The deployment of air bags during car crashes involving pregnant women does not appear to affect pregnancy outcomes, a retrospective cohort analysis has found.

Maternal and fetal outcomes were assessed in pregnant women who were front seat occupants in cars involved in crashes in Washington State from 2002 to 2005. Information about the crash was obtained from police data.

The presence of air bags in the vehicle and, if fitted, whether they deployed was determined. For vehicles without air bags a crash investigator assessed whether an air bag would have deployed had it been fitted.

Duration of pregnancy, preterm labour, abruption, caesarean delivery as well as signs of fetal morbidity and death were recorded.

There were 3,348 non-rollover crashes involving pregnant women over the four-year study period.

This was equivalent to 10.3 crashes per 1,000 deliveries. In those crashes in which seatbelt use was known, seatbelt use was greater than 96%.

No significant differences in maternal or fetal outcome were identified between occupants of cars equipped with an air bag compared with those that were not.

The analysis controlled for small demographic differences, cars with air bags tended to be newer, more expensive, have more safety features and were more likely to be carrying more affluent and educated women and Asian women.

A non-significant trend for a higher risk of preterm labour RR 1.7 (95% CI: 0.9-3.2) and fetal death RR 3.1 (95% CI: 0.4-22.1) was found if an air bag deployed (or would have deployed had it been fitted). The absolute numbers of fetal deaths however were very small; 2 out of 198 in women whose air bags deployed compared with 2 out of 622 in those whose air bags did not.

Although some anecdotal reports have suggested that placental abruption, rupture of the membranes and even of the uterus are risks if air bags deploy during a collision, this study has provided some reassurance.

In general, air bags are an important safety feature of modern cars although they have been responsible for injury and even death particularly in children and adults of small stature. This author's hypothesis, that air bags present a specific hazard to women during pregnancy, seems to have been disproved, at least for women in the second half of pregnancy involved in non-rollover collisions.

Dr Chris Barclay