Whooping cough (pertussis) has been in the news for the past several months. For example, there’s been a sizable epidemic in California over the summer that has caused several deaths in infants. I’ve cared for several cases myself recently, including a couple of infants who needed mechanical ventilators to breathe, and who coughed so much for weeks afterward they could barely eat. It’s a serious and potentially lethal infection for infants, although in older children and adults it mainly causes only a nagging cough.
We have a vaccine for pertussis. It’s combined with other vaccines — diphtheria and tetanus — in a single shot, called DTaP. Each capital letter stands for one of the three vaccines in the shot; the little “a” indicates that the pertussis vaccine component is “acellular,” to distinguish it from the older version (the DPT) made from whole pertussis bacteria. That version caused more reactions (mainly high fever) in children who received it than does our current, acellular variety. A child should receive the DTaP at 2, 4, 6, and 18 months, and again at 4-6 years of age.
Overall, pertussis vaccine has been a success, with the number of annual cases, and death rates, dropping enormously. (You can read about its development here.) But the pertussis vaccine has never been a very good one, compared with most of the other ones, such as those for polio or measles.
The ideal vaccine confers life-long immunity. Generally this requires a series of shots to boost immunity, but, once achieved, the person can never get the disease. In practice, all immunity wanes with age and with the passage of time from the last booster shot. This waning is quite pronounced with pertussis vaccine; probably the majority of adults who received their shots as children are no longer immune. They can get pertussis, although when they do their symptoms are just those of a nagging, chronic cough. Even though their symptoms are not severe, they can still infect others — especially infants. And their disease can be life-threatening.
The persons most vulnerable to getting severe or fatal pertussis are infants, often before they can even get their first vaccine shot. So the only way to protect them is for everyone else to be vaccinated so as to be unable to bring the infection to them. Our current vaccine works in older children. And we now have a pertussis booster shot that restores immunity in adolescents and adults. It is called Tdap, and I strongly encourage every adult to get this, especially if you are exposed to young children.
Pertussis vaccine is unusual in this way. All vaccine strategies for protecting populations involve what is called herd immunity, but the principle for controlling pertussis is primarily based upon keeping those for whom the disease is only mild vaccinated, so that they may be prevented from passing the infection to infants who are too young to join the vaccinated herd. We do it for the common good.
You can find an enormous amount of useful and reliable information here, at the Vaccination Education Center maintained by the Children’s Hospital of Philadelphia.
We badly need effectiveness research — which medical treatments work and which ones don’t. After all, some reasonable estimates are that a third or so what we spend on medical treatments is for things that aren’t known to work, or worse, don’t work. Effectiveness research means comparing two competing therapies to see which works better; if both work the same, our preference should be for the less expensive one. We have very little of that now.
There is another way to think of effectiveness research, a more global one. This approach looks at the bottom line — does a treatment make people’s lives better and/or make them live longer? A key metric for this kind of research is the quality-adjusted life year, or QALY. The idea is simple: to what extent does the treatment add meaningful years of life to people? This notion has been around for decades. It’s particularly useful because it crosses all disease categories, simply comparing life outcomes.
The idea of the QALY, however, can create fears that it will somehow be used to judge lives, one against the other, regarding which has a higher “qualtiy.” Could QALY measurement be a stalking horse for rationing, of allocating heathcare resources to those some committee deems more deserving. The political traction that Sarah Palin’s nonexistent “death panels” received shows the depth of this distrust.
This fear is, in fact, embedded in the recent healthcare reform bill, the Affordable Care Act. The Act established an outcomes research center, called the Patient-Centered Outcomes Research Institute. But, as a recent editorial in the New Journal England of Medicine notes, the Act specifically forbids the use of QALY measurement “as a threshold.” It is not clear at all what that means, but, like the authors of the editorial, I think it reflects a fear that some lives will be judged more worthy than others, leading to unfair (or unethical) rationing of care.
But here’s the problem: QALY analysis is one of the most powerful tools of effectiveness research, and it’s absurd to pretend is should not or will not be used. From the article:
“The antagonism toward cost-per-QALY comparisons also suggests a bit of magical thinking — the notion that the country can avoid the difficult trade-offs that cost-utility analysis helps to illuminate. It pretends that we can avert our eyes from such choices, and it kicks the can of cost-consciousness further down the road. It represents another example of our country’s avoidance of unpleasant truths about our resource constraints.”
I suppose the fear-of-QALY clause made into the bill because the legislative sausage-making machine contains many pet items of the various legislators. But this particular one is absurd.
Everybody knows that doctors sometimes have the annoying habit of using medical jargon when speaking to patients and their families. It’s annoying because it gets in the way of communication, and many people feel inhibited from saying that they just don’t understand what the doctor is talking about. So on and on the doctor drones, assuming the absence of questions from the other side means both are communicating. Some doctors do this more than others, of course, but I think it’s a problem for all of us to some degree.
Lately, though, I’ve wondered if talking in medical jargon means thinking in medical jargon. After all, an old theory among linguists, the Whorf-Sapir Hypothesis, proposes that language literally is the stuff of thought, that we think in language, or at least in linguistic categories. A corollary to this notion is that people with very divergent languages probably think in very divergent ways.
George Orwell, in his famous novel 1984, clearly was a Whorf-Sapir believer. Here is what he has to say about Newspeak, the new language invented by the state (Ingsoc) to control thought by driving out traditional language, or Oldspeak:
“The purpose of Newspeak was not only to provide a medium of expression for the world-view and mental habits proper to the devotees of Ingsoc, but to make all other modes of thought impossible. It was intended that when Newspeak had been adopted once and for all and Oldspeak forgotten, a heretical thought–that is, a thought diverging from the principles of Ingsoc–should be literally unthinkable, at least as far as thought is dependent on words. Its vocabulary was so constructed as to give exact and often very subtle expression to every meaning that a Party member could properly wish to express, while excluding all other meanings and also the possibility of arriving at them by indirect method. This was done partly by the invention of new words and by stripping such words as remained of unorthodox meanings, and so far as possible of all secondary meanings whatever.”
That’s a chilling passage. But I think there is some truth to the notion that corrupting the speech easily leads to corrupting the speaker. Professional vocabulary — jargon — has its place when used among members of the profession. But when I hear doctors speaking entirely in mealy-mouthed or impenetrable jargon, both to patients and to each other, I wonder how clear their thinking is. For a doctor, plain speaking may be more than just good communication; it may be good for his or her brain.
Like many of my colleagues, I teach and supervise students, residents, nurses, and respiratory therapists. I’m also the medical director of a PICU. Overall, I’ve been teaching and doing administration for over 30 years. And, like most of my colleagues, I never received any formal instruction at all in how to do these things. To some extent I got help from my own mentors, primarily by watching what they did, but basically I learned on the job. I hope I am reasonably good at it, but really, I have little way of knowing if I am. So I’ve always had an interest in whatever tidbits I could pick up in teaching theory that might be useful. One particular topic that’s always interested me is the opposing pros and cons of reward versus punishment. I use those tools to train my horse – rebuking bad and praising good behavior. What do we know about applying them to people?
The practical problem, one faced by most teachers, is what to do when a student does a poor job. The tradition in medical teaching, certainly when I was in training, was to lean heavily on the rebuking, punishing side of the equation. Public ridicule was common, and there was more than a little yelling involved. Did fear of that help me avoid doing the wrong thing next time?
Recently I was reading one of my favorite group blogs, Crooked Timber. Most of the contributors are professors of one sort or another, and the topic of effective teaching comes up now and then. This whole reward/punishment tension was the topic of a recent post there. The situation it describes involves military flight instructors, who universally believed that yelling at fledgling pilots when they made mistakes was much more likely to make their next attempt better than was praising a good action. Here’s what one instructor had to say about it:
“On many occasions I have praised flight cadets for clean execution of some aerobatic maneuver, and in general when they try it again they do worse. On the other hand, I have often screamed at cadets for bad execution, and in general they do better. So please don’t tell us that reinforcement works and punishment does not, because the opposite is the case.”
It’s a military example, but training doctors has traditionally been done using that sort of get tough model. I was aware of a statistical principle called the regression to the mean, but this example applies it to teaching in a way I hadn’t thought about. The argument goes like this.
If a given student does a bad job at something, that is more likely to be a low point for them, below their average. Statistically speaking, they are more likely to do better on the next attempt no matter what the teacher does. So the teacher is likely to think whatever he or she did – screaming, for example – as causing the improvement. On the other hand, if a student does an exceptionally good job, the same regression to the mean makes it likely the next attempt won’t be as good, so whatever the teacher does – in this case praising – tends to cast doubt on the usefulness of praise.
For myself, I think praising, in the long run, works much better. I’d be interested in what any professional teachers think about this.
In the comment trail to the article, the classic The Art of Raising a Puppy was cited as a useful source. That was interesting.
We’ve known for a long time that tobacco smoke causes asthma exacerbations in children. And it doesn’t have to be the actual smoke; for children with asthma, simple exposure to rooms where people have smoked, or even to the clothing of smokers, can trigger breathing problems. A principal argument for banning smoking in public places is that these environments can be downright toxic for asthmatics. Do these anti-smoking regulations actually work? The bottom line question would be this: after these laws are passed, is there any change in the number and severity of asthma attacks? An important recent report in the New England Journal of Medicine gives us an answer. Yes — these regulations help children breathe better.
The research was done in Scotland, where all smoking at enclosed public places was banned in 2006. The premise of the study was simple: look at how many children were hospitalized for asthma before the ban and afterward. The results were striking: throughout Scotland, there was a reduction of asthma admissions in children of 15%. That may not seem like much, but the ban had only been in effect for 3 years when the data were collected. The direction of the trend line strongly indicates that there will be further reductions.
It’s important to understand that correlation doesn’t imply causation. That is, there’s no way to know if the drop in asthma attacks was really caused by the ban. But the authors did a large number of adjustments and corrections to the data set to rule out confounding variables (such as rural/urban, socioeconomic status, age, severity of previous asthma attacks). The reduction occurred in children of all ages — toddlers up through school-age.
The link above is only to the abstract of the article (you need a subscription to get the full text), but the key figures are there in the abstract. If anyone wants the complete article and can’t get it, let me know.
I’ve been doing this blog for three years, and by far the post that has provoked the most interest is this one, about the safety of x-rays. The comments, now at 102, keep steadily coming. Google tells me that the most common search string that brings folks here is some variant of the question: “how safe are x-rays?” Recent studies, such as this one, have highlighted the issue of CT scans.
The important thing to understand is that nobody wants to stop doing x-rays and CT scans. The latter in particular represent a quantum leap in our diagnostic ability, and appropriate x-ray studies improve and even save children’s lives. What we want is to strike a balance between doing too few and too many. The question always to consider is this: what is the risk of doing the x-ray or CT (still very, very tiny) versus the risk of not doing the study, of not getting the information the study provides. If the study is needed to rule out the possibility of a serious condition, then the risk/benefit calculation virtually always favors doing the test.
There is another consideration, one highlighted recently by the Alliance for Radiation Safety in Pediatric Imaging, an initiative led by pediatric radiologists — the Image Gently initiative. The concept is simple: use only as much radiation as you need to get a good picture. In the past, CT scanners in particular often used radiation doses more appropriate for adults than children. Using that dose causes risk without adding benefit.
If my child needed a CT scan, I would ask the doctor to lay out the risk/benefit ratio — the risk of doing versus not doing the scan. If the scan is needed, I’d then ask if the radiologist will use the minimum dose required to get a good picture.
This site, from the International Atomic Energy Agency, has lots of useful information about protecting patients from unnecessary radiation.
Here’s a useful short post by a pediatric radiologist about all this.
Football season is upon us again, so it’s time to think about sports injuries. We frequently have children admitted to the PICU (or to what we call the intermediate or step-down unit) for observation, typically overnight, who have struck their head. They have had concussions. What is a concussion, and what does it mean for the child?
The term itself is centuries old, but even thirty-five years ago, when I was in training, the actual definition of concussion was a bit vague. What was usually meant was that the patient got hit on the head and either lost consciousness briefly or at least wasn’t quite himself for some period of time afterward. These days we’re more precise than that, but concussion is still a somewhat inexact term. This is mainly because of our ignorance of the subtleties of how the brain works.
The formal definition of concussion is a transient interruption in brain function. By implication, various scans of the brain, such as CT scans or MRI scans, show no abnormalities. Since all the imaging studies are normal, defining concussion is necessarily inexact. I’m sure one day we’ll have some kind of machine that detects the reason for the symptoms of concussion, but right now we don’t have such a thing — concussion is an entirely clinical diagnosis, meaning there’s specific no test for it.
There are several systems for grading concussions. Here’s how the American Academy of Neurology grades their severity:
Grade I: confusion, no loss of consciousness, symptoms last for < 15 minutes, has memory of the event
Grade II: confusion, may lose memory of the event but no loss of consciousness, symptoms last for > 15 minutes
Grade III: loss of consciousness and no memory of the event
The list of symptoms that can come from a concussion is a long one. Headache, dizziness, vomiting, and ringing in the ears are common. Various behavioral changes are also common, such as lethargy, difficulty concentrating, and irritability.
What are the effects of concussion on a child? Years ago we pooh-poohed the idea that mild concussions cause brain problems. For example, football players were sent right back into the game after experiencing a concussion. We now know that is dangerous. As a general rule, we don’t recommend any contact sports for at least a week (some authorities say longer) after all symptoms have cleared. This is because a repeat blow to the head, even a very mild one, can cause severe injury to a brain that has not fully recovered from the last injury.
What about long term effects of concussions? The overwhelming majority of children who suffer a concussion, especially a mild one, recover completely. But around a fifth or so of children who have had severe concussions continue to have problems many months afterward.
Medicaid is the joint state/federal program that covers low-income families with children, disabled persons, and long-term care for the elderly. It’s particularly important for our PICU patients: although only a quarter of America’s children are on Medicaid, half of PICU patients are (details here).
With the implementation of the new healthcare reform bill, a key feature is that Medicaid coverage will be extended to low-income adults who are not in these categories. The federal government pays at least 50% of the costs of Medicaid, with the individual states picking up the rest.
Some states, however, receive far more help than others. Mississippi pays only 25% of its Medicaid costs, for example. (You can see what each state pays here.) Why the difference? Is that fair?
The answer is that Medicaid was set up so that the poorest states — those with the lowest per capita personal income — got more support from the federal government. The intent, I think, was to reduce disparities in medical care quality from state to state. It’s not clear it has turned out that way.
Medicaid is an enormous financial problem for most states, largely because many are forbidden by their constitutions to run a deficit; so every year they need to find a way to pay their share of the Medicaid bill. In contrast, the federal government is allowed to use deficit spending for its obligations.
One way to make the system fairer between the states would be to federalize it. After all, Medicaid was enacted at the same time as Medicare, and the latter is entirely a federal program. As Maggie Mahar has pointed out, this was actually Ronald Reagan’s preference. It is unfair to demand, as we currently do, that the states finance Medicaid one way while the feds are allowed to finance it another way. Federalizing the program could also ensure that citizens of poor states get the same care opportunities of those living in richer states.
Making Medicaid a federal program should at least be a financial wash to the total economy — new expenses for the federal government would be countered by reduced expenses to the state governments. There should be administrative savings, since right now there are, in effect, fifty separate Medicaid administrations. As someone who has practiced medicine in several states, I can testify that they vary substantially in how (and how well) they operate.
The short answer is that there aren’t any. I’ve written before about how hard it is for parents to know if the PICU there child is in is a good one or not. All they can really do is see if the components of a good PICU are in place: is the facility state-of-the-art, is it staffed by board-certified pediatric intensivists and pediatric surgical specialists, and do the intensivists have the assistance of the full array of subspecialists if your child needs them. But what parents really want to know is how the results of the PICU, the patient outcomes, stack up against similar units. Even we intensivists have little such information, and what we have isn’t available to parents. The biggest problem standing in the way of even collecting such information is the hugely varied mix of children who need the PICU, so in comparing PICUs we would need to be sure that we aren’t comparing apples to oranges.
One aspect of PICU practice that is fairly standard from place to place is heart surgery; the surgery is similar from child to child, and all children who have heart surgery need the PICU afterwards. The skill of the surgeon, of course, is the biggest variable in how the children do, but the competence of the PICU team is also a key factor.
I’m not aware of any data that compares outcomes among PICUs following heart surgery (or any other condition), but now we do have such information for adult heart surgery. Here is a very good recent editorial from The New England Journal of Medicine discussing what the implications are of such information. One important potential negative effect would be cherry-picking — doctors or hospitals avoiding caring for very sick patients out of fear of worsening their report card. From the article:
“There is no question about the need for accountability on the part of health care providers or the central role of measurement in the improvement of health care. Nonetheless, questions remain about the role of public reporting in improving health care. Performance measurements audited by regulators are one alternative, especially in situations in which the information is too complex for patients to use in discriminating among care sites. Insofar as public reporting drives improvement of all outcomes, it benefits everyone; insofar as risk aversion leads to changes in the population receiving an indicated service, the net effect can be nil or even negative.”
I think that this sort of reporting is something all of us in critical care, PICUs included, will ultimately be seeing.
This is a cross-post of an item I wrote for Maggie Mahar’s excellent heathcare blog, Healthbeat
The electronic medical record, the EMR, is upon us. For those of us who learned medicine entirely with paper charts, some have enthusiastically embraced the EMR and some have refused, to the extent they can, to deal with it at all. But most of us have plowed ahead into learning how to use it as best we can. It seems to me that the degree of enthusiasm physicians show for the EMR relates less to the particular version of it we have chosen (or, more commonly, was chosen for us) than it does to the kind of medicine we practice. The old paper records worked reasonably well for all of us; in contrast, the several versions of the EMR I’ve used work very well for some kinds of doctors, but less well for other kinds. I think a good part of this disparity is that the basic purpose of the medical record has changed over the past half-century or so, and some of these new roles can conflict with the old ones.
The oldest repository of continuous patient medical records is at the Mayo Clinic in Rochester, Minnesota. In a very real sense the modern medical record was invented there, in the first decade of the twentieth century, by Dr. Henry Plummer. As the first multi-specialty clinic, it made more sense for each patient to have a single record that traveled from doctor to doctor with the patient, rather than having each doctor keep his or her own record for each patient. The idea was to have a single packet of paper that contained everything that had happened to the patient. From that it’s only a short step to the notion that the record should travel with patients wherever they go, even if it is to physicians not associated with each other. This is a key promise of the EMR.
Medical records began as the possession of the doctor. This paradigm is changing. Very soon, although medical facilities will have copies, the records will essentially belong to the patient, with doctors only using them from time to time as need requires. Of course this could, in theory, happen with paper records, but it would be cumbersome. One of the things that first attracted me to pediatrics was the sheer size of the pile of paper that the medical records clerk would plop in front of the hapless medical student admitting an octogenarian to the hospital; in contrast, a toddler’s chart fits neatly in a small packet. The EMR allows these massive piles to be reduced to disks or microchips. It also allows the record to be organized into searchable form, so important things don’t get missed because they are buried in the disorganized mess of sequential folders.
Those are a couple of the brave promises of the EMR, but we all know we are a long way from realizing them. One huge barrier is that, as of yet, there is no standard platform for the EMR. Like many physicians, I’ve had to learn several because different facilities choose different vendors. In our pluralistic medical system (if one can indeed call it a system), it’s a free-for-all. And each of them has its own maddening quirks.
I think there is a broader problem here: over the ensuing years from Dr. Plummer’s era the medical record has taken on roles unheard of back then. For one thing, now the record is a legal document, a buttress against anyone who accuses us later of bad care. This process began long before the EMR, or course, which is one reason the charts I had to grapple with as a medical student ballooned so much. As a graduate student in history of medicine I had the chance to review many of the Mayo charts from earlier, simpler times. I recall one chart, from the nineteen-forties, describing the course of a very critically ill child. Overnight the child’s condition had markedly deteriorated; it was easy to see this from the recorded blood pressures and heart rates. The physician’s note for the following day analyzed these developments with only four words: “mustard plaster didn’t work.” Now the EMR offers the possibility of recording all we do easily and without getting writer’s cramp.
The medical record has also become something else it wasn’t back then: it is also now a commercial document, proof of what we did and why, used by payers to check up on us to make sure we should be paid for what we bill. Today’s payers want to know what the doctor did and why. They want to know, quite precisely, why that mustard plaster didn’t work and all that we did to make it work.
I think some of the problems with the promise of the EMR are that these legal and commercial roles can clash with the original purpose of the chart, which is taking care of the patient. The computer whizzes who design the software don’t always seem to me to have quite the same goals as we doctors who use it. The old paper charts were easy to adapt to new things, new procedures. All we needed was a different sheet to add to them and stuff in the folder. Upgrades and tweaks to the EMR are much more formidable things.
In spite of all these things I find the EMR to be a powerful addition to my practice. In fact, I think I’m a better doctor for using it. I think a key reason for that is because of what I practice – critical care medicine. In the ICU we love to measure and count things. We want minute-to-minute monitoring of variables, which in the old days resulted in huge paper flow sheets covered with dots and numbers. Rummaging through them to identify key moments in a patient’s care was often difficult. In the ICU, each patient gets a large number of tests each day, results which used to get stuck on clipboards with all the other paper. Important things got missed. Now I can sit at a computer screen and find it all with a mouse click, and the EMR makes it very hard not to notice anything important.
In contrast, I have friends who hate the EMR. It causes them hours of pain in training time, pain for which they aren’t compensated, and is slower for them to use than paper records were. In their minds, it gives them little or no advantage over paper in caring for their patients. I’ve noticed that they practice specialties that are less concerned with number-crunching than mine. They also tend to be office-based, rather then hospital-based, and don’t have to deal with as many other physicians as I do each day in the ICU. Thus many of their notes are written for themselves, not for other members of a large clinical team. Yet now they are asked to conform to how others want their charts to be.
I don’t know how all of this will work out. The EMR is here to stay. On balance, I think this will ultimately be good for doctors and their patients. But we don’t really know yet just what it is and what it should look like. I worry it will end up like one of those military boondoggles – it gets loaded with so many bells and whistles because it is supposed to serve so many purposes that it ends up being an expensive monstrosity that doesn’t perform any of its missions well.
Still, I’m an optimist. I prefer to be excited by the possibilities, rather than discouraged by the obstacles. I think the EMR will be good for patients, and will make us better physicians. For a while though, things will continue to be more than a little messy.