When showing it doesn’t work doesn’t work

I went to Melbourne on the weekend.  To break the time-honoured Moseley tradition of not working on weekends requires some convincing and I was convinced by the nature of the crowd – truly interdisciplinary – and the proximity of venue – a mere 1000 km away.  I went to two talks, so I could get back home in time for Chapter 6 of Rowan of Rin. Of the two, mine was the second best talk I went to. The best was a talk by Prof Rachelle Buchbinder, who did a BiM post not that long ago on tennis elbow.  She walked us through an RCT she did on vertebroplasty for painful osteoporosis-related fractures. Vertebroplasty involves squirting bone-cement into the vertebral body that is fractured.  It is, incidentally, very big business. Rachelle published this RCT in an obscure journal called the New England Journal of Medicine (I haven’t even heard of the old one!). In the same issue was another trial on the same thing.  Both trials clearly show no advantage of vertebroplasty over sham.

Rachelle’s trial was top-shelf. If you want to know how to do an RCT PROPERLY, have a look at this one – the sham intervention was everything the treatment was EXCEPT injecting the cement – local anaesthetic, tapping on the vertebrae and wafting cement-fumes (patients had previously remarked that they could remember the sensation of tapping on their spine and the smell of the cement).  Only the person who did the injecting knew which was which.

These trials were received well in some quarters and not at all well in others. The negative responses were predictable – “you have to select your patients carefully”, “you need to do several vertebral levels”, “you used the wrong cement”, “the operator wasn’t sufficiently well trained”.   Rachelle took us through the criticisms one by one and addressed each one clearly and sensibly. The most sensible conclusion from these two RCT’s that show vertebroplasty is no better than sham is that vertebroplasty is no better than sham.

Four things were particularly notable on Saturday:

(i) Prof Buchbinder is clearly a Truth-seeker: she appears to have no agenda aside from determining what actually works.  Her work is so rigorous, so careful and so persistent that one feels a bit amateur being in the same line-up.  We would all do well to make a habit of asking ‘What would Rachelle think of this?’;

(ii) Some of the room were clearly not having a bar of it and quickly began offering the solution to vertebroplasty that apparently eluded the trials – “you have to select your patients carefully”, “you need to do several vertebral levels”, “you……” (it seems they missed the talk);

(iii) Some of the room were observing it all and with open minds voiced that even here, amongst the keener, more learned, more expert clinicians in the country, showing that it doesn’t work, doesn’t work;

(iv) a grey day in Melbourne feels much like England – I am not sure if I loved it or felt that claustrophobic sky thing coming back.

References

Buchbinder, R., Osborne, R., Ebeling, P., Wark, J., Mitchell, P., Wriedt, C., Graves, S., Staples, M., & Murphy, B. (2009). A Randomized Trial of Vertebroplasty for Painful Osteoporotic Vertebral Fractures New England Journal of Medicine, 361 (6), 557-568 DOI: 10.1056/NEJMoa0900429