Possible role ofHelicobacter pylori serology inreducing endoscopy workload

Summary:Wevalidated acommercial enzyme-linked immunosorbent assay (ELISA), Helico-G, in diagnosing H.pylori in129patients (meanage50years, range15-86). We analysed theresults of endoscopy against serology toseewhether there wasapossibility ofadopting thestrategy ofnot endoscoping dyspeptic subjects under theageof45.H.pylori infection wasconsidered present ifeither histology and/or culture werepositive. TheELISAhadasensitivity of88%,specificity of72%,positive predictive value of85%,negative predictive value of77%andaccuracy of82%indetecting H.pylori. Ina subgroup of52subjects aged45orless (mean age35years, range 15-45), 17outof25patients with positive endoscopic findings wereH.pylori seropositive while 16outof27patients hadnormal endoscopic findings. Eighteen outofthe52patients (35%)wereH.pylori seronegative andnormal endoscopicaily except forfive patients (10%)whohadmildtomoderate oesophagitis andtwowhohadnon-erosive gastritis (4%). Allpatients with duodenal ulcer disease (7)wereseropositive giving predictive values of positive andnegative serology foradiagnosis ofduodenal ulcer disease as28%and100%,respectively. Therefore adopting astrategy ofendoscoping subjects under theageof45only iftheywereH.pylori seropositive wouldhavesaved 35%ofendoscopies inthis agegroup butmissed oesophagitis in10%. Negative serology wouldtendtoexclude duodenal ulcer disease while positive serology discriminates poorly forit. Serology maybeauseful adjunct inscreening toreduce endoscopy workload provided that patients withgastro-oesophageal reflux symptoms areexcluded.