Risk-adjusted Morbidity and Mortality Models to Compare the Performance of two Units after Major Lung Resections
Back to 86th Annual Meeting
Back to Program Outline
Objective:
To develop risk-adjusted morbidity and mortality models to compare the performance of two different thoracic surgery units in patients submitted to major lung resections.
Methods:
743 patients (551 males, 192 females) submitted to lobectomy (611) or pneumonectomy (132) from January 2000 through December 2004 at two European thoracic units (519 cases in unit A and 224 cases in unit B) were analyzed.
Risk-adjusted models of 30-days or in-hospital cardiopulmonary morbidity and mortality were developed by stepwise logistic regression analyses and validated by bootstrap bagging simulation. Pre-operative and operative variables were initially screened by univariate analysis. Those with a p< 0.10 were used as independent ones in the regression analyses. Problems of overfitting and multicolinearity were considered. Variables were at least 95% complete and sporadic missing values were imputed. The regression equations were used to estimate the risk of outcome and the observed and predicted outcome rates of the two units were compared by the z test for comparison of proportions.
Results:
The following regression models were developed:
Predicted morbidity: ln R/1-R= -2.4 + 0.03Xage - 0.02XppoFEV1 + 0.6Xcardiac co-morbidity (Hosmer Lemeshow statistic=6.1 (p=0.6), c-index=0.65).
Predicted mortality: ln R/1-R= -6.97 + 0.095Xage - 0.042XppoFEV1 (Hosmer Lemeshow statistic=2.99 (p=0.9), c-index=0.77). The models proved to be stable at bootstrap analyses.
No differences were noted between observed and predicted outcome rates within each unit, despite an apparent better performance of unit B (Table).
Conclusions:
The use of risk-adjusted outcome models prevented misleading information derived from the unadjusted analysis of performance. Risk modelling is essential for the evaluation of the quality of care.
| Unit | Observed Morbidity | Predicted Morbidity | p-value | Observed Mortality | Predicted Mortality | p-value |
| Unit A | 23% | 22.7% | 0.9 | 4.8% | 4.9% | 0.9 |
| Unit B | 17% | 18.2% | 0.7 | 4.4% | 4.2% | 0.9 |
Back to 86th Annual Meeting
Back to Program Outline
