Lying with (non)statistics: Phasers are set to “kill habit”

Public citizen has “filed a petition”: to end the marketing of a “low-power laser acupuncture” treatment for smoking cessation.

What struck me was not the treatment itself, but the way in which the so-called clinical trials they are running to “prove” their treatment’s efficacy. You can read about it at the link, but to me the following details are very fishy:

  • the company charged smokers to participate in trials (it’s usually the other way around — companies usually pay participants) — in fact, companies charged more to participate in their trials than GlaxoSmithKline does for a month’s supply of Zyban
  • the company did not collect any direct efficacy data, such as the time to smoking cessation (or whether a person stopped smoking at all) (in which case the trial is unethical)
  • the company is conducting trials, but is claiming their efficacy measure is “client referrals”
  • “When we do follow-up phone calls, people don’t call us back,” according to one product’s owner. You know what? If you are going to conduct a clinical trial, you have to control for things like this. Of course, there’s always going to be people who escape tracking, but in the clinical research industry we’ve become very good at designing studies so that people can be followed up. It’s a heavy cost, but well worth it and even necessary to conduct proper clinical research.

Since smoking cessation is a huge market, and effective non-pharmaceuticals is probably very desirable, I can only imagine that a patented medical device with lots of properly-run clinical trials to back it up can make a lot of money.

Another company believes it has enough data to submit to the FDA, but was cited a year and a half ago for failing to properly monitor its trials. And take it from this biostatistician, if a trial is not properly monitored (i.e. someone verifies that the data has been recorded correctly and that the correct data was recorded), it’s hard or next to impossible to make any accurate conclusions. Any results from that trial is weak. For their sakes, I hope that they have several other well-run studies that support their application.

Who knows, maybe the theory of acupuncture works with lasers, too. Maybe endorphins are released when you nail someone with a laser. It shouldn’t be too hard to find out these days. If you want to use the methods of science to back up a claim, use them properly, or else someone will call you out. Or even worse, someone will write you up for a skeptic’s circle post.

%d bloggers like this: