In-hospital mortality rates after CABG by autonomous regions in Spain.

Purpose: The paper's purpose is twofold: to provide a predictive model for estimating in-hospital mortality rates after coronary artery bypass grafting (CABG) in Spanish autonomous regions (AR) after adjusting relevant factors; and to determine whether there is a difference between expected and observed mortality rates.

Design/methodology/approach: All patients registered in a minimum basic data set (MSBD) undergoing CABG between 2000 and 2004 were selected. After bivariate analysis to explore associations between in-hospital death and other variables, a multivariate analysis using logistic regression was conducted. The predictive model was evaluated using calibration and discrimination techniques. Standardized mortality ratios by AR were calculated.

Findings: The expected Spanish in-hospital mortality rate after CABG was 7.68 and the observed rate was 7.69 deaths per 100 operations. Discrimination obtained with the model resulted in an area under the curve of 0.70 (95 per cent CI, 0.69-0.71). When each AR's mortality rate is calculated and compared with the observed rate, some ARs present an observed mortality rate higher or lower than the expected rate according to adjusted variables in the model.

Research Limitations/implications: The MSBD registry does not contain patients' critical data, such as arterial damage severity, or in which hospital procedures were performed.

Practical Implications: There are factors related to individual patient variation, financial resources or healthcare quality in different ARs, which should be investigated in follow-up studies.

Originality/value: The paper shows that, although the global expected mortality rate is almost the same as the observed Spanish mortality rate, this similarity disappears when AR rates are compared.