Balancing the efficacy and safety of misoprostol: a meta-analysis comparing 25 versus 50 micrograms of intravaginal misoprostol for the induction of labour

Authors

  • K McMaster,

    Corresponding author
    1. Department of Obstetrics and Gynecology, University of Florida College of Medicine – Jacksonville, Jacksonville, Florida, USA
    • Correspondence: Dr K McMaster, University of Florida College of Medicine – Jacksonville, 655-1 West 8th Street, Jacksonville, FL 32209, USA. Email kristen.mcmaster@jax.ufl.edu

    Search for more papers by this author
  • L Sanchez-Ramos,

    1. Department of Obstetrics and Gynecology, University of Florida College of Medicine – Jacksonville, Jacksonville, Florida, USA
    Search for more papers by this author
  • AM Kaunitz

    1. Department of Obstetrics and Gynecology, University of Florida College of Medicine – Jacksonville, Jacksonville, Florida, USA
    Search for more papers by this author

Abstract

Background

The optimal dose of misoprostol for the induction of labour remains uncertain.

Objectives

To compare the efficacy and safety of 25 versus 50 micrograms of intravaginal misoprostol tablets for the induction of labour and cervical ripening.

Search strategy

We performed electronic and manual searches to identify relevant randomised trials.

Selection criteria

The efficacy outcomes assessed were rates of vaginal delivery within 24 hours, delivery within one dose, and oxytocin augmentation, and interval to delivery. The safety outcomes assessed were incidences of tachysystole, hyperstimulation, caesarean delivery, cesarean delivery for non-reassuring fetal heart rate (FHR), operative vaginal delivery, abnormal 5-minute Apgar score, abnormal cord gas values, admission to a neonatal intensive care unit (NICU), and meconium passage.

Data collection and analysis

Thirteen studies (1945 women) were included. Relative risk (RR) and 95% confidence intervals (CI) were calculated using fixed-effects and random-effects models.

Main results

We found that 25 micrograms was less efficacious, with lower rates of delivery after one dose (RR 0.59; 95% CI 0.39–0.88) and vaginal delivery within 24 hours (RR 0.88; 95% CI 0.79–0.96), and with increased rates of oxytocin augmentation (RR 1.54, 95% CI 1.36–1.75). We noted an improved safety profile with 25 micrograms, however, with decreased rates of tachysystole (RR 0.46; 95% CI 0.35–0.61), hyperstimulation (RR 0.5; 95% CI 0.31–0.78), caesarean deliveries for non-reassuring FHR (RR 0.67; 95% CI 0.52–0.87), NICU admissions (RR 0.63; 95% CI 0.4–0.98), and meconium passage (RR 0.65; 95% CI 0.45–0.96).

Conclusions

Although 50 micrograms of intravaginal misoprostol may be more efficacious, safety concerns make the 25-microgram dose preferable.

Introduction

Approximately half of all women undergoing labour induction will have an unfavourable cervix, or low Bishop score, which will require ripening. This can be achieved with pharmacological and mechanical techniques.[1] Prostaglandins are often used for pharmacological cervical ripening. One of the benefits of prostaglandins over mechanical methods is that prostaglandins work not only to ripen the cervix, but also to stimulate myometrial contractility.[2] This feature also leads to one of the drawbacks of prostaglandins: the potential for inducing excessive uterine contractility, or tachysystole.[3] The incidence of tachysystole is increased with higher doses of misoprostol; however, the clinical implications of isolated tachysystole are uncertain.[4]

Misoprostol is a synthetic PGE1 analogue, which is used worldwide for both cervical ripening and induction of labour, as it is inexpensive, stable at room temperature, and available in more than 80 countries.[5] The recommendations from the American College of Obstetricians and Gynecologists (ACOG) in 2009 and the International Federation of Gynecology and Obstetrics (FIGO) in 2013 are for a starting dose of 25 micrograms of vaginal misoprostol for labour induction to minimise complications.[4, 5] Despite these recommendations, many obstetricians in the USA and other countries continue to employ 50 micrograms vaginal dosing. This continued use of the 50-micrograms dose of misoprostol may reflect published reports suggesting that 50 micrograms dosing works faster, resulting in a shorter interval to vaginal delivery, more deliveries within 24 hours, and less need for oxytocin augmentation. The comparative safety of 25- and 50-microgram doses has been uncertain, however, with conflicting published results as to the incidence of adverse outcomes with the two doses.[1] A 2010 systematic review comparing various doses of misoprostol found that lower dose regimens were no less effective than higher dose regimens, and resulted in a reduced incidence of adverse effects, including lower rates of uterine hyperstimulation syndrome and a trend towards fewer admissions to neonatal intensive care units.[6] Currently, the two most commonly used intravaginal dosing regimens for cervical ripening are 25 and 50 micrograms. Although a 2002 meta-analysis found that the 50-micrograms intravaginal dose was more efficacious, given the relatively infrequent occurrence of adverse safety outcomes, the limited aggregate sample size was not sufficient to allow for a meaningful comparison of adverse outcomes for the two doses.[7] The purpose of this updated meta-analysis is to allow for a greater sample size, with the goal of optimally comparing the efficacy and safety of 25 versus 50 micrograms of vaginal misoprostol when used for labour induction and cervical ripening.

Methods

Literature search

We developed a protocol with explicitly defined objectives, criteria for study selection, approach to assessing the quality of the studies, primary and secondary outcomes, and statistical methods. We followed the guidelines for meta-analyses and systematic reviews of randomised controlled trials (RCTs) outlined by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement.[8] We sought to answer two primary questions: (1) is there a difference in efficacy between 25 and 50 micrograms of intravaginally administered misoprostol tablets for cervical ripening and labour induction; and (2) is there a difference in maternal and perinatal safety outcomes for these two doses?

We attempted to identify all relevant published and unpublished RCTs that compared the efficacy and safety of 25 versus 50 micrograms of vaginal misoprostol for cervical ripening and labour induction. Without language restrictions, we searched the electronic databases MEDLINE, PubMed, ClinicalTrials.gov, EMBASE, and the Cochrane Library from January 1966 to January 2014. The medical subject headings used included the terms ‘cervical ripening’, ‘labor induction’, ‘misoprostol’, ‘25 versus 50 micrograms’, and ‘randomized trial’, in combination with generic and trade names of individual preparations of misoprostol. In addition, we manually searched textbook chapters and references of identified journal articles to locate additional studies. We also reviewed abstracts from major national and international meetings, and contacted other investigators in the field to learn of possible unpublished studies. Finally, we reviewed the articles analysed in previous reviews assessing the efficacy and safety of misoprostol as a cervical ripening and labour induction agent.

Study selection

Specific and prospectively defined inclusion criteria were used to determine which studies would be included in the meta-analysis. All studies included were required to have participants randomly assigned to either the 25- or 50-micrograms group, and to have direct comparisons to evaluate the efficacy and safety of 25- versus 50-microgram doses of vaginally administered misoprostol tablets used for cervical ripening and labour induction. Although each article did not address all of the outcomes we selected for analysis, the studies included were required to have clearly defined outcomes, and documented at least one of the following outcomes: mode of delivery; vaginal delivery within 24 hours; number of doses required to effect delivery; incidence of hyperstimulation syndrome or tachysystole; need for oxytocin augmentation; or interval to vaginal delivery. In addition, at least one of the following perinatal outcomes was required: Apgar scores at 5 minutes; admissions to neonatal intensive care unit (NICU); abnormal cord gases; or presence of meconium. Studies reporting the use of misoprostol for first- or second-trimester termination of pregnancy were excluded. Non-randomised (observational) studies were also excluded from the analysis, as were RCTs with required outcomes not reported, or RCTs that did not directly compare 25 with 50 micrograms of misoprostol. Finally, we excluded studies that assessed misoprostol administered by routes other than vaginal, misoprostol formulations other than tablets, or treatment regimens limited to one-time administrations. Disagreements between authors over the inclusion and exclusion of studies and interpretation of data were resolved by consensus reached after discussion (for full details of study selection, see Figure 1).

Figure 1.

Study identification flow chart.

Each of the studies was evaluated using the study-quality criteria of Jadad and colleagues to assess the methodological attributes.[9] Two reviewers independently reviewed each article in detail and assigned a quality score, and any disagreements in quality scoring were settled by the third reviewer. A mean quality score was calculated and reported; however, because of the relatively small number of studies available, these scores were not used to exclude any studies. Significant heterogeneity in the quality ratings of each of the studies included was noted. Additionally, in order to evaluate individual methodological parameters separately, the criteria from the Cochrane Handbook for Systematic Reviews of Interventions was used to evaluate each study in respect to sequence generation, allocation concealment, blinding of participants and outcomes, attrition bias, and reporting bias.[10] Each of these aspects were evaluated as high, low, or unclear risk of bias. The results of the quality assessments of all studies included are detailed in Table 1.

Table 1. Quality assessment of included studiesThumbnail image of

Data collection

Data were extracted from each study to evaluate maternal and perinatal outcomes reflecting the efficacy and safety of misoprostol. The primary outcomes assessed were vaginal deliveries within 24 hours and rates of hyperstimulation syndrome. Efficacy outcomes were shown by the rate of vaginal delivery within 24 hours of beginning induction, the number of patients who delivered with one dose of misoprostol, the need for oxytocin augmentation, and the time interval from induction to vaginal delivery. The safety outcomes were measured by determining rates of tachysystole, hyperstimulation syndrome, overall caesarean deliveries, caesarean deliveries resulting from fetal heart rate abnormalities, operative vaginal deliveries, 5-minute Apgar scores of <7, abnormal cord gas values, NICU admissions, and meconium passage. Tachysystole was defined as more than five uterine contractions in a 10-minute period. Hyperstimulation syndrome was present when fetal heart rate abnormalities were noted in the presence of tachysystole. Data were extracted from text, tables, and graphs of each of the selected studies on two separate occasions to confirm accuracy of data collection.

In instances where the terminology of the original articles was unclear, the authors were contacted directly for clarification. For example, in two of the studies, interval to delivery was cited, but it was unclear whether this value was referring to interval to any type of delivery or whether the numbers were referring to interval to vaginal delivery, specifically. Both of these authors were contacted for clarification, and responded with the required information.[2, 11]

Statistical analysis

Meta-analysis was performed using stata 11.0 (StataCorp, College Station, TX, USA). For each study with binary outcomes, we calculated the relative risk (RR) and 95% confidence intervals (95% CIs) of selected outcomes. Estimates of RRs for dichotomous outcomes were calculated using fixed-effects (Mantel–Haenszel) and random-effects (DerSimonian and Laird) models.[12, 13] The null hypothesis underlying the overall test of association was that the overall RR was equal to one. Continuous outcomes were pooled using a variance-weighted average of within-study difference in means.[14] The weighted mean difference meta-analysis combines the differences between two groups to estimate the overall mean difference. The weight assigned to each study was the inverse of the variance of the difference in means for that study.

To determine the combinability of individual studies, we did a formal test of heterogeneity by using the Breslow–Day method and the I2 calculation. In addition, homogeneity across studies was assessed by qualitative visual inspection of L'Abbe plots.[15] Outcomes with significant heterogeneity (P < 0.10 or I2 > 50%) were analysed and reported by using a random-effects model with the DerSimonian–Laird method.[13] Outcomes without significant heterogeneity were assessed with a fixed-effects model, using the Mantel–Haenszel method.[12] To explore potential sources of heterogeneity in study results, we developed a random-effects regression model that included several variables as covariates. In the meta- regression, the dependent variable was the RR for the outcome of interest (vaginal delivery within 24 hours), and the independent variables were year, country of study (USA versus other), quality score, and dose of misoprostol administered.

To further assess the robustness of our results, we performed sensitivity analysis to identify any study that may have exerted a disproportionate influence on the summary treatment effect. To accomplish this, we sequentially omitted each study one at a time, and noted the degree of impact on the pooled results. Publication bias was examined by reviewing the Egger test,[16] and by visual inspection of funnel plots, which plot odds ratio (OR) against study sample size. The visual inspection of funnel plots is a technique that may be helpful to estimate whether additional small studies may have been conducted but not published because of unfavourable or negative results.

Results

Thirteen RCTs met the inclusion criteria for this meta-analysis. Aggregating the data in these trials resulted in a total of 1945 women (973 in the 25-micrograms arm and 972 in the 50-micrograms arm). A summary of the studies included in this meta-analysis is listed in Table 2.

Table 2. Summary of studies included in the meta-analysis
ReferenceYearLocationNo. of patients (25/50 micrograms)Gest. agesMedication administrationDosing interval (max doses)
Loto et al.[2]2012Nigeria80/8037–42Tablet in posterior fornix – until adequate contractionsEvery 6 hours (200 micrograms)
Saxena et al.[11]2011India70/7037+Tablet vaginally – until bishop score >6Every 6 hours (three doses)
Nigam et al.[19]2010India60/6037+Vaginally – until adequate contractions, favorable cervix, rupture of membranesEvery 4 hours (five doses)
Gupta et al.[20]2010India80/6837+Tablet in posterior fornix without any lubricant until adequate contractionsEvery 4 hours (200 micrograms)
Eroglu et al.[21]2007Turkey73/7437+Tablet vaginally – until adequate contractions, or bishop score >or = 8Every 4 hours (six doses)
Elhassan et al.[25]2005Sudan31/32Not specifiedTablet vaginallyEvery 6 hours (four doses)
Meydanli et al.[22]2002Turkey60/6041+Tablet in posterior fornix – until adequate contractions, or bishop score > or = 8Every 4 hours (six doses)
Has et al.[23]2002Turkey58/5637+Reconstituted tablet of similar appearance in posterior fornix – until adequate contractions, bishop score > or = 7, dilation of 3 cm, or rupture of membranesEvery 4 hours (six doses)
El-Sherbiny et al.[18]2000Egypt93/9237+Crushed tablet mixed w/ sterile hydroxyethyl gel in posterior fornix via syringe – until adequate contractions, dilation >3 cm, or rupture of membranesEvery 4 hours (six doses)
Wang et al.[26]1998China27/21TermTablet in posterior fornix – until labor, dilation >2 cm, or rupture of membranesEvery 4–6 hours (200 micrograms)
Diro et al.[17]1999USA125/126Not specifiedTablet in posterior fornix – until onset of labor (regular contractions w/ cervical change)Every 3 hours (eight doses)
Srisombooon and Singchai [24]1998Thailand24/2635+Tablet crushed to powder, mixed w/ 2 ml of sterile 1% carboxymethyl cellulose gel, in posterior vaginal fornix – until adequate contractions, rupture of membranes, dilation >3 cm, or bishop score >6Every 6 hours (four doses)
Farah et al.[3]1997USA192/2073rd trimesterTablet in posterior fornix -until adequate contractionsEvery 3 hours (eight doses)

The rate of women delivering with one dose of misoprostol was 23% in the 25-micrograms group and 44.9% in the 50-micrograms group (RR 0.59; 95% CI 0.39–0.88). In the 25-micrograms group, 66.6% delivered vaginally within 24 hours, as opposed to 74% in the 50-micrograms group (RR 0.88; 95% CI 0.79–0.96). The forest plot in Figure 2 depicts data regarding vaginal delivery within 24 hours for each selected study. Oxytocin augmentation was used for 39.4% of women who received 25 micrograms of misoprostol, and 25.6% of women who received the 50-micrograms dose (RR 1.54; 95% CI 1.36–1.75). Additionally, the mean interval to vaginal delivery was 3.9 hours longer among women who receive the 25-micrograms dose. All of these factors can lead to a significant increase in time and expense spent on the induction process when using the 25-micrograms dose. Table 3 outlines the results with regard to efficacy. With respect to each of the above efficacy outcomes, we found no evidence of publication bias using Egger's test or with visual analysis of funnel plots.[16] The only outcomes for which we found significant heterogeneity between studies were delivery with a single dose of misoprostol (= 0.01; I2 = 69.6%) and interval to vaginal delivery (= 0.00; I2 = 94.4%). This can potentially be explained by different dosing frequencies, variations in labour management practices in the different studies, or differences in baseline characteristics of the groups prior to starting induction, such as Bishop score or gestational age. To further explore the potential sources of heterogeneity, we employed a multivariable meta-regression model. The following study characteristics were included in the meta-regression model: year of publication; country of origin; and quality score. No significant association was found between any of these covariates (all with P > 0.1).

Table 3. Outcome variables reflecting efficacy, comparing 25 with 50 micrograms of vaginal misoprostol
Outcomes25 micrograms group n (%)50 micrograms group n (%)M-H pooled RR RR (95% CI)Heterogeneity P-value, I2Clinical significance
  1. 95% CI, 95% confidence interval; D+L WMD, DerSirmonian and Laird weighted mean difference; M–H, Mantel–Haenszel; RR, relative risk.

  2. a

    DerSirmonian and Laird pooled RR (random-effects model).

Delivery after 1 dose[3, 11, 18, 21, 22]112/488 (23)226/503 (44.9)0.59 (0.39, 0.88)a0.01, 69.6%Favors 50 micrograms
Vaginal delivery within 24 hours [3, 11, 19, 20, 22, 23]285/428 (66.6)299/404 (74)0.88 (0.8, 0.96)0.12, 44.7%Favors 50 micrograms
Oxytocin augmentation[2, 3, 11, 17-26]372/973 (38.2)242/972 (24.9)1.54 (1.36, 1.75)0.14, 30.2%Favors 50 micrograms
Interval to vaginal delivery[2, 3, 11, 17-19, 21-26]D+L pooled WMD 3.928(1.91, 5.94)a0, 94.4%Favors 50 micrograms
Figure 2.

Forest plot for the outcome ‘vaginal delivery within 24 hours’.

When evaluating perinatal safety, we found a statistically significant decrease in the rates of tachysystole (RR 0.46; 95% CI 0.35–0.61) and hyperstimulation syndrome (RR 0.5; 95% CI 0.31–0.78) in the 25-micrograms group. The forest plot in Figure 3 provides study-specific details regarding hyperstimulation syndrome. Furthermore, although there was not a significant decrease in operative vaginal deliveries or the rates of caesarean deliveries overall, we did note a significant decrease in the rates of caesarean deliveries performed for non-reassuring fetal heart tracings in the 25-micrograms group (RR 0.67; 95% CI 0.52–0.87). With lower doses of misoprostol, we also found significantly lower rates of NICU admissions (RR 0.63; 95% CI 0.4–0.98) and meconium passage (RR 0.67; 95% CI 0.47–0.96). We did not find a significant decrease in the rates of low 5-minute Apgar scores or abnormal cord gases with the 25-micrograms dose, however. Table 4 outlines the results with regards to safety. With respect to each of the above safety outcomes we found no evidence of publication bias using Egger's test, nor with visual analysis of funnel plots.[16] Additionally, neither the I2 values nor visual inspection of L'Abbe plots indicated significant heterogeneity between studies.

Table 4. Outcome variables reflecting perinatal safety comparing 25 micrograms with 50 micrograms vaginal misoprostol
Outcomes25 micrograms group n (%)50 micrograms group n (%)M-H pooled RR RR (95% CI)Heterogeneity P-value, I2Clinical significance
  1. NRFHR, non-reassuring fetal heart rate tracings.

  2. Abnormal cord gas values defined as pH <7 or <7.16.

  3. Other abbreviations defined in Table 3.

Tachysystole [2, 3, 18, 20-24, 26]56/687 (8.2)125/684 (18.3)0.46 (0.35, 0.61)0.81, 0%Favors 25 micrograms
Hyperstimulation syndrome[3, 11, 17-22]24/753 (3.2)51/757 (6.7)0.5 (0.31, 0.78)0.53, 0%Favors 25 micrograms
Overall cesarean rate[2, 3, 11, 17-26]190/973 (19.5)200/972 (20.6)0.94 (0.79, 1.12)0.16, 29.2%Not significant
Cesarean for NRFHR[2, 3, 17, 20-23]46/128 (35.9)76/141 (53.9)0.67 (0.52, 0.87)0.33, 13.6%Favors 25 micrograms
Operative vaginal del[3, 21, 22, 24-26]43/407 (10.6)39/420 (9.3)1.16 (0.77, 1.74)0.94, 0%Not significant
5 min Apgar <7[3, 18, 21-23]9/476 (1.9)15/489 (3.1)0.61 (0.27, 1.37)0.39, 0%Not significant
Abnormal cord gas [3, 21-23]20/383 (5.2)34/397 (8.6)0.63 (0.37, 1.06)0.37, 5.3%Not significant
NICU admissions [3, 18, 20-23]29/545 (5.3)46/538 (8.6)0.63 (0.4, 0.98)0.81, 0%Favors 25 micrograms
Meconium[2, 18-22, 24, 25]44/490 (9)63/473 (13.3)0.67 (0.47, 0.96)0.64, 0%Favors 25 micrograms
Figure 3.

Forest plot for the outcome ‘hyperstimulation syndrome’.

Discussion

Main findings

We found that 25 micrograms of vaginal misoprostol is significantly less efficacious than the 50-micrograms dose in regards to the number of women delivering after one dose of misoprostol and the proportion of women delivering vaginally within 24 hours. This decreased efficacy was also demonstrated by a significant increase in the need for oxytocin in the 25-micrograms group. We also found that the safety profile of the 25-micrograms dose is preferable to 50 micrograms. The incidence of tachysystole, hyperstimulation syndrome, caesarean deliveries for non-reassuring fetal tracings, NICU admissions, and meconium passage were all significantly decreased with the 25-micrograms dose.

Strengths and limitations

There are several factors that strengthen our analysis. We performed an extensive search strategy, attempting to identify both published and unpublished work, without regard to language. We also were able to more than double the number of aggregate women in the meta-analysis compared with the previous 2002 review. Additionally, the substantial clinical heterogeneity among the studies included with regards to frequency of misoprostol dosing and labour management after the administration of misoprostol increases the generalisability of our meta-analysis compared with that for single studies. Finally, for each of the variables studied, sensitivity analysis was performed to ensure that no single study dominated our results.

Limitations to our study should be acknowledged. Despite the increase in aggregate numbers of women included, the resulting small sample size makes drawing statistical inferences regarding infrequently occurring outcomes, such as low Apgar scores, challenging. Additionally, not every study included provided clarification for definitions of events such as tachysystole and hyperstimulation syndrome, which may lead to variation in the rates of reporting these outcomes. Furthermore, some studies included participants with known abnormal fetal testing, whereas other studies specifically excluded these women, which may lead to some discrepancy between outcomes. We also would have liked to have been able to evaluate subgroup analysis of the different dosages by parity, but unfortunately this information was not available in the included studies.

Interpretation

Although use of the 25-micrograms dose of misoprostol was associated with a lower incidence of NICU admissions and meconium passage, we were unable to demonstrate a significant decrease in the rates of the other direct perinatal outcome indicators, including low 5-minute Apgar scores or abnormal cord gases. Nonetheless, we did observe a trend towards worsening perinatal outcomes of these parameters in the 50-micrograms group. This higher misoprostol dose increased low 5-minute Apgar scores from 1.9 to 3.1%, and abnormal cord gas rates from 5.2 to 8.6%. This unfavourable trend with the 50-micrograms dose was also observed in the previous meta-analysis on this topic, in which low 5-minute Apgar scores increased from 1.8 to 3.3% in the 50-micrograms group.[7]

Additionally, despite our finding of a significantly decreased incidence of tachysystole in the 25-micrograms group, this finding has unknown clinical significance. It has been reported that there is no significant increase in adverse fetal outcomes with isolated tachysystole when frequent uterine contractions are not associated with non-reassuring fetal heart rate patterns.[4]

Although our overall results found a decreased rate of hyperstimulation syndrome with 25 micrograms of misoprostol, sensitivity analysis for this variable indicates that this finding can largely be attributed to one trial. When the Diro study is omitted, the decrease in rates of hyperstimulation syndrome is no longer statistically significant;[17] however, omission of the Diro study has no effect on the statistical significance of other safety outcome measures, including rates of tachysystole and caesarean delivery for non-reassuring fetal heart tracings. One possible explanation for why the Diro report may have had an outlying result is that more than twice as many women with abnormal antepartum fetal testing were randomised to receive 50 instead of 25 micrograms (24 versus 11 women). Aside from the Diro study, the only other study to include women with abnormal antepartum testing was the small El-Sherbiny study,[18] in which two women received 50 micrograms and three women received 25 micrograms of misoprostol. Of the remaining 11 studies, eight specify that all participants had reassuring fetal testing prior to the initiation of induction.[2, 3, 19-24] Three of the studies included do not specifically mention whether patients with abnormal fetal testing were included, or whether non-reassuring fetal testing was an exclusion criteria.[11, 25, 26]

Our finding of increased incidence of hyperstimulation syndrome in cases of abnormal fetal testing stresses the importance of factoring in fetal wellbeing and fetal reserve when choosing an induction strategy. A compromised fetus with reduced or borderline fetal reserve cannot tolerate excessive uterine activity. Too frequent contractions, or tachysystole, significantly increases the resistance in the uterine and arcuate arteries. The degree of fetal compromise prior to induction of labour seems to correlate with the likelihood of success of induction with misoprostol and fetal tolerance of labour.[22] If the 50-micrograms dose of misoprostol is to be considered at all, such use should be limited to inductions in which fetal wellbeing has been established.

Our review also demonstrated increased rates of meconium passage with the 50 microgram dosing of misoprostol. Although this is conventionally considered to be an indicator for fetal compromise, this occurrence may in fact represent an incidental side-effect of misoprostol on the fetal gastrointestinal tract. The use of higher doses of misoprostol is associated with significantly more meconium-stained amniotic fluid; however, this finding may represent a dose-related side effect of unknown clinical significance.[6] Misoprostol has been found to have a similar impact on the gastrointestinal tract of rats.[27]

Conclusion

Although this meta-analysis reaffirms that a 50-micrograms vaginal dose of misoprostol is more efficacious regarding labour induction and cervical ripening than a 25-micrograms dose, the safety profile of the 25-micrograms dose makes this lower dose preferable. The correlation of pre-existing fetal compromise with the adverse outcome of hyperstimulation syndrome presents a sobering finding, which suggests that if the 50-micrograms dose of misoprostol is to be considered at all, such use should be limited to inductions in which fetal wellbeing has been established.

Although this updated meta-analysis more than doubled the number of women included (from 933 to 1945 participants), the aggregate number of study participants still did not achieve the sample size needed to detect a doubling in the rate of low Apgar scores, as this is such an infrequently occurring event. For example, to detect a doubling in the rate of low 5-minute Apgar scores (from 1.9 to 3.8%), 1306 women would be required in each group. A substantial number of additional studies comparing these two dosages would be needed to allow for a meaningful comparison of such an infrequently occurring adverse event. Performing an indirect meta-analysis might result in aggregating participant numbers sufficient to compare infrequent outcomes with the two doses of misoprostol. Such a meta-analysis would compile safety and efficacy outcomes from all studies comparing 25 micrograms of intravaginal misoprostol with other prostaglandins, and all studies comparing 50 micrograms of intravaginal misoprostol with other prostaglandins. The results of each of these two compilations could then be indirectly compared with one another.

Disclosure of interests

There are no financial, personal, political, or intellectual conflicts of interest to disclose.

Contribution to authorship

KM was the primary author for this work. KM was responsible for the conception of the project idea, in addition to planning, carrying out, and writing the vast majority of the article. LSR was the author of the original 2002 meta-analysis comparing 25 micrograms with 50 micrograms of vaginal misoprostol, which was used a template for this project. LSR was also responsible for directly overseeing the project, and assisted extensively with the statistical analysis once the data were collected. AMK was another mentor whose contributions included extensive editing of this article. All authors approved the final draft of the article.

Details of ethics approval

No ethics approval was required, as this was a meta-analysis.

Funding

There were no external funding sources.

Acknowledgements

Special appreciation goes out to Amy Boots, DO (Doctor of Osteopathic Medicine), for assistance with the review and scoring of each included article. We also thank Yi Zhuang MD, MS, for assistance with translating the Wang article from China.[26]

Ancillary