Thirty-day mortality from colorectal cancer surgery within a deprived population

A recently published model for predicting 30-day mortality has been devised using data from 73 hospitals in the UK. Deprivation category was not one of the data sets used in predicting mortality and was not included in formulating this model. We wished to assess if the model was applicable to a population with high deprivation category scores.

Data on all colorectal cancer operations included in the model (age, ASA, cancer resected, Dukes stage and urgency) had been collected prospectively since January 1999. This data was applied to the model.

Two hundred and thirty patients were included in analysis. Overall predicted mortality was 8.2% and actual mortality 7.8%. Within different categories of deprivation, predicted and actual mortality are similar and not significantly different. 137 patients lived in areas of deprivation category 3-5. Predicted and actual mortality were 8.4% and 8.0%, respectively. In areas of deprivation category 6-7, 91 patients had a predicted and actual mortality of 8.0% and 7.7%.

Although deprivation data was not included in the production of the mathematical model to predict postoperative death, the model appears to predict well in the deprived population. It is likely that poorer socio-economic status is accounted for by increased ASA and cancer stage.