Tuesday, April 26, 2011

More rhythm

In an earlier post, I described the importance of circadian rhythm in organizing basic physiological processes.  The significance of this (and shorter and longer period rhythms) rhythm is usually inversely proportional to its discussion, especially in medicine.  But occasionally there is a 'flyer,'  some paper that just shoots of of the rough and lands on the map.  

"Shift Work in Nurses: Contribution of Phenotypes and Genotypes to Adaptation"

Interesting in that it points out a number of the obvious issues with changing patterns of shift work, but also correlations with particular genetic features, which is pretty novel.  Good read, just for all the background about the health risks with shift work.  

And for a bit more entertainment, a very nice piece by one of the deans of chronobiology (no, I did not just make that up!), commenting on sleep and the training of residents., here.


The utility of this kind of training reduced the residents to such sleep deprivation that patients simply become obstacles to sleep.  Easy to grind the empathy out of people.  Not too sure what to expect with the new training laws.  Maybe I'll find a younger doctor and see.

Friday, April 8, 2011

Lobbyist for the Dead

Medicine, if anything, is a daunting practice.  And the US version is particularly burdensome, as the myth of heroic medicine is reiterated everywhere and by virtually everyone.  A local radio station (otherwise best know for repulsive talk and hiring ex-politicians after they make parole (really)) is running a 'fundraiser' for the local children's hospital

They are running promotional stories from parents with gut wrenching tales of sick children saved by the hospital staff and endearing tales of how individuals at the hospital, from surgeons to clerks, nurses to administrators, provided for their every need.  Of course the heroic stories are inspiring and wonderful for the folks involved and engender a belief that the clinical teams that can perform these heroics are due our gratitude and support.  

Great stories, but really not informative for consumers, as the best of medicine are the events that are way outside of the norm.  The way it always works is that half the folks get above average care and half the folks get below average care.  No way around that.  The challenge for most folks, hoe do you tell the difference?

One of the ways is to look at the number of medical errors that are reported.  That reporting, however is not just shoddy, it is contrary to a culture of medicine that excels at generating heroic stories (one reason I cannot watch lots of evening TV is that I cannot suspend my disbelief long enough to entertain doctors or lawyers as heroes).  In a landmark review, the Institute of Medicine generated this report on medical error in the US.  It was, of course, attacked by institutional and academic medicine, suggesting that about 350 people a day are killed (yup, killed) by medical treatment (no, and they were not going to die anyway) and many more injured.  Later estimates, based on better data from Australia and New Zealand suggested that a comparable rate of error for the US would be 2.5X greater.  The 'cure' for this rampant, hidden disease that was likely killing over 500 patients a day and injuring many more was, get this, to be honest about mistakes!  I know, I know, crazy talk.  But that was the idea and the model was what is done in the airline industry, where reporting mistakes is not met with punitive reactions nor lawsuits.  That method allows the airline industry and regulators to cooperate on understanding potential system-level drivers of error, as well as individual events. The goal was to reduce the level or error 50% in 5 years.  As lots of the errors hinged on simple mistakes aggregated at the same time and the same place, just working on the minor mistakes would reduce the catastrophic mistakes (this is a function of non-linearities in iterative systems and conforms to a power law.  More later).  For lots of conflicting and hidden reasons, this never got (nor gets) done.  But some diligent folks keep at it, at least telling the story.  A recent article in Health Affairs (unfortunately, not about hospital-based TV shows) that really looked at individual patients found that medical error rates are astronomically higher than folks recognized.  So even the estimates in that issue of HA dedicated to quality, are likely to underestimate the problem by 90%.  Updating their figures, medical errors cost the medical system about $170 Billion annually and "social costs" are in the trillions.  Yikes!
Children's Cemetery, Colma, IT (S. Lenz)

But why is this news?  Remember that heroic culture of medicine?  The very language used tells us how to understand the problem.  When young cancer patients at the local children's hospital is treated and the disease does not respond to the therapy, they are said to have 'failed chemotherapy.'    In a place where failure is not a option, those who fail are all sent to the same place, where their own tales of heroism lie dormant, slowly forgotten.  And the tales of perpetual heroics continue to resound, from the media, from our friends and families, from our own medical caregivers.  Just seems to me that the dead need a better lobby.  Then maybe we'd figure out how to keep our medical system from killing so many of us. 

Thursday, April 7, 2011

Lies, Damn lies, Medical reporting....

As I was innocently driving down the FL Turnpike, for the Cleveland Clinic (restarting a cardiovascular clinical trial), I was assaulted by a story on NPR about how a progesterone gel was used to reduce premature births in women 'diagnosed' with a 'short cervix.'  First, I think some of the girls on my basketball team should be diagnosed with 'short stature' and maybe we can get some kind of intervention.  The conversion of physical characteristics into pathology is, itself, a pathology.  But that isn't the part that got me.

In the course of the report, they stated that women with short cervix had rates of premature delivery of up to 50% and those that used the progesterone gel had a 45% reduction in premature births. Sounds great.  So the story goes on to 'highlight' one woman who entered the trial and was in the group that had the active cream and delivered a full-term baby.  The quote from NPR was the therapy 'worked.'  Huh?  How would they know she wasn't going to go to full-term without the drug?  Huh?  Science reporting is the only thing lazier than science itself.  I mean, NPR?  What are they trying to prove, that they are just as bad as commercial outfits? 

Well, I had to look at the real numbers, to see what the likelihood of pre-term delivery in the study was and it was 23% of 229 women (in the control arm), meaning around 52 births.  Also means the majority of women with the diagnosis of 'short cervix' don't have premature babies.  And in the treatment arm (same number of women, I think), about 32 women have pre-term deliveries, despite the drug.  So that means the drug 'prevented' about 20 pre-term births out of 229 total births.  I mean not bad, but hardly earth shattering, especially for the horribly misleading commentary about the woman who they followed.

Good news for the pharmaceutical company is that soon all the women in the US with 'short cervix' will get this drug, even though the vast majority will not benefit and there will still be lots of pre-term births.  Oh, and we will all pay the millions necessary to implement the therapy.  But that's the way we apparently like it.