Comparison of 30-day, 90-day and in-hospital postoperative mortality for eight cancer sites (Br J Surg 2012; 99: 1149–1154)

This paper by Damhuis and colleagues provides a complete picture of the outcomes of major cancer surgery in a modern Western society with a high reputation for its health services. The numbers involved are large and would seem to allow little chance of bias. Few would disagree with the statement that survival after a major cancer operation, such as oesophagectomy, gastrectomy, colectomy or cystectomy, is hardly worthwhile if that survival is less than 12 weeks. Data from the present study seem to support 90-day mortality as possibly the best way to report true postoperative mortality. Ninety-day mortality was increased compared with the in-hospital rate for most cancer sites. Breast cancer was the only one that had an expectedly low postoperative mortality rate, but even here the 90-day mortality rate was sixfold higher than the in-hospital rate. Although this was from only 0·1 per cent to 0·6 per cent, it still represents the fact that, during this interval, close on 100 patients died following breast surgery after discharge from hospital and within the 90-day period. Perhaps the most surprising finding in terms of mortality rates reported is that gastric cancer had a considerably higher mortality rate than oesophageal cancer. This may be because oesophageal cancer tends to be treated in specialist units, whereas gastric cancer is often resected by general surgeons in a non-specialist setting. One imagines that the health service involved might want to look further into reasons for this finding. Of note, most mortality data reported in the literature are from specialist units, which often have mortality rates of less than 1 per cent for major cancer surgery. However, if such units reported 90-day mortality rates, the figures may not be quite as flattering. Importantly, this large Dutch study has reported realistic mortality rates of 9·7 per cent for oesophageal cancer, 8·8 per cent for colonic cancer and a seemingly very high rate of 13·7 per cent for gastric cancer. Surgeons are not the only ones who should be reporting 90-day mortality rates. These should also be reported for medical oncological treatments and, again, it is likely that they would reveal a different picture. There is no doubt that mortality rates for major cancer surgery have slowly fallen, but figures such as these leave no room for complacency. The author declares no conflict of interest.