Abstract
In this article, the accuracy of ultrasonic screening for abdominal cancer (U/S) was analyzed based on available literature, and issues to ascertain its accuracy were discussed. To obtain relevant literatures, we used Ichushi-Web, PubMed, and Journal of Gastroenterological Cancer Screening. For each article selected, sensitivity, specificity, and positive predictive value (PPV) with their confidential intervals were obtained. Based on this analysis, the sensitivity of U/S was approximately 80%, the specificity was approximately 95%, and the PPV was 1 to 9%. The reason for inter-literature variance was considered to be from differences in how the subjects were counted, target organs tested, ultrasound instrument machinery, its operators, scanning methods, time spent on the examinations, recording methods, methods of interpreting, diagnosing criteria, definitions of false negatives, and definitions of cancers found. Standardizing such variances is needed. Also, the methods of ascertaining accuracy were different as well, and improving registration rate to Japanese population-based cancer registry is needed so verification with cancer registry can be done more smoothly.
