Measurement of quality of care. Comparison of voluntary versus obligatory data collection for the hospital based stroke register Hessen, Germany

Background and purpose: The stroke register Hesse started 1997 as a project of voluntary external quality control (VQC). Since 2004 participation is obligatory (OQC) for every department in the country Hesse (6 million inhabitants) treating acute stroke patients. The intention was to obtain a comprehensive image of the quality of care. Quality indicators (QI's) were developed in parallel to provide feed back for the participants and a basis for dialog. We hypothesised to obtain a higher dispersion measuring QI's with obligatory participation.

Methods: Two periods 1999–2002 (VQC) and 2004–2005 (OQC) were compared [21.060 (24.870) ischaemic strokes, 7.345 (10.701) TIA, 2.696 (3.140) intracranial haemorrhages]. QI's pertained early brain imaging, neurovascular imaging, echocardiography, screening for dysphagia, physiotherapy, speech therapy and antithrombotic therapy at discharge. Departments providing less than 20 cases per year were excluded from this study. Departments with secondary admission did not met the criteria for diagnostic QI's.

Results: During the VQC 18 and the OQC 22 departments for Neurology and 13 other departments, mainly for internal medicine (OQC 112) participated. In the OQC period a considerably higher percentage of analysed departments (39 of 88 (44%)) did not met the requirements for early brain imaging as compared to 6 of 27 (22,2%) during VQC. Early imaging was performed in 82,9% (1. quartile 56,5%; 3. quartile 86,3%) during OQC and in 91,1% (1. quartile 76,7%; 3. quartile 95,7%) during VQC. Similar results were obtained for neurovascular imaging (15 of 67 versus 1 of 23 departments). On the other hand there was little change for the QI's echocardiography and speech therapy. An improvement during OQC was noted for the QI's physiotherapy and screening for dysphagia as well as for antithrombotic therapy at discharge.

Conclusion: Structural deficits were disclosed by OQC in the field of demanding diagnostics. Departments participating only in the second OQC period showed a high dispersion in realizing early brain and neurovascular imaging and therefore a worse quality of care compared to participants on voluntary base was documented. The QI's for care with improved overall results were less depending on investment and infrastructure but associated with an increased awareness due to the ongoing QI process. Regarding to our results it was shown that only obligatory data collection is suited to improve general quality of hospital care.