Dr. Andy

Reflections on medicine and biology among other things

Saturday, January 28, 2006

US births

I've been asking fellowship candidates the following question:

"There are approximately 300 million people in the United States. About how many babies are born each year?"

Answers so far have ranged from 20,000 to 100 million!

Absolutely no one knows the answer, which is available in this article or as the first commnent to this post.

Try to make an educated guess, then look to see the actual number.

Hint: it is somewhere between 20,000 and 100 million.

Thursday, January 26, 2006

The case for intelligent design

nicely summarized by a commentator on an earlier post of mine.
I think that what you have to say is just a bunch of CRAP!!! To you there may not be any info to prove the "theory" of intelligent design, but think again, its called a BIBLE!!!!! na durr!! look in it once in a while!! and FYI, if there gonna teach about the gay "theory" (which by the way is a load of crap also) of evolution in schools, then they should at least teach the truth, ABOUT GOD (intelligent design)!!!!!!!!

Oh yeah, this is the anoynomous person again (the first 1), and i have a couple more things to say....ur a freak!!!!! You may think that by saying this you aren't affending anyone, but think again.........YOU'RE OFFENDING ALL CHRISTIANS, which i happen to be!!!! so i suggest that you go to church and read the bible, although I don't think that will help your screwed up head!!!!! All you have said is lies!!!!!! EVOLUTION is NOT a theory, its a lie!!!!!!!
Surprisingly, this eloquent proponent of ID doesn't wish to be identified.

And to my anonymous commenter, I'd point out that I attend my church weekly and don't see any conflict between that and accepting evolution.

Wednesday, January 25, 2006

How to choose a specialty

Good advice from Fat Doctor (hey that's her name for the blog, not mine):
You have to like going to work in the morning, enjoy the intellectual subject matter, like the type of people you work with, tolerate the lifestyle and be qualified to get into that residency. Otherwise it just won't work for you
Pretty good advice for life in general. I'd only add that if you are thinking of one of the more rigorous specialties, make sure about the lifestyle. I've met a lot of bitter surgeons who either left their program or would have done something different if they had it to do over again. It's not like they don't warn you it will be tough.

A good article, by one of the best medical writers around, Peri Klass (no blog, but I guess when you have gigs with the New England Journal and New York Times you don't need one).

She talks about how hard it can be to separate the truly sick kid with meningitis or leukemia out from the sea of viral infections and worried moms.

I posted a similar experience here. I think one of the most important things for a pediatrician is being able to differentiate sick from non-sick kids. And by "sick" I mean really sick, not the flu. I have one more case I'll describe in the future, but when caring for the kid with a viral illness that was really leukemia, I wasn't smart enough to know right away that it was leukemia. I was, however, smart enough to know that something bad might be going on. In the midst of a busy night seeing "non-urgent" kids in the emergency room that was enough.

Tuesday, January 24, 2006

Grand Rounds 2:18

is here.

Saturday, January 21, 2006

The Will Rodgers effect

My last post on ALL (Acute Lymphoblastic Leukemia) reminded me of an interesting statistical phenomenon referred to as the Will Rodgers effect: if you have divided a sample into two groups, one at higher probability of succesful outcome, changing the criteria in order to make more individuals members of the low success group will improve the outcomes of both groups. Obviously you could extend this to more than 2 groups.

As an example in leukemia, some patients are characterized as "high-risk" based on characteristics that make them less likely to be cured by standard therapy (age, T cell, high initial white blood cell count, etc). Changing the criteria so that more patients are characterized as high risk will improve survival in both groups, even as overall survival stays constant.


This chart shows the incredbile progress that has been made against childhood leukemia, specifically acute lymphoblastic leukemia. So if your child is diagnosed with ALL today, she probably has a greater than 90% chance of cure. 40 years ago the chance was more like 10%.

Chart is from last week's NEJM. ALL is the success story in cancer. Progress in other cancers has obviously been less impressive. Still, the success rate in childhood ALL is so great that much current research focuses on identifying "low risk" patients who will be cured with less-intensive regimens that have fewer adverse effects.

Friday, January 20, 2006

Paying for healthcare - III

This will be my final post in the mini-series on paying for healthcare. I'd like to point out I don't really consider myself an expert on the economics of medicine, more of an interested observer, unlike science and medicine which I do think of myself as an expert on.

Hospital as airlines.

Another article in the January Health Affairs considers whether hospitals in the United States are analagous to airlines before deregulation. Incredibly, the authors conclude that the answer may be yes, but this is a bad thing:
But competition can have a dark side. U.S. hospitals can treat Medicare and Medicaid patients at less than cost, care for the uninsured, and provide other money-losing services because they can cross-subsidize. By 2025 the need for general hospitals to cross-subsidize will greatly increase, but their ability to do so will be diminished. U.S. hospitals could begin to resemble U.S. airlines: severely cutting costs, eliminating services, and suffering financial instability.
I break down their argument into two parts: the need for cross-subsidies and the damage to the hospitals.

The cross-subsidy argument is basically that the health care market is so complex and inefficient that we shouldn't rationalize it. Sure right now government payors are probably charged less than fully-allocated cost (although probably more than marginal costs). But cutting costs and increasing efficiency could change that.

Imagine that in 1977 (before air travel was deregulated) the CAB required government employees on official business to be given a 30% discount. Would the government currently be paying more for airfare if the old system had stayed in place or paying full fare under deregulation. It's not even close. Airfares have fallen close to 60% since deregulation (figure from this paper, in PDF form.

And service, at least measured by number of flights and cities served has increased as well (NYTimes article from last year via Daniel Drezner):
At the same time, airlines have vastly expanded their networks, bringing air travel - a relatively infrequent experience [several decades ago] - to people all over the country. For example, American, the biggest airline, flew to just 50 cities in 1975; it now serves more than three times that number. Southwest, which started in 1971 with a single route in Texas, now flies to 61 cities, not counting those it serves through a code-sharing arrangement with ATA.

So the cross-subsidy argument is easily dismissed. Increased transparency and competition in healthcare will almost certainly decrease costs for everyone, even assuming the government will pay relatively more compared to commerical payors.

The second argument is that the change may be bad for hospitals. That may be, but I don't personally worry too much about that. If new, better, more efficient hospitals replace legacy ones is that really bad. Other than the luggage fiasco, flying
Southwest is fine by me. As noted above, the analogy to airlines decreasing their service is flawed. Air service has dramatically increased.

To give an example of the waste in our current health care, what if instead of charging $70 or $100 a day for meals patients had to bring their own? Already the food is so bad, many patients order out and surely restaurants do a better job of providing quality food at a good price. This is analagous to airlines which used to give you bad food as part of your inflated ticket price. I'd much prefer paying $100 less per ticket and eating before getting on the plane than the old way.

In summary, I think real competition among hospitals and in healthcare in general will provide better, cheaper care across the board. Will it be painful for many existing hospitals and health-care workers. Sure, as the article notes senior pilots make up to $250,000, new hires will top out at $100K. That is life in a capitalist society and we'd better get ready for it.

Thursday, January 19, 2006

Paying for healthcare - II

The January issue of Health Affairs has a number of articles about pricing of healthcare. I don't normally read the journal, but the articles are fairly brief, understandable and interesting.

My favorite is this one, by note health economist Uwe Reinhardt about how hospitals price services. The answer is that pricing is a joke. There is a giant price list called a "charge master" that covers every concievable service. It is updated each year by a process that can generously be described as capricious. Don't believe me. California is now requiring hospitals to publish their charge masters. Here is how the "list" price of a chest x-ray compares between different hospitals:

I don't think there are any gas stations in California that charge 10 times more than others (and a chest x-ray is the same from one institution to another; it's not like you are paying for the best surgeon).

Of course almost nobody pays the list price (the sad exception being the uninsured) as the government pays a flat fee by diagnosis for most patients and private insurers have negotiated large discounts. Another article in the series points out, the ration of revenue collected to charges (full price) is 1:2.6!

Read the whole thing as it is interesting and Reinhardt discusses more sensible pricing schemes, including requiring that hospitals post prices and charge every patient and payor the list price (like, say, McDonalds does). This would certainly encourage real competition on price and quality.

Paying for healthcare - I

Why is healthcare so expensive and inefficient? Because no one knows or cares how much it costs. Seriously.

This struck home with me recently when I picked up a prescription for my daughter. She has eczema and has been on a low-potency topical steroid. A couple times I've picked up the medicine and dutifully paid a $10 copayment.

The eczema was a bit worse, so wrote a script for a bit more potent steroid to try to get it under control. I went to pick it up and the copay was only $2. I asked why and discovered it was because the medicine is only $10.75 to start so the insurance company picks up most of it.

Now I prescribe these kinds of medicines for kids with eczema everyday. And I had no idea that one (triamcinalone) was cheaper than the other (desonide). No patient has ever asked me if the medicine I suggest is the cheapest available, they just assume I'll pick the best one. And all the steroid cremes work about the same (given that there are various potencies) so there isn't much difference.

Think about it. I know roughly how much I 'll spend on any given restaurant. McDonalds with the kids about $14. Wendy's a buck or two cheaper. Pizza for the whole family is about $20. Chinese $40-50. Sushi runs $60-80, etc. But I have no idea how much these creams and ointments cost, either as a consumer or prescriber.

To get health care costs under control, consumers will have to know and care what things like steroid creams cost.

Saturday, January 14, 2006

Avian flu- is the pandemic closer?

This would seem to be bad news
Researchers have sequenced the bird flu viruses that killed two people in Turkey in early January, and say that one of them contains a worrying mutation.

This genetic tweak can make the H5N1 virus more adapted to humans than to birds, and more adapted to the nose and throat than to the lungs. This latter effect could help to increase the chances of bird flu being transmitted between people, researchers say
Of course additional mutations will be required before efficient human to human spread can take place.

I would not be particularly worried that the virus is in Turkey or that a relatively large number of people seem to be infected. Many of the identified cases in Turkey did not have severe disease. As a more developed country, these less severe cases were picked up, while in Vietnam, say, less-severe cases probably never came to medical attention.

NSAIDs and COX2 inhibitors

A great, albeit technical, article on the relative risks and benefits of NSAIDs and Cox 2 inhibitors in this months Journal of Clinical Investigation. NSAIDS, such as ibuprofen, naproxen, aspirin inhibit both cyclooxygenase (COX) 1 and 2, although to different extents. Low-dose aspirin preferentially blocks COX2 accounting for much of its cardioprotective effects. COX1 is widely expressed while COX2 is involved primarily in inflammation. COX2 inhibitors were designed to block inflammation (and pain) in those who couldn't tolerate traditional NSAIDS

Some interesting highlights

1. Several "traditional" NSAIDS such as diclofenac and meloxicam are fairly COX2 specific, with diclofenac (marketed as Voltaren) comprable to celecoxib (Celebrex) in both the test tube and actual side effect profile of patients taking it.

2. Naproxen may have some protective effect on cardiovascular disease because of its long half-life and relative specificity for COX1

3. Long term use is probably very important in raising the risk of both traditional NSAIDs and COX2 inhibitors. So use for a few days with an acute injury is probably insignficant for an indivdual

4. More than causing heart disease themselves, COX2 inhibitors block the ability to respond to a stimuli that causes thrombosis (clotting of blood which is the acute trigger of most heart attacks and strokes). Mice that have a defect that mimics the effect of COX2 inhibition don't get spontaneous clots, but are at increased risk when manipulated to cause clotting. Similarly, the greatest risks of COX2 inhibitors in trials were seen in patients with underlying causes of clotting (those undergoing coronary artery bypass or with rheumatoid arthirits)

The article also gives a good history of how the problems with COX2 inhibitors were worked out.

Sunday, January 08, 2006

Economist Christmas issue

The Economist is the only non-medical periodical besides Ultrarunning I subscribe too. Each year I look forward to the year-ending double Christmas issue (which seems no longer to be the Christmas issue, alas), which in addition to the usual news contains a variety of longer feature articles, which I have time to enjoy over the holidays. These are mostly available for free here.

One of the most interesting this year is comparing the poor in the US with the rich in poor countries, specifically, an Appalachin man drawing disability and a doctor in the Democratic Republic of Congo, who make about the same amount of money. One might think it would go a lot farther in Africa, but not really:
A typical poor household in America has two televisions, cable or satellite reception and a VCR or a DVD player.
And living in the US is a lot safer and more conveinent with than in any African countries.

Read the whole thing as they say, but it is facts like this, and that this guy on "disability" hasn't worked in 25 years, despite still not being 65, that makes me think the welfare system in this country is a bit too generous.

Thursday, January 05, 2006

Avian flu marches on

A second Turkish teenager died of bird flu on Thursday as a virus that has killed 74 people in east Asia claimed its first lives far to the west, on the fringes of Europe and the Middle East.

In a sign the disease may have infected people over a wide area of eastern Turkey, six people from a different province were taken to hospital with suspected bird flu. In all, doctors said 18 patients were under scrutiny and two of them very sick.

The most important factor preventing a pandemic is infectivity not geography, so this isn't that important. For a pandemic to occur, the H5N1 virus will have to develop the ability to be efficiently transmitted from human to human. As the Turkish teens lived in close contact with birds, there is no evidence that this has happenned. What this case dose demonstrate is how widespread H5N1 is in wild and domestic birds.

Tuesday, January 03, 2006

Patients as customers

No one would argue that hospitals and doctors should not be responsive to patients' concerns. But a business model that says the consumer is always right can clash with what is best for a patient, especially in the area of mental health.

A good commentary about how always trying to please patients can go against their best interests.

During fellowship I saw a patient who was missing a ton of school for some vague nasal/sinus symptoms. I told him clearly he had to go to school every day and we'd work on helping him feel better.

Near the end of the school year I got a call from a very insistent dad that I sign a form justifying more than 30 days of missed school. I refused, as I'd specifically told him that he should not miss school. The dad went to patient relations who got the attending I saw him with (I was still a fellow) to sign. The attending said straight out it was the wrong thing to do, but it wasn't worth fighting