ascitic fluid to serum bilirubin ratio for differentiation of exudates from transudates

abstract background: regarding the diagnostic errors of the classic criteria including serum ascites albumin gradient (saag), total protein concentration and the adapted light et al’s criteria in distinguishing transudate versus exudates, we evaluated the ascitic fluid to serum bilirubin ratio as a new criteria in this regard. we also evaluated whether the combination of bilirubin ratio with each of these classic criteria improves the diagnostic accuracy.   methods: one-hundred ascitic fluid specimens were analysed prospectively whereas the category of fluid was assessed according to the clinical diagnosis. sensitivity, specificity, positive predictive value (ppv), negative predictive value (npv), overall accuracy, positive likelihood ratio (lr+), negative likelihood ratio (lr-) and the youden’s index of each criterion alone and in combination with the bilirubin ratio were calculated. results: saag seems to be the best criterion (specificity = 0.9090, ppv = 0.97, lr+ = 8.03, youden’s index = 0.64). bilirubin and ldh ratio criteria had equivalent specificity (0.8333 and 0.8205, respectively), accuracy (0.80 and 0.81, respectively), ppv (0.54 and 0.55, respectively), npv (0.90 and 0.93, respectively) and lr+ (4.09 and 4.30, respectively) which generally were better than fluid total protein and total protein ratio but not as predictive as saag. the addition of bilirubin ratio to any criterion did not make any improvement. conclusion: ascitic fluid to serum bilirubin ratio, although not more predictive than saag, can be used as an alternative criterion in distinguishing exudative versus transudative ascitic fluid as it is more cost-effective in terms of laboratory processing and also more available.