Introduction

The ICRS Registry is expanding and gaining worldwide recognition as an unbiased, reputable, cost-free clinical registry.

Content

The ICRS Registry mission: Our mission is to create the best source of unbiased outcomes data for treatments of painful articular cartilage lesions in the world, which is paramount for improvement of existing and discovery of new cartilage repair strategies, ultimately beneficial for millions of patients around the world

The ICRS Registry is perfect for continued objective monitoring of all patients with a cartilage lesion or early osteoarthritis, whether or not the lesion is treated directly. The response of patients to cartilage damage and treatments can be variable, treatments can also be forefront of medical advances, many are expensive. It is vital that a patient’s progress is monitored. All registry users have direct access to their own data and can export their data at any time. Additionally, the ICRS registry pools together large numbers of anonymized patient results to better understand how patients progress after treatment, so that doctors around the world have the most accurate picture of which techniques are working best in which patients. This helps patients of the future with similar injuries or cartilage problems, and rapidly identifies treatments that are showing great benefit, those that may not be performing as well as hoped, and also what happens naturally if nothing is done.It is designed to follow patients with following any cartilage treatment, but can also include patients with ligament reconstruction/repair, osteotomy, injections, bracing or no treatment at all. It can absorb existing datasets, undertake long-term follow-up of clinical trial patients, and ongoing follow-up of all patients.

The Registry now live in 8 languages: English, Japanese, Polish, Italian, Spanish, Portuguese, Dutch and Greek. A further three are in progress: Chinese, Swedish, and German. We now have ICRS Registry users in 50 Countries, so it is truly global. The greatest number of patients registered are from the UK, USA, Canada, China, Italy, Poland, Spain, The Netherlands, Switzerland, Germany, Greece, Brazil, Mexico, Japan and India. The patients range in age from 16-89 years, with older patients more likely to receive injections than cartilage repair surgery. Cartilage surgery was performed on patient with an age range of 16-57 years. The mean BMI of patients in the Registry is 26, and 67% have never smoked tobacco. The age and BMI was similar for both male and female patients. An open approach was used for cartilage repair procedures in two thirds of cases, and an arthroscopic approach in one third. Ninety-three percent of surgical cases employed a tourniquet, and prophylactic antibiotics were given to half of all patients undergoing surgical procedures. One third of patients undergoing cartilage surgery also underwent another concomitant surgical procedure eg: ligament reconstruction, realignment surgery, meniscal surgery etc

The number of patient pathways entered is rising exponentially. Nine percent of the data comes from existing datasets that have been absorbed, and 81% are new entries.

The first Annual Report is published before the ICRS Congress, and I look forward to updating the delegates on the progress of the Registry.

References

www.cartilage.org

Introduction

What can we learn from registries

Tim Spalding.

Co-authors: Christopher Kaeding, Magnus Forssblad, Martin Lind, Greg Maletis

Registries collect data on large numbers of patients with the aim of providing information to effect change on practice. Registries take effort both for data entry and for data analysis. Over the years much has been learnt from Ligament reconstruction and joint replacement registries - summarized in this abstract.

Content

Cohort Studies and Registries: whats the difference?

Prospective cohort studies and national registries are observational studies to assess outcomes after a specific medical intervention, evaluating clinical outcomes and the predictors of those outcomes. Registries determine effectiveness of the intervention in the real world of general use – the poorly controlled multi-factorial world of general clinical practice. Cohort studies with controlled assessments are more able to determine how an intervention performs in the environment of carefully selected patients and this is evaluation of efficacy. Determining factors that predict outcome requires controlled clinical efficacy studies.

The Moon Studies

The MOON or Multicenter Orthopedic Outcomes Network was a prospective cohort study designed to look at the prognosis and predictors of clinically relevant outcomes after an ACL reconstruction. The design required a data base with large numbers, standardized independent variable assessment and data collection technique, validated outcome measures, a high capture rate and high quality data points that could be verified.

The cohort study evaluated 5 factors as below and each surgeon participated in inter and intra reliability studies.

activity level

patient oriented outcomes

signs and symptoms of osteoarthritis

recurrent ligament disruption

need for additional arthroscopic surgery

Surgeons entered >97% of patients and obtained >80% follow up – achieving this by calling patients on a Sunday evening. The result was 82% collection at both 2 and 5 years. Such high levels of data from enthusiasts carries the risk of the output not being generalizable, and potentially not large enough in number for more detailed analysis

Registries and Cohort studies

Large national registries are likely to have more generalizable conclusions, with more variables to analyze. But on the converse registries face a challenge of poor follow up with less commitment by surgeons and supporters, along with variability in selection and rehabilitation factors. Overall, both are complementary. Cohort studies are smaller, more focused, and better controlled with excellent follow-up. Registry studies are larger, more generalizable, and has potential for deeper analysis. Both are expensive and labor-intensive, but can be powerful clinic research tools. Both need to be carefully purposed, designed, executed and their weaknesses understood

Scandinavian Registries

The Scandinavian Ligament Registries were established in Norway in 2004 and in Sweden and Denmark in 2005. They now include data from 110,000 patients and the work from the three registries has led to more than 70 publications. The drive was to understand outcome as determined by patient related factors, treatment related factors and injury factors.

The main headline findings that emerged from the registries were that younger patients have an increased risk of additional ACL reconstruction and females have an increased risk of contralateral injury. Analysis of current graft choice shows that 90% of reconstructions in Sweden involve hamstrings but in Norway it is only 50%. Analysis of failure rates show an increased risk of revision with hamstring v patella tendon and the risk of ACL revision with hamstrings is reduced by 14% for every 0.5mm increase in hamstring graft diameter. It is clear that only large datasets can determine this level of detail.

The strength of the Swedish registry is that there is a greater than 90% uptake rate for surgeons with only 20 questions for the busy surgeon to complete. What has been learnt is that the response rate, as with all registries, is frustratingly low at around 70%, and that the KOOS outcome measure utilised may not be reliable as ACL specific. Recent trends analysis has shown an increase the use of quadriceps tendon, the vancomycin RAMP to reduce infection, and recognition of the value of contribution by group surgeon discussions.

Kaiser Permanente Registry

The Kaiser Permanente Registry had made a large impact on understanding of outcomes following ACL reconstruction in the USA. The basis is input from specific insurance company based hospitals who have collected data and importantly have subsequently shown that analysis and feedback can alter practice. 346 surgeons participate in the registry and currently over 44,000 ACL reconstructions are listed. Data has shown a difference in survival curves between allografts, hamstring autografts and patella tendon grafts. Allograft failure rates were particularly high in younger patients with data published in 2013. Specific analysis identified the Biocleanse processing method as having a higher failure rate as well as irradiation. BTB allografts also had higher failure and a particular at risk group were patients under aged 21.

The group noted that in 2010 44% of ACLs were performed using allografts and with the data identifying this group to be at risk information was disseminated amongst the surgeons such that over subsequent years use of allograft has halved, use of irradiated graft has reduced by 80% and, in particular, use in the high-risk patients under aged 22 has reduced by 80%. The drivers of this change is acknowledgement of the ability to spread the information. Registries can therefore alter practice.

Designing a new registry

When considering designing a new registry the team must understand the challenges and purpose of the registry involving surgeons, defining the exact content of the registry and methods for data entry. Management structure and governance is important, handling data protection (GDPR) and producing adequate feedback for surgeons to maintain enthusiasm. Clinician involvement and buy-in is key trusting the outcome produced and not fearing government control.

A balance must be struck between excessive data collection against inadequate detail. With the metric of patient compliance in outcome form completion this can be painfully assessed. Data entry can be online or through manual paper collection which involves an office to collate the information, however it is likely that a combination of the two systems is most efficient. The usefulness of the output depends on the data collected and the input from design team interpreting the information.

Summary

In summary a registry must define its purpose and acknowledge limitations. It needs to ensure clinician participation and co-ordinate the data content with the developers and collaborators. Outcome measures need to be responsive and validated and the registry needs defined management leadership and data ownership. Registries look at real world practice detecting procedures and devices that result in premature failure. From the large data sets factors can be identified and finally improved improvement in treatment outcomes can be achieved through feedback to surgeons.