Anecdote driven activity

The AP published a story by Jennifer Kerr earlier this week on the activities of the Consumer Product Safety Commission (CPSC) relating to science kits for children. http://news.yahoo.com/s/ap/20100929/ap_on_bi_ge/us_children_s_products_safety
It seems that the CPSC is worried that items such as aper clips and rulers within these kits have not been properly tested for safety. The manufacturers of the kits believe that these common household items should not require specific testing since they don;t require this when sold separately. The commission voted 3-2 to decline a waiver for this application.

I was intrigued by this story, prompting me to explore the CPSC site with various CPSC publications. http://www.cpsc.gov/cpscpub/pubs/pub_idx.html as well as the CDC site for data on death from various causes http://www.cdc.gov/nchs/data/nvsr/nvsr58/nvsr58_19.pdf. My question is what type of data would prompt the CPSC to be mindful of placing tighter regulatory scrutiny upon paper clips and rules. The best answer i could come up with was this was not a data driven endeavor.

At the CPSC I found tables summarizing the number of deaths and accidents that could be attributed to accidents with toys.

I am not particularly happy about any deaths of children from accidental causes but I think we need to look at these numbers within the context of the overall denominator. There are approximately 20 million US citizens under the age of 15, meaning the risk of toy related death in 2008 for this group was literally one in a million. Death in children is more likely related to some sort of congenital condition (primarily cardiovascular), accidental deaths (primarily drowning, motor vehicle), and infections.  To invest such time and effort in the regulatory endeavors which the CPSC builds cannot be justified on any rational basis. It appears to be that their efforts are best justified by the desire to justify their own existence. To that end they find the public relations ploy of scaring parents about the danger of toys and play pays handsome dividends in terms of their own budgets. A few scary anecdotes go a long way to accomplish this goal, whether intentional or not. However, we end up in an absurd place, looking to do safety tests on common household items. I wonder if they will next request to do safety testing on the packing materials, including the air in the box.

This type of behavior is not limited to the CPSC and they should not be singled out for special treatment. Indeed, medicine uses the same approach to highlight "problems" which require scrutiny.  My favorite area in medicine which parallels the CPSC behavior is medication monitoring which is hopelessly driven by anecdote. When the famous TV drama ER was on the air, the producers used the principle that if anything happened in the ER anywhere in the country once, it would occur hourly on the TV drama. Shooting, hostage scenarios, plague, bomb threats, you name it, the implication was these should be our primary concerns.

Likewise, if one patient anywhere has been reported to have a particular bad outcome on a given drug or after a given intervention, whether mechanistically linked or not, we will be called upon to restructure how we manage future scenarios based upon rare anecdote. We are concerned with the numerator and not the denominator. To avoid the one in a million unpredictable bad outcome we alter daily practice in such a way that we are guaranteed to hamstring our practices, mis-allocate scare resources and compromise patient care.  We end up doing the equivalent of the CPSC attempts to regulate paper clips and rulers.

Many within the medical profession have lamented the role of the legal profession as drivers of this. I must agree that the legal profession has not been shy about exploiting the environment of fear. However, they could not have made such headway if the health care world did not leave them with such opportunity in terms of operating in such a data poor realm. Beyond that, we have not cultivated a world within health care which is driven by data. We have resisted creating meaningful and measurable objective milestones to be held accountable to not realizing that these can be our best defense against capricious legal intrusions. It has also created openings for other regulatory entities to come in and fill the vacuum, making us measure the measurable and not necessarily what is important.

Those who take care of patients need to step up and realize that part of our job is to address how we can put systems in place which permits us to make important measurements which allow for important mid-course corrections. It is an important part of patient care. We can no longer use a reactionary system driven by rare anecdote.