Continuation: A Meta-Analysis 193 production.35-38 Although we did not distinguish between the 2 in our analyses, the type of cesarean delivery (planned or emergency) has been found to influence breastfeeding initiation and duration. Although mothers who underwent emergency cesarean deliveries had greater difficulty initiating breastfeeding compared with vaginal deliveries,36,39 several studies have shown that planned cesarean delivery in particular is associated with a significant decrease in breastfeeding initiation.39-41 Women undergoing a planned cesarean delivery were less likely to intend to breastfeed, initiate breastfeeding, or seek lactation support.39 Infants born by planned cesarean deliveries are more likely to suffer from factors associated with lower gestational age that impact breastfeeding initiation, such as poor sucking skills and decreased alertness.35,39,42 A large degree of heterogeneity was noted in the individual studies examining the effect of multiparity on breastfeeding initiation. Some studies showed a positive association with multiparity and others showed a negative effect, and the summary RR for breastfeeding initiation was nonsignificant. However, multiparity was positively related to continuation of breastfeeding. Prior work has shown that challenging breastfeeding experiences, unsuccessful attempts, and failure to initiate breastfeeding with the first child have been associated with failure to initiate breastfeeding with subsequent births.39,43 A composite metric for positive dyad connections (skinto-skin vs not or rooming-in vs not) was associated most strongly with increased breastfeeding initiation but also with breastfeeding continuation. Keeping the mother and infant dyad together during their hospital stay promotes attachment within the dyad, which is a likely mechanism to improve breast feeding initiation and duration.44 A delay in breastfeeding initiation can result in reduced suckling ability and receptivity of the infant resulting in reduced or insufficient milk supply.45-53 These results are consistent with the World Health Organization’s Baby-friendly Hospital Initiative that recommends that the mothers are supported and encouraged to initiate breastfeeding with the first hour after birth (step 4) and that infants and mothers remain together 24 hours a day (step 7).35,42,54 Despite measurement differences between study populations from different countries, the highest level of education vs the lowest was consistently associated with a higher likelihood of both breastfeeding initiation and continuation. Other sociodemographic characteristics such as lower maternal age and household income were similarly associated with decreased probability of breastfeeding initiation and continuation in another study.55 These factors are likely to be at least partially accounted for within our factor of maternal education. More highly educated mothers may have more control over their schedule or work environment, which may provide the support needed to breastfeed for a longer time.56 Mothers who received education on breastfeeding were 41% more likely to initiate and continue breastfeeding than women who received no such educational opportunities. Attendance at prenatal breastfeeding classes likely provides women with strategies to cope with the challenges associated with the first few weeks of breastfeeding (eg, perception of insufficient milk supply, breast engorgement, and cracked nipples)57-60 as well as longer term strategies such as establishing breastfeeding or pumping routines.61,62 Our grouping of breastfeeding education generally aligns with steps 3 (inform all pregnant women about the benefits and management of breastfeeding) and 10 (foster the establishment of breastfeeding support groups and refer mothers to them on discharge from the hospital or birth center) of the Baby-friendly Hospital Initiative. The studies we included in our review are more recent than the original research cited by the Baby-friendly Hospital Initiative 10 steps,54 but our meta-analysis comes to the same conclusion as the World Health Organization; there is significant evidence that increased education and support for mothers or parents during pregnancy or soon after birth improve breastfeeding initiation and continuation. A recent systematic review employed stringent inclusion and exclusion factors to examine a large number of sociodemographic, physical, maternal, and social factors in relation to breastfeeding in the first 6 months but did not include a quantitative analysis.63 Although our review and meta-analysis focused on a more limited number of factors that had a relatively large volume of high-quality literature, many factors were not included. In our initial search, we expected to identify infant birth factors such as hypoglycemia, jaundice, birth weight, and gestational age, but found very little published data meeting our inclusion criteria. These factors are often considered clinically important predictors of a successful early breastfeeding relationship but additional well-designed studies to substantiate this assumption appear warranted based on the lack of peer-reviewed evidence. We also did not examine associations between breastfeeding and formula discharge packs provided to mothers leaving the hospital; this practice was shown