The bedside assessment of splenic enlargement.

STUDY OBJECTIVE To evaluate the clinical assessment of splenic enlargement using specific bedside maneuvers including Traube's space percussion, the splenic percussion sign, Middleton's maneuver, supine palpation, and right lateral decubitus palpation. DESIGN Quasi-experimental prospective study of cases and controls selected according to the results of abdominal ultrasonographic examinations. SETTING Selected inpatients of a tertiary care hospital. MAIN RESULTS Comparing the areas under the receiver operating characteristic curves for each bedside maneuver demonstrated that Traube's space percussion and palpation were significant discriminators (p less than 0.001) of splenic enlargement with respective areas of 0.70 +/- 0.04 and 0.76 +/- 0.04. No one palpation maneuver was superior to another, and right lateral decubitus palpation was not useful when performed after supine palpation. The splenic percussion sign (sensitivity 79%, specificity 46%) was no better than Traube's space percussion (sensitivity 62% and specificity 72%) in assessing splenic enlargement. The palpation maneuvers appeared more sensitive and more specific than Traube's space percussion. Palpation was a significant clinical discriminator when performed on patients who exhibited percussion dullness of Traube's space (area = 0.87 +/- 0.04, p less than 0.0001) but was of little value among those without percussion dullness (area = 0.55 +/- 0.08). Also, palpation was significantly more accurate when performed on lean patients versus obese patients (areas = 0.83 +/- 0.04 versus 0.65 +/- 0.08, p less than 0.05). When a positive bedside examination was defined as positive palpation and positive percussion (concordant-positive), the combined test sensitivity and specificity were 46% and 97% respectively. CONCLUSIONS The optimal clinical assessment of splenic enlargement includes the percussion of Traube's space. If Traube's space is dull, palpation of the spleen is warranted. This assessment is most accurate in lean patients.