supplemented with existing administrative data from our electronic health records and other independent databases (Fig. 1). Electronic screening algorithms are applied to administrative databases to detect additional complications, reoperations, and revisions. Independent electronic health record operative modules and anesthesia databases are used to validate registry cases. All registry outcomes are validated with use of a chart review methodology following CDC (Centers for Disease Control and Prevention) and AHRQ guidelines26-28.
Data Elements
Patient Outcomes
The main outcome of interest for the total joint replacement study was revision surgery, defined as removal or exchange of at least one prosthetic component. Aseptic revision was defined as revision for any reason, excluding infection. For the anterior cruciate ligament reconstruction study, reoperation, revision of ligament reconstruction, meniscal injury, infection, deep-vein thrombosis, and pulmonary embolism were the main outcome variables. Infections were defined according to the CDC guidelines27, and deep-vein thrombosis and pulmonary embolism were defined according to AHRQ guidelines26.
Patient Characteristics
The patient characteristics that were assessed in relation to total joint replacement survival included age (less than fifty-five years of age versus fifty-five years of age or more), American Society of Anesthesiologists (ASA) score (<3 versus ≥3), body-mass index, diabetes status, primary diagnosis (osteoarthritis versus other), and sex. In the anterior cruciate ligament reconstruction study, sex, age, and associated concurrent cartilage and meniscal injuries were the main patient characteristics evaluated in relation to study outcomes.
Hospital and Surgeon-Related Variables
The main hospital and surgeon-related variables assessed in the total joint replacement study were annual hospital case volume, geographical region, and surgeon annual case volume. For total knee arthroplasty, annual hospital volume was defined as small (fewer than fifty cases), medium (fifty to ninety-nine cases), and large (100 cases or more). For total hip arthroplasty, annual surgeon volume was defined as small (fewer than ten cases), medium (ten to forty-nine cases), and large (fifty cases or more). These variables were not evaluated in the anterior cruciate ligament reconstruction study.
Implant Characteristics
The total knee arthroplasty implant variables consisted of cement fixation and platform design (mobile versus fixed bearing, cruciate retained versus posterior stabilized). Total hip arthroplasty implant characteristics included cement fixation, femoral head size (≤28 mm, 32 mm, and ≥36 mm), and bearing surfaces (conventional polyethylene, metal-on-metal surfaces, and highly cross-linked polyethylene). The anterior cruciate ligament reconstruction study assessed the type of graft implanted (autograft versus allograft, hamstring versus patellar tendon autograft).
Analyses
Descriptive statistics such as means, standard deviations, and proportions were used to describe the study samples. Chi-square, Fisher exact, and independent t tests were applied to evaluate group differences in demographic characteristics. Kaplan-Meier survival curves with revision as the end point and log-rank tests were used to evaluate implant survival. Death and membership termination were used to censor data in the survival analyses. Multivariate Cox regression models were used to assess total joint replacement relative risk of revision. SAS (Version 9.1.3; SAS Institute, Cary, North Carolina) was used to analyze the data, with p < 0.05 as the statistical threshold.
Source of Funding
The registries are funded through the health plans in each region. There was no external funding for these studies.
Total Knee Arthroplasty
Patient Demographics
As of March 31, 2008, there were 39,286 primary (94.1%) and 2458 revision (5.9%) total knee arthroplasty cases registered. Females accounted for 62.5% of the primary and 53.5% of the revision total knee arthroplasty procedures. Patients younger than sixty-five years of age accounted for 37.3% of primary and 38.4% of revision cases. Osteoarthritis was the most common diagnosis among patients undergoing primary total knee arthroplasty (96.6%) (see Appendix). The prevalence of diabetes in this population was 20.2% among patients undergoing primary total knee arthroplasty and 25.5% among those undergoing revision total knee arthroplasty. Most surgeons used the parapatellar approach (75%) and cement fixation (84%) when performing primary total knee arthroplasty procedures. Patients undergoing revision total knee arthroplasty had higher ASA scores than those undergoing primary total knee arthroplasty (p < 0.001). Infection (26.2%) and aseptic loosening (24.3%) were the most common diagnoses among patients undergoing revision total knee arthroplasty (see Appendix).
Kaplan-Meier Survival Curves for Total Knee Arthroplasty
Of the 39,286 primary total knee arthroplasties in the registry during this time period, 667 (1.7%) were revised, with 281 revisions due to infection (0.7%). During the seven-year study, 3213 patients (8.2%) left the health plan membership, and 1217 (3.1%) died. Patients who died and those who left the health plan membership were considered to be lost to follow-up (N = 4430; 11.3%) and were censored in the survival analysis. The cumulative survival rate at 5.5 years was 97.1% (95% confidence interval, 96.8% to 97.4%). Infection accounted for 42.6% of the primary total knee arthroplasty revisions. Instability was the reason for revision in 16.2% of the cases. The cumulative survival rate at 5.5 years was 97.1% (95% confidence interval, 96.8% to 97.4%). Kaplan-Meier survival curves and log-rank tests indicated lower total joint replacement survival function for the following variables: an age of less than fifty-five years (p < 0.001), an ASA score of ≥3 (p < 0.001), a higher body-mass index (p < 0.001), comorbid diabetes (p < 0.001), diagnoses other than osteoarthritis (p < 0.001), male sex (p < 0.001), uncemented fixation (p < 0.0001), and mobile-bearing platforms (p < 0.001) (Fig. 2) (see Appendix).
Risk Factors for Total Knee Arthroplasty Revision
Variables that were included in the multivariate Cox regression model included age, ASA score, sex, body-mass index, primary diagnosis, diabetes, surgeon and hospital annual case volumes, geographical region, fixation technique, and type of platform (mobile bearing versus fixed, cruciate retaining versus posterior stabilized). After adjustment for all variables, the risk factors for total knee arthroplasty revision included an age of less than fifty-five years (relative risk = 2.56; 95% confidence interval, 1.97 to 3.33; p < 0.001), an ASA score of ≥3 (relative risk = 1.29; 95% confidence interval, 1.04 to 1.60; p = 0.02), diabetes (relative risk = 1.28; 95% confidence interval, 1.01 to 1.62; p = 0.04), implant fixation without cement (relative risk = 1.92; 95% confidence interval, 1.23 to 3.3; p = 0.004), and a mobile-bearing platform (relative risk = 1.52; 95% confidence interval, 1.13 to 2.05; p = 0.006) (Fig. 3).
Total Hip Arthoplasty
Demographic Characteristics
As of March 31, 2008, there were 21,548 (88.5%) primary and 2809 (11.5%) revision total hip arthroplasty cases registered. Females accounted for 56.1% of the primary and 55.4% of the revision cases. Patients younger than sixty-five years of age accounted for 44.7% of primary and 38.5% of revision cases. The prevalence of diabetes was 8.7% among patients undergoing primary total hip arthroplasty and 7.9% among those undergoing revision total hip arthroplasty. Osteoarthritis was the most common diagnosis among patients undergoing primary total hip arthroplasty (89.7%), whereas aseptic loosening (28.3%) and instability (31.1%) were the most common diagnoses among those undergoing revision total hip arthroplasty (see Appendix). Patients undergoing revision total hip arthroplasty had higher ASA scores than those undergoing primary total hip arthroplasty (p < 0.001); 38.6% of those undergoing a revision procedure had an ASA score ≥3, compared with 31.9% of those undergoing a primary procedure.
Implants
Metal-on-highly cross-linked polyethylene accounted for the majority (52%) of the primary total hip arthroplasty bearing surfaces. Metal-on-metal and ceramic-on-ceramic accounted for 8.4% and 1.9% of the primary bearing surfaces, respectively. The use of femoral head sizes of ≥36 mm increased over the years, from 4.8% in 2002 to 54.4% in 2008. Metal femoral heads were used in the majority of cases for both primary (73.9%) and revision (69.6%) procedures. For primary procedures, cementless fixation (76.2%) was used most commonly, with hybrid fixation being used for an additional 15.9%.
Kaplan-Meier Survival Curves for Total Hip Arthroplasty
Of the 21,550 primary total hip arthroplasties in the registry during this seven-year study, 409 (1.9%) were revised, with seventy-eight implants (0.4%) being revised because of infection. The most common reasons for revision after primary total hip arthroplasty cases were instability (49.1%) and infection (19.6%). Within the study period, 1923 (8.9%) patients left the health plan membership, and 876 (4.1%) died. Patients who died and those who left the health plan membership were considered to be lost to follow-up (N = 2799; 12.9%) and were censored. The cumulative survival rate at 5.5 years was 97.3% (95% confidence interval, 96.8% to 98.2%).
Kaplan-Meier survival curves and log-rank tests indicated differences in survival function according to region (p = 0.02), annual surgeon volume (p = 0.02), insert type (p < 0.001), femoral head size (p < 0.001), and bearing surface (p < 0.01) (Figs. 4 and 5). No differences in survival function were observed by fixation type (p = 0.0758) (Fig. 6).
Risk Factors for Total Hip Arthroplasty Revision
Variables included in the multivariate Cox regression models included age, ASA score, sex, body-mass index, primary diagnosis, diabetes, surgeon and hospital annual case volumes, geographical regions, femoral head size, implant fixation, and bearing surface. After adjustment for all variables, the following risks factors for total hip arthroplasty revision were identified: female sex (relative risk = 1.29; 95% confidence interval, = 1.02 to 1.62; p = 0.03), geographical region (relative risk = 1.33; 95% confidence interval = 1.04 to 1.70; p = 0.02), a surgeon annual volume of fewer than thirty cases (relative risk = 1.28; 95% confidence interval = 1.02 to 1.61; p = 0.03), conventional polyethylene insert (relative risk = 1.78; 95% confidence interval = 1.32 to 2.39; p < 0.001), and a femoral head size of ≤28 mm (relative risk = 1.58; 95% confidence interval = 1.06 to 2.35; p = 0.03). Hybrid implant fixation was found to be a protective variable for total hip arthroplasty revision (relative risk = 0.71; 95% confidence interval = 0.52 to 0.97; p = 0.03) (Fig. 7).
Anterior Cruciate Ligament Reconstruction
Demographic Characteristics
As of June 2008, 4025 primary (95.2%) and 205 revision (4.8%) cases were registered. Males (mean age [and standard deviation], 30 ± 10 years) accounted for the majority (65.1%) of the primary cases, and females (mean age, 28 ± 12 years) represented 34.9%. When we examined epidemiology trends from 2001 to 2005, we observed that the incidence of anterior cruciate ligament reconstruction increased for males who were fourteen to seventeen, twenty-six to twenty-nine, and forty to forty-nine years old and increased for females who were fourteen to twenty-one years old, with the most dramatic increase for females who were fourteen to seventeen years old29. Our anterior cruciate ligament reconstruction registry demonstrated that fourteen to seventeen-year-old girls accounted for the highest percentage of females undergoing anterior cruciate ligament reconstruction (N = 419; 30.6%), whereas males had a more even distribution between age groups (see Appendix).
The majority (62.3%) of patients undergoing primary anterior cruciate ligament reconstruction had a meniscal injury, with 41.5% having an injury of the medial meniscus only, 33.6% having lateral meniscal injury only, and 25% having an injury of both menisci (see Appendix). Additionally, chondral injuries were found in 20.8% of the cases. Among patients with a meniscal injury who had a primary procedure, 33.1% had a meniscal repair (including 39% of those with a medial meniscal injury and 25% of those with a lateral meniscal injury), 51.7% had a partial meniscectomy, and 13.7% had the meniscus left in situ. Among patients who had a revision procedure, 29.8% had a meniscal repair, 61.5% had a partial meniscectomy, and 7.7% had the meniscus left in situ. The most common meniscal fixation was a third-generation all-inside suture-based repair system. Males (mean age, 31 ± 10 years) accounted for 55.4% of the revision cases, and females (mean age, 28 ± 12 years) accounted for 44.6%.
Graft Types for Anterior Cruciate Ligament Reconstruction
Autografts were used in 61.2% of the primary procedures, whereas allografts were used in 83.3% of the revision procedures. Of the primary procedures, 35.4% involved bone-patellar tendon-bone grafts, 25.2% involved hamstrings grafts, and 36.4% involved allografts (see Appendix). Patellar tendon autografts were used slightly more frequently in males (37.2%) as compared with females (31.8%). Of the procedures involving allograft, 29.5% involved tibialis anterior tendon graft, 23.5% involved Achilles tendon graft, 19.3% involved bone-patellar tendon-bone graft, and 19.0% involved tibialis posterior tendon graft.
Complication Rates for Anterior Cruciate Ligament Reconstruction
The overall rate of surgical site infection for the 4230 patients undergoing anterior cruciate ligament reconstruction was 0.7% (N = 29), with six deep infections (0.14%) and twenty-three superficial infections (0.54%). Within ninety days postoperatively, there were five cases of deep-vein thrombosis (0.1%) and five cases of pulmonary embolism (0.1%) (Fig. 8).
Revision and Reoperation Rates for Anterior Cruciate Ligament Reconstruction
Of the 4230 patients in the registry who were managed with primary or revision anterior cruciate ligament reconstruction, 12% left the health-plan membership and were considered lost to follow-up. Of the 4025 patients in the registry who were managed with primary anterior cruciate ligament reconstruction, 178 (4.4%) had a reoperation and twenty-four (0.6%) had revision ligament surgery. The three most common reasons for reoperation on the index (ipsilateral) knee were meniscal or cartilage injury (34.3%), stiffness (18.9%), and infection (14%) (see Appendix). Operations on the contralateral knee were due to anterior cruciate ligament reconstruction or other ligament surgery in 65.7% of cases and for meniscal injuries in 28.6% of the cases (see Appendix). For females, 90.0% of the contralateral procedures were performed for reconstruction of the anterior cruciate ligament or other ligaments and 10.0% were performed because of meniscal injuries. For males, 56% of the contralateral procedures were performed for reconstruction of the anterior cruciate ligament and 36.0% were performed because of meniscal injuries. The cumulative survival rate for anterior cruciate ligament reconstruction at four years was 90.7% (95% confidence interval, 86.9% to 93.4%). Kaplan-Meier survival curves and log-rank tests indicated differences in survival function following procedures performed with allograft, bone-patellar tendon-bone autograft, and hamstrings autograft (p = 0.0055) (Fig. 9).
Time to Surgery
The anterior cruciate ligament reconstruction registry identified a difference in the rate of medial meniscal injury in association with increased time to surgery. Longer time from injury to surgery (six to twelve months and more than twelve months) was associated with an increased rate of medial meniscal injury when compared with surgery within three months after the injury (odds ratio = 1.97 [p < 0.001] and 2.42 [p < 0.001], respectively).
Total Joint Registry
In the present study, we provide total joint replacement revision rates from a large community-based practice in the United States. Cumulative total joint survival rates for total knee arthroplasty and total hip arthroplasty at 5.5 years were 97.1% (95% confidence interval, 96.8% to 97.4%) and 97.3% (95% confidence interval, 96.8% to 98.2%), respectively. These rates are similar to those reported by other national registries at five to six years of follow-up8,16. Total joint replacement survival rates in our patients with an age of sixty-five years or more are similar to those reported in United States Medicare studies30-32.
Geographical region, female sex, lower annual surgeon case volume (fewer than thirty cases), a conventional insert, and a femoral head size of ≤28 mm were found to be important independent risk factors for aseptic revision following total hip arthroplasty. Similar to the findings reported in other studies, the uncemented technique was associated with a higher relative risk of revision than hybrid fixation. Geographical variation in the incidence of surgical procedures has been described in the United States population31. The Swedish total hip registry demonstrated variation in outcome among the various counties reporting8, mirroring our findings. Similar findings with regard to surgeon volume were reported by Katz et al.32,33. Investigators from the Mayo Clinic recently reported a lower risk of hip instability following total hip arthroplasty performed with use of larger femoral head size34. This clinical finding was unique and underscores the need for a large sample size to detect a previously intuitive yet unproved finding. Our large registry once again takes this clinical finding further, showing that larger femoral head size reduces the risk of revision surgery.
Similar to the findings regarding total hip arthroplasty, variations in geographical regions were associated with a higher relative risk of revision following total knee arthroplasty, consistent with early findings of the Swedish registry8. Findings also identified younger age, a higher ASA score, diabetes, uncemented fixation, and a mobile-bearing design as independent risk factors for aseptic revision following total knee arthroplasty. The findings that younger age and uncemented implant fixation were risk factors for total knee arthroplasty revision support findings that have been reported previously35,36. In contrast to previous smaller, single-center studies that demonstrated good intermediate to long-term survival of mobile-bearing designs37,38, our multicenter study suggests that mobile-bearing designs may contribute to an increased risk of revision. This may be related to the complexity of the design or early detection of a potential problem identified by the larger sample size of the registry. Other risk factors for total knee arthroplasty revision included higher ASA scores and diabetic status. Consideration should be given for counseling patients in these high-risk groups prior to performing total knee arthroplasty.
Similar to other total joint replacement registries, our findings suggest that feedback to surgeons can positively influence clinical practice. The Swedish national hip registry has influenced cement techniques and reduced the variation in implants to decrease their revision rates. The Mayo Clinic has also identified improvements in care associated with registry feedback. The HealthEast registry has demonstrated findings similar to our study in terms of unicompartmental knee arthroplasty and uncemented techniques, resulting in reduction in these practices. These findings suggest that a national registry could potentially improve the quality of care for orthopaedic patients and reduce United States national revision rates.
Changes in Practice
Feedback to surgeons on total joint surgical techniques and implant performance, along with other published studies, has resulted in changes in orthopaedic practice. For example, the registry identified significant differences in revision rates between unicompartmental and total knee replacements (Fig. 10) and feedback from the registry and the findings of other studies resulted in a reduction in unicompartmental knee arthroplasty volume (Fig. 11). The finding of higher revision rates in association with smaller femoral head sizes, along with the findings of other published studies, was another example of feedback to participating surgeons that resulted in a reduction in the use of small femoral head sizes (Figs. 4 and 12). The registry also identified higher revision rates for cementless total knee arthroplasty fixation, leading to a reduction of cementless total knee arthroplasty (Figs. 2 and 13)28. Specific surgeon profiles are also confidentially provided to participating surgeons for practice assessment and comparison of their own personal revision and complication rates with medical center and regional benchmarks.
Patient Safety
The registries also play a critical role in patient safety. Patients affected by implant recalls and advisories are identified and notified immediately. During the study period, the orthopaedic implant registries were used to track and monitor eight recalls and advisories (Table I). More recently, the total joint replacement registry has been used to identify and monitor patients with the ASR Hip Resurfacing System and the ASR XL Acetabular System (DePuy, Warsaw, Indiana). Another benefit of the registries is the potential for early identification of defective devices. Although a randomized controlled trial design may provide a higher level of scientific evidence, randomized controlled trials can take a long time to recruit an adequate sample size for the required power to detect rare adverse events. Problems with implants may be identified early in a registry as a result of large sample sizes.
In addition to the key role in implant device monitoring, the registries are also important for identifying patient risk factors associated with revisions, infections23, and venous thromboembolic events. This information is used by our surgeons and patients for evaluating treatment options and setting expectations regarding surgical outcomes. Prognostic statistical models have been used to develop web-based revision risk calculators for use by our surgeons in clinical decision-making.
Cost Effectiveness
The registries provide an important role in evaluating new technology and marketing claims. New, more advanced implant technologies are frequently introduced into the marketplace at higher costs, despite the lack of evidence of superior performance. The registries have been used to evaluate new technologies such as high-flex and sex-specific knee implants with use of the collective experience of community-based surgeons. The registries also have been used to assess the comparative effectiveness of implants, with the focus on selecting the best implants for our patients in value-based contracting. Another area that demonstrates cost effectiveness is the establishment of virtual visits, with use of the registries to identify patients who are due for follow-up. The registries automatically generate a list of patients who are due for follow-up and e-mail the list to providers for postoperative follow-up. Providers can then follow up with patients by phone or e-mail, reducing the need for clinical visits.
Research
Finally, the registries provide a foundation for more in-depth research projects. Initial publications have focused on (1) early failures and postoperative complications20,23,24, (2) epidemiology of total joint replacements and anterior cruciate ligament reconstruction procedures21,29, (3) health care utilization after discharge19, and (4) antibiotic-loaded cement in knee replacements and postoperative infections23. As the registry grows and matures, the opportunities for research will also increase.
Anterior Cruciate Ligament Reconstruction Registry
Our anterior cruciate ligament reconstruction registry provides the first reported revision and reoperation rates from a large community practice within the United States with >200 participating surgeons. The primary anterior cruciate ligament reconstructions that were included in our registry during the four-year study period were associated with a revision rate of 0.6% and a reoperation rate of 4.4%. This compares similarly to the 6.6% reoperation rate reported by Lyman et al. in a statewide study that evaluated reoperation rates within one year following procedures coded for anterior cruciate ligament reconstruction in the state of New York39. The advantage of our registry is that it provides further details, including reasons for reoperation, laterality, the incidence and treatment of associated injuries, and graft and fixation types, while also providing comparative graft survival rates for patellar tendon autografts, hamstrings autografts, and allografts. Early results suggest differential survival rates between the three graft types, and we are currently investigating this finding.
Our study revealed that the three most common reasons for reoperations on the index knee following primary anterior cruciate ligament reconstruction were for meniscal injury (34.3%), stiffness (18.9%), and infections (14.0%). Reoperation on the index knee was three times more likely than reoperation on the contralateral knee. For the contralateral knee, the most common procedures were ligament reconstruction (65.7%) and meniscal surgery (28.6%).
Of interest is our finding that contralateral reoperations in our female population were more likely to be for ligament injuries, whereas meniscal injuries were more common for males. Another common cause for reoperations was stiffness, and our plans are to investigate variances across medical centers to see if factors such as postoperative rehabilitation protocols have an impact on this reoperation rate. Removal of fixation hardware was the fourth-most-common reason for reoperation, and as we gather additional data, we aim to determine whether specific implant designs or materials lead to lower reoperation rates.
The finding that females between the ages of fourteen and seventeen years had the highest incidence and an increased annual rate of anterior cruciate ligament reconstruction have led us to focus on this group for injury prevention such as preseason intervention with an online educational video targeted at athletes in this high-risk group. Our findings also have suggested that longer time from injury to surgery (six to twelve months and more than twelve months) was associated with an increased rate of medial meniscal injury. These findings compare similarly with those from the Norwegian registry40. These findings emphasize the need to improve on the surveillance of anterior cruciate ligament injuries with appropriate interventions to prevent further injury, particularly in our young athletes.
The overall rates of deep-vein thrombosis and pulmonary embolism following anterior cruciate ligament reconstruction in this study were low. Therefore, we have not made specific recommendations regarding prophylaxis against deep-vein thrombosis and pulmonary embolism at this time.
Strengths and Weaknesses
The strengths of this study are the large community-based sample with contemporary implants. A large prospective cohort study such as this is valuable for identifying rare outcomes with a large sample size. The findings of associations between variables and these outcomes are many times unique to this type of population-based design; however, it cannot replace randomized controlled trials for identifying causation of variable to outcome. For many clinical questions, randomized controlled trials are not feasible, are unethical, or are prohibitive in cost. A registry allows for the collection of clinically relevant outcomes and reflects real clinical practice. The barriers to receiving health care are limited in an integrated health maintenance organization. Our integrated system allows for the capture of patient, surgeon, implant, hospital variables, and outcomes that occur in any inpatient and outpatient setting, allowing longitudinal follow-up data collection. Our system also has low attrition rates, with <12% of patients lost to follow-up over a seven-year period. These features offer a distinct advantage over similar United States registries.
Our study has several limitations. The registry is a large prospective cohort, whereas some of the outcome data are collected retrospectively from administrative databases. As coding accuracy is important in outcome analysis, our registry data are validated with chart reviews, improving the overall accuracy of the data captured. Another study limitation is the short to intermediate-term length of follow-up. Although total joint replacements can survive for many years, evaluation of early failures can provide important information about technical failures. In addition to short-to-intermediate-term follow-up, the registries are also limited by a focus on revision, reoperations, and complications as the key end points. Although we have implemented the patient-reported outcomes at specific centers, participation rates are <50%. Web-based surveys and virtual visits are in the process of implementation to increase response rates prior to reporting these patient-reported outcomes.
Future Areas of Focus
Future directions for the registries program include interactive patient web sites with patient-reported outcomes, automated post-market surveillance with real-time signal detection of adverse events, expansion of our revision risk calculator to assess other complications and outcomes (i.e., deep-vein thrombosis, pulmonary embolism, hospital readmissions, and postoperative medical complications), and radiographic assessment. We are also in the process of automating the virtual visits process with automated orders for radiographs and a web-based questionnaire.
The Need for a National United States Registry
While other countries have established national registries to track orthopaedic implant volume, utilization, and revision rates8,12-14,16,41, national United States studies have relied on hospital discharge (e.g., National Hospital Discharge Survey, Nationwide Inpatient Sample)2,4,5 and CMS (Centers for Medicare and Medicaid Services) data30,31,42 to provide estimates of orthopaedic procedure volumes, complications, and revision rates. Although these studies provide important estimates of complications and revisions, they are limited by the reliance on ICD-9-CM (International Classification of Diseases, Ninth Revision, Clinical Modification) coding developed for billing purposes. Inherent problems associated with ICD-9-CM codes include hospital variation in coding practices, inaccurate coding, problems with sensitivity and specificity, and inadequate clinical and implant data. Medicare data are further limited by the exclusion of patients under the age of sixty-five years, who are at highest risk for revision surgery36,43, and lack of laterality to distinguish joints and associated outcomes. These limitations emphasize the importance of a national United States registry with standardized coding and terminology, detailed implant information, and laterality necessary to fully assess United States utilization rates, complications, and revision rates.
The development of a United States national registry is critical for assessing revision rates and implant utilization, for tracking implanted devices during recalls and advisories, for conducting post-market surveillance of new technologies, for early identification of defective devices, and for assessing the comparative effectiveness of devices. Ideally, a United States registry could provide feedback on techniques and devices similar to the Swedish registry and could reduce national revision rates. A national registry may be feasible through changes in policies and regulations or through collaborative efforts of the AAOS, surgeons, manufacturers, payers, and hospitals. The FDA and AHRQ have launched efforts in support of national networks and multicenter grants that may provide support for initial efforts in this direction. A national registry could also be developed through collaboration with distributed networks of established institutional and regional registries. Regardless of how it is accomplished, the need for a national registry for specific orthopaedic procedures has been well substantiated.
Cumulative survival rates of total joint replacement and anterior cruciate ligament procedures from a large community-based practice are similar to rates reported by other regional and national registries.
Registries provide an alternative solution for comparative research when randomized controlled trials are impractical, are unethical, or require long-term follow-up with large sample sizes to detect rare complications.
Registries provide real-world results that may be more generalizable than randomized controlled trials and can positively impact clinical practices, safety, cost effectiveness, and research.
There is a need for a national United States registry to monitor device performance. Medicare data seem to provide a potential solution to the national need for a United States registry, but our results indicate that a large portion of the patients at highest risk for revision would not be accounted for with use of Medicare as the primary data source.
Although our registries provide a model for a United States national registry, the reliance of the registries on our integrated system may not be transferable to other health-care systems.
A network of established institutional and regional registries may provide a method for the development of a United States national registry.
Figures showing additional survival curves by the distribution of diagnoses, procedures, and ages are available with the electronic version of this article on our web site at jbjs.org (go to the article citation and click on "Supporting Data").
Note: The authors would like to acknowledge Don Fithian, MD, our orthopaedic surgeons, and the registry staff for their important contributions.
Organization for Economic Co-operation and Development. OECD health data 2009—selected data: health expenditure. .
Kurtz
S;
Ong
K;
Lau
E;
Mowat
F;
Halpern
M. Projections of primary and revision hip and knee arthroplasty in the United States from 2005 to 2030. J Bone Joint Surg Am.
2007;89:780-5.[PubMed][CrossRef]
Statistical brief #34. Agency for Healthcare Research and Quality, 2007. .
Kurtz
SM;
Lau
E;
Ong
K;
Zhao
K;
Kelly
M;
Bozic
KJ. Future young patient demand for primary and revision joint replacement: national projections from 2010 to 2030. Clin Orthop Relat Res.
2009;467:2606-12.[PubMed][CrossRef]
Kurtz
SM;
Ong
KL;
Schmier
J;
Mowat
F;
Saleh
K;
Dybvik
E;
Kärrholm
J;
Garellick
G;
Havelin
LI;
Furnes
O;
Malchau
H;
Lau
E. Future clinical and economic impact of revision total hip and knee arthroplasty. J Bone Joint Surg Am.
2007;89
Suppl 3:144-51.[PubMed][CrossRef]
Wilson
NA;
Schneller
ES;
Montgomery
K;
Bozic
KJ. Hip and knee implants: current trends and policy considerations. Health Aff (Millwood).
2008;27:1587-98.[PubMed][CrossRef]
British Orthopaedic Association and British Hip Society. Metal on metal hip replacement and hip resurfacing arthroplasty: what does the MHRA Medical Device Alert mean?.
Malchau
H;
Herberts
P;
Eisler
T;
Garellick
G;
Söderman
P. The Swedish Total Hip Replacement Register. J Bone Joint Surg Am.
2002;84
Suppl 2:2-20.[PubMed]
Malchau
H;
Herberts
P;
Ahnfelt
L. Prognosis of total hip replacement in Sweden. Follow-up of 92,675 operations performed 1978-1990. Acta Orthop Scand.
1993;64:497-506.[PubMed][CrossRef]
Herberts
P;
Malchau
H. Long-term registration has improved the quality of hip replacement: a review of the Swedish THR Register comparing 160,000 cases. Acta Orthop Scand.
2000;71:111-21.[PubMed][CrossRef]
Herberts
P;
Malchau
H. How outcome studies have changed total hip arthroplasty practices in Sweden. Clin Orthop Relat Res.
1997;344:44-60.[PubMed][CrossRef]
Graves
SE;
Davidson
D;
Ingerson
L;
Ryan
P;
Griffith
EC;
McDermott
BF;
McElroy
HJ;
Pratt
NL. The Australian Orthopaedic Association National Joint Replacement Registry. Med J Aust.
2004;180(
5 Suppl):S31-4.[PubMed]
Lucht
U. The Danish Hip Arthroplasty Register. Acta Orthop Scand.
2000;71:433-9.[PubMed][CrossRef]
Pitto
RP;
Lang
I;
Kienapfel
H;
Willert
HG. The German Arthroplasty Register. Acta Orthop Scand Suppl.
2002;73:30-3.[PubMed]
Bourne
RB. The planning and implementation of the Canadian Joint Replacement Registry. Bull Hosp Jt Dis.
1999;58:128-32.[PubMed]
Havelin
LI;
Engesaeter
LB;
Espehaug
B;
Furnes
O;
Lie
SA;
Vollset
SE. The Norwegian Arthroplasty Register: 11 years and 73,000 arthroplasties. Acta Orthop Scand.
2000;71:337-53.[PubMed][CrossRef]
Gioe
TJ;
Killeen
KK;
Hoeffel
DP;
Bert
JM;
Comfort
TK;
Scheltema
K;
Mehle
S;
Grimm
K. Analysis of unicompartmental knee arthroplasty in a community-based implant registry. Clin Orthop Relat Res.
2003;416:111-9.[PubMed][CrossRef]
Gioe
TJ;
Killeen
KK;
Mehle
S;
Grimm
K. Implementation and application of a community total joint registry: a twelve-year history. J Bone Joint Surg Am.
2006;88:1399-404.[PubMed][CrossRef]
Bini
SA;
Fithian
DC;
Paxton
LW;
Khatod
MX;
Inacio
MC;
Namba
RS. Does discharge disposition after primary total joint arthroplasty affect readmission rates?J Arthroplasty.
2010;25:114-7.[PubMed][CrossRef]
Khatod
M;
Barber
T;
Paxton
E;
Namba
R;
Fithian
D. An analysis of the risk of hip dislocation with a contemporary total joint registry. Clin Orthop Relat Res.
2006;447:19-23.[PubMed][CrossRef]
Khatod
M;
Inacio
M;
Paxton
EW;
Bini
SA;
Namba
RS;
Burchette
RJ;
Fithian
DC. Knee replacement: epidemiology, outcomes, and trends in Southern California: 17,080 replacements from 1995 through 2004. Acta Orthop.
2008;79:812-9.[PubMed][CrossRef]
Namba
RS;
Paxton
EW;
Barkan
HB;
Shah
SB;
Slipchenko
TS;
Fithian
DC. Predictors of total knee arthroplasty deep infections in 23,089 cases. :Proceedings of the 75th Annual Meeting of the American Academy of Orthopaedic Surgeons; 2008 Mar 5-9. Rosemont, IL: American Academy of Orthopaedic Surgeons; 2008. .
Namba
RS;
Chen
Y;
Paxton
EW;
Slipchenko
T;
Fithian
DC. Outcomes of routine use of antibiotic-loaded cement in primary total knee arthroplasty. J Arthroplasty.
2009;24(
6 Suppl):44-7.[PubMed][CrossRef]
Namba
RS;
Inacio
M. Early and late manipulation improve flexion after total knee arthroplasty. J Arthroplasty.
2007;22(
6 Suppl 2):58-61.[PubMed][CrossRef]
Namba
RS;
Paxton
L;
Fithian
DC;
Stone
ML. Obesity and perioperative morbidity in total hip and total knee arthroplasty patients. J Arthroplasty.
2005;20(
7 Suppl 3):46-50.[PubMed][CrossRef]
AHRQ. Agency for Healthcare Research and Quality, Quality Indicators. : Patient Safety Indicators Download; 2007. .
Mangram
AJ;
Horan
TC;
Pearson
ML;
Silver
LC;
Jarvis
WR. Guideline for prevention of surgical site infection, 1999. Hospital Infection Control Practices Advisory Committee. Infect Control Hosp Epidemiol.
1999;20:250-80.[PubMed][CrossRef]
Paxton
EW;
Inacio
M;
Slipchenko
T;
Fithian
DC. The Kaiser Permanente National Total Joint Replacement Registry. Permanente J.
2008;12:12-6.
Csintalan
RP;
Inacio
MCS;
Funahashi
TT. Incidence rate of anterior cruciate ligament reconstructions. Permanente J.
2008;12:17-21.
Mahomed
NN;
Barrett
JA;
Katz
JN;
Phillips
CB;
Losina
E;
Lew
RA;
Guadagnoli
E;
Harris
WH;
Poss
R;
Baron
JA. Rates and outcomes of primary and revision total hip replacement in the United States Medicare population. J Bone Joint Surg Am.
2003;85:27-32.[PubMed]
Mahomed
NN;
Barrett
J;
Katz
JN;
Baron
JA;
Wright
J;
Losina
E. Epidemiology of total knee replacement in the United States Medicare population. J Bone Joint Surg Am.
2005;87:1222-8.[PubMed][CrossRef]
Katz
JN;
Mahomed
NN;
Baron
JA;
Barrett
JA;
Fossel
AH;
Creel
AH;
Wright
J;
Wright
EA;
Losina
E. Association of hospital and surgeon procedure volume with patient-centered outcomes of total knee replacement in a population-based cohort of patients age 65 years and older. Arthritis Rheum.
2007;56:568-74.[PubMed][CrossRef]
Katz
JN;
Phillips
CB;
Baron
JA;
Fossel
AH;
Mahomed
NN;
Barrett
J;
Lingard
EA;
Harris
WH;
Poss
R;
Lew
RA;
Guadagnoli
E;
Wright
EA;
Losina
E. Association of hospital and surgeon volume of total hip replacement with functional status and satisfaction three years following surgery. Arthritis Rheum.
2003;48:560-8.[PubMed][CrossRef]
Berry
DJ;
von Knoch
M;
Schleck
CD;
Harmsen
WS. Effect of femoral head diameter and operative approach on risk of dislocation after primary total hip arthroplasty. J Bone Joint Surg Am.
2005;87:2456-63.[PubMed][CrossRef]
Duffy
GP;
Crowder
AR;
Trousdale
RR;
Berry
DJ. Cemented total knee arthroplasty using a modern prosthesis in young patients with osteoarthritis. J Arthroplasty.
2007;22(
6 Suppl 2):67-70.[PubMed][CrossRef]
Gioe
TJ;
Killeen
KK;
Grimm
K;
Mehle
S;
Scheltema
K. Why are total knee replacements revised?: analysis of early revision in a community knee implant registry. Clin Orthop Relat Res.
2004;428:100-6.[PubMed][CrossRef]
Sorrells
RB;
Voorhorst
PE;
Murphy
JA;
Bauschka
MP;
Greenwald
AS. Uncemented rotating-platform total knee replacement: a five to twelve-year follow-up study. J Bone Joint Surg Am.
2004;86:2156-62.[PubMed]
Callaghan
JJ;
Squire
MW;
Goetz
DD;
Sullivan
PM;
Johnston
RC. Cemented rotating-platform total knee replacement. A nine to twelve-year follow-up study. J Bone Joint Surg Am.
2000;82:705-11.[PubMed][CrossRef]
Lyman
S;
Koulouvaris
P;
Sherman
S;
Do
H;
Mandl
LA;
Marx
RG. Epidemiology of anterior cruciate ligament reconstruction: trends, readmissions, and subsequent knee surgery. J Bone Joint Surg Am.
2009;91:2321-8.[PubMed][CrossRef]
Granan
LP;
Bahr
R;
Lie
SA;
Engebretsen
L. Timing of anterior cruciate ligament reconstructive surgery and risk of cartilage lesions and meniscal tears: a cohort study based on the Norwegian National Knee Ligament Registry. Am J Sports Med.
2009;37:955-61.[PubMed][CrossRef]
Kolling
C;
Simmen
BR;
Labek
G;
Goldhahn
J. Key factors for a successful National Arthroplasty Register. J Bone Joint Surg Br.
2007;89:1567-73.[PubMed][CrossRef]
Kurtz
SM;
Ong
KL;
Lau
E;
Bozic
KJ;
Berry
D;
Parvizi
J. Prosthetic joint infection risk after TKA in the Medicare population. Clin Orthop Relat Res.
2010;468:52-6.[PubMed] [CrossRef]
Coyte
PC;
Hawker
G;
Croxford
R;
Wright
JG. Rates of revision knee replacement in Ontario, Canada. J Bone Joint Surg Am.
1999;81:773-82.[PubMed]