Table 5 Comparison of randomisation methods and intervention of studies evaluating peer counseling programmes

Kistin et al 1994 [13]

Compared women who planned to breastfeed and received support from counsellors (n = 59) with those who requested counsellors but, owing to inadequate counsellors, did not have a counsellor (n = 43).

Trained counsellors matched by race if possible to low-income pregnant or postpartum women. Contact prior to delivery encouraged and then telephone contact every 1 to 2 weeks until two months and then "as needed".

Breastfeeding initiation, exclusivity and duration.

Schafer et al 1998 [14]

2 "intervention counties" and 6 "control counties". All women referred to Women, Infants and Children's Nutrition programme centres in these counties were recruited.

The assignment of trained volunteers with previous successful personal experience with breastfeeding as peer counsellors to low-income pregnant women. Peer counselor met mother antenatally. Maintained telephone contact between visits.

Initiation and duration of breastfeeding

Arlotti et al 1998 [19]

Convenience sample from prenatal and postpartum clients who were assigned to counsellors based on their desire to have a counsellor and availability of counsellors. Mothers who were not matched with a counsellor were the control group.

Counsellors contacted mothers within a few days of delivery and again at 2 w, 1 m, 2 m and 3 m. Further contacts by telephone, letter or in person at clinic

Exclusive breastfeeding at 2 weeks, 1 month, 2 months and 3 months and duration of exclusive breastfeeding

Morrow et al 1999 [8]

Cluster randomisation before the recruitment of study mothers. City blocks were the unit of randomisation.

3 or 6 counselling home visits antenatally and early postpartum

Exclusive breastfeeding at 3 months and duration of breastfeeding

Haider et al. 2000 [7]

Cluster randomisation. city area divided into zones and random number tables used to randomly select zones into 2 study groups Thereafter, mothers were approached in the community and invited to participate.

15 home-based counselling visits scheduled with two visits in last trimester, three early postpartum and then every two weeks til 5 months

Exclusive breastfeeding at 5 months

McInnes at al 2000 [15].

Mothers recruited antenatally by a clerical officer in two culturally and socially similar but geographically separated communities in Glasgow.

An offer to the intervention group of having visits from a "helper" with a minimum of 4 visits (2 antenatally and 2 postnatally), irrespective of breastfeeding intention.

Breastfeeding at 6 weeks

Dennis et al. 2002 [5]

Randomised eligible mothers after they had been recruited and had signed written informed consent. Used randomly generated numbers and sequentially numbered sealed opaque envelopes. Mothers recruited from 2 semi-urban community hospitals near Toronto.

Telephone-based support, within 48 hrs of hospital discharge

Breastfeeding (any or exclusive) at 1,2, and 3 months