PLATELET COUNT AND SPLEEN DIAMETER RATIO FOR NON INVASIVE DIAGNOSIS OF ESOPHAGEAL VARICES

Abstract

BACKGROUND: Portal Hypertension and its consequences mainly, Esophageal Varies (EVs) is one of the most important causes of morbidity and mortality in patients with cirrhosis of liver. Upper GI endoscopy is the investigation of choice for diagnosis of EVs and periodic endoscopies have been recommended for monitoring of varices. There is a need for non-invasive parameters to detect the presence of EVs. Identication of noninvasive predictors of EVs will help us to carry out EGD in selected groups of patients. Unnecessary endoscopies can be avoided and at the same time, patients who require endoscopy can be referred to a higher center, where facilities for endoscopy are available. Among the non-invasive modalities, the platelet count and bipolar spleen diameter ratio has shown promising results in terms of its accuracy in predicting the presence of Esophageal Varices in many studies MATERIALS AND METHODS: Patients with chronic liver disease diagnosed using clinical, Laboratory and ultrasound parameters were assessed using esophagogastroduodenoscopy for the presence or absence of esophageal varices. USG abdomen was done to assess for bipolar splenic diameter and the presence or absence of EV's was correlated with the platelet count/ splenic diameter ratio, CHILD SCORE, MELD score, Platelet count alone and splenic diameter alone. Platelet count/SD ratio of 909 based on previous studies was correlated with the presence or absence of varices. Statistical analysis was done using IBM SPSS software version 20.0 and variables showing statistically signicant correlations with presence of arices were used for plotting ROC curves to assess the cut of points which could be used for non invasive prediction of varices. RESULTS: The PC/SD ratio cut off (909), based on previous studies for non invasive diagnosis of Esophageal Varices gave sensitivity and specicity of 97.9% and 91.7% respectively, in our study, which was statistically signicant (P value <0.001). The positive predicitive value and negative predictive value of the PC/SD ratio (909) was 96.5% and 94.8% respectively and the accuracy of the test was 96%. ROC curve for Platelet count and Splenic diameter ratio area under the curve is 97.8% with P value < 0.001 and cutoff value 895.02 with sensitivity 96.6% and specicity 96.5%. The Positive predictive value and negative predictive value of PC/SD ratio of 895 was found to be 98.6% and 91.8% respectively and the accuracy of the test 96.5%. ROC curve for Child score in our study, area under the curve 71% with a signicant P value < 0.001, and cut-off value obtained for Child score was 8.50 with sensitivity 64.8% and specicity 63.8 %. ROC curve for MELD score revealed area under the curve was 74.3% with P value as < 0.001, and the cut-off value was 15.5 with sensitivity 67.6 % and specicity 67.2%. ROC curve for Platelet count in our study, the area under the curve was 94.5% with P value as < 0.001, and the cut-off value was 108500 with sensitivity and specicity of 89.7% and 89.4% respectively. The ROC curve for Spleen diameter in our study revealed that the area under the curve was 86.8%% with P value as < 0.001, and the cutoff value was 121 with sensitivity and specicity of 78.9% and 81.0% respectively. CONCLUSION: Among the variables studied for non-invasive diagnosis of Esophageal varices, the Platelet count / Spleen diameter ratio had the best sensitivity and specicity for diagnosing EVs. In view of low sensitivities and specicities for the cut off values obtained for Child score, MELD score, platelet count and spleen diameter, these indices may not be useful as PC/SD ratio in the non-invasive prediction of EV's. The Platelet count / Splenic diameter ratio may be proposed as a safe parameter for diagnosing Esophageal Varices in Chronic Liver disease noninvasively, where resources are limited and endoscopy facilities are not available, to select the patients with probable Esophageal Varices who can be referred to higher centres