Comparison between sonographic and plain radiography in the diagnosis of small bowel obstruction at Mulago Hospital, Uganda.

OBJECTIVE To evaluate the role of sonography compared to plain radiography in the diagnosis of patients with suspected small bowel obstruction as well as to determine their specificity, sensitivity and accuracy in the Ugandan setting. DESIGN Prospective comparative study. SETTING Mulago Hospital, Uganda's main referal Hospital and Makerere University Medical School. SUBJECTS Seventy patients with suspected small bowel obstruction (SBO) were evaluated at Mulago Hospital between December 1998 and January 2000, 42 patients were males and 28 females. Patients were aged three days to two years and had a clinical suspicion of SBO. RESULTS Of the 70 patients, 55 were confirmed to have had SBO. Four had ileus and 11 had no SBO. Sixty four per cent were managed surgically while 36% were managed conservatively. Sonography made a diagnosis of SBO in 92.7% and detected strangulation in 71% of patients. Plain radiography made a diagnosis in 85.5% of patients and did not detect strangulation. Specificity was 100% for both, sensitivity 93%, PPV 100%, NPV 73% for sonography. Sensitivity was 85%, PPV 100%, NPV 58% for plain radiography. MAIN OUTCOME MEASURES The accuracy of sonography was 93% as compared to 87% for plain radiography. The level of obstruction was correctly predicted in 81% by sonography and 64% with plain radiography. CONCLUSION It was concluded that sonography is as accurate, specific and sensitive as plain radiography in the diagnosis of SBO as well as determining the level of obstruction in a Ugandan setting. Sonography was found to be more accurate at determining the cause of obstruction and detecting strangulation. It is recommended that sonography should be the initial imaging modality for patients presenting with suspected SBO and plain radiography should only be used as a complimentary investigation.