Objective: to determine the effectiveness of recommending Hoffman's nipple stretching exercises or breast shells (or both) to pregnant women with inverted or non-protractile nipples who intend to breast feed.
Design: randomized controlled trial with a two treatment by two level factorial design.
Setting: in the UK, antenatal clinics in hospital and community settings in 10 centres and the antenatal network of the National Childbirth Trust; in Ontario, Canada, antenatal clinics in six hospital centres and one public health unit.
Participants: 463 women with at least one inverted or non-protractile nipple and a singleton pregnancy, recruited between 25 completed and 35 completed weeks of pregnancy.
Primary outcome measure: rate of breast feeding as reported by postal questionnaire six weeks postnatally.
Findings: 107 out of 234 (46%) women allocated to use Hoffman's exercises compared with 100/229 (44%) women not allocated to use exercises were breast feeding at six weeks after delivery (difference 2%, 95% confidence interval -7% to 11%). One hundred and three out of 230 (45%) women allocated to use shells compared with 104/233 (45%) women not allocated to use breast shells were breast feeding at six weeks after delivery (difference 0%, 95% confidence interval -9% to 9%).
Conclusions: in the light of the findings from this and a previous single centre trial, there is not basis for recommending the use of either Hoffman's nipple stretching exercises or breast shells as antenatal preparation for women with inverted and nonprotractile nipples who wish to breast feed. Given the lack of evidence to support these and other antenatal preparations there are no grounds for midwives to continue routine breast examination in pregnancy for this purpose.