Primary review outcome

In our protocol the primary outcome was 'Postpartum contraceptive prevalence at the first postpartum check‐up visit'. This has been re‐worded as 'Rate of contraceptive implant initiation at the first postpartum check‐up visit'.

Unit of analysis issues

We intended to include studies where individual postpartum women were randomised and cluster‐randomized studies where, for example, the hospital was the unit of randomisation. However, we did not find any related cluster‐randomised study. In future updates of this review, for studies that use a cluster‐randomised design but did not have any information related to the design effect, we will estimate the design effect based on a fairly large assumed intra‐cluster correlation of 0.10. We will base this assumption by analogy on studies about implementation research (Ukoumunne 1999; Campbell 2000).

Assessment of reporting biases

As only three RCTs met the inclusion criteria, we were unable to construct funnel plots to determine the possibility of publication bias as previously stated in the protocol (Sothornwit 2015). In future updates of this review, we will construct funnel plots that correspond to the findings of the meta‐analysis of the primary outcome that this review needs to address if we identify a sufficient number of included studies (i.e. more than 10). We will also perform sensitivity analyses to investigate the effect on the pooled results of the funnel plots (Sterne 2011).

Subgroup analysis and investigation of heterogeneity

We did not perform subgroup analyses, as we stated in the Cochrane protocol, due to the limited number of included studies. All included studies were conducted in the USA. In future versions of this review, we will perform subgroup analysis to investigate heterogeneity, if feasible, and we will employ the following methods for subgroup analysis.

• We will consider whether an overall summary is meaningful, and if it is, use a random‐effects analysis to produce it.

• We plan to perform subgroup analyses according to age of participants (teenagers versus non‐teenagers) and population differences (high‐income versus low‐ and middle‐income countries).

• We will assess subgroup differences by interaction tests available within RevMan 5 (RevMan 2014). We will report the results of subgroup analyses quoting the Chi² statistic and P value, the interaction test, and I² statistic value.

Sensitivity analysis

Only three published RCTs were available for this review. which were at low risk of selection bias. Furthermore, we did not find any unpublished studies. Thus we were unable to perform sensitivity analyses as we stated in the Cochrane protocol (Sothornwit 2015). In future versions of the review we will perform sensitivity analyses, if feasible, in order to determine the impact of the following factors on effect size.

• Repeating the analysis by excluding unpublished studies (if any).

• Repeating the analysis by excluding trials rated as at 'high’ or ’unclear’ risk of bias for allocation concealment.