Numbers That Look Like an Emergency, a Patient Who Isn't Bleeding
A septic patient's coagulation panel reads like a transfusion emergency, but nothing on his skin or in his output backs that up — the team has to decide whether to treat the labs or the patient.
F.N., a 54-year-old man whose wife has been sleeping in the recliner beside his bed for two nights rather than go home, was admitted with a perforated diverticulitis and started on source-directed antibiotics and IV fluids after a CT confirmed a walled-off perforation not requiring emergent surgery. He supervises a warehouse floor of eighteen people who, by his account, have been texting his phone in relays for updates. His vital signs have stayed within normal range throughout — blood pressure holding without vasopressors, urine output adequate, a lactate that was normal at 1.4 on admission and is normal again at 0.9 today — the absence of a deterioration signal rather than evidence of a response — but this morning's labs show a platelet count of 62,000/µL, an INR of 1.9, a fibrinogen of 92 mg/dL, and a markedly elevated D-dimer, a pattern consistent with disseminated intravascular coagulation triggered by his underlying sepsis. On exam, there is no new bruising, no oozing from his IV sites, no blood in his drain output, and no mucosal bleeding of any kind — his coagulation panel describes a patient who should be bleeding, and the patient in the bed is not.
That mismatch is exactly what the team is arguing about. ISTH guidance on DIC management is explicit that prophylactic replacement of platelets, cryoprecipitate, or plasma is not recommended in the absence of active bleeding or a high-risk invasive procedure — the abnormal numbers reflect an underlying process that treating the sepsis itself should correct, and correcting the numbers directly without correcting the trigger risks obscuring the actual trend that matters, which is whether his coagulopathy improves as his infection responds to antibiotics. His fibrinogen of 92, though low, is not below the level some clinicians treat as an independent trigger for cryoprecipitate regardless of bleeding status — a threshold more commonly invoked in trauma or peripartum hemorrhage than in a source-controlled, hemodynamically stable medical patient like him.
Rounding on hospital day 2, reading a coagulation panel against an unremarkable exam
Nothing here meets the bar for replacement. ISTH's own DIC guidance is explicit: platelets, cryoprecipitate, and plasma are for active bleeding or an imminent high-risk procedure, neither of which applies to him right now. His actual therapy is the antibiotics already running — the coagulation panel should track his sepsis response, not get corrected independently of it.
I'll push on the fibrinogen specifically — 92 is low enough that I've watched patients look exactly this stable right up until they weren't. Some protocols treat sub-100 fibrinogen as its own trigger for cryoprecipitate regardless of what the rest of the picture shows, precisely to get ahead of that kind of sudden turn.
I take the guideline point seriously, but a guideline written for the average DIC patient isn't automatically right for every individual trajectory, and this one's direction is what worries me.
I'd name where that sub-100 threshold actually comes from, though — it's largely trauma and postpartum hemorrhage literature, populations already bleeding or about to be. He isn't either. Applying a threshold built for active massive hemorrhage to a stable, source-controlled medical patient is importing the wrong comparison, not a more cautious reading of the same evidence.
What I'd agree to is close monitoring rather than prophylactic correction — repeat fibrinogen, platelets, and INR in twelve hours rather than the standard next-morning draw, so a real deteriorating trend gets caught fast, without treating today's snapshot as an emergency it isn't yet.
Agreed: no prophylactic transfusion today, antibiotics unchanged, repeat coagulation panel at 12 hours rather than the standard interval, and an explicit standing order that any new bleeding of any kind triggers immediate replacement without waiting for that recheck.