Self-Reported Syndromic Surveillance

I've put off linking to Whoissick.org because it got pickup on Boingboing and elsewhere. But I thought again after I realized that no one I'd read had applied a little Epidemiology 101 to the site. So it's a cool idea: using a Google Maps mashup, let the hoi polloi post when they're feeling ill, checking off a few symptoms and other info, and plot it by location. Get enough contributors, and you have an interesting surveillance map of possible infectious outbreaks. Add a little more Web 2.0 magic - tagging, analytics - and you've got perfect blog fodder. The idea is basically a self-reported version of what's called 'syndromic surveillance,' which is basically the effort to systematically track disease outbreaks on a geospatial level, by pulling from a host of data inputs - hospital admissions, drug store purchases, school illnesses, and so forth. It's been used for infectious disease for years, but it's gotten attention more recently because of it's utility in possible terrorist or bioterrorist events. That, plus the fear of avian flu, make it quite a trendy concept.

So on the face of it, whoissick.org is especially enticing, because it combines that thinking with a dash of Web 2.0 special sauce. But I knew there was something a bit, well, imprecise about it. And that comes down to bias. Quick refresher: bias - the notion that systematic error can be built into a study design - is one of three fundamental mistakes in epidemiological studies, along with confounding (the unmeasured influence of an extraneous factor from the study) and random chance (the likelihood that the study result was due to chance rather than any true association). Basically, Whoissick.org succumbs to one of the most common forms of epidemiological bias - reporting bias. This bias holds that those who are aware of being ill are sometimes more (and sometimes less) likely to pipe up and participate in a study than those who aren't ill - therefore giving a disproportionate measure of the true level of illness. Unless you're measuring the true rate of illness against a true rate of non-illness in a population, you're going to have a biased result.

So Whoissick.org suffers from the same non-statistically valid flaw that those self-selected polls on CNN.com do - there's no way to determine whether the results are in any way an accurate reflection of the actual population in question.

There's also a host of other problems with the site - starting with its categories of illness (stomach ache and muscle ache are hardly medically exact categories). So while I'm inclined to give it props as an experiment and a fun thing to have around, I sure hope it doesn't get overblown.

DNA Day!

Evidently today - April 25 - is DNA Day. Uh, "a unique day when students, teachers and the public can learn more about genetics and genomics!" according to the National Human Genome Research Institute. I say that we can do that any day, but sure - happy DNA Day to you all.

This Week in the MMWR: Cesarean Nation

The US has hit a new high in cesarean births - more than 30% of all births are now done surgically. This news is especially timely to me - last week my wife gave birth to our first child, and after a lengthy but never alarming labor she gave birth to a fine boy early Tues morning. But a few hours later, word was that we were far closer to a cesarean (or rather, closer to the doctors recommending caesearean) than we'd ever imagined.

I find the geographical distribution of cesarean rates especially fascinating - highest in the South and lowest in the upper Midwest and West. Bearing in mind the story in today's NYTimes that infant mortality rates are actually increasing in the deep South, it raises some questions about the inequities and disparities in how we treat childbirth and postnatal care in this country.

This Week in the MMWR: Those Gol' Darned Nail Guns

First in a series: From this week's Morbidity and Mortality Weekly Report, the CDC's chronicle of reports from state health departments: "Nail-Gun Injuries Treated in Emergency Departments --- United States, 2001--2005". The, uh, upshot: Nail gun injuries among consumers have tripled since 1990, with nearly 15,000 misfires in 2005. Not surprisingly, 98% of those injured are men. Here are the gory details:

Certain puncture wounds resulted from a nail going through construction material into a person; in others, a nail was shot completely through a body part, or a person removed the nail before seeking treatment. Approximately 4% of nail-gun injuries among workers resulted in fractured bones.¶ Injuries to upper extremities, primarily hands and fingers, accounted for 75% of all consumer nail-gun injuries and 66% of all worker nail-gun injuries. Lower extremities also were injured frequently, accounting for 17% of consumer injuries and 24% of worker injuries. Examples of other nail-gun injuries among either workers or consumers included eye injuries from foreign bodies and corneal abrasions; dental injuries; musculoskeletal injuries such as sprains, strains, tendonitis, nerve damage from tool use, and finger dislocation from reaching and lifting a tool; lacerations; electrical burns; and noise-induced hearing difficulty.

I don't really have anything to add, other than: Ouch.

The Problem With False Positives

Following up my post last week about computerized breast cancer screening, I found this report today about the perils of false-positives intriguing. Women who received a positive mammogram for breast cancer - and then later were told the test was wrong, and they don't have cancer - suffered high levels of stress for long periods of time. This was a meta-analysis of 23 different screening studies - meaning it's not a one-off, but much more likely to provide a clear consensus finding. Here's a link to the study abstract. A good reminder that as we embark into an era of more and more screening - for more and more biomarkers and genetic signatures and whatnot - there's a very clear side effect, and a great need to keep the tests' sensitivity and specificity - the accuracy, in other words - as high as possible.(It's not just cancer, false positives are a problem for screening in many diseases).

Which also reminds me of a post by the aforementioned Andy Kessler, who's blogging over at the NYTimes (alas, behind the Times Select wall). His piece today neatly summarizes the argument in his End Of Medicine book - that we're embarking on a new era of high-fidelity screening technologies. If you have Times Select access, look especially, though, at the comments, the first four of which all seem to be from physicians who astutely note the perils of increased screening - both in terms of health care costs and false-positives.

Your Genetic Future

Last summer, Andy Kessler came out with new book, The End of Medicine, which despite its decisive and severe title, is in fact a nifty ramble around the frontiers of medical scanning. A former VC, Kessler was particularly compelling in assessing the business prospects of the store-front scanning businesses - BodyScan and others - and how they have tumbled in the last few years. The problem, as Kessler described it, was in the business model: Once you get people in the door for a scan, if there's something wrong they go to a doctor. If there's nothing wrong, they don't come back. No repeat business. And so the sector crashed, with the SEC sniffing for corpses. Such are the perils of trying to bring the frontiers of science to the consumer market. What works for hospitals and insurance companies may not be such a sure sell to patients/citizens. The tricky nature of direct-to-consumer high-tech medicine come to mind whenever I hear about some new genetic profiling startup. They've been around for a couple years now; DNA Direct was one of the first one on the block, I think they started in 2005. Recently there's word of the awkwardly named 23andMe, a startup that looks to be backed by Google, and has generated much scuttlebutt accordingly. The NIH/National Human Genome Institute and the FTC have issued some cautions over the past couple years.

This is indeed a quandary: On the one hand, we should all know all we want about our genome (Harvard's George Church has initiated the audacious Personal Genome Project to undertake just such an effort - Church also happens to be an adviser to DNA Direct). On the other hand, the FTC is right to caution that "having a particular gene doesn’t necessarily mean that a disease will develop; not having a particular gene doesn’t necessarily mean that the disease will not."

Still: how many of us don't want to know what our genome might or might not say about us?

A Blow to Computerized Diagnostics

A new study has found that computerized mammogram readers are significantly less accurate than human readers. The technique - known as computer-aided detection (or CAD, not to be confused with computer-aided design or computer-assisted dispatch or coronary artery disease) is expensive; according to Bloomberg, a CAD system can cost $675,000. Despite that cost (or perhaps because of it? Medicare covers the procedure - an important consideration for any facility considering an investment), many cancer centers have adopted the technology. It's a big setback for what many investors saw as a hugegrowtharea. Gonna be interesting to watch these stocks (link is to a Yahoo sector snap) react next week.

An Epidemic of Cancer? Nope.

Oh boy. So this headline is spinning around the world today: Cancer cases are expected to double by 2030, according to a UN agency. What agency? Why, the International Agency for Research on Cancer.

This is a classic piece of epidemiological sleight of hand - as the article points out, the increase is largely due to 1) people living longer, worldwide (the older you get, the greater your risk of developing one cancer or another), and 2) the worldwide population increasing from 6 billion now to perhaps 9 billion by 2030.

So this isn't in fact what a quick scan of the headline would imply - that there's some alarming upswing in cancer rates (that is, per capita cancers over a certain period of time). It's simply an increase in the overall prevalence in cancer rates, for demographic rather than disease related reasons.

So it's a dramatic headline that really has little drama behind it. But given that it comes from an agency on cancer research, makes me wonder if there's not an implicit bid for a little more research money.

A Vaccine Resurgence?

A nice post from Derek Lowe about a couple new cancer vaccines in the works. The post is pretty sophisticated - if you're not a molecular chemist or a pharma expert (which I'm not) his analysis is pretty dense - but it's enough for me that Lowe finds these things credible. Here's a more consumer-centric report on Cell Genesys's GVAX vaccine from the American Cancer Society, and another on Dendreon's Provenge from the Wall Street Journal. What's interesting to me is that these things aren't conventional drug therapies - they are vaccines derived from a patient’s own cells. The vaccine protein mobilizes the immune system to attack cancer.

Reminds me of some promising vaccine work in a totally different field - tuberculosis. The Aeras Foundation is currently field testing several potential TB vaccines in Africa; Aeras' Lawrence Geiter recently told me he expects one or even two of them to pan out and be available within the next decade.

Add in the recent hubub over the HPV vaccine and I wonder if we're not entering some new moment in immunology, where we're learning to commandeer the body's defenses for those things that it, heretofore, hasn't been able to fend off. Reminds me of the flurry of vaccines from Pasteur and Koch, who thought they'd be able to concoct a vaccine for every disease before the end of the century. The end of the 19th century, that is.

My Science Fair Story

Trolling through Google news, it became apparent to me that it's Science Fair season... in Illinois

and Vermont

and Kentucky

and everywhere else in the US and Canada.

Call me sentimental, but this warms my heart: all these kids doing their experiments with biofuels and rockets and making their posters. Science Fair was one of my favorite times of year as a kid - I'd spend all year thinking up different ideas (electromagnetism? volcanoes? computers?) hoping to make it to the Minnesota State finals. One year, I got there, with a project on Sickle Cell Anemia.

Like a lot of 8th graders I was fascinated by blood, and so I borrowed a blood test kit from my dad (a doctor) and bought some slides from a local science supply store. I convinced most of my classmates to convince most of their parents to sign permission slips so that I could draw a few drops of blood, put them on slides, and stick them under a microscope and look for sickle cells. I was 12 years old. No latex gloves. No supervision. I did swab their fingers with an alcohol pad first, but clearly, this was long before Institutional Review Boards, HIPPA, and HIV.

Problem was, I had a bit of a sample problem. Sickle cell, of course, is a genetic disease where the red blood cells have mutated from round shapes into sickle shapes that can get caught in blood vessels, and interfere in oxygen transmission in the blood because of a corresponding low count of red blood cells (the anemia). It's thought that this emerged as a preventative mutation in populations with malaria - mainly in Africa. Thus, in the US, the disease is predominant in the African American population. Thus my problem: As a student at a Catholic gradeschool in Minneapolis, I had all of one black classmate. Thankfully, he agreed to participate in my "study" - and thankfully for him, he did not have sickle cell.

But that meant that my study consisted of a bunch of white kids of Germanic, Irish, and Polish descent - and one African American boy. Zero positive results. Not a great study. And as my poster could attest, my skills with Elmer's Glue and construction paper didn't exactly make me a budding graphic designer. No surprise, then, that I left the State Science Fair with an Honorable Mention. Still: I was there.

That's my Science Fair story.