In last week's NEJM, there is an article about the purported perils of foregoing vaccinations for your kids. The article is here: http://content.nejm.org/cgi/content/full/360/19/1981 .
There are a few points that I think deserve to be made about this issue. First, I digress to outline briefly the idea of "The Tragedy of the Commons."
The Tragedy of the Commons refers to the notion that "commons" such as parks or more traditionally "grazing areas" will be more fruitfully enjoyed by all if they are used responsibly. If everybody grazes as many sheep as s/he pleases on the commons, soon enough, there will be no grass for the sheep to eat. So it stands to reason that one should graze his sheep responsibly and sparingly on the commons. Paradoxically, there is little incentive to exercise such restraint. Because insomuch as you do, your neighbor does not, and the sparing of the commons effected by you is obliterated by your neighbor, or his neighbor, etc. As when passing a sign enjoining you to not walk on the grass and you are want to say "ah, but what difference will it make?", your neighbor might respond "yeah, but if we all did that....". The sign is there to regulate the commons that would be depradated were it not for some social policy forcing restraint. So long as the MAJORITY refrains from treading on the lush monocultured turf, it will remain lush. But after a certain threshold number of defectors trammels it, the commons is lost.
And such, I will demonstrate, is the issue with refusing vaccinations. The threat that results is not so much to the unvaccinated child, but rather to the commons - to the herd immunity. So far, it seems to me, the medical and public health establishments have sought to appeal to the sensitivities of parents to their own children's welfare rather than to supplicate them to "do what's right for society." To me, this is a overtly disingenuous approach. The vaccination of any indivudual child, when the baseline vaccination rate is above some critical threshold is an act of social responsibility much more than it is something essential for the health of the individual child. I suspect that some vaccine-refusing parents (let's call them Refusniks, shall we?) recognize this, and this recognition, combined with a tendency for rebellion, creates an impetus for refusal, especially if they think that the vaccine may cause autism or some other untoward effect. Let's look at some numbers.
First let's start with an estimate of the incidence of Measles with and without vaccination (if you take issue with these estimates and the resulting conclusions, please furnish your own numbers with a reference):
Measles with vaccination:
0.0000010000000 per annum
Measles without vaccination:
0.0002500000000
Even though this is a 250x increase, it is still only an absolute increase of:
0.0002490000000
So, if you fail to vaccinate your child, you increase his/her risk of measles by only 0.024%.
But the case fataility rate for measles is only about 0.3%. So, you increase your child's risk of death from measles by only:
0.0000004230000
That's a very small number, my friends.
Now let's also say that you're concerned about the risk of autism, for whatever reason, even a specious one. And you ask your pediatrician who is skeptical, so s/he refers you to the most recent good quality epidemiological data, the Danish data from NEJM in 2002: http://content.nejm.org/cgi/content/abstract/347/19/1477 .
In this study, the upper 95% CI for an association of MMR with Autism was 1.24. Thus, a 24% increase in the risk of autism is certainly within the range of plausibility based on these data. The base rate of autism in this study was:
Base rate of autism:
0.0005880000000
Rate of autism with a 24% increase (assuming it may be as hight as the UCI):
0.0007290000000
Absolute increase in autism rate:
0.0001410000000
Now, I realise that autism may not be as bad as death for a child, but this POTENTIAL increase in autism, consistent with good data, far overshadows the risk of death from Measles attributable to failure to vaccinate your child.
So it stands to reason that, if a person has, for whatever reasons, a value system that makes autism a grave concern for them, they are NOT acting terribly far outside the bounds of rationality by refusing vaccination for their individual child.
Now if their child has siblings, and/or they live in a community where there is a high rate of vaccination refusal, these numbers are out the window and the individual child risk is much harder to calculate and probably much higher.
(I recognize also that I have used data on the ANNUAL measles risk which may be cumulative and this may sway the numbers in favor of vaccination since presumably the risk of autism from vaccine exposure is a one-time event.)
I do not mean to imply here that I am against vaccination (I am not), nor that I believe that autism is caused by MMR or other vaccines (I do not), but I think 4 points are germane to this conversation which may be emblematic of other issues in public health where officials are apt to take a paternalistic stance:
1.) The individual child's absolute risk of death from Measles is VERY small, as is the increase in risk from failure to be vaccinated.
2.) The risk of autism from MMR based on the Madsen data has a wide confidence interval which does not exclude what some parents may think is a meaningful increased risk of 24%. The meaningfulness of this risk may be especially important in the context of comparing it with another very small risk, such as that of death or diasbility from measles, or motor vehicle accidents.
3.) The refusal to vaccinate is more of a social responsibility issue, a Tragedy of the Commons, than it is an individual patient safety and health issue. (Such is also the case with PPDs, TB, and INH prophylaxis, but don't get me started on that.)
4.) The risks that parents take for their children through vaccination refusal is similar risks they take via motor vehicle travel. We are not encouraging parents to cut in half the number of miles they drive with their children per annum to reduce the risk of death from MVAs from 0.000145 to half of that, so why are we so adamant about their getting MMR? Because it's an issue of the commons, not the individual.
And if it is an issue of civic responsibilty, we should frame it as such, rather than guilt-tripping parents about exposing their children to risk via neglect. Just like driving a massive Ford Excursion, where your children may be safer but everybody else's are worse off (because of the size of your projectile or its impact on the environment), vaccination is better for the commons, if not for your own children.
This is discussion forum for physicians, researchers, and other healthcare professionals interested in the epistemology of medical knowledge, the limitations of the evidence, how clinical trials evidence is generated, disseminated, and incorporated into clinical practice, how the evidence should optimally be incorporated into practice, and what the value of the evidence is to science, individual patients, and society.
Subscribe to:
Post Comments (Atom)
Hey Dr Aberegg. Like the post. But you put forth that the 95%CI on autism includes 1.24 would I be correct in saying that is also includes 0.68? That means that there is also a chance that MMR vaccine could provide a 32% protection from autism and thus may provide direct benefit?
ReplyDeleteOverall it is a great point about civic duty versus parental duty.
I'm looking through your old posts for a good review of Bayes' calculations - is there one?
Ian Myles
This is an interesting post. I'm not sure exactly where to start. Vaccination certainly is directed at the aid of society as a whole rather than the individual. However, herd immunity just wouldn't be good enough for my own kid. I would want my kid to be immune. During a peds residency you just see too much morbidity that could have been prevented. I think the thing that makes it so easy to say is that I truly don't believe there is any link between vaccines and autism. I just don't believe it. I don't know what makes a child autistic, but I don't think it is MMR vaccine. And if there is no downside, and there is a tangible upside, you go for it. I guess you are arguing that the upside isn't tangible for an individual, but it is tangible for me.
ReplyDeleteThanks for your comments!
ReplyDeleteIanMD, you are indeed correct that the 95% CI includes values for which MMR would be PROTECTIVE against autism, and that is indeed a good point.
Jaymez - I too think that there is no link, but I also think that the individual failure of vaccination poses an almost negligible risk for a singular child, AND that the debate has unfairly characterized parents who refuse to vaccinate as whackos. When probabilities are very very small and hard to estimate, it is difficult to make rational informed decisions. For some parents, the fear of autism, however irrational, may be so inflated as to require factoring in to the expected utility equation for them. And the result is less a concern for them than it is for the herd. Also, it's important to recognize that an inpatient sample biases toward the numerator (severe cases) while neglecting the denominator (mild cases) in a VERy large population of children....
What a refreshing post. I have been looking for some great quotes on the social responsibility of vaccinations so if you have some that I could use, I would really appreciate it (for sharing with other parent groups).
ReplyDeleteImmunological Freeloaders: Measles on the Rise
ReplyDelete