Acute Gastrointestinal Injury in Critically Ill Patients in a South Indian Intensive Care Unit: A Prospective, Observational, Preliminary Study.

Abstract

Introduction The acute gastrointestinal injury (AGI) score was proposed by the Working Group on Abdominal Problems of the European Society of Intensive Care Medicine (ESICM) as a tool to define and grade gut dysfunction. There have not been any studies in India to validate this tool. The objectiveof this preliminary study was primarily to study the frequency of AGI in the first week of ICU stay in critically ill patients in ourintensive care unit (ICU). We also sought to determine the risk factors predisposing to the development of AGIand to determine the prognostic implication of gastrointestinal (GI) injury in critically ill patients. Materials and methods A prospective, observational, preliminary, single-center study was conducted on critically ill patients (APACHE II > 8) who were on enteral tube feeds and admitted to a mixed ICU of a tertiary care hospital. Anthropometric data, admission diagnosis, APACHE II score, and comorbidities were recorded. Data of daily heart rate, mean arterial pressure, dose of vasopressors, intra-abdominal pressure, fluid balance, feeding intolerance, mechanical ventilation, and laboratory tests were noted for the first seven days of ICUstay or till ICU discharge, whichever was earlier. The occurrence of AGIscore (1-4) during the first seven days of critical illness was the primary outcome of interest. Patient outcome at 28 days was recorded and the impact of the occurrence of AGI on patient outcome was analyzed using the Chi-square test. The patient characteristics associated with AGI were characterized as risk factors and analyzed using a multivariate model. Results Data were collected from 33 patients over 201 patient days. The frequency of acute GI dysfunction in the first seven days of ICU stay in our group of patients was 45.45% (15/33). APACHE II, fluid balance, creatinine, and lactate were identified as possible predictors ofGI injury based on existing literature. These four variables were entered into an ordinal logistic regression model to assess their ability to predict the occurrence ofGI Injury. When fitted into a predictive model, only fluid balance and creatinine were predictive of the final model (p-value < 0.05). A greater fluid balance was predictive in the final model of the development ofGI injury;however,it showed negligible clinical significance (OR: 1.00033, 95% CI: 1.000051-1.00061). Lower creatinine levels were predictive in the final model of the development of AGI Injury,as demonstrated by the negative coefficient. Creatinine also had a greater clinical significance (OR: 0.63, 95% CI: 0.44-0.90) in the development ofAGI. The impact of the AGI scores on mortality was analyzed. The number of patient days with higher AGI scores was significantly associated with increased mortality at 28 days (p-value < 0.001). Conclusion The study showed that nearly half of the critically ill patients included in the study developed acute GI dysfunction.We could not identify any predictors ofGI injury based on our results. The result suggested an association between the severity of GI dysfunction and mortality at 28 days.