Development of quality indicators and data assessment strategies for the prevention of central venous catheter-related bloodstream infections (CRBSI).

Background: The number of catheter related bloodstream infections (CRBSI) could be reduced and the outcome improved if specific standards in the quality of care were maintained. Therefore, the development of quality assurance (QA) procedures was commissioned to be included in the national mandatory QA programme in Germany.

Methods: Indicators representing quality deficiencies and potential for improvement of quality in relation to prevention and management of central venous catheters (CVC) were developed by (1) evidence-based literature searches and the compiling of an indicator register; (2) a multi-professional expert panel including patient representatives who selected indicators from this register by using a modified RAND/UCLA Appropriateness Method; (3) defining methods for data assessment, risk adjustment and feedback of indicator results to service providers; and (4) consulting all relevant medical societies and other stakeholders with regard to the QA procedures that had been developed.

Results: Thirty-two indicators for CRBSI prevention and management were eventually approved by the expert panel. These indicators represent quality of care at predefined points with regard to indication, insertion and care of CVCs, management of sepsis, general hygiene and training of health care personnel. Fourteen indicators represent processes, together with 7 representing structures and 11 outcomes. For assessing these indicators, data was obtained from four sources: claims data from health insurance funds, routine claims data from hospital electronic information systems, case specific longitudinal documentation from service providers and cross-sectional annual assessment of structures.

Conclusions: It was possible to develop indicators for mandatory QA procedures on CRBSI that take into account the different perspectives of all stakeholders involved. Despite efforts to use routine data for documentation wherever possible, most indicators required extra documentation.