Objectives: Recently the KCE published a report on “the volume of surgical interventions and its impact on the outcome: feasibility study based on Belgian data”. We want to put this report into perspective to the BACTS data.
Methods: The study on volume - outcome relationship is critically analysed and compared to the BACTS data.
Results: Methodological issues: Hospital mortality is only a minor quality indicator and should be expanded to early mortality (3 months for CABG and 6 months or more for complex procedures). The categorical threshold of 200 CABG for high volume is arbitrary. There is no linear relationship between volume and performance, at most an asymptotic relationship. There is an enormous variability in patients morbidity and comorbidity. The Charlson score is inadequate to correct for casemix. Therapeutic and institutional variability is not taken into account. Numerical issues: KCE identified 7071 isolated CABG, 1953 valve and 1655 valve operations, the BACTS registry identified respectively 7422, 2221, 1338 operations. The differences are explained by the primary data selection. KCE uses a combination of NIHDI codes and ICD-9-CM codes. BACTS registry uses CPT codes. Not all patients are identified or classified in the same group by the two registries.
Conclusions: The KCE report has limitations but is nevertheless a good feasibility study to examine the quality of patient care. The analysis is anonymous and no centres or surgeons are identified. In general the quality of patient care regarding to hospital mortality is good. The Belgian mortality seems to be higher, compared to international registries but this can be due to patient and procedure selection.