Prediction of severe pancreatitis in a population with low atmospheric oxygen pressure

Le résumé fourni par la source

To establish the severity of pancreatitis, there are many scoring systems, the most used are the Marshall and APACHE II systems, each one has advantages and disadvantages; but with good relation regarding mortality and prediction of complications. In populations with low barometric pressures produced by a decrease in atmospheric pressure, there is a decrease in partial pressure of oxygen, in these cases scores which take arterial oxygen partial pressure as one of their variables, may be overestimated. A diagnostic trial study was designed to evaluate the performance of APACHE II, Marshall and BISAP in a city 2640 m above sea level. A ROC analysis was performed to estimate the AUC of each of the scores, to evaluate the performance in predicting unfavorable outcomes (defined as the need for percutaneous drainage, surgery, or mortality) and a non-parametric comparison was made between the AUC of each of the scores with the DeLong test. From January 2018 to December 2019, data from 424 patients living in Bogota, with a diagnosis of gallstone pancreatitis was collected consecutively in a hospital in Bogota, Colombia. The ROC analysis showed AUC for predicting adverse outcomes for APACHE II in 0.738 (95% CI 0.647-0.829), Marshall in 0.650 (95% CI 0.554-0.746), and BISAP in 0.744 (95% CI 0.654-0.835). The non-parametric comparison to assess whether there were differences between the different AUC of the different scores showed that there is a statistically significant difference between Marshall and BISAP AUC to predict unfavorable outcomes (p=0.032). The mortality in the group of patients studied was 5.8%. We suggest the use of BISAP to predict clinical outcomes in patients with a diagnosis of biliary pancreatitis in populations with decreased atmospheric pressure because it is an easy-to-use tool and does not require arterial oxygen partial pressure for its calculation.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.