Pregnancy and child

Baby doll simulators may actually increase teen pregnancy rates

"Young girls exposed to electronic babies – designed to simulate the real experience of having a baby and discourage teenage pregnancy – were more likely to get pregnant," The Guardian reports.

"Infant simulators" – dolls that mimic the need of a baby in terms of feeding and nappy changing through crying – are meant to show the challenges of looking after a real baby.

A new Australian study investigated the effect of using Virtual Infant Parenting (VIP) – a type of infant simulator programme – for teenage girls on pregnancy outcomes of birth and induced abortion in Australia.

Results suggest the programme doesn't help prevent teen pregnancies, it actually increases the risk.

Of girls in the intervention group, 17% got pregnant at least once in their teenage years compared with 11% in the control group (who received standard advice).

Some local authorities in England have used VIP type programmes, with varying degrees of success.

The study reinforces the fact that even the most well-meaning interventions, unless backed up by actual evidence, can have the opposite effect to those intended. Most notoriously, previous advice that babies should sleep on their stomach, is now known to be a potential cause of death from Sudden Infant Death Syndrome (SIDS). 

Most UK experts would argue that the most effective methods of preventing teen pregnancy are access to non-judgemental relationship advice and cheap reliable contraception. These methods, as recently reported, may have led to a 50% drop in teen pregnancies since 1998.

Where did the story come from?

The Australian study was carried out by researchers from a number of institutions, including the University of Western Australia, The University of Adelaide and The University of Notre Dame, Australia.

The study was funded by The Health Promotion Research Foundation of Western Australia (Healthway), Lotteries WA, the Western Australian Department of Education and Training, and the Western Australian Department of Health. The study was published in the peer-reviewed medical journal The Lancet.

The Mail Online provides the most accurate summary of the study, with a useful overview of the history of use of infant simulator programmes by local authorities in the UK, such as Birmingham, West Sussex and South Yorkshire.

This good reporting is let down by the fact that the webpage featuring the story contains a promotional video for a US company that sells "virtual infants".

What kind of research was this?

This was a cluster randomised controlled trial (RCT) which aimed to investigate the effect of using Virtual Infant Parenting (VIP) programmes for teenage girls on pregnancy outcomes of birth and induced abortion in Australia.

Randomised controlled trials are considered the gold standard for assessing whether an intervention is effective. "Cluster" means that groups of participants, rather than individuals are randomised to each intervention arm. The nature of this trial means that the participants and health professionals involved were not able to be blinded, however, it is unclear whether the researchers who analysed the data were.

What did the research involve?

The researchers enrolled 57 eligible schools in Perth, Australia into the trial which were randomly allocated 1:1 to receive the VIP programme (28 schools) or the standard health education curriculum (29 schools).

Between 2003 and 2006, both interventions were administered to girls aged 13-15 (mean age 14.9) in the included schools. A total of 2,834 girls were included in the study (1,267 in the VIP programme and 1,567 in the standard education programme).

Alongside caring for a simulation doll, participants also received a series of education sessions highlighting sexual health, contraception and the financial aspects of having a baby.

The researchers followed the participants until the age of 20 via hospital medical and abortion clinic records, noting the occurrence of pregnancy (defined as live birth, still birth or induced abortion) during the teenage years.

The data was analysed to test for differences in pregnancy rates between the two study groups. Only the first pregnancy was used in this analysis. Potential confounders were adjusted for, including:

  • socioeconomic status
  • family type
  • whether the girl had ever had sex
  • whether she had ever had responsibility caring for a baby
  • educational attainment
  • her level of psychological distress
  • whether she drank alcohol
  • current smoker status

What were the basic results?

Overall, the findings showed that girls who took part in the VIP programme were more likely to have a recorded pregnancy compared to those who received the standard curriculum.

Overall, 378 (13%) of the 2,834 girls in the study got pregnant at least once (birth or abortion) in their teenage years. The proportion of girls recording pregnancy events was higher in the intervention group: 17% (210/1,267) versus 11% (168/1,567) in the control group. This meant that the intervention was associated with significantly higher pregnancy rate (relative risk 1.36, 95% confidence interval 1.10 to 1.67).

Additionally, the proportion of girls in the intervention giving birth was also higher when compared with the control group: 8% (97 of 1,267) and 4% (67 of 1,567), respectively. However, it is important to note that the control group did have 300 more participants than the intervention group, so proportions may have differed had the numbers been equally matched.

Three-quarters of the 378 girls had recorded just one pregnancy event. The remaining 93 recorded more than one pregnancy, with 19 or more having two or more births and 26 having two or more induced abortions.

How did the researchers interpret the results?

The researchers concluded: "This study shows that the infant simulator-based VIP programme did not reduce the risk of pregnancy in teenage girls in Australia, as measured by births and induced abortions. Point estimates for the effect of the intervention were increased, suggesting a higher pregnancy risk in girls who experienced the VIP programme than in those who did not."


This trial investigated the effect of using Virtual Infant Parenting (VIP) programmes for teenage girls on pregnancy outcomes of birth and induced abortion in Australia. Contrary to what may have been expected, it found that girls who took part in the VIP programme were actually more likely to have a recorded pregnancy (birth or induced abortion) compared to those who received the standard curriculum.

This trial had a good study design and a suitable sample size; however, there are a few things to bear in mind:

  • While these findings are interesting, this was an Australian study. Social and lifestyle factors may differ from girls in the UK.
  • This intervention was targeted at girls aged 13-15, and may have had different outcomes had it been trialled on girls of different age.
  • The programme specifically focused on girls, whereas boys play an equal part in teenage pregnancies. In the US, similar programmes are received by both boys and girls.

In an accompanying editorial in The Lancet, health researcher Julie Quinlivan offers a number of suggestions why the Australian scheme had the opposite effect. These include (as mentioned above) "it takes two to tango" so teen boys received no training, and teen girls using dolls may have received positive feedback.

The current approach in this country is based on providing non-judgemental advice on sex and relationships (including how it's always okay to say no) as well as information about and access to contraception.  

NHS Attribution