1. Objectives

To evaluate the effect of mother-infant rooming-in in comparison to separation on the duration of breastfeeding

2. How studies were identified

The following databases were searched in May 2016:

• Cochrane Pregnancy and Childbirth Group’s Trials Register
• Register CENTRAL (The Cochrane Library 2016)
• MEDLINE
• Embase
• CINAHL

Reference lists were also hand-searched

3. Criteria for including studies in the review

3.1 Study type

Randomized controlled trials, including cluster-randomized and quasi-randomized trials

3.2 Study participants

Mothers who are able to care for their healthy newborn infants, whether or not they have initiated breastfeeding

(Studies enrolling populations with health problems such as HIV were excluded)

3.3 Interventions

Rooming-in, where the infant is either placed in a stand-alone cot by the bedside or is bed-sharing by attached side-car crib, in comparison to the infant being placed separate to the mother in a hospital nursery, or, for home-deliveries, in a separate room

(For hospital-based deliveries, rooming-in was defined as mother and infant being placed in the same room immediately after leaving the delivery suite, and, in the case of caesarean sections, when the mother was able to respond to her infant. For home-deliveries, rooming-in was defined as mother and infant being placed in the same room immediately after birth)

3.4 Primary outcomes

• Duration of exclusive breastfeeding
• Duration of any breastfeeding
• Proportion of infants being exclusively breastfed at six months of age
• Proportion of infants receiving any breastfeeding at six months of age (not pre-specified)

Secondary outcomes included frequency of breastfeeding; exclusive breastfeeding on discharge from hospital or four days postpartum; maternal breast engorgement, sleep duration, adverse events, satisfaction, and confidence in breastfeeding; and neonatal infection, diarrhoea, hypoglycaemia, hypothermia, and hyperbilirubinaemia requiring therapy

4. Main results

4.1 Included studies

Only one trial enrolling 176 mothers was included in this review

• The included trial employed a 4 x 2 factorial design with the following treatment arms: i) skin-to-skin contact at birth with rooming-in, ii) no skin-to-skin contact with rooming-in, iii) no skin-to-skin contact with 24 hours of nursery care, and iv) no skin-to-skin contact with rooming-in delayed up to three hours following birth
• ‘Nearly exclusive breastfeeding’ was reported in the trial, which was included in the review as ‘any breastfeeding’

4.2 Study settings

• Maternity home in St Petersburg, Russian Federation

4.3 Study settings

How the data were analysed
Rooming-in arms (skin-to-skin contact at birth with rooming-in, no skin-to-skin contact with rooming-in, no skin-to-skin contact with rooming-in delayed up to three hours following birth) were compared to the separate care arm (no skin-to-skin contact with 24 hours of nursery care). Dichotomous data were analysed to produce summary risk ratios (RR) and 95% confidence intervals (CI).

Results
Duration of any breastfeeding
As medians were reported, data could not be analysed in this review. The study authors reported no significant difference in the duration of any breastfeeding between rooming-in and separate care groups.

Proportion of infants receiving any breastfeeding at six months of age
No difference between rooming-in and separate care groups was found in the proportion of infants receiving any breastfeeding at six months (RR 0.84, 95% CI [0.51 to 1.39], 137 women). Subgroup analysis by treatment arm did not alter the results in a meaningful way.

Additional outcomes
Although the mean frequency of breastfeeds per day was slightly higher among the rooming-in group at four days postpartum (eight versus seven times per day), the separate care group was fed at a fixed schedule of seven times per day, thereby preventing analysis of this outcome. In comparison to separate care infants, the rooming-in infants were 92% more likely to be exclusively breastfed on day four postpartum before hospital discharge (RR 1.92, 95% CI [1.34 to 2.76], p=0.00038; 153 participants).

No other outcomes were reported on.

5. Additional author observations*

The main limitation of this review is insufficient data. In the single included trial, one of the three rooming-in arms involved a delay of up to 3 hours before rooming-in was initiated, and in two of the three rooming-in arms and in the separate care group, women did not have any skin-to-skin contact with their infant in the labour ward. These factors may have influenced breastfeeding outcomes. Using GRADE to assess the quality of the evidence, findings for both the proportion of infants receiving any breastfeeding at six months and the rate of exclusive breastfeeding at four days postpartum were rated as being of low quality.

Due to a paucity of evidence, no conclusions regarding the effects of rooming-in or separation on the duration of breastfeeding may be drawn. Well-designed randomized controlled trials are therefore needed to investigate whether the practice of rooming-in after delivery results in an increased duration of breastfeeding. Future trials could also qualitatively investigate the process of rooming-in to explore which social and environmental factors affect this practice.