To the Editor:
Dr. Smiley's supportive editorial, Burden of Proof, prompted me to eagerly read the obstetric article by Dr. Ngan Kee et al. , hoping to find the promised “Holy Grail” of obstetric anesthetic blood pressure control.1,2Instead, I found a questionable clinical methodology of high-dose vasopressor therapy for American Society of Anesthesiologists 1–2 delivery under bupivacaine spinals with limited scientific disclosure of data, burdening me to ask, where is the proof and how is it pertinent to standard clinical methods?
Intravenous prehydration was withheld, and total fluid was limited to 2 l during the 27 min to delivery.
Immediately, preoperatively obtained oscillotonometric systolic pressures were chosen as therapeutic goals (but nowhere are these group systolic values or statistical analysis provided).
Infusions of phenylephrine (100 μg/ml—totaling 960–1,690 μg) versus ephedrine (8 mg/ml—totaling 44.8–79.2 mg) were initiated at 1 ml/min to sustain the therapeutic goal “baseline” pressure.
Rescue treatment occurred using phenylephrine 100 μg bolus in both groups without reporting amounts or intervals administered.
Supplemental oxygen was withheld unless saturation decreased below 95% (no indication of number of patients given oxygen in any group is provided).
The vasopressor was apparently simply “stopped if systolic pressure was greater than 120% baseline,” occurring in approximately 40% of all patients.
We find that
7. While the E versus P group had more hypotension (defined as systolic drop to < 80% of baseline) and phenylephrine rescue (22% vs. 2%), respectively, the mean minimum recorded systolic blood pressures were similar and under classic parameters, not typically requiring treatment in either group (101 vs. 104 Torr, P = 0.33).
8. Almost double the “equipotent volume” of P versus E was infused (13 vs. 7.7 ml), respectively, without information of phenylephrine “rescue” injections/amounts.
9. Hypertension was found in 41% versus 47% of P versus E patients, respectively, and these maximum absolute pressures reported differed significantly statistically (P = 0.044).
10. Although no information is presented regarding rates of reduction versus termination time intervals of infusions relative to the point of delivery (when calculation of umbilical vein vs. maternal artery levels occurred), strong argumentation is made for ion trapping of ephedrine in the fetus. Consideration of important pharmacokinetic parameters, including α-redistribution phenomena, diffusion gradients of ephedrine present at levels more than 50 times that of phenylephrine, unreported timing of (recent?) phenylephrine bolus doses effecting hemodynamics, and kinetics or interactions of P + E on the resultant fetal blood gasses obtained, remains unreported and unknown.
As a clinician, it is interesting that the Apgar score of less than 7 was reported only in a phenylephrine patient (subtle warning?). The use of systolic pressure instead of mean arterial pressure measured by oscillotonometry in awake patients is fraught with motion artifact error. Mean arterial pressure is most reliably detected by this method, and mean blood pressure is perfusion pressure. Defining baseline pressure “on the table” without consideration of clinic outpatient values, failing to prehydrate and continuing vasopressor infusion until pressure exceeds 120% of this most probably relatively hypertensive “defined baseline” value, is clinically questionable. The high rates of hypertension, unusual amounts of pressors administered, and minimum recorded values of 101–104 Torr found in this study and previous studies all reflect vasopressor overuse rather than any typical clinical care, in which classically 100 Torr or 10–20% drop in systolic pressure from baseline is deemed appropriate in healthy parturients.3Is hypertension better than hypotension, to what degree and why? Is this “safe” only in healthy patients, and what is the danger of extrapolation to preeclamptics or others?
The hypertension and lower volume of study infusion administered in the ephedrine group may reflect a specific shortcoming of this indirect acting drug as an infusion candidate versus phenylephrine, as indicated previously by Dr. Ngan Kee et al .4Clinically, bolus dosing of either drug is more typical and prevents inadvertent overinfusions. The higher absolute milligram dose per milliliter of ephedrine infusate may have facilitated diffusion across the placental gradient during administration, whereas the lower total volume administered suggests that a tapering (termination) occurred relative to the time of delivery at 27 min; only 7.7 ml (mean) was infused versus the baseline rate of 1.0 ml/min, which would have yielded 27 ml at delivery. The high umbilical vein/maternal artery ratio found may reflect simple reverse fetal-to-maternal diffusion kinetics of previously higher fetal (and maternal) ephedrine levels into the hyperdynamic mother's circulation (after infusion termination) of a drug with 3- to 5-h renal elimination T1/2and a volume of distribution of 3 l/kg (i.e. , standard uptake and elimination kinetics).*What were the time intervals from ephedrine/P infusion termination and rates relative to the umbilical vein/maternal artery blood level ratios in individual patients and in different groups? Was typically P infused/bolus injected up to delivery, thus maintaining the maternal-to-fetal gradient of diffusion found?
Dr. Smiley suggests that this study is further proof of clinical superiority of phenylephrine, and possibly infusion over bolus therapy, perhaps as a “believer” of this Holy Grail. I question this “proof” itself, validity, and values of baseline definitions: What blood pressure was the “desired baseline” and is it appropriate for these two groups from a Chinese population, known to exhibit lower pressures than Americans by diet and predisposition, a baseline determined other than from standard unstressed clinical method? Where is the missing data: no mention of herbal medical treatment as exclusion factor in this Chinese population with “traditional medicine” (i.e. , not Western traditional medicine) trend toward such use, which may have effected ephedrine responses/kinetics, the fact that “equipotent” phenylephrine was apparently infused more continually/longer/later, and what amounts of crystalloid infusion and rescue drug/patient/per group and the relative interval time of last dose to delivery of any/all drugs were administered?
Perhaps, this study simply shows phenylephrine versus ephedrine overtreatment by infusion is easier (but not accurate) to titrate, but is bolus dosing inferior and why? Bolus treatment is certainly more common, typically involves less preparation or total drug injected, may promote more normal pressure, and is effective. Because the blood gas values obtained were within normal limits in both groups, do the statistical differences noted provide any real index of substantial clinical relevance/superiority? Will we ever convert to mean arterial pressure versus systolic pressures in clinical care and investigation? Why not? How does a study of blood pressure and drug dosing/kinetics reach publication, without disclosing absolute values, timing and statistical analysis of the blood pressure, and rescue versus infusion drugs given? Did the 2% and 22% of P versus E group patients, respectively, receive one or multiple doses of rescue phenylephrine and when? Was it the E, or the “P on E” administered in the E group, responsible for the blood gas differences? I look forward to further illumination of this “Holy Grail.” The Burden of Proof has all but been lifted for clinical antihypotensive therapy by this report of hypertensive treatment. Large randomized outcome studies of all complications will be required to prove safety and efficacy of such hypertensive infusion therapy or controlled infusions based in microgram per kilogram per minute and individual responses.
Cleveland Clinic, Cleveland, Ohio. firstname.lastname@example.org