The first 72 hours of acute pancreatitis: patterns of variability in early clinical decision-making

To assess real-world adherence to evidence-based international guidelines in the early management of acute pancreatitis (AP), and to identify how variability in clinical decision-making is distributed during the first 72 h of care.

This multicentre, cross-sectional survey was conducted among consultant physicians and trainees in General Surgery, Gastroenterology, and Emergency Medicine actively practising in Catalonia, Spain. An expert-reviewed 21-item questionnaire assessed self-reported practices across eight key domains of early AP management relative to international consensus recommendations (IAP/APA and WSES). Factors associated with guideline adherence were analysed using chi-square tests and multivariable logistic regression.

A total of 165 clinicians clinicians completed the survey. Self-reported adherence to key recommendations was low, particularly for target-directed fluid resuscitation and early oral feeding, while more than half of respondents reported antibiotic use in the absence of confirmed infection. Crucially, clinical practice variability was not uniformly distributed. The greatest differences were observed in time-sensitive, treatment-related decisions, particularly fluid resuscitation and nutritional strategies, where adherence varied significantly according to clinician experience and hospital size. In contrast, antibiotic use showed no significant variation across subgroups, indicating a consistent and widespread deviation from guideline recommendations. On multivariable analysis, overall adherence was independently associated with clinician experience (p = 0.035) and hospital size (p = 0.009), but not with medical speciality or clinician age.

Early AP management showed substantial variation in self-reported practice across clinical domains. Variability was greatest for time-sensitive therapeutic decisions, whereas antibiotic overuse appeared to represent a widespread pattern of non-concordant practice across subgroups. Structured local pathways and decision-support tools may help reduce these implementation gaps.