[Are hospital administrative data suitable for external quality assurance? Comparison of quality indicators based on separate statutory data collections (BQS) and hospital administrative data].

BACKGROUND

German hospitals are obliged legally to provide clinical data for external comparative quality assurance. Data rely on administrative data and just as on additional data collections for this purpose only. They are used to identify defined quality indicators (so-called BQS data). The Agency for Healthcare Research and Quality (AHRQ) also developed quality indicators that rely on hospital administrative data to evaluate the quality of inpatient care.

METHODS

Six selected quality indicators were computed by both methods. 2007 data from the nationwide external quality assurance program were analyzed and compared to quality information derived from a 2007 10 % nationwide sample of administrative hospital data.

RESULTS

Regarding the indicators "Obstetric trauma", "Mortality of community acquired pneumonia", "Postoperative deep vein thrombosis" and "Postoperative pulmonary embolism" rates are significantly higher in hospital administrative data than in BQS data (p < 0.01). Inversely, rates of the indicator "Decubitus ulcer" are significantly lower (p < 0.001).

CONCLUSION

Possible causes for the results might be divergent motivations for data collection or restrictions in data collection. It remains unclear which method properly reflects the true status. Selected indicators (e. g. obstetric trauma), however, are suitable to be substituted by hospital administrative data.