Secondary Logo

Journal Logo


A Quantitative, Systematic Review of Randomized Controlled Trials of Ephedrine Versus Phenylephrine for the Management of Hypotension During Spinal Anesthesia for Cesarean Delivery

Lee, Anna MPH, PhD; Ngan Kee, Warwick D. MBChB, MD, FANZCA; Gin, Tony MBChB, MD, FANZCA, FRCA

Author Information
doi: 10.1097/00000539-200204000-00028
  • Free


Hypotension during spinal anesthesia for cesarean delivery can have detrimental effects on both mother and neonate; these effects include decreased uteroplacental blood flow, impaired fetal oxygenation with asphyxial stress and fetal acidosis, and maternal symptoms of low cardiac output, such as nausea, vomiting, dizziness, and decreased consciousness (1). Therefore, there has been much attention in the literature to methods of preventing and treating hypotension in obstetric anesthesia. Uterine displacement is routine, whereas the use of IV fluid preload is controversial (2). Despite these conservative measures, a vasopressor drug is often required. The drug usually recommended in this context is ephedrine, which is effective in restoring maternal arterial pressure after hypotension (1).

Despite the wide acceptance of ephedrine as the vasopressor of choice for obstetric anesthesia (1,3), its superiority over other vasopressors has not been clearly defined. Historically, ephedrine was recommended on the basis of observations in pregnant sheep that showed it was more effective for increasing arterial pressure with better preservation of uteroplacental blood flow compared with other vasopressors (4,5). This was explained by ephedrine’s predominant β-effect that caused an increase in arterial pressure by increasing cardiac output rather than by vasoconstriction. Accordingly, the use of pure α-agonists such as phenylephrine has generally been avoided because of concerns about their potential adverse effect on uterine blood flow (4,5).

Extrapolation from animal studies to humans may not be appropriate because there are species differences and differences in dose, titration, and duration of the administration and use of IV prehydration to consider. Results of several trials suggest that phenylephrine (6–9) may have similar efficacy to ephedrine for preventing and treating hypotension during spinal anesthesia. However, the relative effects of these vasopressors on neonatal outcome are unclear.

Therefore, we undertook a quantitative, systematic review of randomized controlled trials (RCTs) of ephedrine versus phenylephrine in women having spinal anesthesia for cesarean delivery. Our objectives were to examine whether ephedrine and phenylephrine were different in their efficacy for managing maternal hypotension and their risk of adverse neonatal outcome. In the absence of a large RCT, which would require considerable time and resources to answer these questions, we chose to evaluate the current available evidence using a quantitative, systematic review.


Information Retrieval

We performed a systematic search on electronic databases (MEDLINE 1966–June 2001, EMBASE 1988–June 2001, Cochrane Controlled Trials Register). The electronic search strategy included the optimal sensitive search strategy (10) for identifying RCTs with the following medical subject heading terms and text words: “spinal anesthesia,” “hypotension,” “cesarean section,” “pregnancy complications,” “pregnancy outcome,” “fetal outcome,” “neonatal outcome,” “umbilical cord blood gases,” “vasopressors,” “phenylephrine,” and “ephedrine.” Additional reports were identified from reference lists of retrieved reports, review articles of hypotension during spinal anesthesia, and review articles of vasopressors. We contacted authors of the RCTs included in this systematic review (6–9,11–13) for further published and unpublished trials that had not been identified in our original search. Although there was no language restriction, all RCTs included in this systematic review were published in English.

Eligibility of Articles

Full reports of RCTs that examined the effect of ephedrine compared with phenylephrine for managing maternal hypotension during spinal anesthesia or combined spinal-epidural anesthesia for cesarean delivery were included in this systematic review. We considered any trials of ephedrine and phenylephrine administered before, during, or immediately after the induction of spinal anesthesia regardless of dose or mode of the administration.

Outcome Measures

The outcomes collected were maternal hypotension, hypertension and bradycardia, uterine and umbilical blood circulation, Apgar scores at 1 and 5 min, and umbilical arterial and venous pH values and standard base excess. We did not attempt to standardize the definitions of the above outcomes but rather used the authors’ primary definition for the meta-analysis. In the paper by Alahuhta et al. (6), there were several definitions of hypotension; we used their primary definition of hypotension (decrease in systolic arterial pressure of >10 mm Hg from baseline) rather than their alternative definition of systolic arterial pressure of <90 mm Hg.

Data Extraction

The selection of trials for inclusion in the systematic review was performed independently by the reviewers (Lee and Ngan Kee) after using the search strategy described above. Trials were examined for duplicate data. Data were abstracted independently by Lee and Ngan Kee using a standardized data collection form. There was no attempt to blind the reviewers to the authors or results of the relevant trials. Details of anesthetic technique, study population, prehydration, uterine displacement, and definition of maternal hypotension were collected. Where appropriate, the primary author of a RCT was contacted for clarification of data. Discrepancies were resolved by discussion, or advice was sought from a third party (Gin).

Quality Assessment

The quality of the eligible trials was assessed independently. The level of allocation concealment, defined as the process used to prevent the foreknowledge of group assignment in a RCT, was graded as A (adequate), B (unclear), or C (inadequate), as previously described (14). Blinding, losses to follow-up, and whether the authors did a sample size calculation before trial commencement were recorded.

Statistical Analysis

For the purpose of this systematic review, ephedrine was considered the control group for all meta-analyses. In one RCT (8), we combined dichotomous data (hypotension, hypertension, bradycardia, and Apgar scores of <7) from the two arms of ephedrine. For continuous data (umbilical arterial and venous pH values and standard base excess), the lowest ephedrine dose arm was chosen for comparison with phenylephrine. In another RCT (13), we combined dichotomous data (hypotension, hypertension, bradycardia, and Apgar scores of <7) from the two arms of phenylephrine. The highest phenylephrine dose arm was chosen for comparison with ephedrine for umbilical venous pH values.

The DerSimonian and Laird random-effects model was used to combine data for both continuous and dichotomous outcomes because the treatment and conditions in these studies were expected to have some heterogeneity. This model incorporates both between-study (different treatment effects) and within-study (sampling error) variability and is more conservative than a fixed-effects model (15). The random-effects model has been recommended as the approach for meta-analysis and is more realistic than the fixed-effects model (15). The pooled relative risk (RR) and 95% confidence intervals (CI) were calculated for dichotomous data. The weighted mean difference (WMD) method was used to pool continuous data. Heterogeneity was analyzed using the Q-statistic with a threshold for the P value of <0.10. If there was significant heterogeneity, the data were not pooled, and reasons for heterogeneity were sought. Sensitivity analyses according to the quality of RCTs were not performed because there were too few trials. All meta-analyses were done using Arcus Quickstat software (version 1.2; Addison Wesley Longman Ltd, Cambridge, England).


Trials Included

Eight trials were found (6–9,11–13,16), but one (16) was excluded because epidural anesthesia was used. Table 1 shows the seven RCTs (n = 292) of ephedrine versus phenylephrine (6–9,11–13) included in this systematic review. No additional trials were identified after contacting the authors of the RCTs included in this systematic review. The mode of ephedrine and phenylephrine administration varied between trials and included IV bolus doses (7,9,11,12), IV bolus followed by infusion (6,8), and IM (13). In all trials, ephedrine and phenylephrine were administered immediately after the induction of spinal anesthesia. Ephedrine and phenylephrine were used for the prevention (6,8,13) and treatment (7,9,11,12) of hypotension (Table 1).

Table 1:
 Characteristics of Included Trials

Characteristics of Patients

In all RCTs, the women were described as healthy or were graded as ASA physical status I. There were no reports of trials recruiting women undergoing emergency cesarean delivery, and therefore, we have assumed that all trials included in this systematic review involved nonlaboring women undergoing elective cesarean delivery. Drugs used for spinal anesthesia included bupivacaine (6,7,9,12) and hyperbaric bupivacaine (8,11,13). All trials specified the use of uterine displacement. Lactated Ringer’s solution was the most common prehydration given (7–9,11–13). Combined spinal-epidural anesthesia was used in one trial (13).

Quality of Trials

There was adequate allocation concealment in two trials (7,13). The other trials had unclear allocation concealment (6,8,9,11,12). All were double-blinded trials. In one trial (11), three women in the Ephedrine group and one in the Phenylephrine group were withdrawn because their umbilical arterial pH value was <7.25 before data-analysis. Sample size was estimated before trial commencement in two trials (7,13).

Maternal Outcome Measures


All trials gave specific definitions of maternal hypotension (Table 1). For the management of hypotension (prevention and treatment), there was no difference between the Phenylephrine and Ephedrine groups (RR of 1.00; 95% CI, 0.96–1.06) (6–9,11–13). After limiting the analysis to trials for treatment of maternal hypotension, there was no difference between the Phenylephrine and Ephedrine groups (RR of 1.00; 95% CI, 0.95–1.05) (7,9,11,12). There was no difference in the risk of hypotension between women given phenylephrine or ephedrine for the prevention of maternal hypotension (RR of 1.09; 95%CI, 0.71–1.69) (6,8,13).


Three trials collected data on hypertension (8,9,13). It was defined as >20% of baseline systolic arterial pressure for 3 min (8), systolic blood pressure of >140 mm Hg (9), and mean arterial blood pressure of >25% of baseline values (13). There was no difference in the risk of hypertension between the Phenylephrine and Ephedrine groups (RR of 0.65; 95% CI, 0.08–5.13;n = 170).


Four trials (6–8,13) collected data on maternal bradycardia. Definitions of bradycardia varied: heart rate <40 bpm (8), heart rate <60 bpm requiring atropine treatment (7), and heart rate <60 bpm (13). We noted that one patient was withdrawn from a trial (6) because of bradycardia (not defined). Bradycardia did not occur in either the Ephedrine or Phenylephrine groups in one trial (13). Atropine was required in 11 of 19 women in the Phenylephrine (median dose of 600 μg) group compared with 2 of 19 in the Ephedrine (median dose of 20 mg) group (7). Patients in the Phenylephrine group were more likely than the Ephedrine group to develop bradycardia (RR of 4.79; 95% CI, 1.47–15.60) (7,8,13).

Uteroplacental Circulation.

Data from two studies of uterine and umbilical Doppler velocimetry were not pooled because of differences in data presentation. In one trial, the mean maternal uterine and placental arcuate pulsatility indices increased significantly in the Phenylephrine group compared with the Ephedrine group (P < 0.01), suggesting that phenylephrine was associated with increased vascular resistance in uterine and umbilical vessels (6). In contrast, there was no change in the uterine arterial pulsatility index during spinal anesthesia in the other trial (7).

Neonatal Outcome Measures

Apgar Scores.

Six trials (6–9,12,13) measured Apgar scores at 1 and 5 min. At 1 min, there was one neonate with an Apgar score of <7 in the Ephedrine group compared with no neonate in the Phenylephrine group (9). At 5 min, no neonate in the Ephedrine or Phenylephrine groups had an Apgar score of <7 (6–9,12,13). There was no difference in the risk of low Apgar scores (<7) between the Phenylephrine and Ephedrine groups at 1 min (RR of 0.77; 95% CI, 0.17–3.51) (6–9,12,13) or at 5 min (RR of 1.00; 95% CI, 0.21–4.83) (6–9,12,13).

Umbilical Cord Blood Gases.

Pooling the six trials (6–9,11,12) (n = 200) showed that women given phenylephrine had neonates with higher umbilical arterial pH values than those given ephedrine (WMD = 0.03; 95% CI, 0.02–0.04, mean ephedrine umbilical arterial pH values ranging from 7.27 to 7.29;Figure 1). This analysis would be expected to be conservative because three patients in the Ephedrine group and one patient in the Phenylephrine group were excluded in the trial by Pierce et al. (11) because of an umbilical arterial pH value of <7.25. Women given phenylephrine had neonates with greater venous pH values than those given ephedrine (WMD = 0.02; 95% CI, 0.01–0.03, mean ephedrine venous pH values ranging from 7.29 to 7.35) (6,8,9,11–13).

Figure 1:
Meta-analysis of trials. The effect of phenylephrine versus ephedrine on umbilical cord arterial blood pH. Data are mean difference with 95% confidence intervals.

Arterial standard base excess was greater in the Phenylephrine group compared with the Ephedrine group (WMD = 1.41; 95% CI, 0.81–2.02, mean ephedrine arterial standard base excess ranging from −2.90 to −1.51) (7–9,12). Three trials (8,9,12) showed that venous standard base excess was greater in the Phenylephrine group compared with the Ephedrine group (WMD = 1.23; 95% CI, 0.77–1.69, mean ephedrine venous standard base excess ranging from −2.00 to 0.14).

Fetal Acidosis.

The risk of true fetal acidosis, which we defined as an umbilical arterial pH value of <7.20 (17), was similar between the Phenylephrine and Ephedrine groups (RR of 0.78; 95% CI, 0.16–3.92) (6,7,12).


In this systematic review, we showed that there was no difference between ephedrine and phenylephrine in their efficacy for managing hypotension in the range of doses that have been studied. However, our review showed that women given phenylephrine had neonates with higher umbilical cord blood pH values than women given ephedrine, although the risk of true fetal acidosis (umbilical arterial pH value of <7.20) was similar. We found no clear evidence that phenylephrine was associated with decreased uterine blood flow because there were few RCTs examining this issue. On the contrary, acidotic changes in umbilical arterial pH are sensitive indicators of reduced uteroplacental perfusion (18). Our finding is indirect evidence that uterine blood flow may in fact be better with phenylephrine compared with ephedrine.

The exact reasons why ephedrine is associated with lower umbilical cord blood pH values compared with phenylephrine are unclear. In many studies of ephedrine, relatively large total doses of ephedrine were required to maintain maternal arterial pressure (8,9,12). This may be related to the fact that ephedrine exhibits marked tachyphylaxis because its sympathomimetic effects are largely indirect, arising from the release of noradrenaline from postganglionic sympathetic nerve endings that may become depleted after repeated dosing (19). Furthermore, ephedrine has a relatively slow onset of action and long duration of action. These factors mean that ephedrine may be difficult to titrate, especially when given by IV infusion compared with direct-acting vasopressors (20), which may contribute to suboptimal control of arterial pressure. Finally, ephedrine crosses the placenta (21); therefore, it is possible that ephedrine may have a direct effect on the fetus that contributes to acidosis.

Of concern was that maternal bradycardia occurred more frequently with phenylephrine than with ephedrine. This is to be expected because an increase in blood pressure with an α-agonist may lead to reactive bradycardia. However, this was responsive to atropine treatment without adverse consequences (7,8). The incidence of isolated phenyle-phrine-related maternal bradycardia (heart rate <60 bpm) was highest (58%) in one trial when large doses of phenylephrine were used (7). The authors suggested that maternal bradycardia was contributed to by cardiac sympathetic denervation because the sensory block was high (7). Therefore, an ephedrine-phenylephrine combination may help prevent maternal bradycardia, as the β-mimetic effect of ephedrine would counteract this mechanism. This may explain why the incidence of phenyle-phrine-related maternal bradycardia was lower when phenylephrine was used as prophylaxis with rescue ephedrine (8,13) than when phenylephrine was used for treating maternal hypotension (7).

The clinical implications of our findings are arguable. Although we have shown that ephedrine was associated with lower umbilical cord blood pH values compared with phenylephrine, both groups had similar efficacy for preventing or treating hypotension, and there was no difference in clinical neonatal outcome as measured by Apgar scores. Brief periods of hypotension can alter maternal and neonatal acid-base values, but this did not affect neurobehavioral performances at 24 hours of age (22). Nonetheless, because an objective of obstetric care should be to deliver the fetus in the best condition possible, we believe that the role of ephedrine as the only vasopressor of choice in obstetric patients should be seriously questioned.

The applicability of the results of this quantitative systematic review was limited to healthy women with term fetuses. We found no trials evaluating the effects of vasopressors in women who were in labor, had a compromised or preterm fetus, or had hypertension. Extrapolation of our findings to these circumstances may not be valid. The quality of the trials included in this systematic review was fair. Only two of seven trials in this review had adequate allocation concealment. Compared with trials with adequate allocation concealment, trials with unclear allocation concealment have been shown to exaggerate the treatment effect by 30%(14). Therefore, it is possible that the presence of publication bias and the quality of trials included in this review may be overestimating any beneficial effects of phenylephrine over ephedrine.

Some caution is required in interpreting the results of this meta-analysis, which is based on results of many small trials, as subsequent large trials have disagreed with meta-analyses 10% to 23% of the time (23). One of the roles of systematic reviews is to highlight areas of further research. Systematic reviews should be regarded as complementary to, not a substitute for, a large RCT. A large RCT with a priori defined clinically significant outcome (fetal acidosis) should to be done to confirm our findings. If the baseline risk of fetal acidosis (pH value of <7.20) is 14%(24), a large RCT of 4638 women would have 80% power with a 0.050 two-sided significance level to detect a smaller risk of fetal acidosis (odds ratio of 0.78) associated with phenylephrine. A trial of this size would require con-siderable time and resources. In the absence of such a large multicentered trial, the best current strategy for appraising the available evidence is to examine the results of this meta-analysis.

In summary, this systematic review does not support the traditional idea that ephedrine is the preferred choice over phenylephrine for the management of maternal hypotension during spinal anesthesia for elective cesarean delivery in healthy, nonlaboring women. The use of phenylephrine was associated with better fetal acid-base status, but the risk of maternal bradycardia (responsive to atropine) was larger than in those women given ephedrine.


1. Rout CC, Rocke DA. Prevention of hypotension following spinal anesthesia for cesarean section. Int Anesthesiol Clin 1994; 32: 117–35.
2. Rout CC, Rocke DA, Levin J, et al. A reevaluation of the role of crystalloid preload in the prevention of hypotension associated with spinal anesthesia for elective cesarean section. Anesthesiology 1993; 79: 262–9.
3. Burns SM, Cowan CM, Wilkes RG. Prevention and management of hypotension during spinal anaesthesia for elective caesarean section: a survey of practice. Anaesthesia 2001; 56: 777–98.
4. Ralston DH, Shnider SM, DeLorimier AA. Effects of equipotent ephedrine, metaraminol, mephentermine, and methoxamine on uterine blood flow in the pregnant ewe. Anesthesiology 1974; 40: 354–70.
5. James FM III, Greiss FC Jr, Kemp RA. An evaluation of vasopressor therapy for maternal hypotension during spinal anesthesia. Anesthesiology 1970; 33: 25–34.
6. Alahuhta S, Räsänen J, Jouppila P, et al. Ephedrine and phenylephrine for avoiding maternal hypotension due to spinal anaesthesia for caesarean section. Int J Obstet Anesth 1992; 1: 129–34.
7. Thomas DG, Robson SC, Redfern N, et al. Randomized trial of bolus phenylephrine or ephedrine for maintenance of arterial pressure during spinal anaesthesia for caesarean section. Br J Anaesth 1996; 76: 61–5.
8. Hall PA, Bennett A, Wilkes MP, Lewis M. Spinal anaesthesia for caesarean section: comparison of infusions of phenylephrine and ephedrine. Br J Anaesth 1994; 73: 471–4.
9. Moran DH, Perillo M, LaPorta RF, et al. Phenylephrine in the prevention of hypotension following spinal anesthesia for cesarean delivery. J Clin Anesth 1991; 3: 301–5.
10. Lefebvre C, Clarke MJ. Identifying randomised trials. In: Egger M, Davey Smith G, Altman DG, eds. Systematic reviews in health care. London: BMJ Publishing Group, 2001: 69–86.
11. Pierce ET, Carr DB, Datta S. Effects of ephedrine and phenylephrine on maternal and fetal atrial natriuretic peptide levels during elective cesarean section. Acta Anaesthesiol Scand 1994; 38: 48–51.
12. LaPorta RF, Arthur GR, Datta S. Phenylephrine in treating maternal hypotension due to spinal anaesthesia for caesarean delivery: effects on neonatal catecholamine concentrations, acid base status and Apgar scores. Acta Anaesthesiol Scand 1995; 39: 901–5.
13. Ayorinde B, Buczkowski P, Brown J, et al. Evaluation of pre-emptive intramuscular phenylephrine and ephedrine for reduction of spinal anaesthesia-induced hypotension during caesarean section. Br J Anaesth 2001; 86: 372–6.
14. Schulz KF, Chalmers I, Hayes RJ, Altman DG. Empirical evidence of bias: dimensions of methodological quality associated with estimates of treatment effects in controlled trials. JAMA 1995; 273: 408–12.
15. Mosteller F, Colditz GA. Understanding research synthesis (meta-analysis). Annu Rev Public Health 1996; 17: 1–23.
16. Ramanathan S, Grant GJ. Vasopressor therapy for hypotension due to epidural anesthesia for cesarean section. Acta Anaesthesiol Scand 1988; 32: 559–65.
17. Roberts SW, Leveno KJ, Sidawi JE, et al. Fetal acidemia associated with regional anesthesia for elective cesarean delivery. Obstet Gynecol 1995; 85: 79–83.
18. Robson SC, Boys RJ, Rodeck D, Morgan B. Maternal and fetal haemodynamic effects of spinal and extradural anaesthesia for elective caesarean section. Br J Anaesth 1992; 68: 54–9.
19. Stoelting RK. Pharmacology and physiology in anesthetic practice. Philadelphia: Lippincott-Raven, 1999.
20. Ngan Kee WD, Lau TK, Khaw KS, Lee BB. A comparison of metaraminol and ephedrine infusions for maintaining arterial pressure during spinal anesthesia for elective cesarean section. Anesthesiology 2001; 95: 307–13.
21. Hughes SC, Ward MG, Levinson G, et al. Placental transfer of ephedrine does not affect neonatal outcome. Anesthesiology 1985; 63: 217–9.
22. Corke BC, Datta S, Ostheimer GW, et al. Spinal anaesthesia for caesarean section: the influence of hypotension on neonatal outcome. Anaesthesia 1982; 37: 658–62.
23. Ioanndis JPA, Cappelleri JC, Lau J. Issues in comparisons between meta-analyses and large trials. JAMA 1998; 279: 1089–93.
24. Mueller MD, Brühwiler H, Schüpfer GK, Lüscher KP. Higher rate of fetal acidemia after regional anesthesia for elective cesarean delivery. Obstet Gynecol 1997; 90: 131–4.

Attention Authors!

Submit Your Papers Online

You can now have your paper processed and reviewed faster by sending it to us through our new, web-based Rapid Review System. Submitting your manuscript online will mean that the time and expense of sending papers through the mail can be eliminated. Moreover, because our reviewers will also be working online, the entire review process will be significantly faster. You can submit manuscripts electronically via There are links to this site from the Anesthesia & Analgesia website (, and the IARS website ( To find out more about Rapid Review, go to and click on “About Rapid Review.”

© 2002 International Anesthesia Research Society