A critical appraisal of the evidence for using cardiotocography plus ECG ST interval analysis for fetal surveillance in labor. Part II: the meta-analyses

Abstract

We appraised the methodology, execution and quality of the five published meta-analyses that are based on the five randomized controlled trials which compared cardiotocography (CTG)+ST analysis to cardiotocography. The meta-analyses contained errors, either created de novo in handling of original data or from a failure to recognize essential differences among the randomized controlled trials, particularly in their inclusion criteria and outcome parameters. No meta-analysis contained complete and relevant data from all five randomized controlled trials. We believe that one randomized controlled trial excluded in two of the meta-analyses should have been included, whereas one randomized controlled trial that was included in all meta-analyses, should have been excluded. After correction of the uncovered errors and exclusion of the randomized controlled trial that we deemed inappropriate, our new meta-analysis showed that CTG+ST monitoring significantly reduces the fetal scalp blood sampling usage (risk ratio 0.64; 95% confidence interval 0.47-0.88), total operative delivery rate (0.93; 0.88-0.99) and metabolic acidosis rate (0.61; 0.41-0.91).

Keywords: Cardiotocography; ST analysis; fetal surveillance; meta-analysis; metabolic acidosis; randomized controlled trial.

Figure 1

Forest plot and details of an aggregate meta-analysis of the usage of fetal scalp blood sampling in labor.

Figure 2

Meta-analysis of total operative delivery (sum of cesarean sections and instrumental vaginal deliveries). Data on total cesarean section from the Plymouth randomized controlled trial (Westgate et al., 1993) were obtained from the Cochrane Review .

Figure 3

Meta-analysis of neonatal metabolic acidosis. Data from the Finnish randomized controlled trial are from Dr. Ojala’s personal communication with Welin et al. , the Swedish (Amer-Wåhlin et al., 2011) and Dutch (Westerhuis et al., 2011) data are from the revised articles ,, while the Plymouth data (Westgate et al., 1993) are from the original article .

Figure 4

Meta-analysis of neonatal encephalopathy stage ≥2 according to Sarnat & Sarnat . The Sarnat & Sarnat classification was used only in the Swedish (Amer-Wåhlin et al., 2001) and Dutch (Westerhuis et al., 2010) trials ,.