Alice Stewart and prenatal X-rays

In the early 1950s childhood leukaemia seemed to be growing more common, possibly in part because antibiotics now cured infections such as pneumonia that had once masked an underlying tumor. Alice Stewart in Oxford set out to find its causes. The disease was so rare that following a group of children forward would have needed an impossibly large study, so she designed a case-control study on a national scale. She obtained the death certificates of the children dying of malignant disease in England and Wales, at first those under 10, and matched each with a healthy control child of the same sex and date of birth drawn from the birth register of the area where the child had died.

Two, and later three, years after each death, medical staff recruited from local authorities interviewed the mothers of both children, usually the same interviewer for each pair, and the mothers' answers were checked against hospital and clinic records. The first deaths studied were those of 1953. A preliminary communication in the Lancet in 1956 by Stewart, Webb, Giles and Hewitt reported a statistically significant excess of antenatal X-ray exposure among the first 547 case-control pairs. Stewart, Webb and Hewitt confirmed it in the British Medical Journal in 1958 with 1,416 traced, matched and interviewed cases who had died in 1953 to 1955.

Many doubted the result. Critics, among them Richard Doll, argued that mothers of dead children might recall hospital visits differently from control mothers, and a cohort study failed to confirm the association. Stewart defended her findings forcefully. When she reached retirement age in 1974 she took a research fellowship at Birmingham University, and the survey later moved there. It eventually covered deaths up to 1981 and older children, and later analyses showed the risk per X-ray film falling in later birth cohorts, most likely because equipment delivered lower doses.

Doll came to accept that the association was probably causal. Reviewing the survey in 2018, C. R. Muirhead wrote that its most notable result, the link between fetal irradiation and childhood cancer, led in large part to the reduced use of antenatal X-raying.