German Aortic Valve Registry

1. Research aims

The increasing prevalence of severe aortic valve defects is a corollary to increases in life expectancy (Lindroos et al. 1993). For many years, surgical valve replacement with extracorporeal circulation has been the gold standard in the treatment of severe aortic valve diseases. Every year, about 12,000 patients in Germany receive isolated aortic valve surgery (Gummert et al. 2009). Several alternatives to established procedures have recently emerged:

• V-TAVI, vascular transcatheter aortic valve implantation and
• A-TAVI, apical transcatheter aortic valve implantation (Cribier et al. 2002, Walther et al. 2008, Webb et al. 2007, Himbert et al. 2008, Walther et al. 2007).

For some time, percutaneous balloon valvuloplasty (Sack et al. 2008) has been used as a palliative measure for specific patients.

New developments have triggered assumptions that there will over the coming years be a certain reorientation to the treatment of patients with aortic valve defects. In a positioning paper published jointly by the German Cardiac Society (DGK) and the German Society for Thoracic and Cardiovascular Surgery (DGTHG), there is a consensus that these new techniques can be used on patients who have either contraindications against conventional surgery or who are at high risk of preoperative mortality, e.g. due to severe comorbidities (Figulla et al. 2009).

As there is as yet only limited experience in this field and neither randomised trials nor long-term-results are available, the use of new implantation techniques is therefore not recommended for younger patients or patients without comorbidities (Figulla et al. 2009). There is accordingly an urgent need for any introduction of this innovative medical procedure to be stringently monitored according to scientific principles. The goal of the planned Germany Aortic Valve Register is to evaluate the new treatments from the point of view of benefits und risks with respect to the gold standard of conventional surgery, with a view to compiling evidence-based indication criteria. The register will furthermore allow for the first time a comparison of various operative procedures, such as Ross procedure, David procedure and various mechanical or biological aortic valve implants.

2. Rationale of the study design Randomised trials in controlled environments are considered best scientific practice for verifying the efficacy of a new method. The disadvantage is that only a small part of potential patients can be included in the trial, and results therefore reflect only a small section of the real world. Furthermore, a randomised study design is for ethical reasons not an acceptable approach to all questions, especially when contraindications exist for a certain branch of treatments and the spectrum of treatments is to be expanded. Nevertheless, new procedures have to be critically and scientifically analysed if the risk-benefit ratio is to be accurately determined. National heath authorities therefore request register data in addition to controlled trials in order to verify the safety and efficacy of new procedures across larger patient populations.

3. Objectives

1. Description of structure, process and outcome quality for the various techniques of aortic valve therapies
2. Definition of indication criteria (e.g. through scoring systems)
3. Collection of information on quality and safety for special medical devices
4. Evaluation of quality of care on the level of participating centres with a view to increasing quality

5. Health economic evaluation of the applied treatments

   4. Study design Prospective, controlled, multicenter register study

   5. Inclusion criteria

   All consecutive patients from participating hospitals with aortic valve defects who have received one of the following therapies:

   1. surgical aortic valve replacement,
   2. aortic valve surgery (Ross procedure, David procedure)
   3. percutaneous transvascular (retrograde) aortic valve implantation
   4. percutaneous transapical aortic valve implantation
   5. aortic valve valvuloplasty

   6. Variables of measurement

   • Clinical short term and long term results (event free survival) after 30 days, one year and after 3 and 5 years
   • Indication for the various procedures
   • Documentation of surgical risk and quality of life
   • Documentation of technical performance and the success of the intervention, including re-interventions
   • Documentation of in-hospital mortality and non-fatal severe complications
   • Documentation of medication at discharge / long-term medication
   • Documentation of implants used
   • Documentation of length of stay in hospital and rehabilitation

   • Documentation of mortality and non-fatal severe complications, re-intervention, re-hospitalisation, as well as medication within 30 days, one year and after 3 and 5 years.

     7. Statistical methods / evaluation plan All collected parameters are analysed descriptively for the entire study population for each of the included therapies (see: 5. Inclusion criteria). The absolute value and the percentage of the quality data are reported. For continuous variables median and quantities will be presented. Event free survival will be visualised through Kaplan-Meier graphs. The descriptive analysis of main measurement variables helps to illustrate the quality of the various techniques of aortic valve replacement from a structure, process and outcome point of view (objective a.).

   Within the different types of therapy, predictors for surgery/intervention without complications, and therefore also mortality, as well as improvement of quality of life will be ascertained through regression models. The results of these calculations will be used to detect criteria for indicating the various techniques (objective b.).

   Within the various techniques, alternative medical devices will be compared in respect to mortality and quality of life. Propensity-score models can be used for adjustment purposes (objective c.).

   All participating centres will receive a benchmark report once a year in which the patients of the respective centre will be compared to those in other hospitals. The report will distinguish between the various techniques of aortic valve replacement.

   In addition to a descriptive comparison, all complications will be reported risk-adjusted for each centre and analysed from the point of view of the centre-related effect (objective d.).

   An adjusted comparison of different types of therapies in respect to complications, cost and patients' quality of life through statistical methods will not be undertaken until the patient population of the various techniques overlap each other at least partially with regard to predictors for complications, costs and quality of life. This is to be examined during the development of the first report.

   The following data will be needed for evaluation from a health-economy point of view (objective e.):

   Demographical data, details regarding intervention, complications in the context of the intervention, clinical outcome parameters, quality of life (on the basis of a standardised questionnaire), medication, consulting of general practitioners and specialists, hospitalisation, rehabilitation measures, dependence on long term care (care level), employment as well as inability to work and reduction of earning capacity.

   A price for the consumption of resources will be determined after data are collected. This will allow a monetary value to be put on consumption of resources and calculation of the incurred expenses. It may in this context be necessary to incorporate a time factor in consumption in the form of a country-specific and sector-specific inflation adjustment of the costs in relation to a base year.

   This provides the possibility of analysing costs with regard to the chosen procedures (see: 5. Inclusion criteria). It is furthermore planned to run a cost-effectiveness-analysis in close cooperation with clinical experts, in which the costs of the procedures will be set against clinical effects achieved in a cost-effectiveness relation (ICER). Additionally, a cost-utility analysis will be conducted wherever possible. The costs of the various treatments will thereby be set in relation to so-called quality adjusted life years (QALYs), a combined endpoint consisting of quality of life and length of life. The survey of quality of life through a structured questionnaire allows quality of life to be presented in an index.

   The goal of the accompanying health economic research is therefore to provide a comparison of monitored therapy alternatives from the point of view of clinical and economic effectiveness.

   It has to be emphasised that a differentiated approach must be taken to any comparison of individual groups due to the absence of randomisation. Matching will be undertaken to minimise selection bias. Due to particularly high differences in the patient collectives, the question of whether it is possible to apply these methods to the anticipated data can not be conclusively answered until the data records have been received.

   8. Data model The register provides a complete survey. The data model is based on three data sources.

   1. Data set of the external performance measurement according to the German Social Code (§ 137 SGB V): The set provides data on pre-, intra- and postoperative processes during hospitalisation at the time of surgery.

   2. Register data set: Additional internet-based survey, e.g. specific indication criteria and indication for used medical devices 3. Follow-up: Survey of follow-up after 30 days, 1, 3 and 5 years by the registry centre.

   The complete data collection is based on the consent of the patients from all hospitals performing surgical and catheter supported aortic valve procedures. From the point of view of data protection regulations, this is the safest and most feasible way to obtain longitudinal data. The completeness of data collection will be tested with an electronic tool analysing reimbursement data. Monitoring of data validity is carried out by a multi-stage plausibility check and a data crosscheck on a randomly selected sample (audit).

   9. Period of data collection The phase of patient-recruiting starts on 5 July 2010 and ends as decided by the Executive Board. Follow-ups take place 30 days, 1 year, 3 and 5 years after the aortic valve procedure. Efforts will be made to achieve a central follow-up, but this can only be conducted if the written consent of the patient has been obtained and the hospital agrees to export data to the registry centre.

   10. Data management The address of the patient needed for follow-up via telephone will be kept separately from the medical data of the patient. Only persons in charge of telephone follow-up will have access to the address of the patient. Only anonymised medical patient data will be used for statistical analysis. Personal data will be anonymised immediately after the database is closed and will be deleted after the last scientific analysis.

   For the data transmission of the quality assurance data according § 137 Social Code book five to the register, a tool will be made available to the participating centres. For this purpose, the participating centre will have to conduct an "internal BQS-export". The file generated by internal-BQS-export includes all identification characteristics by which a combination of the data sets according § 137 Social Code book five and the e-sheet data-sets is possible.

   This will ensure that only data receiving the written consent of the patient will be transmitted to the register.

   11. Benchmark reports During the period of routine operation of the register all participating hospitals will receive benchmark reports at least once a year. The reports will be conducted individually for each hospital and compare data of the respective hospital with data from all other hospitals. The collected data will be analysed in respect to characteristics, therapy and clinical progress of the patients included. This benchmark system enables internal quality assurance for the participating hospitals.

   12. Supporting organisations and funding The register is supported by an Executive Board consisting of members of the two associations involved and various experts without voting right. All members have to disclose possible conflicts of interest. The Executive Board also has the support of the Steering Committee, which has a consultancy role and to which representatives of the industrials sponsors, insurance companies, health economic experts and representatives of political decision-making-bodies are invited.

   The supporting organisation of the register is the non-profit GmbH Deutsches Aortenklappenregister (GARY), the official holder of the data from a legal point of view.

   13. Independence of Register A funding agreement with the manufacturers of medical devices ensures full independence of the Executive Board. A disclosure stipulation guarantees that all evaluations and publications are conducted solely according to scientific criteria. This prevents any manipulative influence on publications or the withholding of publications. The involved societies and the BQS Institute are by virtue of their statutes autonomous bodies from a professional and scientific point of view.