Study	Reason for exclusion
Greenberg 1973	This is a randomised controlled trial of 100 primiparous mothers who had an uncomplicated vaginal delivery. 50 mothers were assigned to room‐in and the infant was brought to the mother within 12‐36 hours after birth and remained in a cot from 9 am to 6 pm and later returned to the nursery. The other group of the mothers were assigned to a nursery care unit where their infants were placed in the open nursery. The outcomes measured in this study were mother's level of confidence and level of competence prior to leaving the hospital. In this study, rooming‐in did not meet our defined rooming‐in criteria.
Sousa 1974	This trial reported only in abstract form, comparing separate care and rooming‐in, is not clear how the mother‐infant pairs were assigned. 200 mother‐infant pairs were recruited after a normal vaginal delivery. 'Successful lactation' up to 2 months was measured.
Sosa 1976	This is a randomised controlled trial of 160 primiparous mothers to determine the physical benefits of early mother‐infant skin‐to‐skin contact. The mothers in the experimental groups had skin‐to‐skin contact with their infants at birth followed by rooming‐in for only 45 minutes before the infants were placed in the nursery for their remaining hospital stay. The outcomes measured in this study were mean duration of breastfeeding in the first year, the rate of breastfeeding up to 1‐year postpartum, affectionate behaviour, rate of infection and infant growth. In this study, rooming‐in did not meet our pre‐defined rooming‐in criteria.
Hales 1977	This is a randomised trial of 60 healthy primiparous mothers to assess the effect of mother‐infant skin‐to‐skin contact after birth on mothers' affectionate behaviour towards the infants. Mothers were assigned to 3 groups, i.e. early skin‐to‐skin contact, delayed contact and controls, The study did not meet our inclusion criteria as it did not study rooming‐in.
Kontos 1978	This non‐randomised trial examined the effect of extended contact in the early postpartum hours and days on maternal attachment behaviour at 1 and 3 months. The study comparison group did not meet our pre‐specified criteria of rooming‐in.
Ali 1981	This randomised trial examined the effects on later behaviour of a 45‐minute period of contact immediately after birth between a woman and her full‐term healthy newborn. A total of 100 mothers from a low‐income group were recruited after uncomplicated vaginal delivery. Half (50) of the mothers were randomly assigned to the study group where they were allowed contact with their infants for 45 minutes, followed by separate care for 9 hours before re‐establishing contact. Both the intervention and comparison did not meet our pre‐specified inclusion criteria.
Grossmann 1981	This randomised controlled trial examined the effect of early and/or extended mother‐infant tactile contact. This study did not meet any of our pre‐specified inclusion criteria.
Elander 1986	The design of this study is quasi‐random. However, the authors did not fully comply with allocation in that some participants were allocated by convenience. A total of 29 infants undergoing phototherapy to treat jaundice were alternately selected into separated care (infant was transferred to pediatric ward for treatment of jaundice and mother in maternity ward) and non‐separated (both mothers and her infants were transferred to the pediatrics ward when bed was available). However, when there was no empty single room available for rooming‐in the infant was assigned to the separate care group. The age of the infants at allocation and the type of care prior to allocation was not described.
Lind 1986	In this study, 344 primiparas after uncomplicated vaginal delivery of term infants with birthweights of 3000 g to 4000 g were recruited and randomly selected to a rooming‐in group (n = 172) where the newborns stayed with the mothers during the day time and returned to nursery during the night. The comparison group (n = 175) had separate care in the nursery until discharge. At discharge mothers completed questionnaires on duration of exclusive and any breastfeeding during the hospital stay. The study did not meet our pre‐specified inclusion criteria for rooming‐in.
Keefe 1986	This randomised trial compared the state of behaviour of newborns who were rooming‐in with their mothers at night with those who were cared for in the traditional nursery at night. During the day all the infants were cared for in the nursery. The study comparison group did not meet our pre‐specified criteria for rooming‐in.
Lindenberg 1990	The study was carried out over 2 time periods. Women were assigned to separate care during the first time period and to rooming‐in for the second time period. Women during the first period were randomised to 2 different types of separate care. The outcome reported measured was incidence of breastfeeding at 1 week and 4 months.
Perez‐Escamilla 1992	This trial was a non‐random controlled comparison of rooming‐in versus nursery care. 58 eligible women who delivered in Hospital A were assigned to nursery care and compared with 107 women assigned to rooming‐in in Hospital B. The women in Hospital B were randomised to rooming‐in and rooming‐in with breastfeeding guidance. The outcome measured was duration of full and any breastfeeding. This study reported a significantly higher breastfeeding rate at 4 months among primiparae who had a childbirth at the rooming‐in hospital and received breastfeeding guidance.
Segura‐Millan 1994	This randomised trial aimed to evaluate a maternity ward breastfeeding promotion programme and to identify factors related to perceived insufficient milk. Women were interviewed in the hospital, at 1 week, 2 months and 4 months postpartum for factors associated with perceived insufficient milk. This study did not meet our inclusion criteria.
Kastner 2005	This randomised trial aimed to evaluate the impact of a first hour postpartum mother‐infant contact on the mother‐child relationship during the puerperium period. Immediately after delivery the mother‐infant pairs were assigned either to a group where mother and child spent the first hour alone and the control group which followed labour room routine practice. This trial did not study rooming‐in versus separate care of mother‐infant pairs.
O'Connor 1980	This 'randomised' trial examined the effect of rooming‐in on the incidence of measures of parenting inadequacy. A total of 301 mother‐infant pairs were studied. The randomisation and allocation was based on the availability of the rooming‐in beds in the postnatal ward. When both type beds were available, there was no information about how mothers were assigned to the postnatal beds. About 143 mothers were assigned by the bed availability to rooming‐in (n = 143), where the infants were roomed‐in with their mothers 7 hours after birth up to 8 hours each day until discharge. The other 158 mother‐infant pairs had separate care after a glimpse of their baby at birth. Parental inadequacy was measured after 17 months by the rate of infestations, rate of accident, exanthematous disease and hospitalisation. Rooming‐in in this study did not meet our definition of rooming‐in where the infants are expected to stay with mother in the same room or in nursery for 24 hours a day.
Ball 2006	This randomised trial examined the effect of 3 rooming‐in practices on the initiation of breastfeeding and infant safety. There was no separate care group.
Ball 2011	This study compared the effect on breastfeeding duration of side‐car cribs with stand alone cots in rooming‐in mother infant pairs. There was no separate care group.