Clinical and biochemical changes during exchange transfusion.

Abstract

<h3>Background</h3> Faecal elastase-1 (FE1) is the only widely available test for pancreatic exocrine insufficiency (PEI). However, FE1 is thought to misclassify approximately 10% of patients. False negatives delay treatment with pancreatic enzyme replacement therapy (PERT). False positives expose patients to unnecessary intervention, and the NHS to unnecessary costs. We studied the practice of repeating FE1 at our trust, its impact on being treated, and the predictors of reclassification of PEI diagnosis on repeat testing. <h3>Methods</h3> We carried out a retrospective study at a London teaching hospital. All outpatients investigated with FE1 between 2012 and 2018 were identified. Demographic and clinical information was retrieved from the electronic medical record. PEI was defined as FE1 &lt;200 μg/g. Where FE1 had been repeated, any change to PEI diagnosis was recorded. Univariable logistic regression was used to explore the dependence of having FE1 repeated and reclassification of PEI diagnosis on age, sex, ethnicity, presenting symptoms, comorbidities, and the initial FE1 result (grouped into FE1&lt;100 μg/g, 100–199 μg/g, 200–299 μg/g and ≥300 μg/g). Exposure variables with significant associations (p&lt;0.05) in the univariable analysis were incorporated into a multivariable logistic regression model. Univariable logistic regression was used to explore the association between having more than one positive FE1 result and being prescribed PERT. Firth’s method of penalized likelihood was used to reduce bias in cases of complete separation. Complete case analysis was used where any data were missing. <h3>Results</h3> 1027 patients were included; mean age 53 years; 42.5% male; 54.5% white ethnicity. In total, 124 patients (12.1%) had their FE1 repeated. The median time to repeat FE1 was 5.4 months. 39 patients (31.5%) had their PEI status reclassified on repeat FE1; 28 patients from PEI to no PEI, and 11 from no PEI to PEI. On univariable analysis, diabetes mellitus, chronic pancreatitis and initial FE1 result were associated with having FE1 repeated. In the multivariable analysis, only initial FE1 result remained a significant predictor of having FE1 repeated (FE1 &lt;100 μg/g: OR 4.66, 95% CI 2.76–7.87; FE1 100–199 μg/g: OR 7.26, 95% CI 4.21–12.5; FE1 200–299 μg/g: OR 3.53, 95% CI 1.88–6.61; all p&lt;0.001). Initial FE1 100–200 μg/g was the only significant predictor of reclassification of PEI diagnosis on repeat testing (OR 6.91, 95% CI 2.39–19.95; p=0.007). Patients with more than one positive FE1 result were almost four times more likely to receive PERT than patients with a single positive result (OR 3.82, 95% CI 1.5–9.75; p=0.005). <h3>Conclusions</h3> False positive and false negative FE1 results are common, and clinicians might be reluctant to prescribe PERT after one positive result. We recommend repeating FE1 routinely in all patients with FE1 &lt;300 μg/g.