Background and aim: Endoscopy is the method of choice to determine the presence of esophageal varices (EV) in cirrhosis, however, it is invasive and expensive. Having a simple and non-invasive method would allow optimizing endoscopic procedures, minimizing costs and complications. The platelet / spleen diameter ratio (P / DB) has shown utility in predicting the presence of EV. The objective of this study is compare the P / DB index with endoscopic findings to predict the presence of EV in cirrhotic patients and to determine the best cut-off point in our population.
Material and methods: Retrospective study conducted from January 2015 to January 2019. Inclusion criteria: adults with a diagnosis of cirrhosis, sent for screening endoscopy of VE for the first time in our hospital and with complete clinic files. Patients with a history of non-cirrhotic portal hypertension, history of variceal hemorrhage and under treatment for primary or secondary prophylaxis were excluded. Variables analyzed: sex, age, etiology, Child Pugh score (CP), endoscopic findings. The P / DB index was determined by dividing the number of platelets by the maximum bipolar diameter of the spleen in millimeters. Statistical analysis was performed using Excel and SPSS.
Results: Of 455 files, 155 met inclusion criteria, 84/155 (54.1%) women and 71/155 (45.8%) men. Mean age: 60 years+29 years. Etiology: alcohol was the most frequent in 45/155 (29.0%) followed by HCV in 33/155 (21.2%). 131/155 (84.5%) had VE, of these 82 (62.5%) were CP A, 36 (27.4%) CP B, and 13 (9.9%) CP C. The P / DB index with cutoff point on 1101 (n / mm3) / mm obtained a sensitivity of 73%, specificity of 61.9%, and AUC of 73%.
Conclusions: Previous studies suggest a cut-off point of the P / DB index of <909 (n / mm3) / mm, in our study the cut-off point identified with the best sensitivity and specificity was <1101 (n / mm3) / mm, undoubtedly it is not ideal. The results could change or be confirmed with a larger study population, in order to have a simple tool that allows optimization of the screening endoscopy.
Conflicts of interest: The authors have no conflicts of interest to declare.
