The prevention of sudden infant death syndrome (SIDS or cot death) has been one of the major success stories in epidemiology. In the 1980s the SIDS mortality rate in New Zealand was extremely high (over 4/1000 live births) that is one in every 250 babies died suddenly, unexpectedly and without explanation.

The New Zealand Cot Death Study was a 3-year case–control study (1987-1990), funded by the Medical Research Council (now Health Research Council). Results from the first year of the study were reported in 1991 in this Journal.1 This identified three modifiable risk factors for SIDS, namely prone sleeping position, maternal smoking and lack of breastfeeding.

A prevention programme was launched with the release of the first year’s results.2 However, the prevalence of prone sleeping position had started falling before the study was completed, and with it a reduction in SIDS mortality.3 Prone sleeping position changed from 43% in the controls, which were a representative sample of all live births, prior to the prevention programme to less than 3%.

The fall in SIDS mortality was dramatic, with a halving of the SIDS mortality rate, within 2 years and this was accompanied by a reduction in total (all causes) postneonatal mortality (1–11 month mortality) showing that the reduction was real.4

Mortality rates have continued to decline, albeit more gradually and this has been attributed to the reduction in the prevalence of side sleeping position,5 which doubles the risk of SIDS compared with supine (back) sleeping position. This occurred without understanding of the pathophysiological mechanisms by which prone sleeping position causes death, although since then this has been the focus of considerable research, debate and speculation.

Other countries rapidly followed our lead, initially in Australia and then in the United Kingdom. The number of deaths in the UK fell from 1500 to 600 by the mid 1990s with a concomitant fall in prone sleeping.6 Since then the numbers have halved again with clear evidence of a reduction in side sleeping position and head covering.7 Indeed as the “Back to Sleep” campaign has been implemented throughout the developed world similar results have been seen.

How many lives has this simple intervention saved in New Zealand? Given that total postneonatal mortality had not changed in the two decades prior to the prevention programme it can be assumed that the SIDS mortality rate would have remained unchanged if the association between infant sleeping position and SIDS had not been identified.

The table shows the number of deaths from SIDS that would be expected to have occurred if the mortality rate had remained unchanged, the number of deaths from SIDS that did occur and thus the number of lives saved each year. Cumulatively more than 3000 lives have been saved.

A similar calculation has been done for England & Wales with over 17,000 lives saved and in the United States more than 40,000 lives saved (Hauck, personal communication, 2011). It is hard to think of any other intervention in the developed world that has had such a dramatic, rapid and clear cut effect.

So should we sit on our laurels or could more be done? The SIDS prevention programme also targeted smoking and promoted breastfeeding. There is considerable effort by the Ministry of Health and many other organisations to discourage smoking and promote breastfeeding. Given the many other health benefits there has been somewhat limited focus on their association with SIDS.

The prevalence of smoking in pregnancy has not been consistently collected in New Zealand. In the first year of the New Zealand Cot Death Study (predominantly 1989) the prevalence of smoking in pregnancy in the controls, which were a representative sample of all livebirths, was 34.1%.1

In Auckland the prevalence of smoking in pregnancy in 2009 was 10.1%.8 This is a fantastic achievement, but these total figures hide marked ethnic differences (Maori 41.0% vs. European 6.6%).

New Zealand breastfeeding rates have been good in comparison with many developed countries and thus there has been relatively little room for improvement. The Royal New Zealand Plunket Society (Plunket) has collected breastfeeding data for many years. Figure 1 shows the breastfeeding rates up to 6 weeks of age, 1985 to 2010 (Nikki Hooper, Plunket, personal communication, 2011). New Zealand’s high rate of breastfeeding has been maintained.

In 1992 we published the association between parents and infants sleeping in the same bed and an increased risk of SIDS.9 The following year we reported that the risk was particularly in infants of mothers who smoked.10 This has been confirmed in many other studies.11

Further studies have shown that the risk of SIDS from bed sharing is especially high in those infants who are under 3 months of age. Two retrospective studies of infant deaths referred to the coroner in Wellington and Auckland show that more than 50% of all sudden unexpected deaths in infancy occur while bed sharing, and this is 90% in the first month of life.12,13

There has been resistance from some quarters about promoting advice not to sleep with baby in the same bed. Some groups have actively encouraged bed sharing to encourage and maintain breastfeeding14 and there is emerging evidence of a complex interdependent relationship between these two infant care practices.15

Although the message about safe sleeping includes the risk from bed sharing, it is so dilute that the message has not been heard16. In particular the specific risks associated with hazardous bed sharing need to be clearly spelt out. Inappropriate sleep surfaces such as soft mattresses and sofas should be avoided and parents need to be reminded to never bring the baby into bed if they have recently consumed alcohol or taken legal or illegal sleep-inducing drugs.7 Surveys in Auckland show that less than 50% of mothers of infants identified the risk of SIDS with bed sharing.17,18

So what can be done? Parents need to be given clear evidence based guidance on the risks. The evidence has been summarised by the International Society for the Prevention and Study of Perinatal and Infant Death (ISPID) and the information concerning bed sharing is shown with permission in the box.19

Parental education is needed at antenatal services, in the obstetric unit and in the community by well child health care workers. Grandparents, child care workers, baby sitters all need to understand what keeps babies safe.

Modelling of appropriate infant care practices in obstetric hospitals is crucial. If parents are encouraged to bed share in obstetric units to facilitate breastfeeding, one cannot be surprised if this practice continues when the mother and baby go home.

Families need to be reminded that sleeping the infant in a cot next to the parental bed is the most risk-free environment and if they bed-share, intentionally or unintentionally, need to be aware of the risks involved. The media also has an important role, and was used effectively when the SIDS prevention programme was launched, now 20 years ago.

Some families cannot afford cots, and cots should be provided or rented to these families, rather like the Plunket Society’s infant car rental scheme. It is somewhat ironic that more infant deaths occur in the parental bed than in car crashes. The disruption to families following the Christchurch earthquakes resulted in some parents having to share their bed with their baby.

The rapid provision of pepi-pods (Figure 2) was not only a humanitarian response, but probably saved infant lives. Similarly the wahakura, a woven flax basket, is able to be taken into the parental bed and provides a safe sleeping space.20 These interventions have face validity, but have not been proven to save lives. However, it is hard to imagine that they have any downside.

The ideal of course is to identify ways that would enable parent and baby to sleep safely together in the same bed. This is the focus of Nationwide SUDI Case-Control Study which has been just been funded by the HRC.21

Unfortunately it will be 3 to 4 years before results are available. In the meantime we need to inform parents about the established risks from bed sharing with their infant.

Author information: Edwin A Mitchell, Professor of Child Health Research, University of Auckland, Auckland; Peter S Blair, Senior Research Fellow, University of Bristol, Bristol, England

Acknowledgements: We thank Chris Lewis, Ministry of Health for supplying the mortality data and The Royal New Zealand Plunket Society for supplying the breastfeeding data. Ed Mitchell is supported by Cure Kids.

Correspondence: Prof Ed Mitchell, Department of Paediatrics, University of Auckland, Private Bag 92019, Auckland 1142, New Zealand. Fax: +64 (0)9 3737486; email: e.mitchell@auckland.ac.nz