There is an interesting article on issues with radiation therapy machines in the NY Times.
You may need to register to read it, but is is free (for now).
The story has extensive anecdotal evidence, but little sense of how extensive such problems are. Indeed one of the thrusts of the article, was that the current patchwork of regulations and non-regulations make it difficult to determine how big the problems are. The video that goes with the article implies the error rate is lower than other medical procedures. I didn't find this totally reassuring. There are also issues with some types of radiation therapy being regulated by the Food and Drug Administration and other types being regulated by the Nuclear Regulatory Commission.
I was disappointed, but not surprised by references to software problems with the therapy machines in some cases. The FDA passed extensive regulations on medical software after the Therac-25 accidents back in the eighties. I read the standards that the FDA published and they required extensive documentation of the design and testing process for software. As with ISO9000, there seems to be a part of the bureaucratic mentality that assumes that you will get a quality product if you document it enough. Indeed, their operational definition of quality seems to be how much paperwork is generated. I'm a little surprised that the NY Times didn't mention the Therac-25 for historical background. Many people were injured and at least two were killed by the Therac-25. There was a software bug, where the machine would sometimes dose people with a 100 times the radiation prescribed. Previous versions of the Therac had the same bug, but there was a hardware interlock that prevented the failure.
0 comments:
Post a Comment