Non invasive parameters- validity for the prediction of esophageal varices in cirrhotic patients

Aim & Backround: Approximately 5–15% of cirrhotics per year develop varices, and it is estimated that the majority of patients with cirrhosis will develop varices over their lifetime. Variceal hemorrhage is an immediate life-threatening problem with a 20–30% mortality associated with each episode of bleeding. Present study is planned to predict the presence of esophageal varices by means of simple biochemical and ultrasonographical parameters which may be cost effective.
Methods: We analyzed four parameters including platelet count alone, portal vein diameter, and spleen diameter and platelet count / spleen diameter ratio in 50 cirrhotic patients (male: female ratio 30:20) for the prediction of esophageal varices. 
Results: We found the sensitivity, specificity, positive predictive value, negative predictive value of each parameter as given below.  For platelet count alone with a cut off value of 90.27×103 /cu.mm is (33%, 100%, 100% and 27%), Portal vein diameter alone with cut off value of 14.23mm is (69.23%, 81.8%, 93.11% and 42.8%), Spleen diameter alone with a cut off value of 143.7 mm is (74.35%, 63%, 87.87% and 41.17%) and Platelet count and spleen diameter ratio with a cut off value of  0.93 (×103) is 79.5%, 72.73%, 91.18% and 50%). 
Conclusion: Even though all the non invasive parameters can predict the presence of esophageal varices, the negative predictive value of all non invasive parameters were very low. These parameters cannot exclude the presence of esophageal varices. So the non invasive parameters cannot be an alternative for upper GI endoscopy.