A blood pressure of 88/52 mmHg and repeated hematemesis demand action. An INR of 1.8 and platelet count of 45 × 10⁹/L may prompt an automatic order for plasma and platelets—but normalizing those numbers is not the same as controlling a variceal bleed. The focused decision is whether either result alone justifies transfusion or delaying endoscopic hemostasis. In suspected portal-hypertensive bleeding, generally, it does not.
Consider a 54-year-old man with ascites and other signs of chronic liver disease who vomits bright red blood. His heart rate is 112/min, hemoglobin 8.2 g/dL, INR 1.8, and platelet count 45 × 10⁹/L. His presentation makes variceal bleeding plausible, but endoscopy is needed to identify the source. The immediate priorities are perfusion, airway assessment, and bleeding control—not a normal INR.
The history is incomplete: there are no prior blood counts or coagulation results, medication list, or prior bleeding history. Establish baseline cognition and function, check for anticoagulants or antiplatelet drugs, and look for possible additional contributors such as infection or acute kidney injury. Do not assume every abnormal value is new or caused only by cirrhosis.
Read the results without turning them into transfusion orders
The INR reflects selected clotting-factor activity in a laboratory assay. In cirrhosis, both procoagulant and anticoagulant factors can be reduced, while changes in platelets, endothelium, and fibrinolysis also affect clotting. The resulting hemostasis is often described as rebalanced—but that balance is fragile and can tip toward either bleeding or thrombosis. An INR alone does not measure the patient’s overall clotting function or predict whether variceal bleeding will stop.
A platelet count is useful context, but it is not a direct test of platelet function or a stand-alone measure of whether band ligation will succeed. The distinction is between identifying an abnormality and proving that correcting it will improve the outcome.
| Finding | What it tells you | What it does not establish |
|---|---|---|
| INR 1.8 | The prothrombin time is prolonged; liver dysfunction may contribute. | That plasma will restore hemostasis, or that an INR target must be reached before endoscopy. |
| Platelets 45 × 10⁹/L | Thrombocytopenia is present and may relate to portal hypertension or other causes. | That platelets must be transfused to exceed 50 × 10⁹/L before variceal treatment. |
| Hemoglobin 8.2 g/dL | The measured concentration at this point in the bleed. | That the patient is stable or that ongoing blood loss can be managed by waiting for a threshold. |
Don’t transfuse plasma or platelets just to cross a number
For acute variceal bleeding, the Baveno VIII consensus does not recommend routine fresh frozen plasma (FFP). FFP does not correct cirrhotic coagulopathy and may cause volume overload or worsen portal hypertension. There is no INR target that must be reached before endoscopic variceal ligation.
The same caution applies to a platelet count of 45 × 10⁹/L. A count below 50 × 10⁹/L is often treated as a trigger in other clinical contexts, but that does not make it a universal transfusion threshold for active variceal bleeding. Baveno VIII states that platelet count and fibrinogen levels do not correlate with the risk of failure to control bleeding or rebleeding, and their transfusion is not routinely recommended. If bleeding control fails, platelet or fibrinogen replacement may be considered case by case; that is not a reason to automatically correct the lab panel before endoscopy.
Viscoelastic testing, such as thromboelastography, may help characterize clot formation when available. A small randomized trial in cirrhotic patients with variceal bleeding found that a test-guided strategy reduced product use without compromising initial endoscopic bleeding control. The evidence does not make such testing a requirement for every patient, and it should not delay source control.
Keep red-cell decisions separate
A restrictive red-cell strategy generally targets a hemoglobin of about 7–8 g/dL in acute variceal bleeding. That is a target for an individualized strategy—not an instruction to withhold blood from a patient with ongoing major hemorrhage, shock, or significant cardiac ischemia until the hemoglobin falls below 7. Hemoglobin may not reflect the full effect of very recent blood loss, so assess perfusion, the bleeding trajectory, and comorbidities alongside the result.
For this patient, the hypotension and recurrent hematemesis matter more than the isolated hemoglobin value. Resuscitate to maintain organ perfusion, prepare compatible blood, and use the institutional major-hemorrhage approach if his clinical course warrants it. Do not confuse that decision with giving plasma to normalize INR or platelets to reach an arbitrary count.
Move toward hemostasis, not laboratory perfection
The team should proceed with parallel care: close hemodynamic monitoring and resuscitation; prompt airway assessment, with attention to active hematemesis or altered consciousness; vasoactive therapy and antibiotic prophylaxis when variceal bleeding is suspected; and urgent endoscopic evaluation and treatment after initial stabilization. If banding is indicated, an INR of 1.8 or platelet count of 45 × 10⁹/L alone should not be used to postpone it while products are given to reach a target.
Common traps are treating a prolonged INR as a bleeding-risk score, applying a platelet threshold from a different procedure as though it were a rule for variceal hemorrhage, and interpreting restrictive red-cell transfusion as a rigid cutoff even during shock. The opposite mistake is just as important: rebalanced hemostasis does not mean cirrhosis protects the patient from bleeding. Portal pressure, the bleeding source, perfusion, and the clinical trajectory remain central.
Practical takeaways
- In suspected variceal bleeding, do not routinely give FFP just to correct the INR; no INR target is required before endoscopic hemostasis.
- A platelet count of 45 × 10⁹/L alone does not establish a need for platelet transfusion before variceal banding.
- Treat red-cell thresholds as a strategy to individualize, not a reason to wait during ongoing hemorrhage or shock.
- Do not delay source control for laboratory normalization; reassess for other contributors if bleeding remains uncontrolled.
