Antibody treatment has cut deaths in newborns
A test for spotting a mismatch between the blood of a pregnant woman and her baby could prevent thousands from undergoing unnecessary treatment.
Around 100,000 pregnant women a year are found to be Rhesus (RhD) negative, which can be dangerous for the baby.
But scientists say an easy, rapid test to assess the baby's RhD status means more than a third of RhD negative women can avoid uncomfortable injections.
Trial results of the test are reported in the British Medical Journal.
Currently, all women who test RhD negative at routine antenatal checks are given one or two antiserum injections during the pregnancy.
Where the baby is RhD positive, the injections prevent the mother making antibodies to the baby's blood, which can pass into the mother's blood at birth.
Without treatment there are usually no problems during a woman's first pregnancy, but if she goes on to have another RhD positive baby, such antibodies can cross the placenta and destroy the baby's red blood cells causing serious and sometimes fatal disease.
However, about 38% of RhD negative women are carrying an RhD negative baby, so they receive antiserum injections unnecessarily.
Researchers at the NHS Blood and Transplant Centre in Bristol assessed the new test for predicting the baby's blood group in almost 2,000 women.
DNA from the baby in blood was taken from the mother at or before the 28 week antenatal visit to test the baby's blood type.
In 96% of cases, the correct RhD status of the baby was predicted by the test, which the NHS researchers hope could be available more widely next year.
The findings are important because there have been concerns that the anti-serum injections, which are derived from blood products, could be a source of infections such as hepatitis C.
Study leader Geoff Daniels, head of molecular diagnostics at NHS Blood and Transplant in Bristol said they were carrying out further research to determine if the test could be done earlier in pregnancy.
"It's good practice not to give treatment to people who don't need it."
Dr Sailesh Kumar, consultant in foetal and maternal medicine at Queen Charlotte's and Chelsea Hospital in London said the test was already being used for select patients.
"This paper shows very nicely it's possible to apply this technique to much larger numbers and that you get reliable results and you can target women who are carrying RhD positive babies.
"It's fair to say women don't like any interventions unless they are absolutely necessary."
But Professor Andrew Shennan, an expert in Obstetrics at King's College London said the injections were not ideal and were expensive and inconvenient for the woman.
"It's not a perfect test and you get an occasional false negative, so the question is can you justify not treating all women."