The Model for End-Stage Liver Disease (MELD) score was originally developed in 2000 to predict 3-month mortality in patients undergoing TIPS (transjugular intrahepatic portosystemic shunting) procedures, then validated as a superior predictor of waitlist mortality compared to the Child-Pugh score. UNOS adopted MELD for organ allocation in 2002, replacing the Child-Pugh/MELD hybrid system. The score uses three objective laboratory variables: serum bilirubin, international normalised ratio (INR), and serum creatinine — all reflecting different aspects of hepatic and renal function. MELD-Na (incorporating sodium) was adopted by UNOS in 2016 as it better predicts 90-day waitlist mortality.

The Child-Pugh score (originally "Child-Turcotte" classification, later modified by Pugh) was developed in 1964 to assess the prognosis and operative mortality risk in patients with portal hypertension undergoing portosystemic shunt surgery. It incorporates five clinical and laboratory variables: bilirubin, albumin, prothrombin time (or INR), degree of ascites, and degree of hepatic encephalopathy — the latter two being subjective. Despite its age and subjective components, the Child-Pugh score remains clinically relevant and is used in HCC staging (BCLC), drug dosing decisions in liver disease, and assessment of surgical risk.

The Glasgow-Blatchford Score (GBS) was developed by Blatchford and colleagues in 2000 using data from 1,748 patients presenting to a Scottish district general hospital with upper GI bleeding. Unlike the Rockall score, GBS can be calculated entirely at the bedside before endoscopy, using only clinical and basic laboratory variables. Its primary utility is to identify patients at sufficiently low risk (score = 0) to be safely discharged without endoscopy — a powerful triage tool in emergency settings. Multiple studies have shown GBS score of 0 carries a negative predictive value >99% for need for intervention.

The Rockall Score was developed from a large UK national audit of 4,185 cases of acute upper GI bleeding in 1996, making it one of the most rigorously derived scoring systems in gastroenterology. The full post-endoscopy Rockall score incorporates endoscopic findings (diagnosis and stigmata of recent haemorrhage) in addition to clinical parameters, allowing mortality and rebleeding prediction after diagnosis has been established. A pre-endoscopy Rockall score (without endoscopic components) is less accurate but useful for initial risk stratification.

The FIB-4 (Fibrosis-4) Index was originally developed in 2006 by Sterling and colleagues in HIV/HCV co-infected patients at the ACTG A5071 trial, using four simple variables: age, AST, ALT, and platelet count. It was subsequently validated in NAFLD and other chronic liver diseases. The FIB-4 index is the most widely recommended non-invasive test for advanced fibrosis assessment in NAFLD/MASLD, endorsed by the American Association for the Study of Liver Diseases (AASLD), EASL, and the National Lipid Association. It avoids the need for liver biopsy in the majority of patients and is practical in primary care settings.

The AIMS65 score was developed by Saltzman and colleagues in 2011 using data from 29,222 hospitalised patients with acute upper GI bleeding at 187 US hospitals (Premier hospital database). It was designed as an even simpler alternative to the Glasgow-Blatchford Score for predicting in-hospital mortality, length of stay, and cost — all using five variables available at initial presentation. The name AIMS65 is a mnemonic for its components: Albumin, INR, altered Mental Status, Systolic BP, and age ≥65. Its particular strength is predicting in-hospital mortality rather than need for intervention.

The BISAP (Bedside Index for Severity in Acute Pancreatitis) was developed by Wu and colleagues in 2008 from a retrospective analysis of over 18,000 pancreatitis admissions from a national US database. It was specifically designed to be a simple 5-point bedside tool that could predict in-hospital mortality within 24 hours of admission — before CT imaging or 48-hour Ranson's criteria become available. BISAP was validated in a separate cohort of over 16,000 patients, demonstrating comparable discriminatory ability to APACHE-II while being far simpler to calculate. Each point of BISAP approximately doubles mortality risk.

Ranson's Criteria were developed by Dr. John Ranson at New York University in 1974 from a retrospective analysis of 100 patients with acute pancreatitis. The original 11-criteria system (5 measured at admission + 6 at 48 hours) was specifically validated for pancreatitis from any aetiology, with separate criteria for alcoholic pancreatitis. For 50 years, Ranson's Criteria remained the most widely taught and used severity assessment tool in acute pancreatitis — though its 48-hour completion requirement and complexity have led to its gradual replacement by BISAP, APACHE-II, and CT severity index in contemporary practice.

The Mayo Score for Ulcerative Colitis was developed by Schroeder and colleagues in 1987 as the primary clinical endpoint in the first RCT of oral 5-aminosalicylic acid for mild-to-moderate UC. It rapidly became the most widely used activity index in UC clinical trials and clinical practice. The score incorporates four subscores (stool frequency, rectal bleeding, endoscopic findings, physician's global assessment), each scored 0-3, giving a total of 0-12. The endoscopic subscore is the most objective and correlates best with mucosal healing. The partial Mayo score (without endoscopy, 0-9) is used for outpatient monitoring.

MELD-Na (MELD with Sodium) was developed to address the limitations of the standard MELD score, which underestimates waitlist mortality in cirrhotic patients with hyponatraemia. Hyponatraemia (serum sodium <135 mEq/L) is independently associated with a 2-3× increased risk of 90-day waitlist mortality in cirrhosis. Biggins and colleagues (2006) demonstrated that incorporating serum sodium significantly improved the prognostic accuracy of MELD. UNOS officially adopted MELD-Na for organ allocation in January 2016, replacing standard MELD as the primary allocation score. Serum sodium is capped at 125-137 mEq/L to prevent gaming by aggressive sodium correction.