The main result of this systematic review suggests that encouraging women to adopt an upright position or to ambulate during the first stage of labor reduces its duration. However, the robustness of this finding is limited, since it is associated with a high level of heterogeneity (assessed by the I2). In fact, the consistency of a meta-analysis depends on the similarity of magnitude of the effects of the studies included, and the assessment of the consistency of effects across studies can be carried out by measuring heterogeneity (i.e, the degree of genuine differences between the studies and their results). Considering the 95% confidence interval, this positive effect of maternal position on the duration of labor may be merely marginal.
To critically appraise these results, the presence of any underlying potential sources of heterogeneity has to be examined. Typically, heterogeneity is associated with reporting bias, differences in: the intensity of interventions, the underlying risk, the effect size according to the study sample size and irregularities of data.
Analysis of Figures 2 to 6 may suggest the occurrence of a "small study effect". The small study effect is the trend for smaller studies in a meta-analysis to show larger treatment effects and it is also associated with reporting bias . In this systematic review, two studies accounted for approximately 76% of the total reported sample (large studies) [10,14], two studies accounted for almost 14% of the sample (intermediate studies) [12,15], while five studies, each with fewer than 100 participants, accounted for approximately 10% of the total reported sample (small studies) [2,8,9,11,13]. Using the study performed by Flynn et al  as an example of the small study effect, this study accounts for approximately 3% of the total meta-analysis sample, but may have a much greater weight in the analysis (Figure 2). There is some controversy regarding how to deal with the small study effect. Simulation of exclusion of the Flynn study would change the conclusion of this meta-analysis (duration of first stage of labor, WMD (random) = -0.65; 95%CI -1.43-0.13).
Another possible source of heterogeneity in the present meta-analysis is the intensity of intervention in the studies included and the occurrence of several degrees of contamination (provision of the intervention to the control group) and co-intervention (provision of unintended additional care to either comparison group), as noted in Table 1. On the other hand, the occurrence of a performance bias in one of the larger studies was observed . In this study, amniotomy was performed later in the study group. Considering that early amniotomy is associated with a reduction in the duration of first stage of labor , it is possible that the delay in performing amniotomy counterbalanced any possible effect of ambulation or standing in the upright position during the first stage of labor in that study.
The adoption of an upright position or walking during labor possible interferes on the performance of other interventions such as amniotomy, analgesia and monitoring during labor. The reverse way of this statement is also valid and this fact makes the isolation of possible effects of upright position or walking during labor a difficult task.
The evaluation of secondary outcomes suggests that the upright maternal position is a safe intervention. At the same time, while it produces no apparent benefit, neither does it appear to do any harm. The effect on maternal comfort is unclear, but freedom of movement may benefit patients individually.