PWE-066\u2005Are we missing cases of pancreatic exocrine insufficiency in hospitalised patients with high alcohol intake?

Abstract

Introduction High alcohol intake is an established risk factor for pancreatic exocrine insufficiency (PEI) and chronic pancreatitis. PEI is associated with maldigestion, weight loss and malnutrition. Faecal elastase-1 (FEL-1) is the most acceptable first-line test for PEI. We aim to study the current practice and yield of PEI in patients with high alcohol intake by testing with FEL-1. We also aim to identify positive and negative predictive factors and improve the testing yield. Methods We prospectively identified patients admitted to hospital with alcohol related problems and tested with FEL-1. In addition to demographic information, data was collected including body mass index (BMI), alcohol intake per week, pack-year smoking history, haemoglobin, platelet, albumin, bilirubin, alkaline phosphatase, alanine-transferase, C-reactive protein, vitamin B12, folate, ferritin, calcium, magnesium, phosphate, zinc, copper and selenium levels. FEL-1 <200 µg/g was taken as a presence of PEI. Results A total of 47 patients with high alcohol intake were recruited and tested with FEL-1 (28 male, mean age 46.3 ±10.9 years). No patients had previously been tested for PEI. Mean alcohol intake per week was 177.1 ± 94.8 units. An FEL-1 level of <200 µg/g was identified in 23.4% (11/47) patients. Potassium levels showed positive correlations with FEL-1 value (p =0.048, R2 = 0.084). A positive correlation was also seen with serum copper (p = 0.0087, R2= 0.48). A higher proportion of biochemical markers with below normal level was seen in low FEL-1 patients compared to FEL-1 ≥200 but without reaching statistical significance, including albumin 54.5% (6/11) vs 36.1% (13/36) p= 0.31, magnesium 63.6% (7/11) vs 53.3% (16/30) p= 0.73, and phosphate 45.5% (5/11) vs 35.3% (12/34) p =0.5 Thirty-three patients were smoker with no difference in mean pack-year smoking history between patients with low FEL-1 and those with FEL-1 ≥0 200 µg/g (10.8 ± 12.3 vs 14.2 ± 15.7, p =0.5189). Also, there was no difference in mean BMI in patients with a low FEL-1 compared to those with FEL-1 ≥200 (28.4.34 vs 26.3 kg/m2, p= 0.4616). Conclusion PEI is common in patients hospitalised with alcohol related problems and not recognised. PEI patients are overweight suggesting that using a low BMI to target patients is insufficient to prompt testing for PEI or malnutrition in this patient group. Serum markers have a correlation but are unhelpful at predicting PEI. However, given the high yield we recommend follow up studies to see if treatment has an impact on quality of life and hospitalisation of patients.