Co-Author: Paul Rigby, Clinical Educator, West Michigan AirCare, Kalamazoo, Michigan
I Wish I could Make This Stuff Up
In the early hours of a cloudy, overcast Sunday morning, this critical care transport team was called to the Obstetrical floor of a small critical access hospital for a 27-year-old G4P4 female who had given birth via uncomplicated normal vaginal delivery 3 hours earlier. She was now experiencing intractable post-partum hemorrhage. It was reported by referring staff that she had normal prenatal care and no history of complications. This morning, her severe, uncontrollable hemorrhage resulted in cardiac arrest with return of spontaneous circulation (ROSC) 25 minutes later. During the arrest, she had received 1 asynchronous shock, 4 doses of epinephrine, 6 units of uncross matched packed red blood cells (PRBC) and 2 units of fresh frozen plasma (FFP). At this point her estimated blood loss (EBL) is close to 3 liters.
Why I Hate Managing OB patients
No offense to current or expectant mothers but this is exactly why obstetrical patients scare me. In my mind, bleeding of this nature, refractory to fundal massage or some Pitocin, is attributed to one thing until proven otherwise (especially in the setting of peri- or actual cardiac arrest): Amniotic Fluid Embolus (AFE) with disseminated intravascular coagulation (DIC). As we were flying instruments (IFR) in the clouds on our way to this small hospital, that is exactly what we thought we were going to get.
What We Walked In To
Upon our arrival in to the patient’s room on this small, six-bed obstetrical floor, we were met by the nurse and the respiratory therapist, both obviously anxious, the distraught husband and mother crying in the corner of the room. The patient was still attached to the defibrillator. She was intubated and breathing above current ventilatory support. Her mean arterial pressure (MAP) was a soft 50 mmHg with an epinephrine infusion and having just finished her 6th unit of uncross matched PRBC. She was visibly hemorrhaging from her vagina and her fundus was not palpable (a significant point later in the discussion).
After our initial assessment, we attached the patient to our cardiac monitor which displayed the following cardiac rhythm:
At first glance at the monitor, this looked an awful lot like a wide complex tachycardia (peaks of the T waves actually looked higher, giving this initial perception). This could have been confirmed with the rate on the monitor reading 216 bpm. Because this didn’t make sense (palpated pulse was 120 bpm), our second glance prompted us to immediately administer calcium chloride with the resultant rhythm several minutes later:
As the T wave got smaller, we noticed the monitor stopped counting it and the rate on the monitor was confirmed with the palpated femoral pulse.
So now that we briefly stabilized an apparent hyperkalemia, we decided to review the referring unit’s labs and obtain some point of care (POC) laboratory results of our own. As we were doing this we began a plasma transfusion along with starting a norepinephrine infusion (again, all reasonable with a working diagnosis of AFE). Verbal report on laboratory studies included prolonged coagulation times, low hemoglobin and platelet count and liver function that was in the referring nurse’s words “through the roof.”
At this point in our management plan, we were now focusing on HELLP vs. AFE (albeit for us, the management isn’t all that dissimilar). Hemolysis, Elevated Liver Enzymes and Low Platelet count (HELLP) did seem a bit strange for a mother with an uncomplicated pregnancy. AFE with massive DIC and Multi System Organ Failure (MSOF) seemed more logical, especially in the setting of cardiac arrest. As we had a brief period of hemodynamic stability at this point, we attempted to draw some blood to get a baseline set of laboratory values of our own (point of care). Multiple attempts at arterial and venous vasopuncture produced no results and further attempts to obtain recent laboratory values were aborted. It was obvious at this point that the patient was acidotic, hypoperfused and coagulopathic. Hyperkalemia was temporized with IV Calcium. Our priorities quickly shifted to making this patient stable for transport.
Management, Transport and Arrival to Definitive Care
As we were packaging the patient for transport, we transfused the 2 units of liquid plasma and the 2 units of uncrossmatched, type O negative blood that we carry with us on every transport. This was while we waited for the bloodbank at the referring hospital to thaw 2 units of fresh frozen plasma (FFP) and release that, along with an additional 2 units of type O negative packed red blood cells (PRBC) for our 20 minute flight to the receiving hospital. In all, this patient ended up receiving 15 units of uncrossmatched PRBC’s and 11 units of plasma before she made it to the operating room.
In addition to the continuous transfusion of blood products, the patient was on 3 vasoactive substances and intermittent blousing of IV calcium chloride. As it is now widely accepted that an AFE is a massive inflammatory response, treating an apparent distributive shock with norepinephrine and vasopressin is consistent with current Surviving Sepsis guidelines. The patient remained on continuous epinephrine as this was the one continuous infusion she was on when we initially walked through the door.
Ventilator management wise, the patient was on some seemingly benign settings with moderate PEEP, 5 ml/kg of tidal volume, and an FiO2 of 1.0. She was breathing well above the established respiratory rate and was seemingly struggling against an assist control (volume) mode. As these struggles seemed to be counterproductive to the hemodynamic, oxygenation and ventilatory status, we provided cautious chemical sedation and neuromuscular blockade. Following this, we increased ventilatory support to match the minute ventilation that she was providing for herself given her assumed acidotic state.
As we were walking out the door and heading to the aircraft, the referring obstetrician stated that he had just gotten off the phone with the receiving obstetrician who had requested we administer 0.2 mg of atropine, 8 mg of ondansetron and 15 mg of ketorolac for her presumed AFE. After scratching our heads as to why this interesting cocktail, we administered it as ordered and proceeded to the helicopter (more on this below).
Final Disposition
Other than feverishly working to rapidly infuse blood products into this patient, the transport was relatively uneventful. The obstetrical team, the surgical team and the emergency department staff met us on the helipad of the receiving hospital and escorted us to an awaiting trauma bay in order to provide additional stabilization including arterial and central venous access and continued blood product resuscitation. The obstetrician taking charge of the patient performed a vaginal exam and quickly raised the temperature of an already warm trauma room with the simple phrase: “the uterus is inverted, we have to go to the OR Now.”
Again, as the team escorted our patient to the OR for an emergent total hysterectomy, we were left scratching our heads. While this certainly explained why we could not appreciate the fundus (since it wasn’t where it was supposed to be), it didn’t explain how the uterus became inverted. What we later learned was that while attempting to manually dislodge the placenta (in a patient with placenta acreta by the way), the referring obstetrician inverted the uterus while pulling “gentle” traction on the umbilical cord. While it could easily take another entire blog to discuss how a uterus might become inverted, this Khan Academy video explains it in succinct terms.
Take Home Points (or at least what we learned)
For this case, there were numerous learning points. However, what resonated with our crew the most was the following:
- Broadening Differentials: With our limited clinical acumen for the obstetrical population, knowing the diagnosis may not have changed our treatment plan. With that said, this knowledge would have more than likely prompted a call to the receiving medical staff to perhaps get some additional guidance that is not part of our current treatment protocols, possibly reducing the amount of hemorrhage and certainly hastening the patient’s arrival in the OR.
- The AOK Treatment Regimen for an AFE: The administration of atropine, ondansetron and ketorolac has shown some anecdotal evidence in reversing three contributors to death from AFE. Rezai et al. (2017) provide compelling data suggesting that this regimen has specific mechanisms by which to combat coagulopathy and vagal stimulation that can lead to cardiovascular collapse.
Some authors have suggested ventricular dysfunction from AFE is due to either pulmonary hypertension caused by serotonin and thromboxane or systemic hypotension caused by vagal stimulation. It has been proposed that Atropine and Ondansetron may act to block serotonin and vagal stimulation improving cardiovascular function rather than simply providing cardiovascular support which has been mainstay therapy in AFE. Additionally, the A-OK regimen rather than replacing consumed coagulation factors blocks the proposed cause of coagulopathy by inhibiting thromboxane with Ketorolac (schematic of these mechanisms and where the medications act is in the diagram below, taken from the Rezai et. Al. article).
- Do not Discount ANY Abnormal Clinical Finding: In this case, we passed over the fact that we couldn’t appreciate the fundus because we were focused on the overt external hemorrhage, presumed coagulopathy and reason for the coagulopathy. Again, while it perhaps would not have changed our management, it could have conceivably changed the course of this patient’s care, starting with the referring obstetrician’s.
- In this Case, Blood Over Plasma: While recent trauma literature discusses the change in practice and rationale for prioritizing plasma over PRBCs initially, this is a case where we were managing overt blood loss in lieu of the challenges with coagulation as seen in other forms of trauma. While both blood products were important (as evidenced by the number of units of each administered within a relatively short period of time), PRBCs are required for oxygen carrying capacity. The current ability to provide this for our patient was limited due to frank blood loss.
References
Cheng, S. H., & Baldisseri, M. R. (2020). Management of amniotic fluid embolism. In Evidence-Based Critical Care (pp. 739-742). Springer, Cham.
Lisonkova, S., & Kramer, M. S. (2019). Amniotic fluid embolism: A puzzling and dangerous obstetric problem. PLoS medicine, 16(11), e1002976.
Pacheco, L. D., Saade, G., Hankins, G. D., Clark, S. L., & Society for Maternal-Fetal Medicine (SMFM. (2016). Amniotic fluid embolism: diagnosis and management. American Journal of Obstetrics and Gynecology, 215(2), B16-B24.
Rezai, S., Hughes, A. C., Larsen, T. B., Fuller, P. N., & Henderson, C. E. (2017). Atypical amniotic fluid embolism managed with a novel therapeutic regimen. Case reports in obstetrics and gynecology, 2017.