A spine procedure for fractures was no better than fake

Researchers in Australia put it to a placebo-controlled test. Seventy-eight people with painful osteoporotic spine fractures were randomly assigned to receive real vertebroplasty or a sham procedure, where the needle was placed against the bone but no cement was injected.

Both groups reported similar drops in pain at every check from one week to six months. At three months, the vertebroplasty group's pain score fell by about 2.6 points on a 10-point scale; the sham group's fell by about 1.9, a difference too small to call real.

The finding stunned a field that had been performing the procedure for over a decade on the strength of uncontrolled studies. A companion trial in the same issue of the journal reached the same conclusion. Vertebroplasty did not disappear, since some patients and doctors still reported feeling helped, but insurers and guideline panels began asking for evidence beyond what people believed had worked.

A ritual with a needle, not a cure.