David R. Henderson  

The Risks of Kidney Donation

PRINT
Valuing Financial Services... Shorter Dierdre McCloskey...

In a recent New York Times op/ed that is creating a buzz on the blogosphere, Alexander Berger tries to persuade people that donating a kidney while you're alive is no big deal. He says that donating a kidney is easy, and writes:

When I first told some friends and family that I wanted to donate a kidney, they assumed I'd gone off my rocker. They saw it as a crazy act of self-sacrifice, rather than what it is -- one of the many ways a reasonably altruistic person can help others.

Yet he gives one key datum that I found incredibly sobering. If he's right about this datum, then he successfully talked me out of ever donating a kidney to a stranger. Here's the datum:
I am 21, but even for someone decades older, the risk of death during surgery is about 1 in 3,000.

Whoa! Really? That's huge. Let's put it in perspective. 1 in 3,000 is 33 in 100,000. So look at Table 2 in this Concise Encyclopedia article by Aaron and Adam Wildavsky.

The probability of death in a year in some of the riskiest occupations is the same order of magnitude. Per 100,000 people at risk in farming it's 28, in mining 27, and in police work 20.

When, as senior economist for health policy with the Council of Economic Advisers, I wrote my memo on why kidney sales should be legal, I mistakenly left out this risk.

Berger and I agree that kidney sales should be legal and people should be allowed to take that risk (although he advocates screwy restrictions on sales and I don't, and I think he way overstates the market-clearing price of a kidney), so that's not the issue here. The issue is that a risk that Berger dismisses as small is, if he has the numbers right, actually substantial.

Question: did he mean to write "1 in 30,000?" That's much easier to take.

HT to Danny Shapiro.


Comments and Sharing





COMMENTS (27 to date)
jh writes:

"Small risk" and "substantial risk" are subjective. It could seem substantial to you and small to him. I assume he'd say the risk in farming, mining, and police work is also small.

Of course, there is probably much more variation in the risk in those jobs than in kidney donation surgery. Police work in NYC is going to provide a much different risk than Small Town, USA. I assume surgery in hospital A is pretty similar to surgery in hospital B.

Yancey Ward writes:

Deaths from general anasthesia alone are about 10-20/100,000. So the surgery itself is probably in the range Berger described.

John Goodman writes:

If the numbers are right, that IS a significant risk.

Ted Levy writes:

The risk of death from screening colonoscopy is apprx. 1 in 5000, and this is generally recommended every 5 years or so in everyone over 50.

GD writes:

"I assume surgery in hospital A is pretty similar to surgery in hospital B."

Do you really believe that? Your assumption couldn't be more incorrect. The quality of medical care, and in particular surgical care, varies dramatically from facility to facility, and often there is even considerable variation between the care provided by different physicians within the same institution.

Sadly, I'm afraid many decisions makes such as legislators who don't have a medical background share a similar, erroneous assumption. Unfortunately, they're in a position to enact broad policy based on such flawed logic to the detriment of millions.

jh -- I don't know what you do for work, but think about the variability in the quality of the work done by people within your field. Every workplace has workers at the (relative) extremes of the talent spectrum...and medicine is no exception.

Yancey Ward writes:

And I still wonder just how sure we are that people with one kidney really don't have higher mortality rates than those with two, all else being equal.

David R. Henderson writes:

@jh,
"Small risk" and "substantial risk" are subjective. It could seem substantial to you and small to him.
Exactly. That’s why I’m glad he wrote the piece: he talked me out of something that, after Virginia Postrel’s wonderfully benevolent act, I had been toying with doing.
@John Goodman,
You and I think the same (subjective) way.
@Ted Levy,
Thanks. I had a colonoscopy about 7 or 8 years ago and was putting off my next one. I think I’ll put it off a little more.

David R. Henderson writes:

@Yancey Ward,
Actually, we’re sure that they do. Read my post that I linked to.

Finch writes:

> The risk of death from screening colonoscopy is
> apprx. 1 in 5000, and this is generally
> recommended every 5 years or so in everyone over
> 50.

The sort of risks appropriate for a 50 year-old may not be appropriate for a 20 year-old.

Tom West writes:

1 in 5000 for a colonoscopy? That sounds high given that's it's usually done under sedation rather than anesthesia.

rpl writes:
The sort of risks appropriate for a 50 year-old may not be appropriate for a 20 year-old.
To expand a little further on this, the purpose of a colonoscopy is to head off another risk, namely colon cancer. So, to evaluate the overall advisability of the procedure, you have to consider the risks of not performing it, as well as those of performing it. Donating a kidney, on the other hand, has no corresponding risk associated with not donating.
1 in 5000 for a colonoscopy? That sounds high given that's it's usually done under sedation rather than anesthesia.
I thought it was a little high too, but the Wikipedia article on the subject provides references that confirm that it's in that ballpark. In fact, there's a controversy surrounding colonoscopy recommendations because it's not clear that the risk of colon cancer is high enough in most people to offset the risk of having the procedure done.
Carl C writes:

To put it in perspective...

1 in 3000 risk is roughly equivalent to flipping a coin and getting heads 11 times in a row.

1 in 5000 risk is roughly 12 heads in a row.

1 in 10000 risk is roughly 13 heads in a row.

Richard H writes:

Why would evolution create a kidney you don't need?

happyjuggler0 writes:

Richard H,

Not withstanding the fact that evolution doesn't create species for a purpose, but rather that mutated species survive on the basis of survival of the fittest, my answer to your question is:

Trauma.

Scotty Frampton writes:

Clearly, it should be legal to sell a kidney, and the discussion on risk is interesting. Let's up the ante on the discussion though. What is YOUR personal price that would entice you to sell one of your kidneys? Me? $1,000,000 to even consider the idea.

Alex Tabarrok writes:

The risk to a live donor of a kidney is primarily in the 90 days after the operation and is a one-time risk of approximately 3 per 10,000 (per operation). After the operation there appears to be no additional risk to a kidney donor (donors can live perfectly well with one kidney.) See

http://www.scientificamerican.com/article.cfm?id=kidney-donor-mortality

I too think that is high but note that Commercial fishing fatalities in Alaska (1991-1996) were 140 per 100,000 or 14 per 10,000 per year and so quite a bit greater than donating a kidney.

http://www.cdc.gov/niosh/fishfat.html

Finch writes:

> What is YOUR personal price that would entice you
> to sell one of your kidneys?

Well, 1/3000 of my life expectancy is something like a week.

So would a week's income be a good first approximation? That makes me think risk-of-dying is not the biggest factor in how I'd price it.

Tom West writes:

> 1 in 5000 risk is roughly 12 heads in a row.

Yeah, but 1 in 8000 is like rolling a 1 on a 20 sided dice three times in a row.

I do that every other gaming session.

(Which may be why all the other players try to keep about 15' between their characters and mine at all times.)

Matt Flipago writes:

I also agree, 1 in 3000 is really LOW. If you gave like 2075 kidneys, you would have a 50% of still being alive at that rate. I'd like to know the mortality rate for a 40 year, how many die each year out of 100000, that number would put it in perspective.

joeftansey writes:

Giving up a kidney is a once in a lifetime event though. The reason those professions are so risky is because they roll the dice year over year for 30 years. To donate a kidney, you only roll the dice once.

Bryan Willman writes:

Perspective.

40,000 traffic deaths per year over 330MM people yields a fatality rate of about 12/100K per year.
(Of course that's an average - your real risk will vary wildly due to all sorts of things, some of which you can control.)

And of course, most people spend a fair amount of time moving about in the transportation system, so just like being a police officer, it's an ongoing risk, rather than a one shot risk.

The CDC web page reports the death rate from corectal cancer at 53K on the US population, so in fact on average you are more likely to die of colon cancer than to die in a traffic accident.

All of this ignores individual risk and age. How many deaths due to colonoscopies are in patients who are already very ill? How many in rogue clinics that don't practice good medicine (they make the news every so often)?

So our intuition is that giving up a kidney, at about 33/100K is really pretty bad, given it's a one shot event, normally involving very healthy people.

And finally, traveling in the transportation system or having a job (even a dangerous one) have compensations. A colonoscopy that prevents ever getting colon cancer at all, let alone dying from it, it is an obvious win.

Giving up a kidney - for the donor - not so much. Which is probably an argument that donors should get compensated just like the doctors, staff, hospitals, and insurance company involved in the process.

Henry writes:

1/3000 may not seem that high, but the downside is very large so "putting it into perspective" the way some people have done it is misleading.

Here's another way of putting it into perspective. How much do you value your life, statistically speaking? It varies from person to person but $6 million is a reasonable conservative estimate. That suggests you'd be willing to pay about $2000 to prevent a 1 in 3000 chance of death. That's by no means trivial: indeed, if you think altruism could easily supply this, consider that there are plenty of other lives (mainly in third world countries) that could be saved for $2000 a pop.

Scott Sumner writes:

You are about to attend a football game at the Rose Bowl. You are told that a crazed gunman with an Uzi will show up and 33 people will be massacred. Do you attend that game?

I might decide to watch it on TV instead.

Henry writes:

@Scott Sumner - there's guaranteed to be a lot of disruption even if you're not killed. For one, they're probably not going to see the game to completion after the massacre.

Of course, some people might perversely attend so they could say they "had been there".

Dan writes:

Top 10 most dangerous jobs:

http://jobs.aol.com/articles/2011/03/01/top-10-most-dangerous-jobs/

You can make $40k for a 1 in 500 shot at death as a fisherman or a similar amount for a 1 in 1,600 shot as a logger.

Noah Yetter writes:

@Finch:
"Well, 1/3000 of my life expectancy is something like a week."

A perfect illustration of how poor "expected value" is as a model for human decision making in the face of risk.

livingdonor101 writes:

[Comment removed pending confirmation of email address. We have tried to reach you previously. This is your second and final warning. Email the webmaster@econlib.org to request restoring this comment. A valid email address is required to post comments on EconLog and EconTalk.--Econlib Ed.]

Comments for this entry have been closed
Return to top