AIMS65 Score Calculator

AIMS65 is a five-item mortality score for acute upper gastrointestinal bleeding, built so that every input is available from the first set of emergency department bloods and observations, before any endoscopy. The components are albumin below 3.0 g/dL, INR above 1.5, altered mental status, systolic blood pressure of 90 mmHg or less, and age 65 or older, each worth a single point. Notably, none of them measures the bleed directly. Albumin and INR act as markers of chronic liver disease and synthetic reserve, altered mentation and hypotension mark systemic decompensation, and age captures physiological reserve, so the score is really quantifying how poorly the patient can tolerate a haemorrhage rather than how large it is. In-hospital mortality climbs steeply across the range, from well under 1 percent at zero points to roughly a quarter of patients at five, and a threshold of 2 or more is the usual trigger for higher-acuity monitoring, earlier endoscopy, and correction of coagulopathy. The score also correlates with length of stay and cost. Its limitations follow from its design: because it predicts death rather than the need for intervention, it should not be used to decide who can be discharged without endoscopy, a question for which the Glasgow-Blatchford score is better suited and considerably more sensitive at its zero-to-one threshold. It has not been validated for lower gastrointestinal bleeding, and the albumin criterion can misclassify patients whose hypoalbuminaemia reflects malnutrition or nephrotic syndrome rather than liver disease.

Risk stratification in upper gastrointestinal bleeding began with the score published by Timothy Rockall and colleagues in Gut in 1996, derived from a national audit in the United Kingdom. It predicted mortality well but required endoscopic findings for the complete score, so it could not guide decisions made in the emergency department. Oliver Blatchford and coworkers addressed that gap in a Lancet paper in 2000 with a purely clinical and laboratory score aimed at identifying patients needing transfusion, endoscopy, or surgery, and it became the standard tool for deciding who could safely avoid admission. AIMS65 was derived by John Saltzman and colleagues and published in Gastrointestinal Endoscopy in 2011. Rather than proposing candidate variables from clinical intuition, the group applied recursive partitioning to a large national database of tens of thousands of admissions for upper gastrointestinal haemorrhage, letting the analysis select the smallest set of routinely captured fields that predicted in-hospital death, then validated the resulting model on a separate cohort of comparable size. The five surviving variables gave the mnemonic that names the score. Subsequent comparative studies, including a multicentre international validation published in 2015, generally found AIMS65 at least as accurate as Rockall or Glasgow-Blatchford for predicting mortality while being simpler to compute, and correspondingly weaker for predicting the need for intervention. Current practice in many centres uses the two scores together for different questions.