Commentary

Although the umbilical cord plays a vital part in fetal development, the newborn's cord stump is generally thought to be a redundant feature. It receives much attention, however, from maternity staff and parents.¹ As the only “open” area of the baby's body, infection can develop. Many prescriptions and proscriptions for newborn cord care have achieved ritual status in response to fears of infections, cross contamination, and parent anxiety; however, many have been based on good intention rather than on the results of rigorous scientific trials.

Medves and O'Brien studied cord care using common agents (water and isopropyl alcohol); Zupan and Garner have analysed data from another 10 original studies from developed countries on topical agents for cord care. Medves and O'Brien found that the mean time to cord separation when using sterile water for cord care was about 10.5 days, approximately 2.5 days shorter than when using isopropyl alcohol, with no differences in colonisation rates. Zupan and Garner found that the shortest time to cord separation occurred with powder (7 d), with no intervention (8 d), and with alcohol (10 d). Antiseptics (strong chlorhexidine, silver sulphadiazine, and triple dye) reduced bacterial colonisation and prolonged time to cord separation.

Both the original study and the systematic review are methodologically sound. Medves and O'Brien calculated sample size to ensure that they would find a clinically important difference if it existed. They used an intention to treat analysis (ie. the infants remained in the treatment group to which they were assigned for analysis; even if the parents did not use the assigned topical agent). To determine colonisation, microbial growth was recorded by a technologist who was unaware of which topical agent was used. All infants were bathed in Hibitaine (chlorhexidine gluconate 2%) after birth. This cleansing of the skin of bacteria required for colonisation may explain why so many specimens had no microbial growth on the third day. A description of the instructions given to parents about the frequency of cord care in this trial would have been helpful in applying the results or replicating the study. The very “rough” calculation of annual cost savings if umbilical cord care was discontinued was CN$16 000 in a hospital with about 4800 births a year. Detailed collection of costings data and a proper economic analysis would have been preferable.

The Zupan and Garner overview was limited to randomised controlled studies. They included studies of various topical agents and were comprehensive in the outcomes they examined. They abstracted data from published reports only which have biased the results. It has been shown that studies which report positive findings (ie, that an intervention had beneficial or harmful effects for patients) are more likely to be published than studies that report no differences.² Studies which found that topical agents did not reduce cord separation time or colonisation may therefore be underrepresented in the literature. No studies were from developing countries; follow up was generally to cord separation and therefore, outcomes subsequent to this, such as local or general bacterial infection, would not be detected; sample sizes were small; and cost savings were not measured. Interventions were tested on well, term infants and so findings cannot be generalised to preterm infants who are at higher risk of infection.

With respect to clinical practice, the following points may be considered. Firstly, few interventions for cord care have been evaluated using randomised controlled trials. This is surprising given the millions of newborns in developed countries who receive cord care. Secondly, no studies are from developing countries where neonates have a much higher risk of infection. Thirdly, given the importance of avoiding a delay to cord separation because the unhealed umbilicus is a potential site for infection, Medves and O'Brien recommend the discontinuation of alcohol for routine umbilical cord care, monitoring the progress of cord separation, and intervening as necessary. Zupan and Garner also note that no evidence exists to suggest that doing nothing or maintaining clean care (cleansing with water or gentle soap and water and ensuring a dry site) is harmful. One study suggested that doing nothing caused parent concern. This is not surprising in light of how cord care has been committed to the historically long list of maternity routines. Fourthly, although Zupan and Garner found reduced colonisation with antiseptic use, they explain that the clinical significance of skin colonisation is not known. Because use of antiseptics is unlikely to be harmful, however, there appears to be no good reason to stop the use of antiseptics in situations where the risk of bacterial infection seems high (eg, preterm babies, babies in intensive care units, and those in institutions in developing countries).