Two studies in Sweden by the same group have found high rates of defects in ultrasound machines used to scan pregnant women. The line of research began when doctors at the Karolinska Hospital discovered that many of their ultrasound machines were malfunctioning.
The first study examined about 700 machines from 7 manufacturers. About 40% had defective transducers — the only element tested. “The high error frequency and the risk for incorrect medical decisions when using a defective transducer indicate an urgent need for increased testing of the transducers in clinical departments,” the authors concluded.
The second study tried to find out how fast the machines break. The researchers examined about 300 machines all of which were working correctly (or had been fixed to work correctly) a year earlier. The retest showed that about 30% were defective. Apparently they break easily — and nobody notices.
These are lower bounds on the defect rate of the whole machine because only one part of it was tested. Because one part of the machine breaks so easily without detection, it makes me worry about the part of the machine that determines the strength of the ultrasound. How often is the actual strength much higher than the intended strength?
Caroline Rodgers has argued that ultrasound is a plausible cause of autism. A 2010 epidemiological study found no link between autism and ultrasound but these high defect rates call that study into question. In that study, almost all children (about 90%) had had experienced ultrasound before birth. This means that lack of correlation with autism means lack of correlation of number of scans (1 or more) with autism. This leaves open the possibility that it only takes one scan from a defective machine to produce autism. If that were true, and each child scanned repeatedly by the same machine (e.g., if a child has three scans, all are from the same machine), there would be no correlation between autism rate and number of scans.
These results scare me. It isn’t just ultrasound and autism. I blogged a few weeks ago about gross dosage errors in CT scans, some patients getting 10 times the intended dose of radiation. Here is another example where the operators of dangerous medical devices had no clue about appropriate testing and maintenance.
Thanks to Emily Williams who sent me a paper that mentioned these studies.
More about the ultrasound/autism link. Association between prenatal ultrasound and lefthandedness, which implies that ultrasound affects fetal brain development.