Hepatic Staging in Cirrhosis: The Evidence Behind the Pathway

A 46-year-old lorry driver is admitted at night, deeply jaundiced, his abdomen tight with ascites. He has been drinking heavily for fifteen years. For a week he has been taking diclofenac from the pharmacy for back pain. His sodium is 127 mEq/L. His prothrombin time is 24 seconds against a control of 13, and his bilirubin is 11 mg/dL. Over the last two days his creatinine has gone from 0.9 to 1.6 mg/dL. This is an illustrative case, not a real patient.

By the end of this article you will know what his sodium is telling you. You will also know whether he should have steroids tonight, and what to make of his kidneys.

1. What the Sodium Is Telling You

“His sodium is low. Should I give him salt?” Almost never, and the reason is the reason sodium sits inside the MELD-Na score at all. The tool used to explain it in its Evidence tab, as its section 1:

1. Pathophysiology: Why Sodium Is in MELD-Na

The original MELD score used Bilirubin, INR and Creatinine alone. Sodium was added because a low sodium in cirrhosis is not really a sodium problem. In advanced cirrhosis, portal hypertension drives the release of systemic vasodilators, nitric oxide chief among them. The splanchnic bed dilates, the effective arterial blood volume falls, the baroreceptors read that as underfilling, and non-osmotic Anti-Diuretic Hormone (ADH) is released whatever the plasma osmolality is doing. The ADH holds on to free water, and the sodium falls by dilution. So a serum sodium < 135 mEq/L is a reading of how far the portal hypertension has gone, and it predicts survival on its own, over and above the other three terms. That is what earned it a place in the allocation formula.

Let us take that paragraph a step at a time. MELD is the Model for End-Stage Liver Disease. Splanchnic means the blood vessels of the gut. When they open wide, the blood pools there. The arteries elsewhere are underfilled, even though the body holds more water and salt than it needs. The brain reads underfilling and releases antidiuretic hormone, the hormone that tells the kidney to hold water. The kidney obeys. The water dilutes the sodium.

So our patient’s sodium of 127 is not a shortage of salt. It is a measure of how far his portal hypertension has gone. That is why it makes his MELD-Na worse, and why it predicts his survival beyond the bilirubin, the INR and the creatinine. The tool prints his MELD-Na, which estimates 90-day mortality, beside his Child-Pugh class, which is what operative risk and most drug dosing advice were written against. The two scores answer different questions and will not always point the same way.

2. Steroids Tonight? First the Maddrey Function

A jaundiced heavy drinker may have alcoholic hepatitis on top of his cirrhosis. Maddrey’s discriminant function, mDF, weighs it: 4.6 times the prolongation of the prothrombin time, plus the bilirubin. For our patient, that is 4.6 x (24 - 13) + 11 = 4.6 x 11 + 11 = 61.6.

A Maddrey’s Discriminant Function (mDF) of 32 or more defines severe alcoholic hepatitis with a high 1-month mortality. That is the threshold at which corticosteroids, prednisolone 40 mg/day, are classically given. His 61.6 is far above it.

“Then I start prednisolone now?” Not yet. The arithmetic cannot tell whether the patient is also infected. Before those steroids are started, exclude the occult infections that present quietly in a cirrhotic: UTI, SBP and pneumonia. Think about latent tuberculosis, which in this population is a live question and not a formality. A cirrhotic patient with untreated sepsis has little defence left, and steroids take away what remains. Once steroids are given, the tool’s result asks for the Lille score on day 7: at 0.45 or above, the patient is not responding, and the steroid is stopped.

3. His Kidneys: Three Illness Scripts

A creatinine rise of 0.3 mg/dL or more within 48 hours is acute kidney injury. His went up by 0.7. In cirrhosis, three causes are worth separating, and the tool sets them side by side:

Notice that hepatorenal syndrome is the same splanchnic vasodilation that lowered his sodium, taken one step further. The kidney is starved, not damaged.

“So he has hepatorenal syndrome?” You cannot call it that tonight. The International Club of Ascites criteria, which the tool’s second tab checks, ask for cirrhosis with ascites and an acute kidney injury. They also ask for no improvement after at least 2 days of diuretic withdrawal and volume expansion with albumin, 1 g/kg/day. And they ask for no shock, no nephrotoxins and no macroscopic kidney disease. He has had a week of diclofenac. Until that is stopped and the albumin challenge has run its two days, the diagnosis is not open.

4. Why the Diclofenac Matters

The tool’s instruction is plain: avoid NSAIDs and ACE inhibitors in cirrhosis. Either one can tip a compensated patient into acute renal failure, and it happens quickly. The reason used to sit inside that instruction:

In cirrhosis the kidney is already vasoconstricted, and what holds the afferent arteriole open is locally made prostaglandin. That is a single point of failure. An NSAID blocks the prostaglandin and the afferent arteriole closes down; an ACE inhibitor removes the efferent tone that was propping up the filtration pressure.

The afferent arteriole is the small artery that brings blood into the glomerulus. The efferent arteriole carries it out. In a cirrhotic kidney, prostaglandin keeps the inflow open and angiotensin keeps the outflow tight. Between them they hold up the pressure that drives filtration. An NSAID shuts the inflow. An ACE inhibitor relaxes the outflow. Either way the pressure falls, and so does the filtration.

5. The Common Mistake

The common mistake is to treat the numbers one at a time: salt for the sodium, steroids for the Maddrey function, fluids for the creatinine. It is easy to see why. Each number looks like its own problem, and each has its own familiar remedy. But in this man they are one problem, a dilated splanchnic circulation, seen from three sides. And the most dangerous drug on his chart is the painkiller he bought himself.

Back to Our Patient

The diclofenac is stopped. His diuretics are held and he receives albumin, 1 g/kg/day, while his creatinine is watched. Blood and urine cultures, a diagnostic ascitic tap and a chest X-ray go off before any steroid, and latent tuberculosis is considered. His sodium is not chased with salt. If the cultures are clean and the tap shows no SBP, he is a candidate for prednisolone. His mDF is 61.6, and the Lille score follows on day 7.

Try This

A 52-year-old woman with alcohol-related liver disease has a prothrombin time of 17 seconds against a control of 12, and a bilirubin of 6 mg/dL. Does she meet the threshold for steroids? Illustrative case.

Answer. Her mDF is 4.6 x (17 - 12) + 6 = 23 + 6 = 29. That is below 32. Corticosteroids are generally not indicated below this threshold. What changes her outcome is nutritional support and strict abstinence.

Take-Home Points

• A low sodium in cirrhosis is dilution by antidiuretic hormone, a reading of portal hypertension; do not chase it with salt.
• An mDF of 32 or more is severe alcoholic hepatitis: exclude UTI, SBP, pneumonia and latent TB before prednisolone 40 mg/day.
• Hepatorenal syndrome does not improve with volume; pre-renal azotaemia does.
• Hepatorenal syndrome cannot be called until nephrotoxins are stopped and albumin has been given for 2 days.
• Avoid NSAIDs and ACE inhibitors in cirrhosis: either can tip the kidney into acute failure.

If the liver disease is alcohol-related, two more tools apply: the alcohol withdrawal pathway, because he will withdraw, usually on day 2 or 3, and the alcohol use assessment. For patients and families, the author’s The Liver Dialogue tells the same story from the liver’s side.

Declaration: the author built and maintains the tool. It is free to use. The figures in this article are those the tool carried in its Evidence tab on 5 November 2026, each from the source the tool names. The illustrative cases follow the tool’s rules; check the tool’s result for a real patient.

References

1. Pugh RN, et al. Transection of the oesophagus for bleeding oesophageal varices. Br J Surg. 1973;60(8):646-649.
2. Organ Procurement and Transplantation Network. OPTN Policies, Policy 9: Allocation of Livers and Liver-Intestines. Rockville, MD: US Department of Health and Human Services, Health Resources and Services Administration; 2023.
3. Maddrey WC, et al. Corticosteroid therapy of alcoholic hepatitis. Gastroenterology. 1978;75(2):193-199.
4. Angeli P, et al. Diagnosis and management of acute kidney injury in patients with cirrhosis: revised consensus recommendations of the International Club of Ascites. J Hepatol. 2015;62(4):968-974.
5. Kamath PS, Wiesner RH, Malinchoc M, et al. A model to predict survival in patients with end-stage liver disease. Hepatology. 2001;33(2):464-470.
6. Kim WR, Biggins SW, Kremers WK, et al. Hyponatremia and mortality among patients on the liver-transplant waiting list. N Engl J Med. 2008;359(10):1018-1026.
7. Kim WR, Mannalithara A, Heimbach JK, et al. MELD 3.0: The Model for End-Stage Liver Disease updated for the modern era. Gastroenterology. 2021;161(6):1887-1895.e4.
8. Organ Procurement and Transplantation Network. Updates to medical urgency scoring of liver transplant candidates in effect. Effective 13 July 2023. Rockville, MD: HRSA; 2023.
9. Thursz MR, Richardson P, Allison M, et al. Prednisolone or pentoxifylline for alcoholic hepatitis (STOPAH). N Engl J Med. 2015;372(17):1619-1628.
10. Louvet A, Naveau S, Abdelnour M, et al. The Lille model: a new tool for therapeutic strategy in patients with severe alcoholic hepatitis treated with steroids. Hepatology. 2007;45(6):1348-1354.
11. Sarin SK, Choudhury A, Sharma MK, et al. Acute-on-chronic liver failure: consensus recommendations of the Asian Pacific Association for the Study of the Liver (APASL): an update. Hepatol Int. 2019;13(4):353-390.
12. European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of patients with decompensated cirrhosis. J Hepatol. 2018;69(2):406-460.