The case mortality for acute pancreatitis in the UK has remained large
ly unchanged, at just under 10% for the last 30 years. This is despite
improvements in intensive therapy, radiology and surgical interventio
ns. Acute pancreatitis also remains unpredictable. Organ system failur
e and pancreatic collections may develop either suddenly, or insidious
ly, and thus go undetected. Reducing mortality depends on better under
standing of pathophysiology and more specific therapeutic approaches,
but prognostic systems for the early identification of severe attacks
may improve the success of current supportive therapies. Prognostic sy
stems are also useful to compare clinical series and stratify severity
in therapeutic trials. The multiple criteria of Ranson and Imrie pred
ominate, but offer limited accuracy, involve delay, are cumbersome, an
d provide a one-off overall prediction. Internationally agreed definit
ions of complications demand accurate risk assessment for individual c
omplications, while serial monitoring of severity is needed to assess
progress and to detect subtle changes after therapeutic intervention.
Laboratory methods now provide equal, or improved, accuracy and speed
compared to traditional criteria, and may be repeated serially. Growin
g knowledge of the systemic inflammatory response syndrome (SIRS) and
the availability of response modifiers suggests that inflammatory medi
ators may prove to be the most useful and accurate means of assessment
of severity.