Monday, August 10, 2009

Response to NEJM Articles

Last week, there were two studies published in the New England Journal of Medicine that posited that vertebroplasty is no more effective than a 'sham' procedure. One study was from the Mayo Clinic (Kallmes et al) and the other was from Australia. I'll focus on the former

I recommend the following analysis debunking the paper. My comments follow.

Unfortunately, both studies have major flaws and it's pretty sad that they made it into such a prestigious journal. Unfortunately, that de rigeur for major medical journals--many of them end up publishing studies that have big flaws. It's also common for Medicare and private insurers to use studies like this to make determinations for coverage. Translation: the authors did their job poorly, yet patients may end up suffering as a result.

First, the suggestion that vertebroplasty doesn't work is simple nonsense. However, let me define this in terms that the authors didn't. Vertebroplasty is highly effective at fixing acute pain from vertebral compression fractures (VCF). It doesn't treat any other kind of back pain. Second, patients who get VCFs tend to get additional fractures.

Major problems:
1. Not enough patients. Not only did the studies fail to their target number of patients, too many patients declined to enroll in the study. Translation: the patients who were more likely to benefit from vertebroplasty went on to have vertebroplasty anyway because they didn't want the risk of having to continue living with the pain.
2. Poor patient selection/Wrong patients. The patients that were actually in the study had relatively low pain ratings. The typical patient I see rates their pain from 8-10 on a scale from 1-10.
3. Flawed sham procedure. What the authors describe as a sham procedure actually amounts to a facet block. This is relatively effective procedure for one of the most common causes of midline axial back pain. Take this and the relatively low pain ratings, it is likely that many of the patients had facet mediated pain and the sham procedure was actually a treatment for that.
4. Crossover. Patients were allowed to crossover after one month. Of the control/sham patients, almost half of them chose to have a vertebroplasty. Yet 88% of the vertebroplasty group patients chose not to crossover. This is consistent with previous studies showing efficacy of vertebroplasty around 85-95%.
5. Statistical significance. The authors state that there was no significant difference in pain levels after one month, but that the vertebroplasty group tended to have better pain relief at 3 months compared to the control/sham group. That is congruent with a positive result from a facet injection wearing off.

1. Patients who were more likely to benefit from vertebroplasty went on to have vertebroplasty anyway and weren't included in the study.
2. This was not a properly controlled-study; the sham procedure was essentially a facet block.
3. Almost half of the patients in the 'control' group eventually elected to have vertebroplasty.
4. 88% patients who had vertebroplasty were happy with the results.

Lastly, for those who have had patients who have undergone vertebroplasty or kyphoplasty, you know how effective the procedure is. I find it almost farcical that this paper was published in NEJM. However, this shouldn't surprise any of us as we are heading down a road toward increasing government interference in the doctor-patient relationship.

No comments:

The Blogs I Read

All posts are copyright Musculoskeletal Imaging of Tulsa.