Tranexamic Acid Before a Percutaneous Tract, in a Woman With No Clot Risk to Speak Of
The transfusion-reduction data for tranexamic acid before a percutaneous kidney tract is real and repeated. So is the fact that its safety record comes almost entirely from patients bleeding for a completely different reason.
Naomi R., a 39-year-old woman, is three weeks out from her age group's qualifying race for the national age-group triathlon championship, and has spent the last month treating her training log the way other people treat a second job. A CT ordered after a bout of low-grade fevers and flank pain found a 2.4cm lower-pole stone with a partial staghorn extension into a lower calyx, and her urine has grown Proteus mirabilis twice in the past year despite two courses of antibiotics — a struvite, infection-driven stone that isn't going to dissolve or pass on its own. She has no history of clotting problems, isn't on hormonal contraception, has never had a personal or family venous thromboembolism, and by her own account is fully weight-bearing and moving within hours of anything short of major surgery. Her only real hematologic wrinkle is a baseline hemoglobin of 11.2, mildly low, attributed to heavy periods she's never had formally worked up — not disease, but not a lot of reserve either.
Percutaneous nephrolithotomy for a stone this size carries a real, well-documented transfusion risk, and tranexamic acid's case for reducing it isn't built on one favorable trial. A systematic review in the Annals of the Royal College of Surgeons of England pooled six randomized trials and 1,323 patients and found the odds of needing transfusion cut by 67% with TXA (OR 0.33, 95% CI 0.21–0.52). Against a starting hemoglobin of 11.2, that is the difference between absorbing an average tract bleed and not. The harm side is where the case actually turns, and it is not blank: Cleveland et al.'s 2023 Cochrane review, pooling ten randomized trials of this exact drug in this exact procedure, found systemic TXA may reduce transfusion, major complications and length of stay — and in the same breath concluded it MAY INCREASE adverse events, naming uncertainty over venous thromboembolism as the most important thing still unknown about using it. The reassurance usually offered against that concern comes from somewhere else entirely: CRASH-2 randomized trauma patients already bleeding massively, and WOMAN randomized women with postpartum hemorrhage, both settings whose baseline coagulation state, injury burden, and volume status have little in common with an elective, planned, ambulatory-recovery procedure in an otherwise healthy 39-year-old.
Pre-op huddle, morning of surgery
A staghorn-component stone this size, in a percutaneous tract, carries a real transfusion risk, and the evidence for TXA reducing it isn't thin — a systematic review of six randomized trials, 1,323 patients, found the odds of transfusion cut by two thirds, OR 0.33. She's starting from a hemoglobin of 11.2, which gives her less room to absorb even an average bleed without needing blood she clearly doesn't want.
I'd give it. The benefit side of this literature is about as consistent as anything we have for PCNL.
You're right that the transfusion numbers are real and repeated across multiple reviews — I'm not disputing the benefit side at all. What I'd flag is where the SAFETY reassurance for this drug actually comes from. CRASH-2 randomized trauma patients already hemorrhaging; WOMAN randomized women with postpartum hemorrhage. Both are patients with active, major, ongoing bleeding — a completely different coagulation and volume state than an elective tract in a well 39-year-old.
But I want to correct something before it hardens into the reason we do this. The PCNL-specific literature is not silent on harm. Cleveland's Cochrane review pooled ten trials of TXA in this exact procedure and concluded it may INCREASE adverse events, with VTE risk named as the single most important open question. So "no signal in trauma and postpartum populations" isn't just evidence about a different population — it's being used to answer a question our own population's evidence has already raised and not resolved.
Then let's be accurate about what that Cochrane signal is and isn't. "May increase adverse events" at low certainty across ten small trials is a reason to name the risk out loud to her, not a reason to withhold a replicated transfusion benefit — and the actual question isn't whether TXA is safe in the abstract, it's whether it's safe in her specifically.
Naomi has no personal or family history of clotting, isn't on anything estrogen-containing, and by her own description will be fully weight-bearing within hours — she sits about as far from a plausible VTE candidate as an elective-surgery patient gets. Against a baseline hemoglobin that leaves her little margin, and a benefit signal that's held up across multiple independent reviews, I'd proceed.
Single intraoperative dose, standard protocol. If anything about her presentation changes between now and the OR — a new symptom, a reason to think her coagulation picture is different than it looks today — that changes the calculus, but nothing in front of us does.
Agreed within minutes: tranexamic acid given as a single intraoperative IV dose. The disagreement wasn't resolved so much as named accurately. The Clinical Pharmacologist's correction stood: the PCNL-specific evidence base is not silent on harm, and the Cochrane review's own conclusion that TXA may increase adverse events is a real finding, not a gap. What the group agreed on is that it is a low-certainty population-level signal with no risk factor in Naomi to attach it to — and that she should be told that in those words rather than told the drug is simply safe.
Standard perioperative VTE precautions (sequential compression devices, early ambulation) proceeded unchanged — nobody argued TXA's use should alter routine prophylaxis in either direction.