Platelets Falling by the Hour: Neuraxial Anesthesia in Evolving HELLP

I would not refuse and I would not simply agree - I would first get more information, and I would say why out loud.

What supports proceeding now:

• There is no absolute platelet cutoff for neuraxial anesthesia. A stable count in the high five figures to low six figures is widely accepted as adequate for a spinal or epidural in obstetrics, and 92,000 with normal INR and fibrinogen is above the range at which most practitioners would refuse.
• The alternative - general anesthesia with a difficult airway and a hypertensive crisis at induction - is not obviously safer in a preeclamptic.

What worries me:

• The trend is the story. Falling from 210,000 to 92,000 in 48 hours indicates a consumptive process, not a benign gestational thrombocytopenia. A falling count may be a snapshot on the way down, and platelet function in HELLP can be abnormal even when the number looks acceptable.
• Fibrinogen 280 is not reassuring in pregnancy, where the normal is much higher; it may be an early sign of consumption.

What I do in the next 20–30 minutes:

1. Repeat the platelet count, and check fibrinogen, and order a viscoelastic test if available (TEG/ROTEM) - a normal MA may allow me to proceed with more confidence.
2. Examine for bleeding, petechiae, bruising, and check the blood pressure trend and neurologic status.
3. Verify blood products are available and crossmatched.
4. Parallel-track: tell the team that if the repeat count is falling further, I will convert the plan to general anesthesia; if it is stable at or above the current level with a normal fibrinogen, I will proceed with a single-shot spinal, not a CSE - a spinal catheter is an indwelling foreign body in a patient whose coagulopathy is evolving, and the benefit of incremental dosing does not justify it here.

Decision rule I say out loud: "Stable count and normal fibrinogen - spinal. Falling count or low fibrinogen - general anesthesia, and we prep for a difficult airway."

Magnesium and neuromuscular blockade:

• Magnesium reduces presynaptic acetylcholine release by competing with calcium at the motor nerve terminal, and it also decreases the sensitivity of the postjunctional membrane. The net effect is potentiation and prolongation of nondepolarizing relaxants.
• Therefore: reduce the rocuronium dose (roughly by a third to a half), give it in small increments, and dose to a train-of-four rather than to a number on the vial. Do not give a full intubating dose and walk away.
• For succinylcholine, I give the standard dose (approximately 1–1.5 mg/kg). Magnesium may modestly prolong blockade, but the priority is a reliably secure airway; under-dosing a relaxant to protect the train-of-four is the wrong trade in a patient with a likely difficult airway.
• Reversal: at the end, this patient will still be under the influence of magnesium. I would use sugammadex for rocuronium if there is any doubt rather than relying on neostigmine, and I would not extubate on a train-of-four alone - I would want sustained tetanus and a clear clinical picture.

Concurrent monitoring:

• Magnesium toxicity: respiratory rate, deep tendon reflexes, urine output (magnesium is renally cleared and she has a creatinine of 1.02), and magnesium levels. Keep calcium gluconate drawn up and immediately available; a lost reflex or falling respiratory rate is a stop signal.
• Blood pressure lability: magnesium causes vasodilation and can worsen hypotension at induction; have phenylephrine ready and titrate.
• Uterine tone: magnesium is a tocolytic and increases the risk of postpartum hemorrhage - I expect it and prepare for it.
• Fetal/neonatal: the neonate will be magnesium-loaded - floppy, with poor respiratory effort. I notify the neonatal team before delivery.
• Ongoing coagulopathy: repeat counts and fibrinogen, and have a massive transfusion protocol on standby.

First, treat it as hemorrhage and resuscitate in parallel: two large-bore IVs, a fluid/blood warmer, send CBC, fibrinogen, and coagulation studies, transfuse to a balanced resuscitation target (favor plasma and platelets early; fibrinogen below 200 mg/dL in an obstetric hemorrhage triggers cryoprecipitate or fibrinogen concentrate), and give tranexamic acid 1 g within three hours of the onset of bleeding.

Uterotonics, in order, with the reasoning specific to this patient:

1. Oxytocin - first line. Give it as an infusion rather than a rapid bolus, because bolus oxytocin causes vasodilation and tachycardia, and she is already hypotensive at 88/50. Reason: the safest, most effective first agent, but the route matters in a patient who cannot tolerate an SVR drop.
2. Carboprost (PGF2α) - second line, with a caveat. Contraindicated or used with great caution in asthma, and it can cause bronchospasm and hypoxemia. Ask about asthma specifically.
3. Misoprostol - useful adjunct, but watch for fever and shivering, which mimic other problems in a preeclamptic.
4. AVOID methylergonovine. This is the key judgment point. Ergot alkaloids cause generalized vasoconstriction and hypertension and are contraindicated in hypertension and preeclampsia - giving it to a woman with a blood pressure of 172/108 risks a hypertensive crisis, intracranial hemorrhage, and seizures.

In parallel: escalate to the surgeon for mechanical measures (bimanual compression, balloon, uterine artery embolization, hysterectomy as a last resort), recheck the platelet count and fibrinogen during the resuscitation, and keep the magnesium running while watching for toxicity with an acute kidney injury on top of blood loss.