Within Normal Limits of Reason

China's healthcare reform and clinical trials

2010-05-10T05:40:00.000-07:00

Image by michael40001 via Flickr

Drove up to Foster City this weekend and attended the BioPacific 2010 Conference sponsored by the Chinese-American Biopharmaceutical Society (CABS). Not that long ago at a Bay Area meetup, the question of drug/device regulation in China came up, and it was clear that few people in the U.S. Biotech industry had a clear idea of the regulatory environment in People's Republic. This is unfortunate as China is projected to be the number 3 market for pharmaceuticals in 2010, and soon to surpass Japan and become number 2 a few years after.

One of the featured speakers is Mr. Ruling Song, the Executive President of China Pharma R&D Association. He had participated in the country's healthcare reform, which was first announced in 2009. The regulations and guidelines are now in place, and now it is clearer what the direction of the reform will be.

Echoing the U.S.'s sentiments, Mr. Song reports that the major focus of the government is on "Big Medicine". That is to say, targeting the most common diseases that cause the most morbidities, in particular in chronic diseases. In order to provide medications to 70% of the 1.3 billion population that the reform plans to cover, the government will rely heavily on generic medications.

This is not surprising. For the purpose of cost-control, insurance plans--public or private--have relied on off-patent medications.

What is important with public healthcare is that the government plays the dual roles of the regulator as well as the payer. In the U.S., under the rubric of comparative effectiveness research, large chunks of resources will be directed towards looking for the most cost-effective (read cheapest) treatment protocols in clinically plausible settings. This will make randomized clinical trials an order of magnitude more complex. The NIH will be awarding grants to study how best to do these complex clinical trials.

In China, the government as the payer will be directing its resources towards the development of generic medications. According to Mr. Song, government grants will also be given out to the developers of generics. The government's goal of 44 new medications--with 2 going for global sales--will also include the development of generics.

This was spoken, of course, in a room full of biotech industry professions, most of whose work involve R&D of novel biochemical entities.

In the U.S., generic medications only have to demonstrate bioequivalence to the predicate product, not pharmaceutical equivalence. Clinical trials, then, can be limited to pharmacologic, e.g., demonstration of similar absorption.

In the grand scheme of things, public money may not be a significant contributor when multinational corporations are throwing money and people at the Chinese pharma market. The outcome will depend on how strict the public healthcare plan will be, and how much competition there will be from private insurance plans.

Matching in clinical studies

2010-02-11T23:48:00.000-08:00

Image by ishmagination via Flickr

This blog post is motivated by articles that I read this month in which matching was used inappropriately in clinical studies.

First, some background. The primary purpose of randomization in clinical trials is to hope that the stochastic process of group assignment would, on average, remove the effects of confounders.

In observational studies, group assignments (e.g.., whether a person is a case, or a control) are not under the control of the researcher. To remove the effects of confounders--at least those that are known--one could adjust for it in the analysis stage. Alternatively the groups can be actively matched during the design stage.

The typical matching one thinks of is the 1-to-n case-control study, where for each case, n matching control are obtained. This is termed individual matching. Another often used matching method is frequency matching, in which the distributions of the confounder variable are matched.

However, matching in design stage must be echoed by statistical adjustment in the analysis stage.

We note that for for frequency matching, at least simple stratified analysis should be adopted. To take the extreme example, consider the Simpson's paradox. This is the phenomenon where an effect seen in all subgroups are reversed when all subgroups are considered together.

Crowdsourcing topics in medical research

2006-06-07T01:03:00.000-07:00

...still waiting by alterednate via Flickr

Not much activity here at the Within Normal Limits of Reason blog lately. Looking over the traffic to this blog, I found that my short post on "post-hoc analysis" was getting a lot of referrals from the major search engines. I decided to port that into Wikipedia. Then, seeing the lack of entries on specific topics in medical research, I started to contribute to other articles... and now I'm hooked! I'm adding bits and pieces on entries related to medical research like randomized clinical trials and EBM, and initiating a couple of entries like post-randomized consent and interim analysis--mostly works in progress.

The quality of the Wikipedia entries are generally pretty good, and given the editorial oversight and the informal peer review by the entire English-reading internet there really isn't as much misinformation as one might fear. There is some link-spamming but that is generally obvious and unobtrusive. As Wired magazine put it, Wiki has really harnessed the power of crowdsourcing and proved it can work.

Technorati Tags:
Medicine / Statistics

Evidence-based rock, paper, and scissors

2006-03-30T22:03:00.000-08:00

Non-transitive dice, invented by Stanford statistician Professor Bradley Efron, defies intuition. In one version, it consists of a sequence of 4 dice A, B, C, D, each with non-standard numbers of its 6 faces. The numbers are chosen such that no matter which die of the 4 you pick, all I need to do is to pick the one before it (e.g., if you picked C then I'll pick A, or if you picked A then I'll pick D), and then, on our next throws of the dice, I am guaranteed to beat you 2/3 of the time. There is no trickery. The dice are not mechanically loaded; the numbers work out so. But this non-transitivity is disconcerting and boggles the mind.

We see something similar in the mediacl research literature. In the Feb issue of the Journal of American Psychiatry, Dr. Heres et al reviewed head-to-head comparisons of 2nd-generation antipsychotics. The report, titled Why olanzepine beats risperidone, risperidone beats quetiepine, and quetiepine beats olanzepine, studied 42 trials that were designed to each compare two 2nd-generation antipsychotics. Published results of these trials were assessed by a psychiatrist and an internist who were blinded to the identity of the medications. They evaluated whether the report favored one drug of the other, and a host of other potential sources of bias in these studies.

Some interesting findings:

33 trials (out of 42) were sponsored by pharmaceutical companies: No surprise here. In fact they thought these industry-sponsored trials were generally of better quality as far as design, implementation, and sample size are concerned.

90 percent of these trials favored the medication sold by the sponsoring pharm company

The wording of the abstract is also systematically different. In those abtracts in which the medication sold by the sponsoring company was shown to be superior, the findings would be elaborated in detail towards the end of the abstract. On the other hand, those trials with results unfavorable to the sponsor only briefly mentions the ersult at the beginning of the abstracts.

All in all not as surprising as the transitive dice.

Technorati Tags:
Medicine / Statistics / Clinical trials / Psychiatry / Antipsychotics

STAR*D: landmark study in psychiatry

2006-03-29T01:46:00.000-08:00

The Orion Nebula [M42] by NASA via Flickr

The results of Sequenced Treatment Alteratives to Relieve Depression (STAR*D), a landmark study in psychiatry, are beginning to be published. Beyond the important questions it answers, STAR*D is remarkable for:

Large sample size: about 4,000 patients are recruited, a mammoth study by the standards of psychiatry literature

Broad inclusion criterion: any patient presenting for care with nonpsychotic major depressive disorder (defined using the HAM-D scale) for whom his/her clinician deems outpatient anti-depressant therpay would be appropriate. Most anti-depressant trials advertise for patients, which would arguably select for a qualitatively different patient population.

Participation of primary care physicians: in addition to psychiatrists in outpatient psych clinics, PCPs were managing these patients in their daily practice. Titration of anti-depressant dosages were based on validated psychometric instruments (for example see the HRS-D and QIDS-C form used by the STAR*D researchers)

Use of depression remission as endpoint: this is closer to the goals of treatment in clinicial practice. Quantitative abatement of depressive symptoms--the outcome measure of most previous anti-depressant trials--may be maningful and significant but its clinical value is difficult to assess.

Simply, by virtue of its design, STAR*D has ecological validity and clinicians can incorporate with confidence the findings of STAR*D studies.

What are these findings? So far, they have found:

About 30 percent of patients showed remission when placed on Celexa. Predictors of remission include high levels of education, employment status, Caucasian race, and few psychiatric and medical co-morbidities. (Trivedi et al 2006, American Journal of Psychiatry 163, 28)

Of the patients who showed no sufficient response to Celexa and who immediately switch to another anti-depressant (either Wellbutrin, Effexor, or Zoloft), about 1 in 3 patients will show remission within 14 weeks. The magnitude of the effect is about the same regardless of the class of medication switched to. That is, switching to Zoloft--another SSRI like Celexa--was as effective as switching to either Wellbutrin or Effexor (Rush et al 2006, NEJM 354, 1231).

Of patients who showed no sufficient response to Celexa and whose medication regimen were immediately augmented with Wellbutrin or BuSpar, about 30 percent showed remission. (Trivedi et al 2006, NEJM 354, 1243)

STAR*D is a result of the NIMH's push for "practical clinical trials", trials that by design answer practical questions of great clinical value.

Technorati Tags:
Medicine / Statistics / Psychiatry / STAR*D / Depression / SSRI / NIMH

Overlapping confidence intervals

2006-02-19T07:46:00.000-08:00

tae kwon do bikarmot 050 by Bara via Flickr

From AP's report on patch contraceptives and the risk of venous thromboembolism:

However, because the confidence intervals of the results for the two forms of contraceptive overlap, there actually may be no increased risk from the patch or it may be more than double.

It's a common mistake to use confidence intervals of 2 measures to assess whether they are significantly different. Two measures can have overlapping CIs yet remain statistically significantly different. As an counterexample, let's take this extreme situation:

* Experiment 1 arrives at the conclusion that if the experiment were repeated thousands of times, the outcome measure X will turn out to be 0 with probability 2.5%, 1 with probability 94.5%, 2 with probability 0.5%, and 3 with probability 2.5% (see top graph). Obviously from the graph, the 95% CI for outcome X is from 1 to 2.

* Another (indepedent) experiment 2 arrives at the conclusion that if the experiment were repeated thousands of times, the outcome measure Y will turn out to be 0 with probability 2.5%, 1 with probability 0.5%, 2 with probability 94.5%, and 3 with probability 2.5% (see bottom graph). Again the CI for outcome Y is simple to see and spans 1 and 2.

Now we have 2 CIs that not only overlap but are in fact identical!

Given this fact that X and Y have identifcal CIs (ignoring the tail probabilities for now), can we conclude that they are not statistically significantly different? Since outcome X is so strongly concentrated at value 1 vs the strong weight of outcome Y at value 2, inspecting the graph and relying on our intuition, we are forced to conclude that despite their identical CIs, outcomes X and Y are statistically significantly different.

How does our intuition compare with statistical reality? A simple simulation test found X and Y to be significantly different with p basically equal to 0 (by Mann-Whitney test).

That overlapping CIs do not imply lack of significant difference is true in real-world situations too. More on this later.

Technorati Tags:
Medicine / Statistics / Contraceptives

Paul Krugman - "Drugs, Devices and Doctors"

2005-12-15T21:34:00.000-08:00

show 9 by Mark Duerr via Flickr

Paul Krugman's column explains Vioxx, Merck, and Cleveland Clinic, in terms of the inter-dependent network of pharma, academic research centers, and doctors as they now are.

Above all, the line between medical researcher and medical entrepreneur has been blurred... Usually, [Marcia Angell] says, "both academic researchers and their institutions own equity" in these companies, giving them a strong incentive to make the big drug companies happy.

The point is that the whiff of corruption in our medical system isn't emanating from a few bad apples. The whole system of incentives encourages doctors and researchers to serve the interests of the medical industry.

The good news is that things don't have to be that way. Economic trends gave rise to the medical-industrial complex, but only because those trends interacted with bad policies, which can be fixed.

Prof Krugman will propose some policy changes in future columns.

Technorati Tags:
Medicine / Medical research / Paul Krugman

Two blackeyes for biomedical research

2005-12-15T11:02:00.000-08:00

MUG Noodle(panda version) by +fatman+ via Flickr

As 2005 comes to a close, we first find that Merck had fudged its data on the Vioxx NEJM article.

Now it appears that Prof Hwang Woo-suk of Seoul National University had fudged his data too in his Science article on the creation of new stem cell cell lines.

BBC reports:

At least nine of 11 stem cell colonies used in a landmark research paper by Dr Hwang Woo-suk were faked, said Roh Sung-il, who collaborated on the paper.

Dr Hwang wants the US journal Science to withdraw his paper on stem cell cloning, Mr Roh said.

Technorati Tags:
Medicine / Biomedical research

The Crab Nebula across the millennia

2005-12-15T04:39:00.000-08:00

Composite of Hubble pics of the Crab Nebula by NASA

The Crab Nebula is a six-light-year-wide expanding remnant of a star's supernova explosion. This composite image was assembled from 24 individual exposures taken with the NASA Hubble Space Telescope�s Wide Field and Planetary Camera 2 in October 1999, January 2000, and December 2000. It is one of the largest images taken by Hubble and is the highest resolution image ever made of the entire Crab Nebula.

So here is in fact a snapshot of the nebula as human civilization passes a major milestone. It's an anniversary of sort for the nebula as well, as the Crab Nebula was first observed, with the naked eye, just about a millenium ago in year 1054. From the Messier Catalog:

The supernova was noted on July 4, 1054 A.D. by Chinese astronomers as a new or "guest star," and was about four times brighter than Venus, or about mag -6. According to the records, it was visible in daylight for 23 days, and 653 days to the naked eye in the night sky. It was probably also recorded by Anasazi Indian artists (in present-day Arizona and New Mexico), as findings in Navaho Canyon and White Mesa (both Arizona) as well as in the Chaco Canyon National Park (New Mexico) indicate.

mon oeil by io2 via Flickr

Prof Zhi Cheng Xiao at the Institute of Molecular and Cell Biology and Prof Gavin Dawe et al at the Stem Cell Research Center of the National University of Singapore They showed that stem cells from the fetus in the womb find their way into the maternal brain, where they differentiate into cells of the nervous system and *may* become a part of the mom's thought process. Fascinating stuff.

from article in Stem Cells journal by Prof Zhi Cheng Xiao and Prof Gavin Dawe et al

Technorati Tags:
Astronomy

Mataariki

2005-12-15T03:44:00.000-08:00

via Wiki:
In Polynesian mythology (specifically the Maori of New Zealand), the Matariki ("small eyes") are the seven gods of the Pleiades, deities of agriculture and patron deities of navigators.

Technorati Tags:
Astronomy

Merck lied

2005-12-15T02:31:00.000-08:00

ONE DUCK A'DUCKING by jon fobes via Flickr

Editorial released online by the New England Journal of Medicine, on the paper published in year 2000 by Bombardier et al on Merck's Rofecoxib. The editorial is penned by editors Gregory D. Curfman, M.D., Stephen Morrissey, Ph.D., and Jeffrey M. Drazen, M.D.:

a memorandum... obtained by subpoena in the Vioxx litigation and made available to the Journal, [showed] that at least two of the authors knew about the three additional myocardial infarctions at least two weeks before the authors submitted the first of two revisions and 4 1/2 months before publication of the article.

We determined from a computer diskette that some of these data were deleted from the VIGOR manuscript two days before it was initially submitted to the Journal on May 18, 2000.

NEJM re-ran the main anlysis and found:

before

after

Dr. Curfman, as quoted in the NY Times article by Alex Brenson:

"They did not disclose all they knew," [Dr. Gregory D. Curfman, the journal's executive editor] said. "There were serious negative consequences for the public health as a result of that."

As we can see clearly from these numbers that the exclusion of these 3 cases of myocardial infarction does not change the main conclusion!

The culprits

The authors of the original study implicated are Claire Bombardier, M.D, and Alise Reicin, Merck's Vice President of Clinical Research.

The fall of Merck

The history of Merck is reviweed here. Merck famously commercialized morphine for medical use and with Pfizer was one of the first companies to mass produce penicillin.

According to NY Times:

Merck now faces more than 6,000 lawsuits from people who say they or their family members suffered heart attacks and strokes as a result of taking Vioxx, and tens of thousands more lawsuits are expected.

Despite Merck's recent victories in court, it looks more and more likely that Merck will be brought down by its COXII inhibitor.

Technorati Tags:
Medicine / Randomizaed clinical trial / Merck / Vioxx

Pregnancy and diseases

2005-12-12T21:07:00.000-08:00

Deep In Thoughts by Ya Ya via Flickr

Fascinating summary in JAMA on pregnancy and the associated medically relevant physiological changes, by Profs Risto Kaaja of Helsinki University Hospital and Ian Greer of University of Glasgow.

In order to optimize the delivery of nutrients and oxygen to the fetus, the removal of fetal waste, the preparation for labor, and the peaceful co-existence of fetal tissue within the body, a variety of coordinated physiological changes occur during pregnancy:

These perturbations--traumatic as they already are for the healthy mom--represent an even more severe challange to those with diseases or those with otherwise-hidden propensities towards developing diseases.

For instance, during pregnancy the mom's immune system, in order to develope tolerance to the fetus, shifts towards a TH2-dominant state. Correspondingly, TH1-dominant diseases such as rheumatoid arthritis (RA), multiple sclerosis (MS), and thyroiditis all improve during pregnancy. In constrast, systemic lupus erythematosis (SLE), a TH2-dominant disease, tends to deteriorate.

Other diseases of the endocrine and the cardiovascular systems are also perturbed by the pregnant state:

Based on these findings, Profs Kaaja and Greer, both obstetricians, recommend screening protocols to be performed peri-partum to screen for these diseases.

We note also recent work with stem cells have found neurons of fetal origin to be present in the maternal brain, such that in parts of the maternal brain a portion of the mom's nervous system is contributed by their children! (it's important to note that it's not clear these neurons of fetal origin participate in the maternal brain or are just idle bystanders. See also Parasitic Hairworm Charms Grasshopper Into Taking It for a Swim)

Technorati Tags:
Medicine / Evidence-based parenting / Immune system

Outbreak of Clostridium difficile diarrhea associated with fluoroquinolone use

2005-12-08T16:47:00.000-08:00

Curved steps by ::: mindgraph ::: via Flickr

A New England Journal of Medicine editorial by Drs. John Bartlett and Trish Perl on these 2 papers:

An Epidemic, Toxin Gene-Variant Strain of Clostridium difficile by McDonald et al

A Predominantly Clonal Multi-Institutional Outbreak of Clostridium difficile-Associated Diarrhea with High Morbidity and Mortality by Loo et al

and the CDC report.

Diarrhea secondary to Clostridium difficile colitis is in hospital in-patients characterized clinically by watery diarrhea, abdominal cramps, and white blood cell counts elevated in the 50,000s. Such dramatically high WBC counts are rare, in fact, except in cases of leukemia, and are useful in clinical diagnosis. The typical laboratory diagnosis is done by detection of the C. diff toxins. The assay is in fact based on cross-reactivity between antiserum to Clostridium sordelli against Clostridum difficile cytotoxin (a.k.a. Toxin B) in tissue culture. (Notes on Medical Microbiology by Yifan Douglas Yang, Morag C Timbury, A Christine McCartney, Bishan Thakker). Interestingly, this lab test will not identify this epidemic strain (? more details coming).

In the 1970s, clindamycin was associated with C diff diarrhea. In the 1980s it was cephalosporins. Now it is the use of fluoroquinolones.

Pathogenicity of the new C diff strain

As Warny et al described in the Lancet, this new strain of C diff associated with fluoroquinolone use in hospital in-patients is unique in its rate of toxin production. With mutations in its tcdC locus, it seems to have evolved the ability to produce vastly larger amounts of toxins A and B. In addition, whereas other strains of C diff produce only minimal amounts of toxin while they are busy growing and dividing (i.e. the log phase), the new outbreak strain produces toxins A and B even during this time.

We note that the genome of Clostridium difficile is still being sequenced by the Sanger Institute.

Technorati Tags:
Medicine / Public health / Microbiology / Clostridium difficile

Applying public health measures to HIV infection

2005-12-07T15:22:00.000-08:00

everybody hurts by Cilest via Flickr

Dr. Thomas R. Frieden, the Commissioner of the New York City Department of Health and Mental Hygiene, et al in the New England Journal of Medicine.

25 years into the epidemic, progress is stalled. The number of deaths among people with AIDS has not declined since 1998, and the number of newly diagnosed cases is rising slightly... Late diagnosis of infection is common... Notification of the partners of infected persons is rare.

...given the availability of drugs that can effectively treat HIV infection and progress on antidiscrimination initiatives, perhaps society is ready to adopt traditional disease-control principles and proven interventions that can identify infected persons, interrupt transmission, ensure treatment and case management, and monitor infection and control efforts throughout the population.

Conan & Walker Texas Ranger

I do not know much about applied epidemiology myself, but this is how Dr. Frieden breaks it down:

Case finding and surveillance:

Named reporting of all found with a reportable condition (e.g., Salmonella infections and Reye's syndrome). This is now true for patients found to be HIV positive in Michigan.

Availability of routine testing in health care settings. Dr. Frieden states that this is often not available.

Notification and testing of partners

Interruption of transmission: promotion of condom use, perinatal veritcal transmission prevention, blood product screening, development of vaccine.

Systematic treatment and case management:

Case management of the infected individual and his/her contacts, with the case managers accountable for patient outcome. In Ann Arbor, the non-profit HIV/AIDS Resource Center (HARC) provides this service.

Linkage of social services to medical care compliance

Population-based monitoring:

Contact of treating physician by public health agencies if patient shows inadequate response. This is generally not done for HIV/AIDS.

Monitoring viral resistance. Not done outside of research studies.

Technorati Tags:
Medicine / Public health / HIV / AIDS / Case management

Washington Post on the (inadequate) use of fistula for hemodialysis

2005-12-06T01:11:00.000-08:00

lovers by Daniel M. Shih via Flickr

According to the The National Kidney Foundation�s Kidney Disease Outcomes Quality Initiative (K/DOQI) Clinical Practice Guidelines for
Vascular Access, published in year 2000, Guideline 3:

The wrist [arteriovenous] fistula is the first choice of access type because of the following advantages:

It is simple to create.

It preserves more proximal vessels for future access placement.

It has few complications. Specifically, the incidence of vascular steal is low, and in mature fistulae, thrombosis and infection rates are low.

Yet Gilbert M. Gaul of WaPo notes:

Today, fewer than four in 10 dialysis patients nationwide have a fistula, despite overwhelming evidence that they are safer, cheaper and more effective than grafts and catheters.

...fistula rates remain stubbornly low in the United States, and vary dramatically by geography.

Medicare, which picks up most of the $15 billion annual tab for dialysis treatment, pays surgeons more for grafts and catheters than fistulas -- in effect, rewarding inferior care. That's because its reimbursement system is based on the time and resources needed to do a procedure, not on the outcome.

In 2003, it cost Medicare an average of $52,751 to care for a patient with a fistula, compared with $61,929 for those using a graft and $69,893 for those on a catheter, federal data show.

Patients with grafts and catheters also have a 20 to 70 percent greater chance of dying in the first year of treatment, according to CMS.

From the Rayner et al 2004 paper in the American Journal of Kidney Diseases, "Vascular access results from the Dialysis Outcomes and Practice Patterns Study (DOPPS): performance against Kidney Disease Outcomes Quality Initiative (K/DOQI) Clinical Practice Guidelines", we see that:

The U.S. ranks in the very bottom for fistula use in chronic hemodialysis. The DOPPS studies have produced some very interesting results. Among others, it has shown that American patients on hemodialysis have much higher mortality than those in other countries. I'm not certain if this is due to differences in patient characteristics or medical care, but it has been suggested the difference may be related to the much faster, much higher-flow hemodialysis preferred by patients and dialysis centers alike.

Technorati Tags:
Medicine / Medicare / Hemodialysis / Health policy

Nicholas Kristof on "The Hubris of the Humanities"

2005-12-05T22:08:00.000-08:00

Music Library: Man and woman reading
in "small room" via pantufla via Flickr

In this column, the worldly and profound Nicholas Kristof begins with the intelligent design debate and comments on the American culture's disdain of science as it is, "a profound illiteracy about science and math as a whole".

A larger problem is the arrogance of the liberal arts, the cultural snootiness of, of ... well, of people like me - and probably you.

... In the U.S. and most of the Western world, it's considered barbaric in educated circles to be unfamiliar with Plato or Monet or Dickens, but quite natural to be oblivious of quarks and chi-squares.

Such great thinkers of the Western tradition as Aristotle insisted that women have fewer teeth than men. As Bertrand Russell said, "although he was twice married, it never occurred to him to verify this statement by examining his wives' mouths."

While I'm a fan of Aristotle's On Rhetoric--as I am a bigger fan of Kant's metaphysics--parts of their works have quite limited applicabilities today. As historical works of smart men and women before our time, they demonstrate both the power and the limits of reason, how deeply the mind can make sense of the world and how simply it can fall short.

Indeed,

...don't pin too much faith on the civilizing influence of a liberal education... similar arguments were used in past centuries to assert that all a student needed was Greek, Latin and familiarity with the Bible - or, in [Ming Dynasty] China, to argue that all the elites needed were the Confucian classics.

Increasingly, we face public policy issues - avian flu, stem cells - that require some knowledge of scientific methods, yet the present Congress contains 218 lawyers, and just 12 doctors and 3 biologists. In terms of the skills we need for the 21st century, we're Shakespeare-quoting Philistines.

The infamous Triumph Star Wars interviews

Mr. Kristof quotes work by Jon Miller, the Director of the Center for Biomedical Communications at Northwestern University showing that "in 34 countries, says Turkey is the only one with less support for evolution than the U.S." Couldn't find the original source, so can't comment on the methodology.

I did find this report at National Science Foundation on public understanding of science and tech from 2004:

More Americans now agree with the theory of evolution. The 2001 NSF survey marked the first time that more than half (53 percent) of Americans answered "true" in response to the statement "human beings, as we know them today, developed from earlier species of animals." (In Europe, 69 percent responded "true.")

The data sources of this report:

Other interesting findings from this report:

While belief in pseudoscience is still common--60 percent of surveyed Americans said they believe in extrasensory perception, and 41 percent thought that astrology is at least somewhat scientific., only a small subset (10 percent) of the American population admit no interest in science and technology issues. It's not clear however if this interest is in science and tech in themselves or if it includes antagonism.

As seen on the figure on the right, Americans get most of their information about science and tech from TV (top pie). But when they have specific science & tech questions, they overwhelmingly turn tot he internet (bottom pie).

Technorati Tags:
Science / Humanities / Public understanding of Science / Nicholas Kristof

Diagnostic Clinical Trial: Phased evaluation of medical diagnostic technologies

2005-12-05T00:11:00.000-08:00

Leap of Faith by Jon Lucas

From the Statistics in Medicine special issue on the 25th Annual Conference of the International Society for Clinical Biostatistics, Prof Margaret S. Pepe from University of Washington Dept of Biostatistics and Fred Hutchinson Cancer Research Center's article "Evaluating technologies for classification and prediction in medicine".

To begin with, Prof Pepe points to the CONSORT statement, which in essence gives guidelines for the design, analysis, and reporting of randomized clinical trials of therapeutics. She notes the Standards for Reporting of Diagnostic Accuracy (STARD) initiative which has arisen to specify analogous guidelines for medical diagnostics.

Prof Pepe proposes that the structured investigation of therapeutics into the canonical 4 phases can also be adapted for the investigation of medical diagnostics.

The 4 Phases of Clinical Trials as outlined
by the International Committee of Harmonization

Prof Pepe proposes a 5-phase approach. For each phase she identifies the major objective and the best study design to achieve these goals.

Preclinical exploratory: to identify promising directions by convenience sample case-control studies

Clinical assay and validation: to determine if a lcinical assay detects established disease by population-based case-control studies

Retrospective screening: to determin if the biomarker detects disease before if becomes clinical, and to define a "screen positive" rule, by nested case-control studies

Prospective screening: to determine the extent and characteristics of disease detected by the test and to identify the false referral rate, by cross-sectional studies

Cancer control: to aseess the impact of screening on reducing the burden of disease on the population, by--ideally--randomizaed trial

Prof Pepe also identifies a number of interesting statistical issues that apply distinctly to medical diagnostics. Will review.

Technorati Tags:
Medicine / Statistics / Medical diagnostics / Margaret S. Pepe / International Committee of Harmonization

DNA microarray for speciation of clinical fungal pathogen isolates

2005-12-04T22:31:00.000-08:00

Myra Ake and Ethel Spence via pantufla via Flickr

From the Journal of Clinical Microbiology, Leinberger et al uses microarray technology to rapidly speciate human fungal pathogens.

Susceptibility testing is already quite unreliable for bacteria, and it is even less useful for fungal infections. In fact antifungal agent choice is mostly based on the speciation of the fungus. Yet speciation is typically performed by examining the shape of the fungus and performing biochemical tests on fungal colonies. This requires a relatively large number of fungi, which will needs a couple of days to grow. This process is of course the rate-limiting step of clinical decision making in treating fungal infections.

Using DNA microarray technology, Leinberger et al produced a spot array chip containing unique sequences of rRNAs from 12 common pathogenic fungi. These were validated against 21 speciated clinical isolates. They found that the microarray chip was able to rapidly (within about 4 hours) and accurately speciate 12 of the most common fungal pathogens in the genus Candida and Aspergillus.

This clinical applications of DNA microarray technology is particularly clinically relevant, as critically ill and immunocompromised patients can be infected by multiple fungal species. Rapid and targeted antifungal selection should help clinicians better balance the treatment of infections against the often serious adverse effects of antifungals.

Technorati Tags:
Medicine / Statistics / Computational biology / Fungal speciation / Gene expression microarray

Tom Cruise buys an ultrasound machine for home use

2005-12-04T14:54:00.000-08:00

23 Weeks, the wonder of 3D
by Craig Cmehil via Flickr

The American College of Radiology (ACR)'s press release regarding Mr. Cruise:

...ultrasound should not be considered completely innocuous. Laboratory studies have shown that diagnostic levels of ultrasound can produce physical effects in tissue, such as mechanical vibrations and rise in temperature, particularly when used for a prolonged period of time.

The FDA's statement in June 2005 warned:

those who subject individuals to ultrasound exposure using a diagnostic ultrasound device (a prescription device) without a physician's order may be in violation of State or local laws or regulations regarding use of a prescription medical device.

This may not be a purely technical violation, as MSNBC's Fran Kritz noted:

Tom Cruise Kills Oprah

The FDA says that while prenatal sonograms are generally done at very low power levels, current machines, which range in price from $15,000 to $200,000, can produce sound wave intensities eight times higher than a decade ago and at that level, especially if used for prolonged periods, there could be a real risk to the developing fetus.

Mr. Cruise should note that there is a new Alberta, Canada law which will allow children to sue their mothers for injuries suffered in the womb.

(via Kevin, MD blog)

Technorati Tags:
Medicine / Ultrasond / Tom Cruise / Consumerism in Medicine

Medical tourism amongst Asian countries and the Globalization of Medical Care

2005-12-04T09:39:00.000-08:00

Face of the world by Jaxpix50 via Flickr

By Steve Mollman of WSJ:

Facing age 40, Paul Luciw decided an extensive health checkup was in order. His local hospital in Hong Kong said it would cost the equivalent of about $1,300.

He got a quote from Bangkok's Bumrungrad International hospital for about $300. So, with a few buddies who also wanted checkups, he made a mini-vacation out of it earlier this year.

The article gives figures quoted by Ruben Toral, marketing director at Bumrungrad International:

Health-care services in Thailand is

about 40 percent cheaper than in Singapore

about 75 percent cheaper than in Japan

Mr. Toral also says the number of foreign patients at his hospital jumped to more than 380,000 this year, from about 168,000 in 2000.

Quality Assurance in Globalization of Medicine

The Joint Commission International is a U.S.-based accreditor of health-care organizations around the world, which

extends [Joint Commission on Accreditation of Healthcare Organizations (JCAHO)]'s mission worldwide. Through international consultation, accreditation, publications and education, Joint Commission International helps to improve the quality of patient care in many nations. Joint Commission International has extensive international experience working with public and private health care organizations and local governments in more than 60 countries.

What this means for medical students

BSB-I want it that way

I personally do not see any significant force or event of sufficent strength and persistence--save an all-out WWIII as in Scorpion's Gate--to reverse this emerging trend. One could make the argument that for U.S. patients, the psychological and financial barriers are a bit higher, given the longer distance required for travel and the stronger baseline suspicion we harbor towards foreign science and technology. But at least within regions--specifically Asia and Europe where medical tourism has been observed by local media to have emerged--this trend will be irreversible. The main question then is how far will this go?

From a medical student's perspective, it does not seem unreasonable to go into a global medical practice. That is, since we have no established practice of yet, this trend is less disruptive than it is promising. As an example, for 6 months out of the year, my family and I could fly to an Asian country and practice, where given the lower local costs I charge lower prices and I live cheaper. Suppose I was a pulmonologist, then I could help deal with the COPD patient population to come given the massive cigarette consumption in Asian countries. For the other 6 months of the year, I could practice in the U.S. in an academic center where I can focus on taking care of American patients and learning the latest developments of my field. For more elective services, I might even be able to refer U.S. patients to my Asian practice.

Some pre-requisite:

Infrastructure: sufficient number of academic centers/practice groups need to be comfortable enough with accepting such international doctors.

Accreditation: it sure would be nice if medical licensure can become global. I don't see this happening anytime soon since this is the major mechanism for controlling the supply of doctors in the U.S.

Well my hypothetical wife and children will have to be able to accept this lifestyle. As far as education for children... I guess I'll just enroll them in MIT's Open CourseWare :)

Technorati Tags:
Medicine / Medical tourism / Globalization of Medicine

WHO Stops Hiring Smokers

2005-12-03T16:53:00.000-08:00

guten tag by Fib via Flickr

Sam Cage via AP via WaPo:

"WHO has taken a very public lead in the fight against tobacco use," spokesman Iain Simpson said. "As a matter of principle, WHO does not want to recruit smokers."

WHO chief Lee Jong-wook wrote a letter to U.N. Secretary-General Kofi Annan asking him to consider whether all United Nations agencies and offices should be made smoke-free areas.

Lucky Strike: Marching Cigarettes Commercial 1948

For example, the WHO job posting for an Epidemiologist in New Delhi has this at the bottom:

Technorati Tags:
Smoking / WHO / Hiring policy / Lee Jong-wook

NYT Editorial on "South Korea's Cloning Crisis"

2005-12-03T15:47:00.000-08:00

Shiny by Noli Novak via Flickr

Continued fall out of Dr. Hwang Woo Suk's use of his underlings' eggs for stem cell research.

How harshly Dr. Hwang should be judged for such transgressions is a matter of dispute... But what really torpedoed Dr. Hwang was the cover-up: his repeated lies to the effect that his eggs were donated by unpaid volunteers. These misrepresentations led his most prominent American collaborator to sever ties because his trust had been shaken.

Actually I believe Dr. Hwang has only admitted to ignorance of these problems.

Christopher Reeve's Last Public Speech
at Rehabilitation Institute of Chicago

The key unresolved issue is whether lying about egg donations suggests that the Korean team may have lied about its scientific results... American collaborators and observers remain confident that the team's achievements were real. But science is an enterprise that relies heavily on trust. The Koreans should not be surprised if their next scientific breakthrough is greeted with extreme caution.

I'm afraid NYT is being a bit harsh here. Is one research group's actions to taint the scientific output of one Department? University? An entire Country?

While their personalities are apparently disparate, I like to compare Prof Hwang with our own Dr. Craig Venter. Both scientists of vision. Both ostracized for the source of their research raw material. (Recall the human genome sequenced by Celera/Applera was Venter's own)

Technorati Tags:
Science / Hwang Woo Suk / Stem cell research / Politics in academia

GAO - Geographic variations in pricing and spending of the Federal Employees Health Benefits Program (FEHBP)

2005-12-02T11:16:00.000-08:00

The Raking by Lawrence Roberts

Government Accountability Office's Health Care Director Bruce Steinwald testifies before the House's Subcommittee on Health, Committee on Ways and Means, titled "Differences in Health Care Prices Across Metropolitan Areas Linked to Competition and Other Factors".

What can this report tell us? According to the GAO, the prices that FEHBP preferred provider organizations (PPOs) negotiate with hospitals and physicians are comparable to those negotiated by PPOs working with private sector employers. And since prices for FEHBP PPO show great geographical variations--prices for hospital stays vary by more than 250 percent and that for physician services by about 100 percent across metropolitan areas, understanding the factors contributing to these price differences may shed some light on price differences ordinary consumers across the U.S. encounter.

Definitions

Market competitiveness was defined as either:

Hospital competition: a measure of concentration of hospital beds across all hospitals in the metro area.

Payer bargaining power: measured as percentage of HMO capitation in the area.

Main results

"Areas with the least competitive markets�that is, areas with a higher percentage of hospital beds concentrated in the two largest hospitals or hospital networks�had prices of hospital stays and physician services that were higher than areas with the most competitive markets."

When other factors are controlled for, the associations between hospital competition and payer bargaining power with lower prices are decreased, but remain statistically significant. And much variation remained unexplained.

In general, price (e.g., price per night of hospital stay) contributed to 1/3 of the geographic differences in spending per enrollee, whereas utilization (e.g., number of days of hospital stay) explained the other 2/3.

This report was not very detailed as far as its methodology. We also note that this testimony is not written with the usual and customary tone of neutrality that characterize other GAO reports I have read.

Technorati Tags:
Medicine / Health care / Health insurance / Government Accountability Office

SLIM, from NCBI comes a fantastic alternative PubMed search interface for clinicians!

2005-12-01T20:23:00.000-08:00

By Charloto via Flickr

A pre-print from BMC Medical Informatics and Decision Making 2005, "SLIM: an alternative Web interface for MEDLINE/PubMed searches - a preliminary study", by Michael Muin, Paul Fontelo, Fang Liu, and Michael Ackerman from NCBI.

You have to give SLIM a try at http://pmi.nlm.nih.gov/slide/.

Muin et al have have given the PubMed "Limits" search capability a complete makeover. Using Javascript and DHTML, they use an interface based on sliders for users to limit search by:

Publication date: going back 1 year? 2 years? ... no need to laboriously type in dates

Journal subset: from all journals to the 120 core clinical journals in the Abridged Index Medicus

Age group

Study design: from case reports to randomized clinical trials to systematic reviews

Search mapping: e.g. text search in Title/Abstract, MeSH term search, etc

Preset values are available for common limits such as recent systematic reviews.

The search results can be displayed in the same page. Individual abstracts can be expanded/collapsed easily.

SLIM was written in PHP and developed on an Apache 2.0.52 server running PHP 4.3.10. The PHP scripts generate a HyperText Markup Language (HTML) and JavaScript search form. JavaScript provides most of the functionality of the search form and search results. Free and open source JavaScript codes were downloaded from the Internet for the slider controls.

Muin et al make use of widgets obtained from WebFx and Dynamic Drive.

Eighteen physicians from the US, Australia and the Philippines participated in the beta-testing phase of the application and provided performance and usability feedback through an online survey.

I love this interface!

As an example, here I search for "paroxetine" and "suicide" and limit my search to the core clinical journals and the age group to adolescents, like so:

The results are displayed as:

Technorati Tags:
Medicine / Medical informatics / PubMed / Medline / NCBI / SLIM

GAO - Medicaid: States� Payments for Outpatient Prescription Drugs

2005-11-30T20:46:00.000-08:00

Niece staring with Suspicious Eyes
by Toshihiro Oshima via Flickr

A Government Accountability Office report to Congressman Edward Whitfield, the vice chair of the House Subcommittee on Oversight and Investigations of the Committee on Energy and Commerce. This report was initiated to investigate the rising costs of Medicare to cover outpatient mediations.

Methods

Five states were studied: Mississippi, Montana, Pennsylvania, South Carolina,
and Utah.

For each drug covered, each state�s payment per unit was calculated as first taking

the lowest of:

The state�s estimate of drug acquisition cost

The pharmacy�s usual and customary charge

The federal upper limit (FUL, if available. The CMS is required by Federal regulations to set specific FULs for drugs that are provided by at least three suppliers.)

The state's maximum allowable cost (MAC, if available. Analogous to the FUL but set by individual states. As of December 2003, 38 states had established MACs for certain drugs at rates below the corresponding FUL or for drugs for which CMS had not set an FUL.)

...and dividing this minimum by the number of units dispensed to yield each state's payment per unit.

Note that dispensing fees are not included.

Each state's payment per unit (PPU) is compared against each of the 3 estimated market prices defined as:

Results

Overall, minimal variation existed among the five states� payments for most drugs.

The five states� payments for 189 brand-name drugs varied less than 7 percent on average

The five states� payments for the 5 generic drugs we reviewed varied 30 percent on average.

The GAO found that Medicare in all 5 states pay higher prices than each of the 3 market prices:

On average, each state�s payments for brand-name drugs exceeded each market-based price by 10 percent or more. Additionally, states� average payments for brand-name drugs were 12 percent higher than AMP, 36 percent higher than Best Price, and 73 percent higher than FSS Price, on average.

The Centers for Medicare and Medicaid Services was given a chance to respond to the GAO:

CMS stated that this report makes it clear that the current payment rules result in overpayments for drugs and emphasizes the need for reform. CMS commented that payments should be determined using accurate acquisition cost data, which it said requires congressional action.

Technorati Tags:
Health care / Medicare / Drug costs / Government Accountability Office

ABC News - Baby Coach Teaches Newborns to Sleep All Night

2005-11-30T18:45:00.000-08:00

Sunflower by Jon Lucas via Flickr

Accompanying video clip.

The first few weeks with a newborn are all a blur, as "Good Morning America Weekend Edition" co-anchor Kate Snow can tell you. Her baby, Abby, is 11 weeks old.

Suzy Giordano, whose book, "Twelve Hours' Sleep by Twelve Weeks Old: A Step-by-Step Plan for Baby Sleep Success," will be out in January, teaches babies how to sleep through the night.

After Giordano leaves, she coaches Snow and Breault by e-mail for weeks. It's a time-consuming effort, but the payoff has been huge. At 11 weeks, Abby sleeps 10 hours every night.

"By the time she's 12 weeks of age, three months, she will be able to sleep 12 hours straight," said Giordano, who has been coaching new parents � mostly professionals in their 30s and 40s in the Washington, D.C., area, where she's based � for 13 years.

According to the Amazon Editorial Review,

Suzy Giordano is the mother of five children, the youngest being fraternal twin boys. Also known as the Baby Coach, Suzy has worked with hundreds of Washington, D.C. area families as a baby sleep specialist for the past ten years.

This is her website.

From a WaPo Article on Suzy Giordano back in 6/2005, by Jennifer Frey,

Suzy is Suzy Giordano, aka "the baby coach," a petite, Brazilian-born woman who is an underground legend in the Washington area for her ability to teach newborn babies how to achieve that parenting nirvana: sleeping through the night.

Interestingly in the WaPo article it is "Good Morning America"'s senior correspondent Claire Shipman and her husband Time magazine correspondent Jay Carney who need help with their baby,

[They] have a 4 1/2 -week-old baby and a 3 1/2 -year-old toddler. They celebrate five-hour sleep nights, they're not so sure where they left things -- maybe it's somewhere over there in the kitchen? -- and they don't know when they'll next have time for each other, let alone themselves.

But they have Suzy.

I have not read the book so I won't comment on the Baby Coach's methods. Since I'm such a fan of baby sleep methods, I'll be reviewing Ms. Girodano's book when it comes out in Jan 2006.

We note that like the Baby Coach, Dr. Harvey Karp is also moving towards a service-oriented approach. He has been touring the U.S. and Canada to promote his new certification program, which will train THB Educators, analogous to Lactation Consultants and childbirth Educators.

Infant Doll
available at The
Happiest Baby
Certification
Store :)

From Dr. Karp's FAQ,

Especially since a lot of Happy Baby is counterintuitive - the worry it will spoil baby, parents think it is a fad, baby may resist the swaddle, shush seems too loud, swinging, colic, parents needing to know when to suspect problem and when to call MD, etc.

Apparently this program is doing pretty well, with hundreds of participants already.

I welcome this trend. From discussions with pediatricians and non-pediatricians I meet during rotations, it appears that Dr. Karp's methods are quite conceptually appealing to medical professionals. And certainty it seems many people who have tried it find it immensely helpful.

Technorati Tags:
Pediatrics / Parenting / Baby sleep coach / Harvey Karp / Certification

In the Pipeline - Ghrelin and Obestatin

2005-11-29T21:40:00.000-08:00

kelp catches the light by Sara Heinrichs via Flickr

Dr. Derek Lowe remarks on 11/11/2005 Science article Obestatin, a Peptide Encoded by the Ghrelin Gene, Opposes Ghrelin's Effects on Food Intake by Jian V. Zhang et al from Prof Aaron Hsueh's lab at Stanford. As Dr. Derek explains, this paper found that the same peptide product which gets cleaved into ghrelin--the famous appetite inducing peptide hormone--has been found to also give rise to obestatin, which does just the opposite and "suppressed food intake, inhibited jejunal contraction, and decreased body-weight gain". Dr. Derek says,

So now we know more about the regulation of appetite than we used to, although researchers in that field probably thought it was complicated enough already, thanks very much. What I find particularly interesting about this discovery is how these two opposing hormones are cut from the same larger protein. That means that they both come from the same gene, you know. Which shows you just how far a pure genome-driven approach to drug discovery will get you: not far enough.

IMHO, this may also demonstrate how far it does take us and all the new opportunities available to us in this post-genomics era.

We note that obestatin would not have been identified without cross-species sequence analysis of the pro-peptide giving rise to ghrelin. As shown in thie figure from the online supplement:

Zhang et al 2005 online supplement

It is upon observing peptide sequence conservation and peptide hormone cleavage signals that Prof Hsueh's lab started the investigations ultimately discovering obestatin. We note that they actually developed an in-house algorithm for this purpose (Avsian-Kretchmer O and Hsueh AJ 2004 in Molecular Endocrinology).

For fun we look at the POMC gene, which like the ghrelin/obestatin precursor is also cleaved into a number of peptide hormones. We use my favorite genome browser Ensembl. The human POMC is on chromosome 2:

We align the POMC transcript with those from the dog, rat, and mouse, and find that they match up pretty well:

A quick peek at the finer details of the alignment, down at the base pair level, shows:

Note that in the first row, corresponding to the human POMC sequence, a base G is highlighted in green. This means that there is a SNP at this location. Clicking on the green part allows us to link to the NCBI dbSNP page explaining this genetic polymorphism. At this point even the existence of the SNP is as yet unvalidated. But if I were a medical researcher studying any of the peptide hormones generated from POMC, I'd be looking into the biochemical implecations of this SNP.

Of course this has been a very cursory look which didn't really yield any useful information. But I hope it demonstrated that genomics can efficiently target and fine-tune wet laboratory investigations.

All Ensembl results can be further explored by following this link.

Ensembl.org is heavily based on web services and is a good example of Science 2.0.

Technorati Tags:
Medicine / Genomics / Bioinformatics / Sequence analysis / Ensembl / Science 2.0

Widowhood and Mental Illness

2005-11-29T03:28:00.000-08:00

By +fatman+ via Flickr

During rotations I have taken care of a couple of elderly widows and widowers. Suffice to say that they are quite a unique group of patients to take care of. So some research is warranted...

In the U.S., 60% of women between ages 75 to 84 are widowed (Merck Manual of Geriatrics). With their longer life expectancy, the majority of women seem destined to share the common experience of widowhood. But the loss of a spouse is an intensely personal experience�after all, the spouse is the most physically and emotionally intimate companion in life. Across age groups and cultures, the death of a spouse is consistently rated as the most stressful life event (Holmes and Rahe 1967). So traumatic is the experience that many new widows fit the diagnosis of PTSD (Zisook et al 1998). But the mental illness the widow is most at risk of, unsurprisingly, is depression.

The spouse is arguably the most important person in one�s social network. The disappearance of the spouse, then, represents a fundamental perturbation on the network. Feelings of isolation and loneliness are common. She must adapt to the loss by re-structuring her social network, substituting roles played by her husband and compensating for her unmet social needs. This substitution may take any of 3 forms:

Formation of new social ties

Rekindling of dormant ties

Intensification of existing ties

In a fascinating cohort study, Zettel and Rook (2004) conducted serial interviews on 322 older women, ages 60 to 85, widowed for 3 to 30 months, over the course of a year. Among some of their findings were the surprising observations that

Widows whose social network included more rekindled ties or more intensified ties developed significantly more depressive symptoms over the course of a year.

Widows whose social network included more new social ties developed significantly higher loneliness ratings over a year.

Unexpected as these findings may be at first glance, they authors present a number of plausible explanations. Most importantly, I think these associations may be useful clinically.

Of course, this was an observational study that measured profound emotional responses with structured interviews. But we could hardly do any better.

Technorati Tags:
Medicine / Geriatrics / Bereavement / Psychiatry / Depression

(The Official) NCBI Toolbar

2005-11-28T23:03:00.000-08:00

Via Alberto

Much like the Google Toolbar and the Yahoo Toolbar, now there's NCBI Toolbar! There's a search box and you can designate the database to search in. Here is an example of a PubMed search:

Note that search terms can be highlighted, just like Google Toolbar. You can search in the Genes DB:

Here is a Flash-based demo.

Ok, while not revolutionary, this is certainly an encouraging sign. I have always been impressed by how technology-friendly NCBI is and how open they are about sharing their data. Another feature I find very useful for getting updates on topics of interest is the PubMed RSS feed:

I'm still getting updates on new articles about acetaminophen toxicity, a RSS feed I set up when when I was taking care of a 16yo girl during pediatrics rotation.

Technorati Tags:
medicine / Science 2.0 / NCBI Toolbar

Paul Krugman - "Age of Anxiety"

2005-11-27T20:31:00.000-08:00

"Do you have an appointment?" by soozclooz via Flickr

NYT's Paul Krugman,

Mr. [Peter] Drucker wrote "The Age of Discontinuity" in the late 1960's, a time when most people assumed that the big corporations of the day, companies like General Motors and U.S. Steel, would dominate the economy for the foreseeable future.

But

New technologies would usher in an era of "turbulence"... corporations can't provide their workers with economic security if the companies' own future is highly insecure.

This, according to Prof Krugman, is an Age of Anxiety. He goes on to relate this to the burden of health insurance on companies and the need for a single-payer national health insurance in the U.S. By framing the financial difficulties of old-time American corporations this way, perhaps Prof Krugman can make the idea of national health insurance palatable to economic conservatives.

The individual in the Age of Anxiety

snaily leaves the family by Sara Heinrichs via Flickr

These are really interesting times. As Thomas Friedman puts it,

When I was growing up, my parents used to say to me, "Tom, finish your dinner. People in China and India are starving." Today I tell my girls, "Finish your homework. People in China and India are starving for your jobs."

In The World Is Flat, Friedman proposes that there are 4 kinds of "untouchables" in this new flat world.

Special, e.g. Michael Jordan, Bill Gates

Specialized, e.g. vascular surgeons

Anchored, e.g. the local barber

Adaptable

Adaptability refers to the ability to learn how to learn.

As parts of your work becomes commoditized and fungible, or turned into vanilla [ice cream], adaptable people will always learn how to make some other part of the sundae... "because job churn will come faster, because innovation will come faster"

I don't think we in health care will be immune to this era of global transformation. We do live in interesting times.

Technorati Tags:
medicine / Paul Krugman / Thomas Friedman / The World Is Flat / Parenting

On Divorce

2005-11-27T00:21:00.000-08:00

Butterfly on Fire by Jeff Kubina via Flickr

Media reports

"More Options to Answer 'What About the Kids?'" by NYT's Mireya Navarro, 11/27/2005,

The shift to more equitable custody arrangements over the last decades, a reflection of the changing role of fathers, has propelled many couples into more contentious divorce proceedings, as they fight over equal time with their children, some divorce lawyers say. Yet the resulting financial and emotional toll of protracted litigation is at the same time creating a powerful incentive for many separating couples to look for ways to agree.

The OurFamilyWizard.com web site is mentioned. This is basically a secure community site encompassing otherwise common tools such as group calendars, message boards, file storage space, etc, but specifically designed for divorced families. This web site received a lot of media coverage some years ago.

"The things that are difficult to discuss - who gets the kids when, who owes who money - that's handled electronically" ... "The kids don't have to listen to talk about all that. Our conversations are focused on the kids. They're about the school conference or Christmas."

Navarro talks about the "bird-nesting" custody arrangement, in which the family's original residence is kept while the 2 parents take turns live in with the kids.

The couple keep most of their belongings and receive mail at their old address. They switch off stays under certain rules: each will leave the house tidy, each is responsible for his or her own groceries, and neither will bring dates. Mr. Bowerman said the arrangement "makes it more difficult" for both parents "because you don't get closure."

But he added, "Right now I see it as the best thing for the kids."

San Francisco Chronicle's Adair Lara writes on bird-nesting,

By all reports, birdnesting requires so much tact and respect that onlookers wonder why, if they can pull this off, don't they stay married?

The kids are grateful... But when she visits friends, and sees their parents, Marie said, she missed not having two parents at home... Many experts worry that birdnesting kids will fantasize about their parents getting back together. "Parents need to remind their kids that's not in the future," said lawyer Sutton.

Pediatricians of children of divorce

One interesting thing of note in the OurFamilyWizard.com service is a tool to keep track of the children's medical records, in particular vaccination records. The American Academy of Pediatrics Task Force on Medical Informatics has a policy statement "Special Requirements for Electronic Medical Record Systems in Pediatrics". Specifically in the context of taking care of children of divorce,

Data representation: growth data, parent identifiers.

Data processing: pediatric dosage calculation (e.g. Dr. Jeremy Adler's Pocket-Doc) incorporated into computerized physician order entry; immunization records with decision support (e.g. how to make up for missed shots); automatically generate customized reports (e.g. shcool sports physicals, documents for family court)

System design: distinction between biological parents vs legal guardians; ability to represent adopted children; links to other family members; links to registries (e.g. newborn screening, immunization)

Most recent statistics on divorce in America

Finally, some statistics. The popular number that many like to quote is that "50% of all marriages end in divorce". However, this estimate is generated cross-sectionally. That is, (I believe), it looks at the number of divorces within a year and divides it by the number of marriages taking place that year. However, while it is by construction a ratio, it is not a proportion, since it is really studying 2 different populations. More nuanced analyses are performed by demographers and we summarize some results here.

Based on Cycle 5 of the National Survey of Family Growth (NSFG), collected in 1995, the CDC published a report in 2002 titled Cohabitation, Marriage, Divorce, and Remarriage in the United States. By dividing marriages by the time period in which they start, i.e. marriage cohorts, it found that for marriages starting in the 70s and 80s, the chance of divorce within 10 years is about 30%.

Now, broken down by duration of marrige, we note that the chance of divorce is almost constant, at an annual rate of about 4% per year. This cannot hold true of course, or else all marriages will end in 25 years. But we see no leveling in this graph.

There are some interesting findings on cohabitation. The chance of a cohabitation breaking up also appears to be constant, at about 6% per year.

... the chance of a cohabitation advancing into marriage yet again appears more or less constant, at about 7% per year.

Interesting.

Technorati Tags:
medicine / pediatrics / divorce / cohabitation / marriage / parenting

Science - Assistance of Microbial Glycolipid Antigen Processing by CD1e (de la Salle, Mariotti, Angenieux et al 2005)

2005-11-26T23:41:00.000-08:00

Mozambique by Steve Evans via Flickr

Over view

Cell surface antigens CD1 A through E are expressed by professional antigen-presenting cells (APCs) for the presentation of non-peptide antigens to T cells. That is, they serve analogous functions as the MHC genes but are specialized in capturing and presenting a whole different class of antigens.

The MHC molecules are a large family of proteins. Each is different in its specificicity of antigen-capture, and most are highly polymorphic between individuals (hence matching of MHC, a.k.a. HLA, genotypes are performed prior to organ transplantation).

Moody et al 2005
in Nature Reviews Immunology

The CD1 genes are rather different. They are much less polymorphic between individuals and each molecule (CD1a through CD1e) binds a diverse set of antigens. Several crystal structures of CD1-antigen binding have been solved. The CD1 genes have been found to contain deeply buried hydrophobic surfaces to bind lipid antigens while exposing carbohydrate- and peptide-moieties for interaction with T-cell receptors. An example of phophotidylinosiol in the CD1b pocket is shown here.

The CD1 genes show differences in structure and function (reviewed by Moody et al 2005, ibid):

The antigen-binding grooves in CD1a function as a 'molecular ruler' that binds lipids with a defined alkyl chain length.

CD1b proteins can bind antigens with a lipid component of variable length. Similiar to the ability of MHC II to bind peptides longer than its binding groove, CD1b lets long antigens to protrude through portals to the outside of the antigen groove.

de la Salle, Mariotti, Angenieux et al 2005

This article in Science showed that

mycobacterial antigens hexamannosylated phosphatidyl-myo-inositols (PIM6) stimulate CD1b-restricted T cells only after partial digestion of the oligomannose moiety by lysosomal a-mannosidase and that soluble CD1e is required for this processing.

Polymorphisms in the CD1 gene gamily

Coming back to the point about the CD1 genes being functionally non-polymorphic. As summarized by Profs Manfred Brigl and Michael B. Brenner in Annual Reviews of Immunology, the few polymorphisms that have been found so far are either silent, amino acid changes outside the antigen-binding groove, or have and no detectable differences in expression or antibody reactivity. The exceptions are two CD1e alleles, CD1E*01 andCD1E*02, which did affect antigen-binding, but so far no association has been found between carrying these alleles and susceptibility to mycobacterial infections.

In general, it is much more difficult to make negative inferences in science(e.g., no crows are white) than it is to make positive ones (e.g., lots of crows are black). we check the dbSNP database of single-nucleotide polymorphisms and find that there are 10 SNPs in CD1e, many of which are validated. I will not be surprised if one day an association is found linking some of these polymorphisms with autoimmune diseases or diseases of chronic inflammation.

SNPs in the human CD1e gene, based on NCBI Map Viewer

Role in the immune system: immunity against Mycobacteria

deferred

CD1 gene family: role in disease

deferred

Lipid antigens and the delivery of dietary lipids: a convergence

In an Oct Nature article, researchers from Prof Michael B. Brenner's lab found that the immune system uses apolipoproteins to deliver lipid antigens to CD1. That is,

Apolipoprotein E mediates the presentation of serum-borne lipid antigens and can be secreted by APCs as a mechanism to survey the local environment to capture antigens or to transfer microbial lipids from infected cells to bystander APCs. Thus, the immune system has co-opted a component of lipid metabolism to develop immunological responses to lipid antigens

By Peter van den Elzen from article in Focus

From Focus, the news magazine for Harvard Longwood Campus,

"This may change the way we look at activation of T cells in asthma,� said Dale Umetsu of Children�s Hospital Boston, whose work has implicated lipid-reactive T cells in asthma and allergy in both mice and humans.

Technorati Tags:
medicine / Immune system / lipid antigen / Apolipoprotein / ApoE / Asthma / Structural biology / Chronic inflammation / Computational biology

Harbin, China, 2005 vs Valdez, Alaska, 1989

2005-11-26T07:52:00.000-08:00

FacePlant by idiottool from Flickr

The supertanker Exxon Mobil Valdez spilled between 11 million and 35 million gallons of crude oil. Take the middle value of 23 million gallons, assuming the density of crude oil to be 915 kg/m^3, this comes out to 915 * 42,000 = 38,430,000 kg of crude oil, or 38,430 tonnes. The Harbin benzene spill was 100 tonnes according to the BBC, which is 100,000 kg of benzene, less than 0.3% the Valdez spill by weight. Benzene is a lot more toxic however, and according to the BBC, this amount of benzene is "equivalent of 10 tanker-loads of lethal substances", all into the Songhua river. Tracing the BBC's reasoning, we note that 10 PANAMAX class tanker can carry up to 790,000 metric tons, hence the Harbin disaster spill is like a 20-super-tanker pile-up!. Ouch! I hope my math is wrong.

From the Wikipedia:

On March 23, 1989, the oil tanker Exxon Valdez departed from the Valdez oil terminal in Valdez, Alaska (on its 28th voyage), heading south through Prince William Sound, with a full load of oil. Captain Joseph Hazelwood radioed to the Coast Guard station that he would be changing course in order to avoid some growlers, small icebergs which had drifted into the sound from the Columbia Glacier... the vessel was not turning sharply enough and at 12:04 a.m. on March 24, the vessel hit Bligh Reef.

Hopefully the investigations into this disaster will yield a report as detailed and precise, to hold the responsible individuals accountable.

The Songhua River, just upstream of Harbin, China (to the right)

What about the people of Harbin?
In the NYT article The Half-Life of Anxiety on 7/10/2005 by Benedict Carey,

In studies of Alaskan communities that were affected by the oil spill from the tanker Exxon Valdez in 1989, and of towns dealing with water contamination in New Jersey and New York, sociologists have found what they call social corrosion. Sustained anxiety breaks down social groups and leads to an increase in mental health problems and potentially to economic downturn, said Lee Clarke, a sociology professor at Rutgers University and author of the forthcoming book, ''Worst Cases,'' an analysis of responses to disaster.

Much recent research analyzes mass disasters in terms of bioterrism, e.g., Facing the Possibility of Bioterrorism by Bill Durodi� in Current Opinion in Biotechnology:

In his [2004 book Therapy Culture: Cultivating Vulnerability in an Uncertain Age], Furedi explores the roots of a growing sense of social and individual vulnerability in what he coins 'therapeutic culture'. By increasingly framing problems through the prism of their emotions, people are actively incited to feel powerless and ill. Accordingly, �the spirit of stoicismand sacrifice�, along with �a sense of common purpose,unity or a commitment to fight� are now rarely in evidence. A powerful consequence of this, along with distorted perceptions and an increase in reported rates of depression, is provided by the phenomenon of mass psychogenic (or sociogenic) illness, numerous instances of which became evident in the aftermath of the anthrax attacks.

While the concept of "therapeutic culture" sounds reasonable and intriguing, it may be oversold here and takes on too much of a Theory of Everything tinge. I do like the basic idea though.

On an individual level,

Dr. Jerome Groopman's take on The Grief Industry in the New Yorker (on Dr. Jeff Mitchell's 7 steps)

A debate on psychological debriefing at British Journal of Psychiatry.

Technorati Tags:
China / Harbin /Benzene / Environment disaster

Science - Hematotoxicity in Workers Exposed to Low Levels of Benzene (Lan et al 2004)

2005-11-26T00:56:00.000-08:00

By langmuir via Flickr

A landmark study on benzene toxicity by a collaboration between the U.S. National Cancer Institute and the Chinese Center for Disease Control, Lan et al 2004 was an observational study comparing hematological changes in workers exposed to varying amounts of benzene vapor.

It demonstrated statistically significant hematological disturbances even for exposure less than 1 ppm, the current U.S. occupational standard.

It found benzene exposure to assocaited with decreases in hematopoetic stem cells in the blood, in a dose-dependent manner.

Here is Table 1 with just the comparison of CBCPD and flow cytometry results between the control group and the less than 1 ppm group. The other 2 exposed groups are 1 to 10 ppm (110 subjects) and greater than 10 ppm (31 subjects).

Here is Figure 1 comparing the amount of hematopoietic stem and progenitor cells in the circulation between subjects with 0, less than 10 ppm, and greater than 10 ppm of benzene exposure. As explained in the online supplement:

Colonies arising from the most primitive, early progenitor cells are called colony-forming-unit�granulocyte, erythroid, macrophage, megakaryocyte (CFU-GEMM) because the progenitors can give rise to any of these mature cells. Colonies derived from more committed progenitor cells that give rise to reticulocytes and erythrocytes are called burst-forming unit�erythroid (BFU-E), whereas those that give rise to granulocytes and macrophages are called colony-forming unit�granulocyte-macrophage (CFU-GM).

Notes on methods:

Levels of exposure were measured repeatedly for up to 16 months before hematological measures

Benzene and toluene air levels were estimated by "the arithmetic mean of an average of two measurements per subject collected during the month prior to phlebotomy" (online supplement)

Notes on stats:

Population heterogeneity examined with random effects model

Tests for trends (e.g., trends in changes in WBC count with respect to leveel of benzene exposure) were adjusted by the subject of the covariates { smoking, EtOH, recent infections, and BMI } that is significant (model selection method unclear? not detailed in online supplement either). Level of toluene exposure is also put in the model for continuous covariates since toluene competitively inhibits benzene metabolism.

Repeated measures adjusted with generalized estimating equation models

Technorati Tags:
China / Harbin /Benzene / Environment disaster / Toxicology / Hematology

NYT - The Deadly Shot

2005-11-26T00:16:00.000-08:00

The Confounded Fork by Duane Romanell via Flickr

NYT Editorial on the problem of neede re-use in poor countries leading to HIV and Hepatitis C infections.

As third-world health problems go, this one seems solvable. Single-use syringes, whose plungers break or are blocked after first use, cost about 6 cents apiece.

The biggest problem is that many poor countries are injection-crazy... patients demand injections because they think the medicine is stronger, and health care workers like to give them because they can charge more.

The Safe Injection Global Network, backed partly by the World Health Organization, is trying to help countries develop educational programs for health care workers and doctors.

Technorati Tags:
medicine / Public health / World Health Organization

Paul Krugman - "Bad for the Country"

2005-11-24T23:10:00.000-08:00

56buick by Mark Duerr of Flickr

NYT's Paul Krugman has been commenting on the need for a National Health Insurance in the U.S. He attributes GM's planned termination of 30,000 jobs to the disproportional cost of healthcare that American companies have to spend to cover health insurance for their employees.

... commentary [on G.M.'s troubles] from some conservatives has an unmistakable tone of satisfaction, a sense that uppity workers who joined a union and made demands are getting what they deserve.

We shouldn't be so complacent.

... If the United States had national health insurance, G.M. would be in much better shape than it is.

I need to make a clarification here... On a previous post on Prof Krugman's column A Private Obsession, I said,

One day, many generations from now, when all medical records become electronic, when every senior is well-informed of the risks and benefits of every medication for every disease that they may develop with known probability, the wisdom of the market will triumph. Oh yes it will.

Some have asked me if I said that in jest. First let me say that I know nothing about theories of economics other than what I learned from a political economics course I took in undergrad--about 8 years ago. Hence I don't know where to find reputable sources for support. What follows is all IMHO...

The market economy is known to arrive at efficient equilibriums without intervention. Since interventions are known in some cases to prevent the market economy from reaching that equilibrium, it is argued that the market should be left alone. Laissez-faire.

However while there is much talk on this nirvana of equilibrium, little attention is paid to the kinetics of market economy. That is, given a system of market economy, assuming that such equilibrium exists, how much time will it take to reach this (presumed) nirvana?

Certainly the fact that we have observed successful cases of market economy suggests the kinetics may be reasonable--within a human lifetime--at least in these isolated cases. The economic theorists in us wish to generalize these observations into a Grand Theory of Market Economy. But we cannot. We do not have the data to make this inference. In fact, the empiricists in us make the observation that millions of Americans are left uninsured. These empircists in us may have no health insurance themselves.

Theoretical generalizations are always be trumped by empirical observations.

Comments welcome.

In the same column, Prof Krugman comments on trade deficit:

The trade deficit isn't sustainable... we'll have to reorient our economy back toward producing things ... pulling a lot of workers back into manufacturing. So the rapid downsizing of manufacturing since 2000 - of which G.M.'s job cuts are a symptom - amounts to dismantling a sector we'll just have to rebuild a few years from now.

How much of that rebuilding will be in Michigan?

Technorati Tags:
evidence-based / Paul Krugman / Health care / Health insurance

AP - Toxic slick flows into major Chinese city

2005-11-24T21:22:00.000-08:00

Forbidden City Panorama
by Jesse Varner of Flickr.

According to the Medical Management Guidelines (MMGs) for benzene published by ASTDR.CDC, benzene is only slightly soluble in water and floats on the water surface. Ingestion of benzene can cause both acute benzene toxicity--CNS depression, light-headedness, headache, euphoria, respiratory depression, apnea, coma, and death--and

a burning sensation of the oral mucous membranes, esophagus, and stomach may occur after ingestion. Nausea, vomiting, and abdominal pain may also result from oral ingestion.

ASTDR's online case study on benzene toxicity, designed for PCPs (CME credits!).

Out of curiosity I queried ToxMap for industrial sources of benzene around Ann Arbor--not surprisingly there's a cluster of 4 sources in Wayne County.

According to Toxics Release Inventory, these 4 sources are:

MARATHON OIL CO (EPA #48174RMLSR28000): 445 lb/year

EQ RESOURCE RECOVERY INC (EPA #48174MCHGN36345): 250 lb/year

FORD MOTOR CO TRUCK PLANT (EPA #48184FRDMT38303): 132 lb/year

FORD MOTOR CO WAYNE ASSEMBLY (EPA #48184FRDMT37625): 60 lb/year

[update: BBC reports that the leak was 100 tonnes, "equivalent of 10 tanker-loads of toxic chemicals." 10 tonnes is about 220,000 lb, or about 250 times the environmental release of benzene in Wayne County, MI per year.]

People in China do not enjoy this kind of government oversight...

Protests have erupted in rural areas throughout China over complaints that pollution is ruining water supplies and damaging crops. Protesters often accuse officials of failing to enforce environmental rules either in exchange for bribes or for fear of hurting local business.
�This is the tip of the iceberg,� said Elizabeth Economy, director of Asia Studies at the Council on Foreign Relations in New York and author of the 2004 book �The River Runs Black: The Environmental Challenge to China�s Future.� �We�ve seen over the past six months or so a number of factory-related protests ... because factories don�t live up to or don�t enforce China�s own environmental regulations and laws,� she said. �So if, in fact, this is a case of that happening, then this is part of a much broader, systemic problem.�

Dr. Elizabeth Economy, a Wolverine, is Director for Asia Studies at Council for Foreign Relations.

China's environmental crisis is evident everywhere. The country's air quality is among the worst in the world: According to the World Bank, 16 of the world's 20 most polluted cities are on the mainland, and acid rain affects one-third of China's agricultural land. The country is already one-quarter desert, and that desert is advancing at a rate of 1,300 square miles per year.
The most serious environmental challenge, however, is providing clean water to the Chinese people: 60 million people have difficulty getting enough water to meet their daily needs and 10 times that many drink contaminated water on a daily basis.

A recent study published in Science found that benzene may have adverse health risks even at levels below currently accepted safety standards. Review pending.

Technorati Tags:
medicine / environment / China / benzene / Harbin / Elizabeth Economy

AP - South Korean cloning pioneer apologizes

2005-11-24T21:07:00.000-08:00

�I am very sorry that I have to tell the public words that are too shameful and horrible,� Hwang said, appearing downcast and solemn before a packed news conference. �I should be here reporting the successful results of our research, but I�m sorry instead to have to apologize.�

�Ethics and science are the two wheels that drive the civilization of mankind,� Hwang said. �Scientific research should be conducted within the boundaries of ethics but in reality, there were some cases in which the ethics regulations backing (quickly developing) science had not been in place.�

�The responsibility for all disputes and controversy lies on me,� Hwang said. �I will not make any excuse.�

Professor (Woo-suk Hwang), a national hero and a de facto celebrity in South Korea, will resign as head of the World Stem Cell Hub �to atone to the public.�

Extensive coverage at the official blog of the American Journal of Bioethics, including a very interesting unsigned comment

A comment from Korea... Through a series of media interviews and lectures, Prof Hwang has been strongly advocating that the development of stem cell therapy is just around the corner and it would save hundreds of thousands of lives.

Despite [the revelations] ... a considerable number of Korean citizens think that that's not an important issue at all. Because .... for them, Prof Hwang's research "IS" a cure for incurable diseases ... for them, his research "IS" a solution for the Korean economy. These people think that the societal debates about complex bioethical issues are unnecessary, if not harmful to the patients and the Korean economy.

It was also reported, via JoonAng Daily, that Professor Hwang's supporters have started a "I love Hwang Woo-suk" website, on which numerous women have already pledged to donate their eggs for stem cell reearch.

Technorati Tags:
medicine / stem cells / bioethics / politics in academia / Professor Woo-suk Hwang

Book: Mommy Knows Worst: Highlights from the Golden Age of Bad Parenting Advice.

2005-11-23T21:23:00.000-08:00

Friday is full of fun and activities.
By ahmedzahid.

(via Samantha Critchell of Chicago Sun-Times)
Review of James Lilek's Mommy Knows Worst : Highlights from the Golden Age of Bad Parenting Advice. Medical recommendations from the 40's and 50s included "advice from the Chicago Board of Health in the 1920s to give children sunbaths -- and then cod oil baths."

Unlike in physics, in medicine the different roles played by theorists and empiricists are not clear-cut. Here, tthe empiricists in us draw statistical conclusions from studies, but it is the theorists in us that make the scientific inferences and generalizations. IMHO it's best to tame these theorists in us, lest we lost the public's credibility.

Technorati Tags:
evidence-based parenting medicine pediatrics parenting

The Doctor's Office - Quality Rankings Can Work, For Patients and Doctors

2005-11-23T19:49:00.000-08:00

Atlantic from Above
By BombDog.

WSJ's Dr. Benjamin Brewer, a private practice doctor, shares his practice's report card,

The research is conducted through the Practice Partner Research Network, a joint effort whose participants include a medical software company and the Medical University of South Carolina, which designs and conducts the research. The university studies data from my electronic medical record on about 75-80 quality indicators ... compared with data from about 100 practices across 37 states.

We've been participating for the last six months and the process is still relatively new to us. This week I received my first big "report card."

Here's Dr. Brewer's report card

Since it is expected that practices with higher scores will attract more patients,

To balance supply and demand, the price for the services of top-performing physicians would have to go up in response to the good ratings -- but so far insurers and the government don't seem to want to pay the price when it comes to primary care.

Technorati Tags:
medicine report card quality-of-care health care

WSJ - In a Stroke Patient, Doctor Sees Power of Brain to Recover

2005-11-23T19:05:00.000-08:00

Based on "zoneplate - Magnolia"
by low resolution of Flickr.

Created with fd's Flickr Toys.

WSJ Page One news item today by Thomas Burton talks about a movement by some in Neurointensive Care to use scientifically reasonable (but still incompletely unproven) methods to treat comatose stroke patients. These are patients that according to current standards of practice are deemed incurable. Some like Dr. Stephan Mayer, Director of Columbia University's Critical Care Neurology, disagree.

They say many of the studies underlying the earlier consensus are out of date, and they believe newer treatments such as one designed to cool the brain may help stroke patients in comas. "Doctors are telling people there's no hope when, in fact, there is," says Dr. Mayer.

Dr. Justin A. Zivin at UCSD Stroke Center, who was instrumental in the development of tPA stroke therapy, responds:

"It would be extraordinarily helpful if they had evidence," he says. "Have they proven that these therapies are better than nothing? I'm not saying it won't ultimately be proven." Dr. Zivin says this is one reason many hospitals have yet to embrace the idea of a separate neurocritical care unit.

Dr. Mayer:

"We have to push the envelope and do things that at least are well-grounded in the scientific evidence that is available."

One such still-experimental treatment is "brain cooling", which we note is also being tested at UCSD's Stroke Center.

The Oxford Vascular Study published in this week's The Lancet found that, at least in the UK, the clinical burden of stroke has surpassed that of heart disease.

In men, the incidence of stroke is comparable to coronary artery disease (CAD), while in women the risk of stroke is significantly higher than the risk of CAD:

The risk of non-fatal (but potentially debilitating) stroke is also much higher than that of non-fatal (likewise debilitating) CAD:

As WSJ notes,

Aggressive treatment of stroke victims can have a serious downside. If a patient is kept alive for a few extra weeks in an intensive-care unit only to die at the end, the cost may be tens of thousands of dollars with no benefit. The American Stroke Association estimates that the annual U.S. cost of stroke care is $35 billion.

Technorati Tags:
medicine stroke evidence-based medicine vascular diseases brain cooling

NYT - Deaths After Abortion Pill to Be Studied by Officials

2005-11-23T17:19:00.000-08:00

...conva(lumin)escence.
By underbunny.

Gardiner Harris of NYT:

Ms. Patterson died [on Sept. 17, 2003] seven days after taking Mifeprex. She lived in Livermore, Calif.

On Dec. 29, 2003, Vivian Tran, 22, of Costa Mesa, Calif., died six days after taking Mifeprex.

On Jan. 14, 2004, Chanelle Bryant, 22, of Pasadena, Calif., died six days after taking Mifeprex.

And on May 24, 2005, Oriane Shevin, 34, of Los Angeles died five days after taking Mifeprex.

In each case, Clostridium sordellii infected the women's uteruses, flourished and then entered their bloodstreams. The bacterium can cause nausea, vomiting, diarrhea and weakness but may not induce fever ... antibiotics [treatments after the onset of infection] are often ineffective ...

No similar deaths have been reported in Europe, where Mifeprex is widely used. But in the United States, most physicians give Mifeprex and an accompanying drug, misoprostol, in a regimen that involves inserting misoprostol vaginally.

Combing through ~~the stacks of UM's Taubman Library~~ Google Book Search on Clostridium sordelli reveals:

Postpartum episiotomy infection beneath the deep facsia leads to myonecrosis. Clostridium perfringens accounts for most cases, but Clostridium sordelli has also been implicated. These cases were unique in presenting with "massive malignant vulvar edema thought to be caused by toxin production and results in death from cardiovascular collapse" (Clinical Maternal-Fetal Medicine by Hung N Winn, John C Hobbins)

Clostridium sordelli has also been reported in pleuropulmonary infections after penetrating chest trauma. Of the 39 cases of such infections due Clostridium reported from 1935 to 1993, 2 cases (5%) were attributed to Clostridium sordelli. (Community-Acquired Pneumonia by Thomas J Marrie)

It has been implicated in neonatal omphalitis (infection of the umbilical stump) in deliveries outside traditional medical facilities and in cultures where cow dung is applied to the umbilical stump postpartum. (The 5-minute Pediatric Consult, Edited by M. William Schwartz)

Interestingly, antiserum to Clostridium sordelli cross-reacts and neutralizes Clostridum difficile cytotoxin (a.k.a. Toxin B) in tissue culture. This forms a component of the laboratory diagnosis of intestinal C. diff overgrowth in patients on antibiotics. (Notes on Medical Microbiology by Yifan Douglas Yang, Morag C Timbury, A Christine McCartney, Bishan Thakker)

That little literature search was completed within 20 minutes. Unlike regular Googling, the reliability of these sources are not in doubt. The importance of Google Book Search in Science 2.0 cannot be denied.

FYI this was on my search results page

Technorati Tags:
medicine Science 2.0 Google Book Search Microbiology Abortion Mifeprex Clostridium sordelli

NEJM - . . . And a Diagnostic Test Was Performed (Greenwald 2005)

2005-11-21T03:14:00.000-08:00

By the Infinite Cat Project

A correspondence from Dr. Robert A. Greenwald, describing a recent case conference during which a patient with IPEX was diagnosed by a fellow in attendance using not her clinical acumen per se but Google.

Where does this lead us? Are we physicians no longer needed? ... Even worse, the Google diagnostician might be linked to an evidence-based medicine database, so a computer could e-mail the prescription to the e-druggist with no human involvement needed. The education of house staff is morphing into computer-search techniques. Surely this is a trend to watch.

Search Google Book Search (formerly Google Print) with "IPEX immunodeficiency"...

Do we even need medical libraries? From my laptop, I can access publication quality, full color, annotated dermatology pictures, and pathology slides (and the entire BluePrint review book series)! Indeed Google Book Search turned out quite useful the other day when we admitted a patient with a strange rash.

Technorati Tags:
medicine Science 2.0 The World Is Flat

Science 2.0 - Peer-reviewed publications

2005-11-20T09:51:00.000-08:00

By alterednate from Flickr

Robert Boynton gives us a cautionary tale of then-University of Chicago political science junior faculty Daniel Drezner. He was denied tenure because part of his academic output had been directed into the blogosphere.

The current antipathy toward blogging may have something to do with the fact that universities have no tools for judging blogs. And most people agree that blogs would need to be evaluated through some kind of peer-review mechanism if they are to be taken into account. "It is utterly absurd to propose giving someone credit for activity with no barriers to entry," [John Holbo, the Editor of The Valve] says.

So, how might a blog be peer-reviewed? The market provides a number of viable models. eBay, for one, has established an efficient rating system for buyers and sellers, based on the number and quality of transactions they execute. In a noncommercial medium, Slashdot uses a "Moderation and Meta Moderation System," in which moderators are awarded higher or lower "karma" according to how well they police the discussions on the site. (The "Meta Moderation System" judges the moderators' moderators.)

An even better model is Kuro5hin.org, which I think has successfully adapted the peer-review process to the Web. The manuscript submission process should sound familiar:

The author contribue articles by formatting the manuscript (or "stories" in the world of Kuro5hin) with respect to a standard recommendation.

The author submits the manuscript to the edit queue, along with plans for editing the manuscript

A subset of Kuro5hin users are selected to be editors of this particular mansucript. They see a link to the manuscript in their Kuro5hin "moderation" pages. At this point, the text is not visible to the rest of the Kuro5hin community.

Editorial commments are visible to the author, who can respond to the comments, make changes to the manuscript, and even rate the usefulness of each individual editorial comment

Editors vote on acceptance of this manuscript

Assume it is accepted, the manuscript becomes an article. Another subset of Kuro5hin users are select to vote on the placement of the article: on the Frontpage or in one of the Sections (e.g., Politics, Technology, Culture, etc).

Now the entire Kuro5hin community can read the article and comment.

Each reader can rate other readers' comments. The author can likewise respond to the questions and comments.

Possible improvements:

Allow editors to tack on metadata to the text. This seems a natural task for editors, who will be doing fact-checking as a part of the editing process.

Allow for multiple authors per manuscript for collaborative efforts.

Anonymous editorial process to encourage frank discussions.

Giving some structure to the comment system, making it easier to track questions that have been answered by the author and those that have not.

As noted by many, a credit system for rewarding editorial work and constructive comments.

John Holbo's post on this on The Valve.

Technorati Tags:
Science 2.0 peer-review

Science 2.0 - How To Guide to CiteULike

2005-11-19T20:32:00.000-08:00

By City Kayak from Flickr

As the world flattens, people-to-people and machine-to-machine communication becomes standardized and efficient, how do we as scientists find new ways to plug into this new infrastructure and collaborate with each other in the 21st century?

An article at OpenWetWare talks about the emergence of Science 2.0. See the accompanying brainstorming Wiki for even more inspiring ideas.

Web1.0-->Web2.0
Britannica Online-->Wikipedia
...
Science 1.0-->2.0

Some drawbacks of the current situation
There is only one main level of information dissemination: peer-reviewed, highly polished summaries of work in journals.

...
Current publishing approaches do not encourage open feedback and reviews of work.

Pubmed --> Faculty of 1000 --> Cite-U-Like

What is CiteULike? Like del.icio.us, CiteULike is a social bookmarking service. But it is specifically tailored to academic discussions. Mechanisms are in place to automatically create records of journal articles, tag them, and discuss them with other users of CiteULike. Connotea is a similar service hosted by the Nature Publishing Group. But the CiteULike website is more responsive, its user interface more intuitive, and its comment system more amenable to public discussions.

I have created groups in CiteULike where we could review and discuss recent articles in evidence-based medicine and randomized clinical trials methodology.

To get started:

Install the bookmarklet on this page appropriate for your machine. Make sure to bookmark it into a toolbar for easy 1-click access.

Go to the PubMed page of the article you want to comment on, and click the bookmarklet button

Tag this post, leave public or private notes, and even upload your PDF file for later access.

Go to the CiteULike groups page to join discussion groups. If you'd like, join the groups Evidence-based-medicine, Randomized-clinical-trials-methodology, and/or Randomized-clinical-trials-review

Technorati Tags:
science 2.0 The World is Flat

NEJM - Sildenafil Citrate Therapy for Pulmonary Arterial Hypertension (Galie et al 2005)

2005-11-19T13:45:00.000-08:00

background
Sildenafil inhibits phosphodiesterase type 5, an enzyme that metabolizes cyclic guanosine monophosphate, thereby enhancing the cyclic guanosine monophosphate�mediated relaxation and growth inhibition of vascular smooth-muscle cells, including those in the lung.

methods
In this double-blind, placebo-controlled study, we randomly assigned 278 patients with symptomatic pulmonary arterial hypertension (either idiopathic or associated with connective-tissue disease or with repaired congenital systemic-to-pulmonary shunts) to placebo or sildenafil (20, 40, or 80 mg) orally three times daily for 12 weeks. The primary end point was the change from baseline to week 12 in the distance walked in six minutes.

This was an impressive international collaboration involving 53 centers in the United States, Mexico, South America, Europe, Asia, Australia, South Africa, and Israel. "Local institutional review boards or independent ethics committees approved the protocol, and written informed consent was obtained from all patients." One may raise the issue that with so many different investigators at different medical centers in different countries using different equipments, are the measurements really comparable? Perhaps there are some systematic differences in their measurements that are not seen because each center on average contributed only 5 patients? I believe that in this case since both the primary outcome (baseline walking distance) and the major baseline characteristic (WHO functional class) are so straightforward to implement, systematic differences, if any, should not be severe (sorry no data to back this one up). The same may not be said of the many secondary hemodynamic outcomes (e.g., mean pulmonary arterial pressure) that are also presented in the paper.

As a trial designed to evaluate the efficacy and tolerability of 3 different doses of sildenafil, the question of multiple comparisons invariably arises--that is, given all the possible statistical tests we could make between the outcomes of patients receiving 0, 20, 40, or 80 mg of sildenafil, how do we control for alpha, the chance that whatever trend is discovered is attributable to chance alone? The statistical solution is to control for the experiment-wise error rate using sequential step-down, closed testing procedure.

A stratified central-randomization scheme was used to assign patients to four treatment groups � those receiving 20, 40, or 80 mg of sildenafil or placebo three times daily � in a 1:1:1:1 ratio. The randomization was stratified with respect to the baseline walking distance (<325 m or >=325 m) and cause of pulmonary arterial hypertension.

Missing data for assessments at week 12 were imputed with the use of the last-observation-carried-forward method.

The change in mean pulmonary-artery pressure and World Health Organization (WHO) functional class and the incidence of clinical worsening were also assessed, but the study was not powered to assess mortality. Patients completing the 12-week randomized study could enter a long-term extension study.

results
The distance walked in six minutes increased from baseline in all sildenafil groups; the mean placebo-corrected treatment effects were 45 m (+13.0 percent), 46 m (+13.3 percent), and 50 m (+14.7 percent) for 20, 40, and 80 mg of sildenafil, respectively (P<0.001 for all comparisons). All sildenafil doses reduced the mean pulmonary-artery pressure (P=0.04, P=0.01, and P<0.001, respectively), improved the WHO functional class (P=0.003, P<0.001, and P<0.001, respectively), and were associated with side effects such as flushing, dyspepsia, and diarrhea. The incidence of clinical worsening did not differ significantly between the patients treated with sildenafil and those treated with placebo. Among the 222 patients completing one year of treatment with sildenafil monotherapy, the improvement from baseline at one year in the distance walked in six minutes was 51 m.

As Dr. Hayward et al proposed in Health Affairs, RCTs should present the effects of treatment stratified with respect to multivariate risk models. This allows decision-makers to evaluate the usefulness of the treatment in the common currency of risks and benefits.

In this case, while no multivariate risk prediction model exists for pulmonary arterial hypertension, baseline 6-minute-walk distance and WHO class are of prognostic value. A stratified analysis shows:

conclusions
Sildenafil improves exercise capacity, WHO functional class, and hemodynamics in patients with symptomatic pulmonary arterial hypertension.

Technorati Tags:
medicine statistics clinical trials dose-response pulmonary arterial hypertension sildenafil

NEJM - Two-year outcomes after conventional or endovascular repair of abdominal aortic aneurysms (DREAM Trial Group 2005)

2005-11-18T09:08:00.000-08:00

A multicenter randomized clinical trial of 351 patients diagnosed with abdominal aortic aneurysms (AAAs) >5 cm who received either endovascular repair or open surgery of their AAA. The primary outcome of interest was perioperative survival.

At 1 year of follow-up, we observe an impressive advantage of endocascular repair:

This difference vanished, however, at 2 years of follow-up:

Just another demonstration of the perils of early RCT termination for benefit. The conclusion would have been very different had the DREAM researchers terminated the trial for observed benefit at 1-year of follow-up.

Note that I'm making an almost surely wrong assumption here, that every subject was enrolled at the same time and followed simultaenously, such that at 1-year interim analysis an exaggerated benefit was observed. In fact patients are enrolle on a rolling basis and each is followed until the patient is lost-to-follow-up or until the planned follow-up time is reached. This subtle difference in timing introduces statistical problems for Cox proportional hazards model..

Dr. Hans van Houwelingen at Leiden Univ Medical Center showed in a recent article in Statistics in Medicine, due to the model assumptions of Cox proportional hazards model, when it is used in interim analysis, it may be estimating the final treatment effect with bias.

Review pending.

Technorati Tags:
medicine statistics clinical trials interim analysis

Paul Krugman - A Private Obsession

2005-11-17T20:22:00.000-08:00

Earlier this year Pennsylvania Senator Rick Santorum proposed bill to forbid the National Weather Service from competing with private weather forecasters in the U.S. (in particular AccuWeather of Pennsylvania)

Although [Senator Santorum] didn't say so explicitly, he wanted the service to funnel that information through private forecasters instead... it was a classic attempt to force gratuitous privatization: involving private corporations in the delivery of public services even when those corporations have no useful role to play.

The Medicare drug benefit is an example of gratuitous privatization on a grand scale.

One day, many generations from now, when all medical records become electronic, when every senior is well-informed of the risks and benefits of every medication for every disease that they may develop with known probability, the wisdom of the market will triumph. Oh yes it will.

[update: please also see post on Prof Krugman's 11/25 column]

Technorati Tags:
medicine health insurance Paul Krugman

Pediatrics - Do Pacifiers Reduce the Risk of Sudden Infant Death Syndrome? A Meta-analysis. (Hauck et al 2005)

2005-11-15T23:17:00.000-08:00

A meta-analysis of 7 case-control studies looked at the association between SIDS and pacifier use.

Based on 4 studies that provided multivariate ORs controlling for a variety of factors including sleeping position, usual pacifier use was associated with a significant reduced risk of SIDS (Summary OR: 0.71 [95% CI: 0.59�0.85]):

Pacifier use during last sleep (in cases) or reference sleep (in controls) showed a stronger and more consistent association with SIDS protection:

This meta-analysis is a basis for the American Academy of Pediatricians's new recommendations on SIDS protection, which encouraged pacifier use at nap time or bed time.

Technorati Tags:
pediatrics evidence-based medicine evidence-based parenting evidence-based medicine meta-analysis

AAP - The Changing Concept of SIDS

2005-11-15T22:35:00.000-08:00

American Academy of Pediatrics's new policy statement on SIDS: "The Changing Concept of SIDS: Diagnostic Coding Shifts, Controversies Regarding the Sleeping Environment, and New Variables to Consider in Reducing Risk"

SIDS is defined as

The sudden death of an infant under 1 year of age, which remains unexplained after a thorough case investigation, including performance of a complete autopsy, examination of the death scene, and review of the clinical history.

The incidence of SIDS has been going down...

But note the concomittent rise in the incidence of sudden unexpected infant death (SUID), especially after year 2000. SUID is a unexpected death with attributable cause, such as accidental suffocation in bed, accidental poisoning, assault and homicide, etc. But as we can appreciate from the graph this increase in SUID does not account for the fall in the rate of SIDS.

The rate of SIDS has indeed been going down, even if some of this progress may be attributable to changes in coding. Based on this encouraging trend and new research findings, the AAP recommends (among others):

A separate but proximate sleeping environment is recommended such as a separate crib in the parent�s bedroom. Bed sharing during sleep is not recommended.

Consider offering a pacifier at nap time and bedtime: The pacifier should be used when placing infant down for sleep and not be reinserted once the infant falls asleep.

Avoid commercial devices marketed to reduce the risk of SIDS: Although various devices have been developed to maintain sleep position or reduce the risk of rebreathing, none have been tested sufficiently to show efficacy or safety.

Do not use home monitors as a strategy to reduce the risk of SIDS: There is no evidence that use of such home monitors decreases the risk of SIDS.

Avoid development of positional plagiocephaly (flat back of head): Encourage �tummy time.� * Avoid having the infant spend excessive time in car-seat carriers and �bouncers.� Place the infant to sleep with the head to one side for a week and then changing to the other.

Technorati Tags:
pediatrics evidence-base parenting SIDS AAP

WSJ - New Advice on Getting Babies to Sleep; Dr. Ferber Softens Stance On Letting Infants Cry

2005-11-15T21:58:00.000-08:00

Link goes to free summary by UPI:

Longtime polar opposites -- pediatricians Richard Ferber and William Sears -- have softened their stances on getting a crying baby to fall asleep.
Ferber, in an update of his landmark 1985 book "Solve Your Child's Sleep Problems" that is due in spring, will back off his position that a child should be allowed to cry untouched until he or she falls asleep. ...
At the same time, Sears in his just-released "The Baby Sleep Book" says parents need to consider their own rest and not necessarily respond to their crying baby day or night.

The Wall Street Journal:

These shifts are among the latest in a flurry of new guidance that seems aimed at offering practical solutions for parents -- neither too harsh nor too permissive.

Empirical validation is the key

Whole Language Movement sounded good in theory when it was proposed in the 70s--until it was carelessly and prematurely adopted nationwide, producing a generation of children with difficulty reading.

If inhibiting COX1 causes gastric ulcers, selectively blocking COX2 with Vioxx then indeed seems like a reasonable idea for targeted pain control.

If menopause is characterized most prominently by decreased female hormones, estrogen replacement therapy surely would correct the undesirable effects of menopause.

Technorati Tags:
within normal limits of reason pediatrics evidence-based parenting evidence-based medicine

GAO - FDA: Decision Process to Deny Initial Application for OTC Marketing of the Emergency Contraceptive Drug Plan B Was (gasp!) Unusual

2005-11-14T18:02:00.000-08:00

(via The Health Care Blog, link goes to original GAO report.)

The Plan B decision was not typical of the other 67 proposed prescription-to-OTC switch decisions made by FDA from 1994 through 2004. The Plan B OTC switch application was the only one during this period that was not approved after the advisory committees recommended approval. The Plan B action letter was the only one signed by someone other than the officials who would normally sign the letter.

The Acting Director acknowledged to GAO that considering adolescents� cognitive development as a rationale for a not-approvable decision was unprecedented for an OTC application, and other FDA officials told GAO that the rationale differed from FDA�s traditional practices.

Small, but significant victory for Science.

Technorati Tags:
medicine FDA Plan B politics

Sci Am - Bugs and Drugs: Gut bacteria could determine how well medicines work

2005-11-14T09:30:00.000-08:00

This Scientific American story refers to this paper in Toxicological Sciences by Lora Robosky et al at Pfizer Global Research.

Robosky et al found that, surprisingly, mice that are genetically identical may metabolize chemicals differently based on variations in their gut microbial flora. The study of bacterial colonizers in the gut is a recent development.

Through genomic sequencing ... Paul Eckburg at the David Relman Lab at Stanford estimates at least 400 species in our gut.

... Scientists have only just begun to elucidate how these mysterious bugs influence health... Two years ago, for example, David G. Binion of the Medical College of Wisconsin showed that the sodium butyrate produced by gut microorganisms could inhibit blood vessel growth in the intestine by blocking COX-2--an enzyme implicated in many inflammatory disorders (and the target of drugs such as Vioxx). Other studies have shown that specific strains of Escherichia coli can metabolize dimethylarsine, a derivative of arsenic, to produce potentially toxic compounds, which may underlie the carcinogenicity of arsenic in the gut... the discovery may partly explain why data from presumably identical animals sometimes conflict.

The is Gene x Environment interaction at its best. Recall this famous experiment in 1999 which was demonstrated the absolute difficulty in controlling and holding constant the effect of the environment on genetically identical mice.

By altering our response to medications and environmental chemicals, gut flora may in fact have non-ignorable effects on our behavior--not unlike "parasitic hairworm charming grasshopper into taking it for a swim".

Technorati Tags:
biology neuroscience

Paul Krugman - Health Economics 101

2005-11-14T00:33:00.000-08:00

NYT's Krugman:

We rely on free markets to deliver most goods and services, so why shouldn't we do the same thing for health care? ... It comes down to three things: risk, [adverse selection] and social justice.

Insurance is a process of risk-sharing amongst a group of individuals. In adverse selection, those people who can save money by buying insurance--ie, those who are more likely to use more health care resources--are more likely to do so, since it is clearly to their advantage. At a large scale, however, these high-risk individuals can overwhelm the risk-sharing scheme by shifting their health care costs onto healthy individuals. What health insurance companies do to counter this imbalance is screening:

This screening process is the main reason private health insurers spend a much higher share of their revenue on administrative costs than do government insurance programs like Medicare, which doesn't try to screen anyone out. That is, private insurance companies spend large sums not on providing medical care, but on denying insurance to those who need it most.

An article in this month's Health Affairs by James Kahn, Richard Kronick, Mary Kreger, and David Gans found that in California, billing and insurance paperwork alone constitute about 20% health care spending.

Technorati Tags:
medicine health care health insurance paul krugman

NYT - For City Kept Sleepless by Colic, No End to Cures in Melting Pot

2005-11-13T14:00:00.000-08:00

NYT's Nina Bernstein:

...in this city where 6 of 10 babies have at least one foreign-born parent and pediatricians come from every corner of the world, a cornucopia of colic cures serves as a kind of Rorschach test of child-rearing culture in migration.

The most recent review of infantile colic I could locate, in an article in The Journal of Pediatrics by Michelle Garrison and Dimitri Christakis, concludes:

Four of the interventions studied had data of adequate quality and statistically significant numbers needed to treat (NNT): hypoallergenic diet (NNT = 6), soy formula (NNT = 2), reduced stimulation (NNT = 2), and herbal tea (NNT = 3).

Notably reduced stimulation was found in a 1991 article in the Archives of Disease in Childhood. The full text is NA online but from the review:

In 1 RCT, 93% of infants whose parents were advised to reduce stimulation improved, as opposed to 50% of those in the control group (RR 5 1.87; 95% CI 5 1.04�3.34).

There are several methodological issues that Garrison and Christakis identified:

First, the case definition of colic was highly subjective, which may have led to inclusion of infants with considerably milder symptoms ... Second, parental diaries as a means of assessing treatment benefits are inherently more subjective than the unbiased assessments by study investigators used in some trials. Third, as a behavioral trial, the study was not double-blinded.

This runs contrary to Dr. Karp's suggestion of LOUD SHUSHING to simulate blood flow in the abdominal aorta in utero. I wish someone would perform a RCT on the effectiveness of Dr. Karp's 5 S's.

Technorati Tags:
pediatrics evidence-based parenting

Subgroup analysis and post hoc analysis

2005-11-13T05:00:00.000-08:00

In reference to the last post, Dr. James Gaulte said...

Could you explain how risk stratified analysis avoid the problem(s) of post hoc sub group analysis namely the multi-comparison problem of false positives and reduced power of the subgroup(s) to show a real difference.

I believe a distinction should be made between subgroup analysis and post hoc analysis. (What follows is all IMHO. Comments welcome)

Post hoc analysis--the more infamous of the 2--refers an exploration of the data that is both unplanned and potentially unlimited, yielding misleading results that are merely superficially significant. It is misleading because the conventional p-value cutoff of 0.05 is no longer valid in the face of repeated testing. It is as if we flipped a coin until we observed 5 heads in a roll, and reported this finding by itself.

Dr. Pocock has proposed the use of interaction test, a rather conservative test, for post hoc analysis.

I think of post hoc analysis as an optimization of the Results section with respect to the Methods section of a paper. This is in itself not problematic as long as we recognize its exploratory nature.

Subgroup analysis refers to an investigation into whether the treatment effect is different in a subset of the patients defined by some baseline characteristic. Subgroup analysis may be exploratory or confirmatory in nature. It may be a priori or a posteriori. For example, investigators may plan, ahead of a trial on heart failure, to analyze the subgroup of patients in NYHA Class IV at baseline.

Nevertheless, subgroup analysis is problematic in that by dividing the patient population into smaller sets, it loses statistical power.

What Hayward et al is proposing is in my opinion a priori subgroup analysis of RCTs. By stipulating that RCTs incorporate stratified analysis in terms of commonly used risk categories, they in effect are requiring investigators to plan to study their treatment with respect to pre-defined subgroups. In fact they make the explicit distinction from �one-variable-at-a-time� post hoc subgroup analysis.

To answer Dr. Gaulte�s concerns, then, I believe Hayward et al�s stipulation of subgroup analysis is a priori and not exploratory. Hence it is not prone to problems associated with post hoc analysis.

A priori or a posteriori, subgroup analysis suffers from reduced statistical power. Hayward et al suggests that

Even for small studies that have marginal statistical power, risk-stratified analysis will still be valuable for comparing results between different studies or conducting meta-analyses.

In addition, as an a priori analysis, investigators are in good position to design their RCTs to endow it with sufficient power to detect treatment differences with respect of risk-stratified subgroups.

References: Pocock 2002.

Comments welcome.

Technorati Tags:
medicine statistics clinical trials subgroup analysis

Health Affairs - Reporting Clinical Trial Results To Inform Providers, Payers, And Consumers (Hayward et al 2005)

2005-11-10T20:55:00.000-08:00

(via How to prevent another Vioxx-like situation)

In this month's Health Affairs, Drs. Rodney Hayward, David Kent, Sandeep Vijan, and Timothy Hofer proposed that the analysis of benefits and risks in RCTs should always include a component in which patients with high risks are analyzed separately from those at low risks. That is to say, in making health policy decisions, we need to know the individual differences in the risk-vs-benefit balance at baseline and present our findings accordingly. For instance, whether or not a medication like Vioxx should be given to a patient will depend on his/her pre-existing risk of developing stomach ulcers. RCTs on Vioxx then should include subgroup analysis that investigate the its effect on high ulcer risk patients and low ulcer risk patients separately.

However,

The main results of clinical trials are presented as the average benefit across all people in the trial. ... [Risk-stratified analysis] could overcome some of the limitations inherent in conventional approaches to analyzing and reporting clinical trials.

With respect to what do we stratify our patients?

Multivariable risk prediction models/tools use statistical formulas that estimate individual risk based on combining information from multiple independent risk factors.

(For example, APACHE to predict mortality in critically ill patients, APGAR to predict risk of adverse outcomes in neonates, and BODE to predict 1-year mortality of patients with COPD.)

Finally, the meat of the article is a review of recent RCTs to see how prevalent is such stratified analysis in RCTs. Hayward et al studied the subset of all RCTs published in the year 2001 in major medical journals (JAMA, NEJM, and Lancet) which used as primary endpoints mortality or major morbidity. Of these 108 RCTs, 42 (39%) did not report subgroup analysis. Only 4 (4%) studies performed risk-stratified analysis, and of these only 1 used statistically robust methods.

As an example of the risk-stratified analysis they have in mind, they point to this
RCT comparing early invasive intervention vs conventional intervention in patients presenting with acute coronary syndrome (Cannon et al 2001).

In this figure, abstracted from the article's Figure 2, multiple stratified analyses are shown. Those on top are "single-variable stratifications", each based on an individual variable (in this case age, gender, history of prior MI). The bottom analysis is Hayward et al's risk-stratified analysis, in this case with respect to TIMI score for unstable angina/non-ST elevation myocardial infarction. This figure makes clear that according to this RCT, early revascularization is benefitial only in those with high risk for adverse cardiac outcomes. It is Hayward et al's thesis that such stratified analyses facilitate decision-making by patients, healthcare providers, insurance companies, and regulatory agencies.

Finally, they propose that CONSORT incorporate risk-stratified analyses in irs RCT guidelines, that the FDA mandate the use of risk-stratified analyses in RCTs, and--given such individualized estimates of expected risks vs benefits--

We propose ... intermediate steps for our policy decisions. For example, instead of deciding whether coverage of a treatment should be zero or 100 percent, we could adjust copayments based on the amount of expected benefit (instead of basing copayments on the cost of treatment alone).

Technorati Tags:
medicine statistics clinical trials risk-stratification

Reviewd by an established medical blog!

2005-11-10T20:50:00.000-08:00

Thanks to Dr. John Ford from UCLA for a generous review of WNL Reason! This really bumped the site traffic up and inspired me... I'll be wrapping up a couple of new blogs tonight!

Technorati Tags:
medicine medblogs

JAMA - Randomized Trials Stopped Early for Benefit (Montori et al 2005)

2005-11-09T02:23:00.000-08:00

After reviewing 143 RCTs stopped early for benefit:

RCTs stopped early for benefit are becoming more common, often fail to adequately report relevant information about the decision to stop early, and show implausibly large treatment effects, particularly when the number of events is small. These findings suggest clinicians should view the results of such trials with skepticism.

Accompanying editorial by Dr. Stuart Pocock, a statistics researcher specialized in RCT monitoring.

Generally speaking, in RCTs, allowing nature enough time to take its course is important for us to sort out variations due to chance alone from variations due to our intervention. Here is a heuristic argument: When estimating a quantity, say the average weight of 1-month-old babies, precision is inversely proportional to the square root of the number of subjects. When estimating the probability of developing an event, the precision is inversely proportional to the length of the observation itself (and not the square root).

Another issue that arises with repeated interim analysis is that the more we look, the more likely we're to be fooled by apparently impressive variations that are due to chance alone. There is a lot research in statistics on sequential methods which allow for interim analysis of clinical trials while controlling type I and II errors.

A watched pot never boils. Or not.

Technorati Tags:
medicine statistics clinical trials interim analysis JAMA

Paul Krugman's "Pride, Prejudice, Insurance" and Taiwan's National Health Insurance

2005-11-06T20:25:00.000-08:00

To establish National Health Insurance in the US,

we'll have to overcome pride - the unwarranted belief that America has nothing to learn from other countries - and prejudice - the equally unwarranted belief, driven by ideology, that private insurance is more efficient than public insurance.

As an example, Krugman points to the Taiwanese National Health Insurance:

Taiwan, which moved 10 years ago from a U.S.-style system to a Canadian-style single-payer system, offers an object lesson in the economic advantages of universal coverage. In 1995 less than 60 percent of Taiwan's residents had health insurance; by 2001 the number was 97 percent. Yet according to a careful study published in Health Affairs two years ago, this huge expansion in coverage came virtually free: it led to little if any increase in overall health care spending beyond normal growth due to rising population and incomes.

Taiwan's NHI first planned in 1987. The original plan was to implement the system in the year 2000. However, in the mid 90s, with Taiwan's political system becoming more democratic and with the first direct democratic election of the president around the corner, both the ruling and the minority parties felt the pressure to pass popular social programs. The originally NHI plan was quickly (and some might say hastily) re-done so as to start in 1995, one year before the 1996 presidential election. The NHI program came finally came into existence in March 1995, less than one month after the legislature finally sorted out NHI's financing details.

This is how a 1997 article in Health Policy summed up Taiwan's experience with NHI up to that point:

What lessons can be learned from Taiwan's experiences, especially for countries facing a large uninsured population and rapidly rising health care costs?

Economic and political forces, rather than of health needs, brought about the forced introduction of Taiwan NHI

The revolution was initiated at a time when the economy was healthy and health expenditure modest

The responses of the health care system, of the previously insured population, and the newly-insured population were not what the Taiwan NHI Bureau expected.

More review of Taiwan'sNHI to come.

Technorati Tags:
medicine national health insurance taiwan

Nature Biotech - Pessimistic response to FDA leadership change

2005-11-05T22:04:00.000-08:00

Dr. Kaitin, Director of the Tufts Center for the Study of Drug Development:

I think the agency has left a period of uncertainty, and has now entered an era of instability.

In September, Dr. Lester Crawford resigned from his post as the commissioner of the FDA. He was replaced by acting commissioner Andrew von Eschenbach, who was the 12th Director of the National Cancer Institute, and continues to be affiliated with it. Given his most recent initiative at NCI--unveiled with great fanfare--of making cancer a managable disease by 2015, some in the biomedical industry have expressed concerns that Dr. Eschenbach may steer the FDA according to his agenda as a cancer scientist.

The image of the FDA has already been severely tarnished: SSRIs have been linked with suicide in children, COX2 inhibitors with cardiovascular complications, and the non-approval of the post-coital contraceptive Plan B with frank political intervention. The FDA has already shown a tendency to be overly conservative in the drug approval process.

What these all add up to, I think, is a re-newed focus of drug research on cancer therapeutics. Cancer drug research is a relatively straightforward area of pharm research. Cancer patients, out of options and desparate, self-refer to clinical trials. Risk-to-benefit calculus is easily dominated by the patients' imminent deathes. These medications are given only briefly, and long-term adverse effects are not technically considerations unless the medication in fact works. But in my opinion, some of the most exciting drug developments in recent years have been things like SSRI, statins, ARBs, etc. The focus on oncology merely will give the agency a chance to lay low and play it safe for a few years--hampering advancement in other areas of applied biomedical research in the process.

Kaitin predicts much more codevelopment and many more licensing agreements between small biotechs and big pharmas. In the uncertain environment of the next couple of years, he believes that few biotechs will be able to transform themselves into large companies that market their own products, the way companies like Genentech and Amgen did.

Technorati Tags:
medicine fda pharmaceutical

BMC Medical Informatics - Pretest probability assessment derived from attribute matching (Kline et al 2005)

2005-11-03T22:50:00.000-08:00

A patient comes into the ED with the complaint of acute chest pain. How do we assess the chance that the patient has something potentially lethal like acute coronary syndrome (ACS)?

Currently we rely on the experience and acumen of doctors with decades of training under their belts. No doubt these clinical judgments will outperform any system of automated patient assessment. But such gut-feelings are often difficult to quantify, and when clinicians disagree, it boils down to a matter of opinion. The Principle of Defensive Medicine kicks in, and our patient stays in the ED for a full night of tests. So what exactly is the probability of ACS given the patient's presentation? Having an objective, numerical estimate based on the collective medical records can assist clincians in making difficult decisions. Clinical acumen can be supplemented with the indisputable evidence of the electronic medical records database.

Typically pretest probabilities are generated using logistic regression models. This is a parametric model for modelling probabilities of binary outcomee, such as the probability of having a myocardial infarction and the risk of adverse outcome given UA/NSTEMI (TIMI score)). A member of the Generalized Linear Models family of parametric models, logistic regression has a number of convenient statistical properties. However as Kline et al noted,

From the perspective of probability estimation for acute disease, one of the main drawbacks to logistic regression equations is that they seldom output a pretest probability in the very low (0�5%) range ... this is the range where the clinician must decide whether or not to use the resources required for formal testing by a chest pain protocol.

In addition,

Additive risk in logit scale, an assumption of logistic regression, may be easily violated and is difficult to validate

Complex interactions between clinical variables are difficult to model and to assess with logistic regression.

Given the gigabytes of data in our collective electronic medical records databases, we have the freedom to adopt less parametric methods. Kline et al uses a database of 14,796 emergency department (ED) patients (the i*trACS study) evaluated for possible ACS at 8 hospitals (7 in US, 1 in India). Each patient record was associated with 70 clinical variables. To find the most informative clinical variables, Kline et al used classification and regression tree analysis (CART). This is a relatively intuitive statisitical method that makes no distributional assumptions. It constructs a binary tree recursively. At each stage, it divides the group into 2 subgroups using the variable that can best distinguish the subjects with respect to the classification variable.

This is analogous to seuqentially cutting a cake into 2 halves, 4 halves-of-halves, and 8 halves-of-halves-of-halves, and so on. Quick intuitive isn't it?

Kline et al built a CART tree using this database, and yielded 8 clinical variables of predictive value:

Age (<35, 35-38, 39-50, >50 years)

Gender

Race (white or asian, nonwhite and non-asian)

Patient report or physician observation of sweating with symptoms

Patient report of a prior history of coronary artery disease or myocardial infarction

Chest pain worsened with manual palpation on physical examination

STD >-0.5 mm in any two leads
T wave inversion >-0.5 mm in any two leads

This CART tree, which they call attribute matching, was validated in a separate population of 8,120 patients presenting at the UCSD and UPenn EDs with suspected ACS. (Table of validation results vs logistic regression model).

While this result is not unexpected (given the large database of patient records, the more flexible, nonparametric CART approach not surprisingly gave a better fit of the data and was better at predicting the risk of ACS), it demonstrated the clinically important strength of CART in identifying low-risk patients:

The sensitivity and specificity of the very low-risk designation for the detection of acute coronary syndrome at 30 days was 95.3% and 25.4% for attribute matching, versus 99.3% and 3.7% for the logistic regression method.

This subset of patients are clinically most important to identify objectively and quantitatively to bolster clinical decision making:

In a 1997 multicenter study, Graff and colleagues found that only 2.5% of patients evaluated by a chest pain protocol were diagnosed with acute myocardial infarction. In our experience, this "rule-in rate" is declining.

A potential application of the present system would be use of the combination of a pretest probability <2.0%, and one negative biomarker of cardiac ischemia or necrosis, in conjunction with the patient's risk tolerance to prevent unnecessary chest pain protocol evaluation

A user-friendly interface was also developed by the authors for clinical application.

Some comments. What Kline et al did not address was the most undesirable aspect of CART: the instability of the results. As you can imagine, since each step depends on all of the cuts made previously, small variations in data leading to a different choice of cut at an early stage may result in progressively amplified differences in all subsequent stages. A statistical solution is the technique of bagging.

We note that a very useful characteristic of CART--endowed by its inherent flexibility--is the ease with which it can accomodate missing data, which we know is all too common in clinical databases.

We note that nonparametric does not mean the absence of modeling assumptions. Specifically, independence of observations is still assumed in building the CART tree.

For review of statistical classification methods see The Elements of Statistical Learning by Hastie, Tibshirani, and Friedman 2001.

Technorati Tags:
medicine statistics medical informatics clinical decision making CART

Dr. Harvey Karp - Happiest Baby/Toddler on the Block

2005-11-03T21:32:00.000-08:00

There was a lecture at the UMHS today by Dr. Harvey Karp on how to calm babies and how to communicate with toddlers.

I read his books while on my pediatrics rotation. He is impressive in his claims and his use of scientific reasoning. In particular at his talk today he gave a number of research literature references from prominent pediatrics journals. In time I will get to review these papers. For now, I will say that the one idea I did not like in The Happiest Toddler on the Block is the invocation of "ontogeny recapitulates phylogeny", which has been shown to be--at least in the original form proposed by Haeckel--false.

Overall however I like his ideas, particularly regarding calming babies. He places our current practice of parenting in the historical context of American medicine and compares against parenting in other cultures--some of which have no concept of infant colic. He also makes a good argument that being able to effectively calm babies is crucial for the health of the baby and the mother (in terms of shaken-baby syndrome, post-partum depression, martital discord, etc). I actually tried some of his techniques during my pediatrics rotation, without much success. Today during the talk however a baby sitting in the back would not stop crying, and Dr. Karp demonstrated--live--his "5 S's" and indeed calmed the baby down in seconds. These techniques were similarly effective in the DVD clips he showed during the talk. Impressive.

Technorati Tags:
evidence-based parenting parenting

Medscape News - Presence of Heparin Antibodies Predicts Morbidity After Cardiac Surgery (Kress et al 2005)

2005-11-02T19:57:00.000-08:00

Dr. David Kress, a thoracic surgeon at Aurora St. Luke's Medical Center in Milwaukee, Wisconsin, reported at the American College of Chest Physicians annual meeting:

Testing positive for heparin/platelet factor 4 (HPF 4) [antibody] is an independent predictor of complications associated with cardiac surgery, with or without heparin-induced thrombocytopenia (HIT), according to a retrospective study presented here at a late-breaking session during CHEST 2005, the annual meeting of the American College of Chest Physicians.

This was a retrospective study of 1114 patients at Aurora St. Luke's who underwent cardiopulmonary bypass.

[The 60 patients (5.4%) who tested positive] had an elevated incidence of acute limb ischemia (5.0% vs 0.9%; P = .03) and a higher incidence of renal dialysis (11.7% vs 4.3%; P = .02), gastrointestinal complications (15.0% vs 5.7%; P = .01), and prolonged ventilation exceeding 96 hours (21.7% vs 8.8%; P = .01).

Some of the weaknesses of this study is of course the retrospective design. The presence of HPF4 antibody may be associated with higher risk simply because it is a marker for higher co-morbidity.

Previously HPF4 antibody is associated with the development of HIT Type 2. Typically these cases arise 5 to 10 days after initiation of heparin anti-coagulation, and come to attention by the dramatic 30% to 50% drop of platelet levels. The disturbance of the intricate balance in the coagulation system can be lead to the formation of lethal arterial or venous clots.

There are isolated reported cases of HIT associated with mere IV heparin flushes (Heparin-induced thrombocytopenia (HIT) due to heparin flushes: a report of three cases by Kadidal et al in Journal of Internal Medicine, 2001.)

Currently, HIT is a "clinicopathological" diagnosis. That is, it is made "most confidently" when the patients exhibit an otherwise unexplained episode of thrombocytopenia and has serum HPF4 antibodies. Dr. Theodore Warkentin has a detailed article on the diagnosis of HIT. It also provides a nomogram for diagnosis appropriate for post-cardiac surgery patients only.

While the treatment of HIT in immediate discontinuation of heparin and replacement with substitutes such as Lepirudin, Argatroban, and Danaparoid. However, as noted in the Medscape article, there is not sufficient data on the use of these alternatives in cardiopulmonary bypass.

This report is remarkable in suggesting an adverse effect of HPF4 antibody in the absence of the salient signs of HIT. A search through PubMed revealed that a case-control study of patients in a Pediatric ICU also showed an analgous finding. Specifically, it found an association between having anti-HPF4 antibody levels and developing thrombosis, in the absence of thrombocytopenia. (no time to read original article, sorry)

By let me re-iterate: Some of the weaknesses of [the Kress et al study] is of course the retrospective design. The presence of HPF4 antibody may be associated with higher risk simply because it is a marker for higher co-morbidity. I'm looking forward to reading the article when it gets published to see how this is accounted for.

Technorati Tags:
medicine statistics surgery new hypothesis diagnosis heparin-induced thrombocytopenia

JAMIA - Health IT and Physician-Patient Interactions: Impact of Computers on Communication during Outpatient Primary Care Visits

2005-11-02T15:58:00.000-08:00

(via Dr. Barron Lerner of NYT - The Computer Will See You Now (Feel Better?))
From the Journal of the American Medical Informatics Association:

The examination room computers appeared to have positive effects on physician-patient interactions related to medical communication without significant negative effects on other areas such as time available for patient concerns. Further study is needed to better understand [health information technology] use during outpatient visits.

This is a single-center, before-and-after comparison of patient perception, at 2 months before, 1 month after, and 7 months after installation of an electronic health medical records in the exam room. 8 physicians at the center, a part of the Kaiser Permanente-Northwest system, took part. 4 were FPs and 4 were IMs. 313 patients completed the questionnaires. This longitudinal data was analyzed using Generalized Estimating Equations (GEE).
The EHR appears to be a custom job by Epic Systems.
At 7 months, there were significant increases in

patient overall satisfaction (OR=1.5, 1.01-2.22)

patients' satisfaction with physician's familarity with patients (OR=1.60, 1.01-2.52)

comprehension with decisions made during the visit (OR=1.63, 1.06-2.50)

At 7 months, no significant changes in

patient satisfaction with comprehension of self-care responsibilities
communication about psychosocial issues

available visit time

It's somewhat curious that the ORs of some or these findings are barely and marginally statistically significant. But no matter. This is the most important finding:

By seven months post-introduction, patients were more likely to report that the computer helped the visit run in a more timely manner (OR = 1.76; 95% CI: 1.28�2.42) compared with the first month after introduction.

BTW,
The Association of Medical Directors of Medical Informatics (AMDIS):

Since its founding in 1997, the Association of Medical Directors of Information Systems is the premier professional organization for physicians interested in and responsible for healthcare information technology.

medicine medical informatics

NYT - Parasitic Hairworm Charms Grasshopper Into Taking It for a Swim

2005-11-02T09:50:00.000-08:00

Nicholas Wade (NYT Science Desk):

The parasite, known as a hairworm, lives and breeds in fresh water. But it spends the early part of its life cycle eating away the innards of the grasshoppers and crickets it infects.

When it is fully grown, it faces a difficult problem, that of returning to water. So it has evolved a clever way of influencing its host to deliver just one further service -- the stricken grasshopper looks for water and dives in.

The original article Biron et al 2005 in Proceedings of the Royal Society Series B. Note that this was studied with a proteomics approach.
The larva of the wasp Hymenoepimecis spp, a parasite of the orb-weaving spider, somehow convinces its host to weave a unique and durable platform for the larva's cocoon (Eberhard 2000 in Nature, a fascinating article with pics).

The parasite Toxoplasma gondii has a unique, seemingly paradoxical cycle involving the cat and the mouse. In order to accelerate this process, it alters the olfaction of the host mouse, such that the mouse is drawn to the smell of cats. This is specific and does not disturb the rest of the mouse host's olfaction. (See Robert Sapolsky article in Scientific American, also collected in his latest book Monkey Luv).

Finally, something to think about is the recent finding of fetal microchimerism in the brain. This is a jaw-dropping study from the National University of Singapore, Chinese Pharmaceutical University, Osaka University, and The Institute of MCB in Singapore. In this study, Tan et al 2005 in Stem Cells showed that in the pregnant lab mouse, fetal cells find their way into the maternal circulation, take residence in the mother's brain, and grow into neuron-like or glia-like cells. In some regions of the mother mouse's brain as much as 1/1000 of the neurons are of fetal origin. I will not be surprised if in a follow-up study these neurons and glia are found to be participate in maternal CNS function, long after the fetus has been delivered.

Bill Cosby:

"No matter how calmly you try to referee, parenting will eventually produce bizarre behavior, and I'm not talking about the kids. Their behavior is always normal."

medicine neurbiology parenting

NYT - China's Little Green Book (Thomas Friedman)

2005-11-01T23:28:00.000-08:00

Thomas Friedman from Beijing:

In China, conservation is not a "personal virtue," as Dick Cheney would say. Today it is a necessity. It was so polluted in Beijing the other day you could not make out buildings six blocks away. That's the bad news. Here's the good news: China's leaders and business community know it.

Green technology will decide whether China continues on its current growth path or chokes itself to death. So green innovation is starting to mushroom in China.

A study by the Chinese Academy on Environmental Planning, found that in the year 2003, 300,000 people in China died from outdoor pollution, while 111,000 died from indoor pollution.

china environment pollution smoking cancer

Flickr's Jerome Chen - Taipei 101 in Sunset

2005-11-01T03:11:00.000-08:00

Flickr Taipei 101 pics
The tallest building in the world, for now (WikiPedia)

NEJM - The BMI, Airflow Obstruction, Dyspnea, and Exercise Capacity Index in COPD (Celli et al 2004)

2005-11-01T01:30:00.000-08:00

Since I'm on the Pulmonary Service this month, thought I'll review some recent Pulmonary papers.

This prospective study consisted of 2 phases. In the first phase, 207 patients enrolled from 1995 to 1997 were assessed. Their baseline characteristics were studied for their ability to predict one-year mortality. A forward step-wise multiple logistic regression model was built, which contained 4 baseline variables (BODE):

Obstruction: FEV1 post-bronchodilator treatment

Dypnea: MMRC dyspnea sclae

Eercise tolerance: 6-minute walk distance

This model was then validated in a cohort of 625 patients rolled from 1997 to 2003. Their baseline BODE characteristics were recorded, and they were then followed for at least 2 years or until death. The ability of the BODE index to predict the hazard of death (either all-cause or from respiratory illnesses) was assessed with the Cox proportional hazards model. They conclude that:

For each 1-point increase in BODE index, a 34% increase in hazard of all-cause death (26%-42%, p<0.001)

For each 1-point increase in BODE index, a 62% increase in hazard of respiratory-cause death (48%-77%, p<0.001)

The BODE index was found to be a much better prognostic factor than FEV1 alone.

One can raise the reasonable objection that since FEV1 is a component of the BODE index, it is not surprising that BODE is a better predictor of death. Celli et al quantified the incremental improvement provided by BODE using the C-statistic, based on the area under the ROC, and found a substantial difference (0.74 vs 0.65). More on this later...

Another objection may be that the 3 variables besides FEV1 simply measured the overall state of health of the patient. As such, the worse one's health was to begin with, the higher risk of death was expected. Certainly this argument must to true in principle. After all, all physiological or functional assessments reflect some aspect of "overall state of health". Celli et al used regression to assess whether BODE provides information that quantitatively captures the important aspects of overall state of health. They made use of the Charlson index, which quantifies the degree of co-morbidity. In a Cox proportional hazards model, the Charlson index was not a significant predictor of death from respiratory illnesses when BODE score is also included. That is, the Charlson index does not provide prognostic information above and beyond that contained in the BODE index. The BODE index has successfully captured important prognostic imformation. Since BODE is focused on cardiopulmonary health, it is not surprising that the Charlson index did have small but marginally significant prognostic value for predicting all-cause mortality.

As the authors noted themselves, there are significant differences between patients from the US (N=348), Venezuela (N=54), and Spain (N=223). One would like to have seen an analysis restricted to the US patients and non-US patients to assess the rebustness of their conclusion.

The BODE Calculator from ICU Medicus is fun to play with.

ACP Journal Club review

medicine statistics pulmonology COPD

California Medicine Man - A brand new technology for distinguishing between viral and bacterial meningitis?

2005-11-01T00:58:00.000-08:00

California Medicine Man has a post on a fascinating new technology to distinguish bacterial from viral meningitis. When I was on General Pediatric Service, I admitted a number of febrile newborns (a few weeks old). Since bacterial meningitis in neonates cannot be easily ruled out--no matter how unlikely--we generally admitted the babies and performed the-whole-package work-ups which included blood culture, urine culture, and lumbar puncture. These tests take 2 to 3 days to come back, necessitating an inpatient stay for the babies and their families. While all of us wait anxiously for the results, we start empirical antibiotics therapy to cover common bacterial causes of neonatal bacterial meningitis, just in case. There are clnical guidelines for estimating the probability of neonatal bacterial meningitis based on parameters that can be obtained earlier in the course (eg Spanos et al 1989 and Nigrovic et al 2002). While helpful they do not allow us to comfortably send the babies home before all the the test results are finalized...

The unfortunate consequence of this conservatism is a very high cost associated with admitting many patients with viral disease, treating them with expensive, potentially toxic antibiotics and needlessly scaring the bejesus out of them and their families. Wouldn't it be great if there was a highly accurate and very rapid test for discriminating these different disease entities?

Metabonomics using NMR:

Metabonomics to the rescue. No this isn't some failed theory of language education. As one definition I got from the internet, metabonomics is: "The quantitative measurement of the dynamic multiparametric metabolic response of living systems to pathophysiological stimuli or genetic modification".

Original article by Coen et al 2005 in Clinical Infectious Diseases.

computational biology medicine stiatistics pediatrics meningitis diagnostics metabonomics omics

Flickr's alterednate - Absorbed

2005-10-31T15:23:00.000-08:00

alterednate:

This is on Belle-Ile, at a famous spot called "Port Coton." The sky is certianly real, however sometimes when I look through my shots I wonder myself.

New journal: Public Library of Science Clinical Trials

2005-10-30T19:28:00.000-08:00

Public Library of Science has a new journal for clinical trials:

Each published paper in PLoS Clinical Trials will be linked to its corresponding entry in the relevant registry. PLoS is collaborating with Global Trial Bank (GTB), a non�profit subsidiary of the American Medical Informatics Association, to ensure that trial results are captured and stored in a computer�readable, standardized format.

According to the editors,

Publication decisions will not be affected by the direction of results, size or perceived importance of the trial.

This should help mitigate publication bias.
What does this mean for researchers? Manuscripts will have to follow the CONSORT structure:

We intend to go beyond CONSORT (a tool developed to improve the quality of reporting of randomized trials; http://www.consort-statement.org/), asking authors to not only submit a CONSORT checklist and flow diagram, but also to organize their papers according to the CONSORT structure. Readers will be able to quickly identify where in the paper they need to look to find out about a particular aspect of the design.

Interestingly, peer review will also be a bit atypical:

Rather than making recommendations about acceptance or rejection, peer reviewers of papers submitted to PLoS Clinical Trials will be asked to focus on improving the quality and transparency of trial reporting. Each trial report will be accompanied by an editorial summary of its strengths and weaknesses, including what it adds to current scientific knowledge. Readers will have the opportunity to post comments.

I really like this idea. Perhaps PoLS Clinical Trials will one day set the standard for publishable RCTs?
(via Microarray and Bioinformatics Blog)

medicine statistics clinical trials journal

PubMed Assistant (PuMA), the PubMed front-end of your dreams!

2005-10-30T13:12:00.000-08:00

What I like the most about this program:

Small and fast to load
Ability to save complex searches and their results
Breaks down complex query strings hierarchically and visually
Displays Outlook-style the list of results, the abstracts, and the MeSH terms

My only gripe, in my limited experience with it, is the sometimes-awkward arrangement of the UI. Also a few keyboard shortcuts would be nice. Oh yes and in this Web 2.0 era, perhaps a central site of storage of my search results. (hint hint, Google)

medical informatics bioinformatics

Science - Recurrent Fusion of TMPRSS2 and ETS Transcription Factor Genes in Prostate Cancer (Tomlins et al 2005)

2005-10-29T13:23:00.000-07:00

In this week's journal Science, by Dr. Tomlins et al:

We used a bioinformatics approach to discover candidate oncogenic chromosomal aberrations on the basis of outlier gene expression.

We identified recurrent gene fusions of the 5' untranslated region of TMPRSS2 to ERG or ETV1 in prostate cancer tissues with outlier expression.

By using fluorescence in situ hybridization, we demonstrated that 23 of 29 prostate cancer samples harbor rearrangements in ERG or ETV1. Cell line experiments suggest that the androgen-responsive promoter elements of TMPRSS2 mediate the overexpression of ETS family members in prostate cancer.

This is a beautiful piece of computational biology with significant clinical implications. More review of the methods and of the ETS transcription factor family (which I encountered in grad school) to follow...

computational biology medicine statistics cancer transcription regulation

Arch Int Med - Efficacy of Bupropion and Nortriptyline for Smoking Cessation Among People at Risk for or With COPD (Wagena et al 2005)

2005-10-27T19:22:00.000-07:00

Bupropion SR is known to be effective in helping smokers quit. Smoking cessation is particular important for patients suffering from COPD, since this is the only things that is known to slow the progression of the disease. This RCT studied the effectiveness of bupropion SR for smoking cessation in patiens with COPD or at risk for it. This is a well-crafted study that enrolled 255 patients, with an impressive less than 5% lost to follow-up at 26 weeks.

Some interesting findings:
- Amongst patients with stages I, II, and III COPD, 27% (12 of 44) of those on bupripion SR vs 8% (4 of 48) of those on placebo remained abstinent from week 4 to week 26 (chi-square p-value 0.02). NNT = 5.3 patients with COPD. (Note that COPD stratification was initially done using the European Respiratory Society definition, but presented with according to the GOLD guideline)
- Amongst depressed patients, 35% (8 of 23) of those on bupropion SR vs 11% (2 of 19) of those on placebo remained abstinent from week 4 to week 26 (Fisher's exact test p-value 0.08... this can be seen as statistically significant because of the exactness and discreteness of Fisher's test). NNT = 4.1 patients at risk for COPD or with COPD who have depression. (Depression was defined as Beck Depression Inventory score <= 15, which is associated with having clinical depression)

Arch Intern Med -- Effect of a Clinical Trial Alert System on Physician Participation in Trial Recruitment (Embi et al 2005)

2005-10-27T18:52:00.000-07:00

Patient referral and recruitment are important rate-limiting steps of clinical trials. It is not surprising that electronic health records (EHRs) can automate some of the mechanical steps involved and remind busy practicing physicians that the patient they are seeing may qualify for a particular trial.

In this study, a clinical trial alert (CTA) system was built on top of an existing EHR, and the rates of patient referral and recruitment before and after implementation was compared. They found:
- 10x increase in the number of referrals made by physicians who referred This effect was seen not only at sites that previously did not refer, but also at sites that previously generated the bulk of the referrals.
- 2x the patient enrollment rate With an EHR in place, researchers can determine the eligibility of referred patients directly on their computers. This simplifies the process and allows them to use lowered thresholds for screening. While this raises false-positive rate, it also tends to increase recruitment rate as well (Bell-Syer et al 2000).

accompanyin editorial

Molecular Medicine - Molecular Identification of Simian Virus 40 Infection in Healthy Italian Subjects by Birth Cohort (Paracchini et al 2005)

2005-10-27T13:21:00.000-07:00

It is not disputed that polio vaccines manufactured from 1954 to 1963 were contaminated with the polyomavirus SV40. However there has been accusations that the vaccine made from monkey kidney cells continue to be in part contaminated, and that the virus is responsible for certain types of cancer (eg malignant mesothelioma) based on animal studies. Epidemiological studies of humans have been inconclusive as a whole. (for a more-than-likley biased review see The Virus and the Vaccine).

Even if current batches of vaccines continue to be potential sources of SV40 contamination, and even if SV40 virus indeed causes cancer in a subset of people infected, an important question had thus far been unanswered--is the vaccine the only potential source of SV40?

This study studies the prevalence of SV40 in donors in Italy who had no known cancer. Using the peripheral blood lymphocytes of 134 solid organ donors (all anonymous) they screened for SV40 using PCR and then confirmed it with DNA sequencing. They found 15 of these (11%) to contain de facto SV40 genome sequences. If indeed the contaminated polio vaccine was capable of infecting its receipients, then we would expect those donors who were born between that period of time would be more likely to be infected with SV40. This was not found to be the case.

One major drawback is the small sample size. But I see this as the beginning of a very important new line of research. At the very least it should settle the SV40-vaccine debate, if not rendering it moot.

Technorati Tags:
medicine / vaccine / statistics / polio / SV40

BMJ - Mortality associated with passive smoking in Hong Kong (McGhee et al 2005)

2005-10-27T11:35:00.000-07:00

There's plenty of evidence for the association between cigarette smoke on one hand and lung cancer, coronary heart disease, and stroke on the other. This study, based on retrospective study of Hong Kongese patients, confirmed again the association of passive smoking with stroke, COPD, and lung cancer. The does-response curve is that much more dramatic in this particular study perhaps due to the crowded living conditions not atypical for urban cities in Asia.

We note that China, with 20% of the world's population, smokes 30% of the world's cigarettes! (sorry can't locate original source)

This is a follow-up to their 2001 case-control study:

In the general population of Hong Kong in 1998 tobacco caused about 33% of all male deaths at ages 35�69 plus 5% of all female deaths, and hence 25% of all deaths at these ages.

In the male smokers tobacco caused about half of all deaths at ages 35 to 69

Thomas Friedman - Living Hand to Mouth

2005-10-27T11:20:00.000-07:00

Thomas Friedman on the increasing awareness in China of the environment and pollution.

Recent article in New England Journal of Medicine on major causes of death in China. Both vascular diseases and cancer have become leading causes of death.

A study by the Chinese Academy on Environmental Planning, found that in the year 2003, 300,000 people died from outdoor pollution, while 111,000 people died from indoor pollution each year.

Combination Therapy: The Future of Medical Management for PAH

2005-10-27T10:58:00.000-07:00

Will review...

NEJM -- Trajectories of Growth among Children Who Have Coronary Events as Adults

2005-10-27T10:57:00.000-07:00

Growth trajectories of 8760 people born in Helsinki from 1934 through 1944 are associated with outcome as adults.

Accompanying editorial has a nice review to other studies of similar spirit.

But how generalizable is this particular result? The Finland population is rather genetically homogeneous due to small founder population.

medicine statistics pediatrics growth velocity

NYT - Scare Yourself Silly, but the Real Terrors Are at Your Feet

2005-10-24T23:22:00.000-07:00

From Dr. Zuger's essay

"I'm really scared about Lyme disease," she said. "I really need to get treated."
"If you want to be scared, how about that untreated AIDS of yours?"

The Observer - Sue the Lancet!?

2005-10-24T19:56:00.000-07:00

Nick Cohen suggests:

The editor of the Lancet is a more tempting target. Wakefield's original research was based on a sample of just 12 children, which was too small to be meaningful, as the Lancet ought to have known. Medical journals are not the richest of institutions, however, and it would probably take only a couple of thousand single jab bills to close the Lancet down.

via blacktriangle

Wakefield's "research" is largely responsible for an entire cohort of British children not receiving the MMR vaccine.

I prefer holding the editors responsible.

In the Pipeline - The Tar Pit Beckons.

2005-10-23T22:05:00.000-07:00

"The short version: (1) Pfizer's ever-increasing size means that most everything scales up except what they need the most: research productivity. (2) Billion-dollar drugs must roll off their conveyer belt, one after the other, and that's something that no one has ever figured out how to do. (3) Lipitor, mighty monster that it is, is the main thing keeping the music playing. But it will go away, and there is nothing to replace it. Perhaps nothing ever could. (4) While a massive sales force is quite a thing to have, they do need things to sell, don't they?"

Roche AmpliChip Video

2005-10-23T21:39:00.000-07:00

Through Personal Genome blog:

"Roche has posted a video describing their Amplichip CYP450 test, a DNA
microarray that can identify 2D6 and 2C19 gene variants...."

Annals of IM - Prevention of Exacerbations of COPD with Tiotropium, a Once-Daily Inhaled Anticholinergic Bronchodilator

2005-10-23T20:22:00.000-07:00

This multi-center, double-blinded RCT of 26 VAs compared the effects of inhaled tiotropium QD versus placebo in patients with COPD, in terms of frequency of exacerbations and hospitalizations.

Patients are:

> 40yo
Cigarette smoking hx >= 10 pack-years
Clinical Dx of COPD
FEV1 <>
Free of significant co-morbidities or other pulmonary diseases such as asthma.

...

JAMA - Effect of Muraglitazar on Death and Major Adverse Cardiovascular Events in Patients With Type 2 Diabetes Mellitus

2005-10-23T17:39:00.000-07:00

In the Pipeline's Dr. Derek Lowe has a interesting post on this. He noted that in the FDA briefing document submitted by GlaxoSmithKline, a noticable association of CV adverse effects could be seen.

His suspicion was confirmed when JAMA rushed out the current article which points out the true magnitude of the CV adverse effect. The accompanying editorial explained that CV risk estimate was diluted by the inclusion of low-dose (hense low-risk) groups. It also points out some other potential sources of bias:

- Selecting a study population unlikely to have adverse outcomes but nonrepresentative of potential future users (eg, exclusion of elderly patients, even though more than one third of type 2 diabetes occurs in this group)

- Conducting underpowered studies increasing the failure rate to detect meaningful safety differences (ie, maximizing rather than minimizing type II errors)

- In contrast to efficacy determinations, reporting individual rather than composite safety outcomes to decrease the likelihood of establishing statistical significance (eg, separate cardiovascular events from CHF)

- Limiting preapproval peer-review publication of results so as to minimize scrutiny and debate of both methods and results (eg, of all submitted data only 1 study of 340 patients has been published8)

- Evoking biological implausibility of safety concerns by the use of surrogate measures (eg, treatment reduces C-reactive protein [CRP]) implying safety, despite no proof that CRP reduction is clinically correlated with improved safety)

- Recording outcomes only in patients who are fully compliant with prescribed treatment because this self-selected group will likely have fewer adverse events (eg, unknown impact of the nonanalysis of the 15% discontinued cases)

- Ignoring the totality of the evidence by excluding consideration of confirmatory safety signals seen in studies of similar molecules (eg, CHF and bladder cancer outcomes with pioglitazone)

- Diverting attention to unproven but potential benefits by concentrating on reductions in surrogate laboratory values (eg, hemoglobin A1C) rather than in meaningful patient health outcomes.

Very intersting....

The Cutter Incident

2005-10-23T15:27:00.000-07:00

Dr. Paul A. Offit from U Penn writes about the Cutter Incident. Soon after the Salk vaccine was shown to effectively immunize children against the polio virus in the 50s, pharmaceutical companies around the country started to produce the vaccine. Sadly, the vaccines produced by Cutter Laboratories contained live viruses, making thousands of vaccinated children ill, and 200 children paralyzed. As Dr. Offit explains, many of the problems with today's vaccines, such as frequent shortages and lack of new development, can be traced to the legal and political aftermath of this unfortunate incident.

Note that Dr.Offit is a patent-holder on the rotavirus vaccine. It was withdrawn after post-market surveillance showed statistically significantly increased occurrence of intussuception.

He also has an article in the May/June issue of Health Affiars on vaccines.

Review pending.

Technorati Tags:
Pediatrics /
Vaccines /
Paul Offit

NEJM - KCLIP: Incidence and Outcomes of Acute Lung Injury (ALI)

2005-10-23T10:16:00.000-07:00

A prospective population-based study of King County in Seattle, the 13th largest county in the US with > 1.8 million residents. During the period from 4/1999 to 7/2000, a total of 1,113 King County residents >15yo who underwent mechanical ventilation met the American-European Consensus Conference on ARDS definition of ALI.

Only arterial blood gas measurements obtained while the patient was intubated
were assessed. The arterial blood gas values with the worst PaO2:FiO2 ratio,
regardless of the positive end-expiratory pressure, were used to assess
oxygenation for each 24-hour period.
Because the PaO2: FiO2 ratio becomes an increasingly unreliable assessment of shunt when the FiO2 is below 0.40, the ratio was used to assess oxygenation only when the FiO2 was 0.40 or more.

Risk factors present during the four days before the onset of ALI: severe sepsis (sepsis with a suspected pulmonary source, sepsis with a suspected nonpulmonary source or an unidentified source), severe trauma (with an Injury Severity Score > 15), witnessed aspiration, transfusion of > 15 units of blood within a 24-hour period, drug overdose, pancreatitis, near-drowning, and inhalation injury.

Outcomes:
In-hospital mortality, ICU LOS, etc.

Analysis:
Pearson chi-square, Mann-Whitney, t-test.

Conclusions
1. The incidence of ALI is 2.5~5.0x as high as previously believed, with age-adjusted incidence of 86 /100,000 person-years. Mortality is about 39%.
2. The incidence of ARDS is 2~40x as high as previously believed, with age-adjusted incidence of 64 /100,000 person-years. Mortality is about 41%.
3. The most common risk factor for developing ALI is severe sepsis with pneumonia (46%), f/b severe sepsis with suspected nonpulmonary source (33%).

Accompanying editorial raises 2 points worth considering:

Why is the incidence in this study seemingly higher than in others? Is this possibly a reflection of US Hospitals overusing ICU beds?
While the main outcome measured was in-hospital mortality, post D/C morbidity is high and the logistics of care need to be carefully arranged.

medicine statistics pulmonology ARDS ALI

Tagging with BlogThis! - Freshblog

2005-10-23T00:47:00.000-07:00

Tagging with BlogThis! - Freshblog

Filed in: blog

HCPro: Hospitals Vary in Spending, Costs for Overseas-Based Radiology Services: Financial News - Yahoo! Finance

2005-10-22T23:46:00.000-07:00

Press release from HCPro,

Twenty-five percent of the 133 hospitals HCPro surveyed use overseas-based radiologists, while 14% said they are strongly considering outsourcing MRI and CT preliminary reads by 2007.

From April 2005 in Washington Post

A 2003 NYT article "Who's Reading Your X-Ray?".

How Not To Give a Presentation

2005-10-22T23:24:00.000-07:00

From Clinical Cases and Images blog: How Not To Give a Presentation: " Some advice how NOT to do a good job when presenting:- How not to give a presentation by the former BMJ editor Richard Smith. This article lists truly invaluable pearls of wisdom like:'Bad slides are the traditional standby of a bad presentation. There must be far too many. They must contain too much information and be too small for even those in the front row to read. Flash them up as fast as you can, ensuring that they are in the wrong order...' BMJ 2000;321:1570-1571 (23 Dec)- Maxims for Malfeasant Speakers - by Norman Ramsey at the Harvard Electrical Engineering and Computer Science And finally some real advice how TO give a presentation:- Tips on... Grand round presentations - BMJ Career Focus 2005"

CNN.com - FDA approves first brain stem cell transplant - Oct 21, 2005

2005-10-22T22:27:00.000-07:00

Stem Cells Inc. from Palo Alto is proposing to transplant neural crest cells into children with Batten disease.

Science - [LETTERS] An Open Letter to Cancer Researchers

2005-10-22T22:07:00.000-07:00

Stephen J. Elledge, Gregory J. Hannon

As cancer researchers, we have a special responsibility with respect to guiding resource allocation to fight cancer. We need to be able to look cancer patients and their families in the eye and say, "We are spending your money in the best way we know to find a cure for you." We must apply this standard in judging any large-scale proposal for dedicated research funding allocations. As currently configured, the [Human Cancer Genome Prokect (HCGP)] needs to be reconsidered and reprioritized to produce a program that gives us the best chance for fighting this disease. Therefore, because the most productive direction of research is still a debatable question, we propose that (i) sequencing be delayed until advances in sequencing technology are achieved; (ii) objective criteria be established to allow a go/no go decision for continued DNA sequencing based on pilot studies; and (iii) large-scale genetic screening to identify targets whose inhibition kills cancer cells should be incorporated into the HCGP.

$1.5 billion over 10 years!

Science - [REPORTS] Interlinked Fast and Slow Positive Feedback Loops Drive Reliable Cell Decisions

2005-10-22T21:59:00.000-07:00

"Many cell signaling networks consist of dual feedback loops, enabling them to combine rapid response with reliable performance.Authors: Onn Brandman, James E. Ferrell Jr., Rong Li, Tobias Meyer"

Positive feedback is a ubiquitous signal transduction motif that allows systems to convert graded inputs into decisive, all-or-none outputs. Here we investigate why the positive feedback switches that regulate polarization of budding yeast, calcium signaling, Xenopus oocyte maturation, and various other processes use multiple interlinked loops rather than single positive feedback loops. Mathematical simulations revealed that linking fast and slow positive feedback loops creates a "dual-time" switch that is both rapidly inducible and resistant to noise in the upstream signaling system.

Of interest are platelet activation and coagulation cascade activation (E Beltrami and J Jesty 1995).

computational biology genetic network

Science - [PERSPECTIVES] CHEMISTRY: The Renaissance of Natural Products as Drug Candidates

2005-10-22T21:17:00.000-07:00

"In recent years, the use of natural products for drug discovery has declined in favor of combinatorial methods and the rapid generation of large libraries of potential lead compounds. In their Perspective, Paterson and Anderson suggest that it may be time to revisit the prevailing dogma and consider ways in which natural products could continue to inspire the development of new drugs. Natural products offer high potency and selectivity as a result of long evolutionary selection. Taking bioactive natural substances as a starting point, researchers can then use the methods of organic synthesis to design targeted modifications of specific structures to create new therapeutic agents.Authors: Ian Paterson, Edward A. Anderson"

Instapundit talks about portable defibrillators an...

2005-10-22T21:13:00.000-07:00

Instapundit talks about portable defibrillators an...: "Instapundit talks about portable defibrillators and a recent Popular Mechanics story: There's an article in the latest Popular Mechanics -- not on their website yet -- saying that the home defibrillators really do save lives. As they get cheaper and more ubiquitous, it's likely to make a real difference. A lot more people die from sudden cardiac death, where a defibrillator will save them but nothing else much will, than is generally realized. Likewise, inexpensive blood pressure monitors mean that -- since you don't have to go to a doctor -- more people will track their blood pressure. Just another way technology is empowering ordinary people.I wrote about this awhile back, looking at the data showing a minimal survival benefit. For some, that may be worth the $1,295: To put that in perspective, in a low-incidence practice (which would still have a higher incidence of sudden death than an average home), over 1900 AEDs (automatic defibrillators) would have to be purchased to treat 16 cardiac arrests over a 7 year period.Despite any form of advertising, the prognosis of sudden cardiac death is quite poor, whether a defibrillator is available publicly or not. A retrospective study suggested that targeted public placement of AEDs increased overall survival rate from a cardiac arrest from 5.0 to 6.3 percent.Certainly some may"

A Lifetime in Recovery From the Cultural Revolution

2005-10-22T21:05:00.000-07:00

A Lifetime in Recovery From the Cultural Revolution: "Xu Tian, a scientist being mentioned by his peers as a potential Nobel laureate, is among those Chinese whose transition to adulthood was hijacked by the Cultural Revolution."

Cornell President Condemns Teaching Intelligent Design as Science

2005-10-22T21:04:00.001-07:00

Cornell President Condemns Teaching Intelligent Design as Science: "Interim president Hunter R. Rawlings III denounced the theory as 'a religious belief masquerading as a secular idea.'"

Rain over Manhattan skyline

2005-10-22T21:04:00.000-07:00

Rain over Manhattan skyline: "mihay has added a photo to the pool: "