Document resource
Introduction Pancreatic exocrine insufficiency (PEI) is classically associated with established pancreatic pathology; however, it is increasingly recognised in patients with non-specific gastrointestinal (GI) symptoms in the absence of pancreatic disease. We therefore evaluated the use, diagnostic yield, outcomes and clinician requesting patterns of faecal elastase-1 (FE-1) in an unselected ‘all-comers’ cohort attending secondary-care gastroenterology clinics.Methods A retrospective dual-centre service evaluation was performed in patients attending general gastroenterology clinics who underwent FE-1 testing between 2018–2023 at one centre and 2022–2024 at the other. Patients with established pancreatic or GI disease were excluded. PEI was defined as FE-1 <200µg/g, with severe PEI <100µg/g. Clinical data were extracted from electronic health records. Requesting patterns were assessed by clinician code. Univariable associations were assessed using chi-squared or Fisher’s exact tests.Results A total of 1,486 patients underwent FE-1 testing (median age 59 [IQR 45–70]; 58.2% female). Low FE-1 (<200µg/g) was identified in 155 patients (10.4%), including 73 (4.9%) with FE-1 <100µg/g, and was more common in men (13.0% vs 8.6%; OR 1.60; p=0.007). Increasing age was associated with low FE-1: for every 10-year increase in age, the odds of low FE-1 increased by 22% (OR 1.22 per decade, p<0.001).Weight loss was associated with low FE-1 at both thresholds (<200µg/g: OR 1.56; p=0.014; <100µg/g: OR 2.14; p=0.002). Steatorrhoea showed a non-significant trend towards association with severe PEI (FE-1 <100µg/g; OR 2.48; p=0.052). No other GI symptoms were associated. Diabetes mellitus (type 1 & 2) was associated with low FE-1 (<200µg/g; OR 2.55; p<0.001).Among patients with low FE-1, 133/155 (85.8%) were formally diagnosed with PEI, of whom 128/133 (96.2%) received PERT. Follow up data were available for 84/128 (65.6%). Of these, symptomatic improvement occurred in 69/84 (82.1%), with complete recovery in 13/84 (15.5%). Similar symptom outcomes were observed for FE-1 <100µg/g.Across 41 clinician codes (unique requester identifiers), marked heterogeneity in FE-1 utilisation was observed (range 0–305 requests). The top four codes (9.8%) accounted for 61.6% of FE-1 requests, and the top ten (24.4%) for 83.0%, consistent with a Pareto-type distribution.Conclusion PEI is relatively common in unselected patients attending secondary-care gastroenterology clinics even in the absence of pancreatic disease, and prevalence increases with age. While FE-1 is not a definitive diagnostic test, a substantial proportion of patients benefit from PERT, supporting its pragmatic clinical utility. Highly concentrated requesting patterns suggest FE-1 testing is under-utilised; broader, more consistent use may identify otherwise unrecognised PEI.