[Clinical utility of ultrasound in the identification of dengue hemorrhagic fever].

OBJECTIVE: To evaluate the diagnostic utility of sonography in cases of dengue hemorrhagic fever (DHF) and to describe the correlation grade existing between the most frequent clinical and sonographic findings in DHF.

MATERIAL AND METHODS: Prospective, cross-sectional and descriptive study applied to 132 patients, 21 of them with typical dengue fever and 111 with DHF. They went through ultrasound, searching for thickening of the gallbladder wall, pleural effusion and ascites. Dengue fever diagnosis was confirmed by positive IgM serology in all patients. Descriptive statistics, sensitivity and specificity tests, and phi contingency coefficient were used to obtain the correlation grade between clinical and sonographic findings present in DHF. chi2 was used to determine the statistics significance.

RESULTS: Fever, retroocular pain and cephalea were found in more than 90% of the cases. The positive tourniquet test and petechiae were present in 70%, and thrombocytopenia, leukopenia and elevation of seric transaminases were found in 100% of the cases. Gallbladder thickening was present in 86% of the patients, pleural effusion in 66%, ascites in 60% and acute alithiasic cholecystitis in 36%. Thickening < 3 mm had a sensitivity of 87%, specificity of 48%, positive predictive value of 90% and negative predictive value of 40%. Certain correlation degree was found (r phi = 0.3, p < de 0.05) between thickening of the gallbladder wall > 5 mm and the presence of alithiasic cholecystitis, and a very slight correlation (r phi = 0.2, p < de 0.05) between thickening > 5 mm and free fluid in the abdominal cavity.

CONCLUSIONS: Thickening of the gallbladder wall > 3 mm is a useful sonographic finding to confirm suspicious cases of DHF. The thickening of the gallbladder > 5 mm was just very slightly related with alithiasic cholecystitis and the presence of free fluid in the abdominal cavity, without any relationship with the most frequent laboratory data.