6.21.2010

Dept of Lazy Reportage

Surprise, surprise. It's from Slate.

Honestly, Slate is second only to HuffPo in its rate of eye-roll-inducing, "why do I bother" stories? (And yet, I still go back, largely for Dahlia Lithwick.) Its obvious pride in its ersatz edginess and counter-intuitive take on [insert topic] annoy me on a regular basis.

Today's arglebargle is about acid-reducing medications for infants. Writes Darshak Sanghavi:
And now the war on stomach acid has a new front: babies with colic. Normal infants reflux all the time (the average preemie, for example, has 71 minor spit-ups daily), but fewer than one in 300 has any evidence of damage to the esophagus. Randomized studies regularly show that acid blockers do nothing to help baby reflux. Worse, drugs like Nexium or Zantac (which blocks acid in a slightly different manner) may increase brain bleeds and gut damage in preterm infants as well as the risk of food allergies in older infants.

[snip]

The overuse of drugs like Nexium isn't a new problem. For years, doctors prescribed COX-2 inhibitors like Vioxx and Celebrex, though they worked no better than Tylenol and ibuprofen. Pediatricians use antibiotics to treat almost half of all kids with colds, though the powerful drugs do no good and drive up the number of resistant microbes. Can we really blame the drug companies for these messes? Malcolm Gladwell, writing on the prescription drug crisis, notes, "For sellers to behave responsibly, buyers must first behave intelligently." For many adults with heartburn and for many babies with colic, doctors regularly write useless prescriptions and insurers happily pay to fill them.

[snip]

In the meantime, the next time your pediatrician suggests Nexium, Prilosec, or Zantac for your baby's mild spit-ups, ask her to explain their necessity better. Tell her you don't need any fancy pills. You would prefer a far more powerful therapy—words of explanation.

Dear God, the douchebaggeryon display here is breathtaking.

I'll start by conceding a point. Nexium is ridiculously over-prescribed. It is no more effective than the (now-generic) Prilosec, and its persistent use by lazy prescribers irks the living daylights out of me. Sanghavi's point about that is a fair one.

But egads, one would really expect a pediatric cardiologist to write a better, more intelligent article about medical science. I'll start with the last paragraph first. I do not prescribe anything for "mild spit-ups, "fancy pills" or otherwise. (Typically newborns have trouble with pills, no matter the degree of fanciness.) For "mild spit-ups" I typically use "words of explanation" about the mechanics of newborn stomachs, and how common it is for babies to spit up. I do sometimes prescribe Zantac (or rather, its generic equivalent) for babies that seem very fussy after feeds. This is because reflux can be very painful, and babies can't tell us what they're feeling. If the medication makes no difference, I stop it.

Are Zantac et al overprescribed? Probably. But let's dial down the rhetoric about pediatricians as pill-happy under-explainers.

The first paragraph quoted above is so intellectually dishonest and so rife with fear-mongering, it disgusts me. Not one of the embedded links supports the contention Sanghavi associates with it. "Brain bleeds" links to an abstract for a study about whether acid-blocking medication protects the lung, and which mentions the brain not at all. "Gut damage" links to one study about acid-blockers possibly being related to a very common complication for very low-birth-weight preterm infants, and it is flagrantly inappropriate to either make a broad conclusion from one study or to extrapolate about healthy infants from a much different, far sicker population. "Food allergies" links to (again) one study in mice and humans, which was entirely lab-based with nothing to do with actual patients experiencing actual allergies. Sanghavi grossly misrepresents what the science is saying.

Finally, the thing that sent me over the edge -- "[p]ediatricians use antibiotics to treat almost half of all kids with colds." Really? That link yields yet another abstract. The setting for the study was merely "physician offices," not specifically pediatricians. Thus, there is no way of knowing how many providers were family practitioners vs. pediatricians. Beyond that, the article is 12 years old!! So, what we have here is one snapshot of medical providers from a dozen years ago used to denigrate what pediatricians "do."

This makes my blood boil. This kind of sloppy laziness would be bad enough from a lay writer, but there is no excuse for it coming from a pediatrician. Slate should be ashamed of publishing such a poor excuse for writing.

10 comments:

  1. "Sloppy laziness" is not reading the studies beyond the abstracts before bloviating about them on a blog. The brain bleed study was a randomized trial of acid blockers in preterm infants stopped early because of a higher incidence of severe interventricular hemhorrage (that means a bleed in the brain) in the babies treated with acid blockers. The same with your comment about the allergy study--which when I clicked on it says humans seemed more likely to have a nut allergy after taking Zantac (an observation which is seen in many reports in Medline). Your post is superifical and pissy. Man up and stop being so sensitive when people criticize bad medical practice a little.

    James G., Baltimore, MD

    ReplyDelete
  2. Dan, your post makes the following argument: 1) You are such a great doctor that you never over prescribe Zantac, ergo 2) The Slate article sucks since it claims many doctors over prescribe acid blockers. Your stupid logic makes my "blood boil."

    ReplyDelete
  3. James, thanks for stopping by.

    1) It is not my obligation as a reader to justify the overblown claims predicated on preliminary study made in this article. If the author has a claim to make, the burden of supporting his claim is his, not mine.

    2) No matter what the fuller articles MAY possibly say, they are isolated studies. They MAY indicate some POSSIBLE association between the outcomes described and H2 blockers.

    3) Your mention of Medline hints that you are a medical provider, which means you have access to fuller citations than a lay reader would. This article was pitched to lay readers who, presumably, do not have such access and thus cannot properly evaluate the inflammatory claims made. The only way one would come across the brain hemorrhage outcome would be if one had Pubmed access of similar. If Sanghavi has a case to make, which is a trite one at best, he should support it with more established and accessible science.

    Cheers!

    ReplyDelete
  4. Um... no, Anonymous. My post makes the following argument -- that Sanghavi makes numerous inflammatory, overblown claims on flimsy evidence. The links he provides to a non-medical audience fail to adequately support the adverse effects attributed to the medications in question. James, above, seems to think that isolated findings in specialized populations are somehow adequate justification for this kind of sloppy writing, but I do not.

    Thanks for your helpful and insightful opinion, though. Always nice to have a reader.

    ReplyDelete
  5. In defense of Dr. Sanghavi, and his obvious concern for babies with colic - in a society in which (it's safe to say)both many medical providers and the general public are oft considered much too dependent on the use of drugs for any or every discomfort - and given that side effects and untoward responses are what they are - "First do no harm" is not such a bad thing to publish, no matter how well or poorly it is written. Buttons do get pushed, don't they.

    ReplyDelete
  6. This discussion shows why humans aren't very good with statistical inference. Correlation is not causation, but if you are a prehistoric hunter-gatherer, logical consistency doesn't put game and berries on the table. Assuming causation is a caveman win, so it gets wired into our instincts.

    Somewhere, a medical study of ingesting water will be associated with an increased frequency of brain bleeds. Medical ethics says terminate the study until a better explanation is developed, but surely Dr. Dan is correct that one study ought not inform policy, or responsible journalism for that matter, without a plausible physical causal mechanism to explain the result. Multiple findings of correlation? Different story.

    Or at least, so says my caveman understanding of medical research. Brickbats and corrections welcomed.

    ReplyDelete
  7. "the 'douchebaggery' on display..." deriding SLATE and a writing by Darshak Sanghavi. Seems that word choice: "douchebaggery," may (possibly?)be unintentionally derisive of women?

    ReplyDelete
  8. I think one would have to seriously over-read the word choice to reach that conclusion, Anonymous.

    ReplyDelete
  9. Thanks, Dan. I will think about a possible-vs-probable over-read of word choice, especially in regard to its use in your call for fact over feeling in medical reporting (and I will not now launch into Sanghavi's use of generalization, or the annoying changing of presumed offending subjects/objects, etc). However, a leap or not, I'm not yet close to letting go my thinking that the word belongs with several other words commonly used to marginalize both men and women who walk to their own God-given drummer. Not realistic(most, if not all, would agree) but I wait the obsolescence of such slanguage, including, of course, this possibly over-read descriptive but possibly derisive word.

    ReplyDelete
  10. O.K. Dan. Your criticism of Sanghavi's article has now been read and re-read - again. Yes, an over-read. I just don't like the word, nor do I like its equation with "sloppy laziness," with what apparently caused marked disgust. Something that to me generically relates to "clean" - by association became what should cause shame. However, boiling blood may expect and ask of readers a bit of time and space, and from that perspective, I confess an over-read.

    ReplyDelete