Impact of clinical symptoms and referral volume on endoscopy for detecting peptic ulcer and gastric neoplasms

Background: We investigated the volume of dyspeptic patients referred by general practitioners (GPs) to upper gastrointestinal endoscopy and the impact on endoscopic findings. We also examined the correlation between clinical symptoms and endoscopic findings. Methods: We collected data on patients sent for upper gastrointestinal endoscopy by GPs of 30 healthcare centres in 1996 in our hospital referral area of 260,000 inhabitants. In addition, national and local cancer registries were used to enumerate the gastric cancer cases detected in 1996. Results: The study population consisted of 3378 patients, mean age 58 years (interquartile range 25 years, male:female 1:1.3). Among the 30 healthcare centres, referral volumes for upper gastrointestinal endoscopy varied from 0.6 to 9.2 per 1000 inhabitants per year (median 3.3/1000/year). In healthcare units with `high' (greater than or equal to 3.3/1000/year, 15 healthcare units, 1297 patients) and `low' ( < 3.3/1000/year, 15 healthcare units, 2065 patients) referral volumes, the detection rates were as follows: duodenal ulcer (DU) 3.5% (n = 46) versus 4.0% (n = 83, P = 0.5), gastric ulcer (GU) 4.9% (n = 64) versus 5.3% (n = 110, P = 0.6), gastropathy 43.8% (n = 568) versus 35.6% (n = 736, P < 0.001), gastric cancer 0.5% (n=6) versus 0.5% (n= 11, P=0.8), gastric polyps 2.4% (n=31) versus 1.5% (n = 30, P < 0.05). Independent risk factors for gastric cancer were age (OR 6.5 per decade, 95% CI 2.4-17.9), male sex (OR 5.5, 95% CI 1.8-17.1) and alarming symptoms and/or signs (OR 3.6, 95% CI 1.2-10.7); for GU, Helicobacter pylori (OR 2.6, 95% CI 1.9-3.5) and alarming symptoms (OR 2.0, 95% CI 1.4-2.7); for DU, male sex (OR 1.6, 95% CI 1.1-2.2) and H. pylori (OR 3.9, 95% CI 2.7-5.5); and for gastric polyp(s), age (OR 2.0 per decade, 95% CI 1.1-3.5) and high referral volume (OR 1.7, 95% CI 1.0-2.0). A high referral volume did not associate positively either with the number of peptic ulcers or gastric cancer. Conclusions: Alarm symptoms associate strongly with significant gastric lesions such as GU and cancer. Increased referral volume results in an increased number of gastropathy and gastric polyp(s), but not of peptic ulcer or cancer.