Impact of endoscopic ultrasonography (EUS) on surgical decision-making in upper gastrointestinal tract cancer:: An international multicenter study

Background: Endoscopic ultrasonography (EUS) is an integrated part of the pretherapeutic evaluation program for patients with upper gastrointestinal (GI) tract cancer. Whether the clinical impact of EUS differs between surgeons from different countries is unknown. The same applies to the potential clinical influence of EUS misinterpretations. The aim of this study was to evaluate the interobserver agreement on predefined treatment strategies between surgeons from four different countries, with and without EUS, and to evaluate the clinical consequences of EUS misinterpretations. Methods: One hundred patients with upper GI tract cancer were randomly selected from all upper GI tract cancer patients treated at Odense University Hospital between 1997 and 2000. Based on patient records and EUS database results, a case story was created with and without the EUS result for each patient. Four surgeons were asked to select the relevant treatment strategy in each case, at first without knowledge of the EUS and thereafter with the EUS result available. Interobserver agreement and impact of EUS misinterpretations were evaluated using the actual final treatment of each patient as reference. Results: Three of four or all four surgeons agreed on the same treatment strategy for nearly 60% of the patients with and without the EUS results. Treatment decisions were changed in 34% based on the EUS results, and the majority of these changes were toward nonsurgical and palliative treatments (85%). Interobserver agreement was relatively low, but overall EUS increased kappa values from 0.16 ("poor") to 0.33 ("fair"), thus indicating increased overall agreement after the EUS results were available. EUS conclusion regarding stage or resectability was wrong in 17% of the cases, but only one serious event would have been the clinical result of EUS misinterpretations. Conclusion: Despite being used in different ways by different surgeons, EUS did change patient management in one third of the cases. The impact of EUS misinterpretations seemed very low, and this study confirmed one of the strongest clinical possibilities of EUS, i.e., the ability to detect nonresectable cases. EUS is an important imaging modality for oncosurgeons from different countries.