Page updated on April 1st, 2009

Outcome measurement requires much more detailed and selective basic data than are needed for procedure and structure quality to be able to achieve optimum results in patient medical services.
The datasets must comply with the requirements of the specific pathology, therefore the options to standardize these datasets are limited.
With respect to methodology, further standardization is necessary and possible regarding medical devices and pharmaceuticals as well as surgical interventions and drug therapy since they have specific individual requirements.

Specific objectives of the orthopaedic pilot
The specific aims of the orthopaedic pilot are the following:

  1. To develop outcome indicators for arthroplasty based on the existing national projects according to the requirements of ongoing European Commission projects.
  2. To summarize the existing projects and the essential issues successfully.
  3. To define best practice procedures for the development and operation of arthroplasty registers.
  4. To validate the potential contribution of different means used in outcome measurement and quality monitoring of medical devices (i.e. registers, meta-analyses of clinical studies, implant failure monitoring systems by public health institutions, quality control and complaint handling systems by the manufacturers) for a structured outcome measurement and quality control system at the EU level.
  5. To present a detailed description of the outcome related registers and similar datasets in two countries (e.g. Sweden, Finland) that have a long established and advanced system in Europe in order to study the organization and function of the entire outcome and quality monitoring system at a national level.

Indicators in Orthopaedics
From an outcome measurement perspective, the indicators E8 (Revision Rate) and E9 (Revision Burden) from the survey phase are of the greatest relevance.
Indicator E3 (In-hospital Mortality after Hip Fracture) shows a relevant frequency of mentions in a subgroup pertinent to public health: the therapy of acute fractures.
Other indicators are only considered in individual cases.
The indicators collected in the field of orthopaedics during the survey phase focus on procedure and structure quality and, in the majority of cases, can only be used indirectly for outcome measurement.
The use of inadequate indicators for comparative outcome analyses can have negative effects on the treatment of patients. Measures of improvement in a benchmarking process might lead to imbalanced patient selection resulting in a deterioration in treatment access on an individual level.

Dataset Quality and Outcome Measurement
Outcome registers are an essential factor in outcome research, in terms of

  •  core dataset, and
  • source of reference for further data sources as regards validity.

Outcome registers for implanted medical devices such as total endoprostheses are defined as follows:

  • Registration of ALL primary and revision operations in a defined area in a central database.
  • Follow the implant until it has to be revised, the patient dies or emigrates.
  • Definition of Revision (=Failure): at least one part of the implant has to be revised during revision surgery.

The main advantage of registers is their ability to enable systematic longitudinal analyses and a multitude of data linkings potentially leading to clear procedure recommendations after specialist analysis.
Analysis and discussion are most efficient when carried out via medical specialist societies.
Apart from a democratic decision-making procedure on a high professional level, this ensures the disclosure of information to physicians and hence the consideration of the results during the treatment process.

Quality of Datasets

Outcome Measurement
Sample-based clinical studies exhibit highly relevant and significant bias factors and thus have only very limited usability as a data basis for evaluations and conclusions.
The data are highly influenced by the authors of the clinical literature as regards the number of cases published. These data constitute a bias factor per se, with varying distributions world-wide though.
Structured surveys show better compliance with register data, but they are inferior to outcome registers in data quality and organisation.
Experimental studies show only low correlation with the clinical outcome and are thus no suitable basis for outcome assessment.
Registers monitor a considerably larger collective under more specified, standardized and comparable conditions and therefore are superior as a data source. Compared to sample-based studies and surveys they allow for statements of considerably higher validity, and they allow robust statements to be made much earlier.
In surgical interventions involving the implantation of medical devices, randomized controlled trials show no essential improvement in the quality of statements as compared to other sample-based study designs.
At present, the usual categorization regarding the quality of the literature and bases of evaluations appears to be inadequate for endoprostheses and similar medical devices.
Therefore, based on the data available, a modification in data quality classification should be taken into consideration.
1. Comprehensive data collections such as registers are to be rated superior.
2. Randomized controlled trials should be assessed with respect to the endpoint.
a. In the case of objective endpoints, such as measurement results (e.g. implant migration as an early indicator of loosening), a randomized controlled study is to be regarded as equal according to the relevant guidelines.
b. In the case of subjective endpoints it has to be checked whether post-operative examinations could possibly break blinding. In such a case, a compromising of results should be assumed.
To be able to make optimum use of the advantages described, publication procedures and basic data such as implant recording in registers should be standardized.

Market Monitoring and Post-Marketing Surveillance
The data currently available from manufacturers and public health authorities are insufficient for the handling of outcome measurement issues, market monitoring, and the detection of serious product deficiencies.
It would thus make sense to consider arthroplasty registers as an additional tool in market monitoring and post-marketing surveillance.

  • The data should be examined retrospectively with regard to irregularities, such as a striking frequency of revision operations with certain medical devices or a cumulation of certain reasons of revision such as implant fracture clearly indicating product failure or requiring measures to be taken with respect to the application guidelines.
  • Manufacturers should be involved in the process, either directly or by means of requesting for statements.
  • Information about patients and departments concerned would be available, for instance, in the case of product recalls, or for vigilance control.

Information procedures should be improved in detail.


Technical reports and Deliverables

  • Deliverable N. 9: “Protocol for the Orthopaedic Pilot study” (pdf)