Background: The aim of the study was to identify all patients who presented with oesophagogastric malignancy within a single National Health Service region (Wales) over 1 year, and to follow the cohort for 5 years. Management and outcome were analysed to identify current practice and draft guidelines for Wales.
Methods: Patients were identified from hospital records. Details were recorded in structured format for analysis.
Results: Analysable data were obtained for 910 of 916 patients. The overall incidence was 31.4 per 100 000 population. Treatment was by resection 298 (33 per cent), palliation 397 (44 per cent) or no treatment 215 (24 per cent). The 30-day mortality rate was 12 per cent and the in-hospital mortality rate was 13 per cent. Some 226 patients (25 per cent) were alive at 2 years. Resection conferred a significant survival advantage over palliation (P < 0.001) and no treatment. Anastomotic leakage occurred in 16 patients (5 per cent), of whom eight died in hospital. 'Open and close' operations were common (23 per cent), laparoscopy was infrequent (16 per cent), and many surgeons undertook small caseloads. Operating on fewer than six patients per year increased the mortality rate after partial gastrectomy (P < 0.05) and was associated with a trend to a higher mortality rate after mediastinal and cardia surgery. Operating on more than 70 per cent of patients seen resulted in a significantly higher mortality rate (P < 0.01) irrespective of case volume.
Conclusion: Tumour resection conferred a survival advantage. Wider use of laparoscopy is advocated. Improved selection for surgery should result in a lower mortality rate.