The myth of choice in medicine


I’m probably not the only one who has noticed a shift in the way we talk about health care these days. It’s no longer about patients taking the advice of their doctors. It’s about “consumers” making “choices” about care.

We’re shifting away from the old model of medicine, in which doctors guided medical decisions because they knew what was best for us. Nowadays we come to our doctor’s appointments armed with printouts from the Internet, try to understand the true risks and benefits of every procedure, and, supposedly, make our own informed decisions.

But the model of consumers and choice isn’t always a perfect fit for medicine, and the ongoing controversy over the Maine home birth study shows why. This month, the American Journal of Obstetrics and Gynecology will publish a series of six letters critiquing the study, released last year, that claimed to find that babies born at home were more likely to die than those born in hospitals. The journal also published the authors’ responses to the critiques.

The Maine study’s methods and conclusions have already been debated in many quarters – including the comments section of this blog – and the critiques that have just been published aren’t surprising. What I can’t help thinking about, though, as I read through the critiques and the authors’ replies, is how all of this would look to a family trying to make a choice about where to have a baby.

The journal that published the Maine paper says that it is airing the critiques because a review panel found that the criticisms were “subjective, and should be debated openly.” This is commendable, but I’m not sure how much it will help our hypothetical mother-to-be to hear the point-counterpoint argument on each of the study’s flaws.

For instance, a major point of controversy is why the Maine doctors did not include data from a large Dutch study in their analysis of newborn, or neonatal, deaths. Many critics have pointed out that inclusion of this data would have likely erased the increase in deaths among the babies born at home, because the large Dutch study didn’t find any difference in newborn deaths among home and hospital births.

The Maine doctors write that they excluded the Dutch study from their analysis because they wanted to analyze newborn deaths up to 28 days after birth, while the Dutch study only analyzed newborn deaths up to 7 days after birth. But when I contacted Ank de Jonge, the lead author of the Dutch study, she said that her group has data on newborn deaths up to 28 days after birth, and would have happily analyzed it for the Maine doctors’ study if they had asked.

Furthermore, she said, the analysis would probably show that newborn deaths were not higher among the home-birthed babies.

“We saw no difference in perinatal [around birth] mortality up to 7 days and therefore it is unlikely that the neonatal mortality up to 28 days would have shown a difference between the planned home and planned hospital group,” de Jonge wrote in an email.

In other words, a single study might have completely changed the frightening outcome of the Maine analysis. Or it might not have; we won’t know until de Jonge runs the numbers, as she hopes to do soon.

Yet the language of “consumers” and “choice” is based on the idea that we can make rational choices about health care, guided by high-quality and definitive information. It doesn’t account for the reality: that often, there is no clear-cut answer – and sometimes, there’s just a morass of technical arguments that leave regular folks anguished and uncertain.


Photo credit Gimnyeong Maze Park, Souh Korea. golbenge (골뱅이)/flickr.

Share Button

11 thoughts on “The myth of choice in medicine

  1. The quote — “We saw no difference in perinatal [before birth] mortality up to 7 days and therefore it is unlikely that the neonatal mortality up to 28 days would have shown a difference between the planned home and planned hospital group,” — confuses me, Erika. Unless I’m missing something obvious. Babies weren’t more likely to die before they were born and up to 7 days after they were born? And how would that predict that babies were also unlikely to die up to 28 days?

  2. Thanks, Ann! I should have said “around” rather than “before” birth; “perinatal” encompasses the whole period from just before to just after birth.
    de Jonge is basically saying that there was no difference in mortality up to 7 days of life, so she would expect no difference in mortality up to 28 days of life as well.

  3. Do you happen to know whether Joanna was right in suggesting the home birth statistics include emergency births? In that case, it wouldn’t be as informative for expectant mothers who are often less interested in home versus hospital than they are in midwives versus doctors and nurses.

  4. Hi Jessa, that’s a very important question. It’s true that “emergency” home births are riskier than either planned home births or hospital births, so the better studies now look at “planned” vs “unplanned” home births.

    The Maine study was intended to analyze planned home births only. However, there’s some debate about whether the study unintentionally included data on unplanned home births, and this is one of the point-counterpoint arguments in the documents that have just been released.

  5. Ya — even if it’s planned but unattended, that would have to skew the stats. I can’t imagine there would be much debate that it helps to have someone there who knows what’s going on (beyond the whole “your body knows best” idea).

  6. Absolutely. And the Maine study actually also found that if the home births were attended by a certified nurse midwife, there was no increased risk of newborn death.

  7. Hey Erika, great post. A big part of a regular person’s ability to understand and make a decision on the available information also I think has to do with the filters the information goes through before it gets to the general public. How did the media portray the results? How well were the methods and limitations of systematic reviews explained? I deal with this all the time at work– so many scientists are interested in the pursuit of science and don’t quite process how the media and the general public interpret their results. It’s like you say– for scientists, it’s the last word on nothing because it’s never conclusive. But as we’ve seen (worst-case scenario) with the vaccine-autism issue, people will consider a single study to be absolutely conclusive even after it’s PI has admitted to falsifying conclusions.

  8. Hi Pietra – that’s an excellent point – few people reading about these studies even have free access to them, or would able to parse them on their own, so most of us rely on media reports to interpret them.

    It’s especially interesting to think about how this will change with the contraction of traditional media. When the media filter disappears, scientists will have a lot more control over their own messages, but as with this situation, I don’t think that will always mean that the nuances will become clearer to the general public.

  9. nice presentation, erika, and good non-ideological discussion. i remember many years ago reading a column by (i think) ellen goodman. she wrote that she went to a new pcp and he introduced himself this way: “hi, i’ll be the junior partner in your healthcare.” she was somewhat taken aback — she didn’t want a junior partner, she wanted someone who knew what was going on. it’s a strange time in medicine – doctors range from those who still follow the authoritative model to those who want to help the patient to make their own choices, with every shade in between. i remember about 20 years ago i interviewed a female ob/gyn named mona shangold who bemoaned fat cigar-smoking ob’s who told women it was bad for their uterus to exericse and at any rate they should stop exercising when they got pregnant. shangold said that she herself showed up at the starting line in an obvious state of protuberance. the real problem is not that patients are not able to process complex data, although that is certainly the typical case. the real problem is that doctors also are often not able to process complex data, especially the complex genetic data that are becoming so common nowadays. in fact, sometimes data resist being definitively processed. i would paraphrase your closing sentence to read:

    “Often, there is no clear-cut answer – and sometimes, there’s just a morass of technical arguments that leave regular folks and their doctors anguished and uncertain.”

  10. The pain was incredible when i went to the emergency room. The likely diagnosis was a gall bladder attack, but not conclusive. I was stabilized with anti-inflamatory, antibiotic, and pain killer. No choice was given, just informed. The next day, the gall bladder diagnosis was confirmed. The doctor came by and said that, since i was stabilized, i could go home, and schedule removal for some more convenient time, or do it the next morning. I thought the next morning was most convenient since i already hadn’t eaten in days. But then he also noted that “sometimes the gall bladder will become infected”, which is BAD. Well, i wanted it out anyway.

    After surgery, it was discovered that it was already gangrenous. Instead of outpatient surgery, i’d need to stay in the hospital for a few more days, to stay on an antibiotic drip, and frequent megadoses of fever reducer.

    I’ve no idea what the chances are that it would turn out the way it did. I’d had no prior symptoms, which is unusual. From my sample size of one, it’s clear that no choice should ever be given to delay surgery. Patients don’t know how to judge risk anyway.

  11. something that the study is nearly certain to have ignored is the issue of Immediate Cord Clamping.

    I hope that external links are allowed???
    You can read about it here.

    How the Cord Clamp Injures Your Baby’s Brain


Comments are closed.

Categorized in: Erika, Health/Medicine

Tags: , , , , , ,