Review of the applicability of the glasgow score versus the ranson score in predicting the severity of acute pancreatitis in Hospital Universiti Sains Malaysia

• Main Author: Mohsin, Syed Hasanul-Hadi Syed
• Format: Thesis
• Language: English
• Published: 2013
• Subjects: R Medicine (General); RC799-869 Diseases of the digestive system. Gastroenterology
• Online Access: http://eprints.usm.my/60895/1/SYED%20HASANUL%20HADI%20BIN%20SYED%20MOHSIN%20-%20e.pdf; http://eprints.usm.my/60895/

Acute pancreatitis is a fairly common disease managed by most surgical units. Though a vast majority of patients undergo a mild disease course, up to 25% of patients may suffer from a severe attack, leading to persistent organ failure requiring intensive care. Furthermore, up to another 10-15% may succumb to their disease. As such, it is imperative to determine which patients would need early intensive care to guide management. The Ranson score and Glasgow score are two physiological multiple parameter scoring systems widely used in risk stratification. This study aims to identify which scoring system is better suited at predicting outcomes in patients with acute pancreatitis in HUSM from 2007 to 2011. All patients with acute pancreatitis fulfilling the inclusion criteria were reviewed retrospectively. The Ranson scores documented were noted and the Glasgow score was then calculated based on each patients parameters. Both the Ranson and Glasgow score was categorized as mild if the score was less than 3 or severe if the score was more than 3. Outcomes was categorized as good if the patients were discharged well from hospital, or poor if the patients required ventilation in ICU, either surgical or radiological intervention and death. Chi-squared test was used to determine the association of each test and the sensitivity, specificity, PPV and NPV were determined to see which test was better. 156 patients were included in the study. The Ranson score, which has been in practice in HUSM its inception in the mid-eighties, in this study has yielded a sensitivity of 94.1%, and a specificity of 68.3% (PPV 26.6% and NPV 98.9%). The Glasgow score showed a sensitivity of 88.2% and a specificity of 79.9%, with a PPV of 34.9% and NPV of 98.2%, which were all statistically significant (confidence interval < 0.01). Head to head, the Ranson score in this study has proven to be more sensitive, but less specific. The Glasgow score despite being less sensitive, has proven to be more specific and has a better positive predictive value in risk stratification of patients with severe acute pancreatitis in the local setting. Furthermore the score is simple to use and does not require cumbersome calculations or repeated blood investigations.