Wednesday, December 31, 2008
My questions so far:
Are you committed to understanding and facing that which is true, no matter how unpleasant? Will you measure your confidence in the truth of various claims rationally --that is, using tests of corroboration, falsification, logic, and parsimony?
Will you stop the anti-science lie of 'alternative medicine'? Any treatment proven effective with appropriate controls is simply 'medicine'. When the public can't appreciate basic rules of evidence, democracy dies. So help America to wise up!
Will you prove America's commitment to the rule of law? Will you allow the prisoners at Guantanamo the basic legal due process afforded to anyone accused of a crime in a civilized society? Their testimony will hurt us, but we will survive.
Snark (I'm not brave enough to actually type this in):
Will someone tally up the unfounded assertions and batshit notions nested among these questions at change.org? The list may help us understand why things are as they are.
Tuesday, December 30, 2008
BTW, anytime you'd like to hear me rant, just tell me that EMRs will save doctors bunches of time while also saving the US taxpayer billions. Ha! Lies! Lies! Lies!
Don't get me wrong. I'm not anti-EMR, merely anti-EMR hype. The hype hurts the EMR cause because it provokes disappointment.
The rant below concerns the data tracking requirements set by insurers, JCAHO, CARF, state licensing agencies, corporate boards of directors, etc. These people are too easily impressed by the appearance of usefulness.
When I imagine the brilliant minds behind Enron, the subprime mortgage crisis, the Wall Street bailout, the Madoff scandal, the invasion of Iraq, and so on, I see the smiling faces of helpful people not unlike the healthcare third party players I have come to know. The nice people in my mind's eye await the reassuring sound of a number which is to be written in a box on a form that will ultimately be incorporated into someone's quarterly report and/or Powerpoint slide.
Example: New Jersey wants nurses to track psychotropic side effects. Sounds good, right? Heh. Don't get your hopes up.
Here's what New Jersey requires: On the back of the monthly MAR for each day that meds are given, the nurse must record "NSE" meaning "no side effects." Otherwise she must list any side effects reported.
Stupid, stupid, stupid.
Firstly, among my patients, self-report is unreliable. Further, most side effects are not evident at the med pass. They're usually pattern alterations that manifest over time --e.g., irregular menses, constipation, sleep-wake cycle disruptions, appetite changes. Imagine how silly you'd feel asking a patient every morning, "any menstrual irregularity today?"
Better to record a few daily indicators such as:
- menses? yes/no
- poop? yes/no
- wet bed? yes/no
Plot these data on a monthly chart and patterns become obvious and can be checked against dates when meds were changed. As a bonus, this system controls for confirmation bias and selective recall.
Logic is on my side. I am the doctor. But can I win an argument with a state licensing agency? No, I cannot.
Although the facility may make a half-hearted attempt to track things my way, the effort will fizzle. No organization can afford two systems that serve the same purpose.
Moral of my story: tracking stupid stuff screws up the tracking of useful stuff. Beware!
Sunday, December 28, 2008
Tuesday, December 23, 2008
It may represent a sampling problem. There may be an outlier, a single individual who is very aggressive pushing up the numbers. Or there may be a disproportionate number of 45 year-olds in the group.
But let's say we rule out a sampling problem. Then we might have fun speculating about why humans seem to become more aggressive around age 45. Midlife crisis? Brain changes? Hormones? Empty nest syndrome? Seven year itch? When inventing explanations, one's powers of imagination are the only limit.
Hence the reason we don't take post-hoc explanations for apparent correlations very seriously. Such explanations are like a rustling in the leaves - might be the wind, might be an animal, might be a friend come to visit. We can't know without further investigation.
The Texas sharpshooter fires his rifle at the side of a barn repeatedly and randomly. He then searches for a cluster of shots close together and he draws a bullseye around them. We control for this kind of error by insisting that the researcher predict what he expects to find before seeing the data.
The alt med guys are Texas sharpshooters. They order hundreds of lab tests without first predicting what they expect to see or what those results might mean. Order enough tests and it's a statistical certainty that something will come back abnormal.
An honest researcher looking at the graph with the bump in aggressions around age 45 would say, "Could be this; could be that. Can't be sure without further study."
But here's the typical alt med doc:
Our group has found that toxin accumulation in the colon becomes critical by the age of 40. Take a look at this graph BigPharma doesn't want you to see...
Symptoms of CTS or Colonic Toxic Syndrome include:
- low self-esteem
- bad breath
- stinky feet
- a tired feeling
If you haven't had regular colonic irrigations and think you might have CTS ...blah blah blah free consultation bogus testing and so on.
Monday, December 22, 2008
A residential program I visit has a system for monitoring severe behavior problems, such as aggression toward other persons, running away, or self-injury: Staff fill out a form describing the behavior then a data entry person puts the info into a database.
The graph above is simply a count of incidents involving aggression to persons. The x-axis is the age of the client at the time the incident happened. These raw data are from 2008.
The above graph confirms my clinical impression: with respect to aggression, the teens and young adults are champs (I should mention that about one-third of the clients at this facility are under and two-thirds are over 18 years of age).
If I were a clever doctor, I'd try to take new patients of about 21 years. I could straighten these kids out with friggin' Fred Flintstone's Chewable Multivitamins. People would think I was a genius.
See why we need control groups for clinical research involving kids?
Sunday, December 21, 2008
BUSH: One of the major theaters against al Qaeda turns out to have been Iraq. This is where al Qaeda said they were going to take their stand. This is where al Qaeda was hoping to take-
RADDATZ: But not until after the U.S. invaded.
BUSH: Yeah, that's right. So what? The point is that al Qaeda said they're going to take a stand. Well, first of all in the post-9/11 environment Saddam Hussein posed a threat. And then upon removal, al Qaeda decides to take a stand.
Mr. President, at this point most people understand the sequence of events: You used 9/11 as a justification for going after Saddam Hussein. Americans, largely ignorant of Middle East politics and religious divisions, accepted your claim that al Qaeda and Saddam were somehow linked. Although it's possible you were sincere when you asserted the connection, you and your administration remain at fault for the lack of due dilligence in examining the evidence available.
You have not taken responsibility for your failure to rationally examine the evidence for a link between Saddam and al Qaeda prior to the invasion of Iraq.
You have not taken responsibility for your failure to rationally examine the evidence regarding WMDs in Iraq prior to 9/11.
May those words haunt you.
Saturday, December 20, 2008
OMG, it's FAIL!
Firstly, Dr. Bock asserts a number of controversial claims as established facts - e.g., that there's an autism epidemic and that thimerosal in vaccines correlates with this epidemic. Objectivity would dictate that one explain for the reader the evidence on both sides of the argument, non? Dr. Bock doesn't bother.
Throughout the book he says, "in my experience" and "I believe" and "I recommend" without explaining the basis for his personal confidence.
When Dr. Bock provides something that might pass for objective data, I'm actually embarrassed for him. If he were a student in my statistics class, he'd fail. Example:
Chelation therapy has been successfully applied to thousands of children onOkay class, what further information might you want before drawing any conclusions from the survey data above? Yes, the file drawer effect! Good job! It's likely that many chelation patients didn't fill out the ARI evaluation, human laziness being a Universal Law n'all. Before interpreting the stated percentages, we'd need to know the total number of patients treated with chelation and how many didn't do the form.
the autism spectrum, under the supervision of DAN doctors and other integrative
physicians. The results have generally been gratifying and promising. The
positive response rate, as quantified in the ongoing series of Autism Research
Institute parent evaluations, is as follows:
Symptoms improved: 76 percent
No discernible effect: 22 percent
Symptoms worsened: 2 percent
These parent evaluations represent a 38-to-1 ratio of improvement of
symptoms compared to worsening symptoms... No other single element in the
biomedical treatment of autism has been rated this highly. (pg. 305)
Any other problems with Dr. Bock's conclusion that chelation works? Right again! We don't know what might have happened without treatment. There's no comparison group. Imagine if we had a no-treatment group and it looked like this:
Symptoms improved: 96%
No change: 2%
Symptoms worsened: 2%
Compared to these numbers, chelation would seem harmful rather than helpful, eh?
Other obvious problems with Dr. Bock's data: We don't know what "improved" means. We don't know the degree of intra- and inter-rater reliability for "improved."
Now, anecdotal data has its place. There's nothing wrong with doctors sharing case reports with colleagues. By comparing notes, we figure out where to put new research efforts.
But Dr. Bock's been at this chelation stuff for many, many years. He's had plenty of time to put his ideas to the test in a real study with adequate controls. For him to pimp the "chelation works" notion without evidence at this late date is, frankly, shameful.
My conversation with the parent who urged me to read this book was awkward. "These are DAN (Defeat Autism Now!) doctors, the leading autism experts in the country! How can you say they haven't got evidence? Haven't you heard of the Autism Research Institute?"
Dr. Bock mentions setting up a credentialing program, something like an autism treatment sub-specialty. I'm imagining it will be like the ABA (Applied Behavioral Analysis) Board Certification. For anyone at a doctoral level, these certifications are a joke. Yet parents, teachers, and funders have bought the marketing hook, line, and sinker. "Doctor, are you ABA Certified?" If you don't buy the certificate *cough*, you're not an expert.
My fellow physicians, we are in deep shit. The rules of evidence have been broken and the thieves are minding the store.
I'm calling red alert. All hands on deck. Battle stations!
Get those Reiki wankers out of the feckin' hospital, to start.
A federal judge in Honolulu dismissed the lawsuit seeking to stop operation of the giant collider, not because the science fiction writer who filed the suit was a few neurons short of a full compliment, but because CERN is not in Honolulu.
Old news, but I just read it and it made me lol. "CERN is not in Honolulu."
Gotta love Bob Park, the physicist and author of What's New, a weekly report of Washington happenings. No wasted words and always the appropriate amount of snark.
He's got a new book out: Superstition: Belief in the Age of Science (Princeton, 2008). Anyone read it yet?
Sunday, December 14, 2008
Nice example of planing.
Did you know that boats have a top speed determined by their length at the water line? Here's the formula: hull speed in knots = (square root of the hull length in feet at waterline) * 1.34.
When I learned of this limit a few months ago, I was surprised. Imagine if cars were like this. Imagine shopping for a fast car: No need to bother about its width or height or weight. Check its length and there's your max speed.
But unlike cars, boats make waves. The wavelength of these waves imprisons the boat, from bow to stern. Well... so long as the boat is in the water. A boat at the surface of the water, a boat that is planing, can break this hull speed rule.
Like the 5O5, my V15 sits high on the water after a capsize. A few seconds of wind and wave action and it's upside down.
I agree that a capsize is "part of the fun" of sailing a little racing dinghy. But a turtled boat is no fun at all. When turtled in shallow waters, the tip of the mast can bang against the lake or ocean bottom. It can bury itself so hard that the boat won't move. Then you've no option but to call for a motorized rescue. Even so, the effort to pull the boat free may dismast the boat.
A float at the top of the mast can keep the boat from turtling for several minutes - long enough for the crew to get the boat upright. The downside: any weight at the top of the mast makes the boat less stable and may increase drag.
At this point, I'm not taking my boat anywhere without some sort of float on the mast. Four empty 2-liter soda bottles seem to do the trick, albeit without elegance. Three soda bottles --trust me on this-- do not suffice.
I'm shopping for a better answer. The Hobie Bob looks heavy. I'd like to see the Flying Dutchman float, which seems to integrate with the top of the mainsail.
Hobie Tigers surfing.
Before I bought my boat I would run YouTube searches for "sail fast" or similar. Lots of the videos I watched were blurry, thanks to spray on the camera lens. Many had wind artifacts or bad metal music dubbed in. But a few, like this one, were just ace.