Chemotherapy Patients Set Off Subway Alarms

dave writes "Recently, cities such as New York and elsewhere have been installing radiation detectors in subways as an anti-terror precaution. However, as reported in New Scientist, patients who are undergoing radiation treatment are setting off the alarms. From the article, "a 34-year-old patient who had been treated with radioactive iodine for Graves disease, a thyroid disorder, returned to their clinic three weeks later complaining he had been strip-searched twice in Manhattan subway stations.""

Comments

My question is...

[dagg]

How much radiation does it take to make those things go off? Those patients must be emitting the tiniest amount of radiation. There is no way that that amount of radiation is actually hurting any nearby people. But the detectors are going off even though noone could be directly effected.

My guess is that the detectors are set to "go off" even if the tiniest amount of radiation is found. That way, any attempt by terrorists to try to hide the radiation (thick lead, etc) will be thwarted.

-- Just look at your waist

My uncle...

[silvaran]

...got nailed twice. He was driving around the U.S. late at night, heading back into Canada, and a patrol unit pulled him over, threw everything out of the back of his trunk, then interrogated him for a little while. He drank some kind of radioactive fluid to treat his cancer after his surgery, and it had set off an alarm in the patrol car.

Same thing happened once he got to the border. The border guard let him go, then some guy came running out of the customs building screaming at the top of his lungs. They stopped him and he had to read them the same story all over again. This drug is so powerful he can only take it once every six months.

Re:My uncle...

[BernManUNC]

Another good story:

My father is a physician, and I used to hang out in the radiology dep. while he did rounds. One of the techs told me about how they had given a patient an injection of a radioactive isotope for a radioacive imaging of his heart (I can't remember the exact name of this technique). Three weeks later, he walks into the White House on a tour, sets off the alarms, and is pulled out of the crowd and questioned by the Secret Service. This isotope had a half-life of eight hours. Now, I understand the chemomtherapy dose setting off alarms, as that has to have some punch. But eight hours for something that just has to be detected with an insturment three feet away? You do that math, that's some senstive equipment they have in the White House.

Radiation levels

[Simon Field]

Normally thyroid cancer patients are told to stay some distance from family members when they return home. After a few days the levels are lower and such precautions aren't necessary.

I don't know if the levels are lower for Grave's disease, or if this person should not have been on crowded subways. But to detect the levels in a shielded device, you would probably want the sensors to be pretty sensitive. Sensitivity also helps to allow fewer detectors to be used.

Should a strip-search be necessary? I doubt it.
Just hold the detector close to the thyroid to verify the guy's story. Maybe hospitals could give out cards, and the security folks could phone the hospital for confirmation.

Or just call a cab for the poor guy.

It happened to a friend of mine

[fava]

A friend of mine had a summer job at Triumpf a number of years ago. Triumf is a particle and nuclear physics lab. One day he took the morning off to get some medical tests done where they injected him full of tracer isotopes. We he tried to go back to work in the afternoon he set off half the radiation alarms in the place just by walking through the front door.

They gave him the rest of the day off.

how about...

[garcia]

I have been searched at the airport EVERY single time I have flown.

I was just searched for explosives, stopped in the next line, questioned as to why I had prescription blood pressure medicine, and why I had a car stereo in my bag.

I don't know if it is the beard or what, but I should not be picked for the random searches over 80% of the times I board a plane.

United Nations -- Iraq -- Weapons Inspections

[webword]

This doesn't surpise me at all. On National Public Radio today (All Things Considered) a researcher was talking about the best research tool for tracking down weapons of mass destruction: a 4" x 4" cotton swab. They run the swab over almost any surface and can detect radioactive material to the level of 1 part per billion. Geeez.
--
Trade it on Trodo!
http://trodo.com

It's hard to check for dirty bombs

[infolib]

A "dirty bomb" could be made out of alpha-active material. Alpha-radiation (He nuclei) will be stopped by a few pieces of paper. If the material is in a suitcase there is no radiation outside.

When the material is spread by an explosion, a fire or some other way, people will inhale it and it will stick in their lungs, giving them a huge dose of radiation.

Radiation != Chemo

[FakePlasticDubya]

When I submitted this story:

2002-12-06 18:34:29 Radiation Treatment Patients Set Off Subway Alarms (articles,tech) (accepted)

The editors changed it, to Chemotherapy... which is obviously not the same... Oh well.

Re:The dose makes the poison

[The Tyro]

You are more right than you know.

Even though some of the chemotheraputic agents we use these days are related to chemical warfare agents, it's the dose that makes or breaks you.

Virtually all chemo agents have one thing in common... they are some kind of metabolic poison. They are nucleoside analogues, directly denature DNA or proteins... whatever. Because of this, they are quite useful in cancer treatments, primarily because cancer cells divide at an abnormally fast rate, and are very metabolically active... ergo, these drugs will affect such cells to a great extent than normal tissues. Keep in mind, however, that some of your normal tissues are also rapidly dividing: bone marrow, hair follicles, intestinal lining. Ever wonder why cancer patients lose their hair and need blood transfusions? That's why, in a nutshell.

Don't let anyone tell you that chemo is bad/evil... that's bullshit. Unpleasant? Yes. Evil/bad/drug-company-conspiracy? No. Because of chemo, we have very high cure rates on some kinds of cancer... testicular cancer is a good example; very treatable with chemotherapy. But, like anything else, it doesn't work on every cancer, or every person... that's the other edge of the biological diversity sword.

Also, there are some chemo drugs that have a lifetime maximum dose... you get amount X and NO MORE... ever.

The dose really does make the poison, and that's not theory... that's real world.

Base rate fallacy

[hey!]

This is an example of an error in numerical reasoning called the base rate fallacy.

The base rate fallacy is trying to interpret the results of a test without considering how common the thing being tested for is in the population being sampled.

For example, suppose there is a medical test for a disease that has a five percent false positive rate. I then grab somebody off the street and administer the test, and he turns out positive. How certain are we that he has the disease? 95%? No, we cannot say without knowing the probability that any individual pulled off the street has the disease.

Suppose one person in a thosuand has the disease. There are two ways we can get positive results from the test. On the one hand he may actually have the disease (p = 0.001). If we sample 1000 people, one person will test positive for this reason. On the other hand he may not have the disease (p = .999). If we test 1000 people, 5% of the 999 (about 50 people) will be false positives.

So, of the 51 positives we'd expect to get, only one person legitimately has the disease. Instead of there being a 95% probability of the disease, there is actually only a 2% probability that a positive test indicates anything at all when applied to a random population. In order to apply the test usefully, I need some independen reason to suspect the person has the disease.

Even a slight reason for suspicion can alter the interpretation dramatically. For example, suppose I'm about 10% certain a person has the disease. If I tested 1000 people who met this criteria, 100 would test positive because they had the disease, 50. So if I'm 10% certain, then a positive test should make me 66% certain. If I'm 50% certain. then a positive test should make me about 90% certain.

A lot of public security measures suffer from the base rate problem. For example random drug testing doesn't tell you with much certainty that a person is doing drugs -- you really ought to test only peple you have independent reason to believe are using drugs. The only time widespread screening makes sense is if the base rate of the thing being tested is very high relative to the false positive probability.

This cancer patient situation is essentially similar. If we have reason to suspect that somebody is a terrorist, if he sets of radioactivity alarms it is very suspicious. If we have no such reason, then whether or not it is suspicious depends on the base rate of nuclear terrorism in the community.

Now it so happens that the false positive rate for this test is rather small: very few people are walking around radioactive for innocent reasons. ON the other hand, the rate of atomic terrorism in the general population is even smaller by several orders of magnitude.

This means that this particular alarm essentially tells us nothing about the people who set it off. It is probably not significantly better than a policy of randomly strip searching people.

However, this is not the only way to look at the problem. Suppose we knew for a fact that there was going to be a suicide dirty bomber somewhere in the city. Screening people in the subway might effectively prevent it from happening in the subway, either by deterring the bomber, or by catching hime, at the price of also catching hundreds of innocent people.

I think the take home message of this is that we should not use such systems on a routine basis; in cases where we have good reason to do so, we should remember that while if there is a terrorist he'll be culled out by the system, any particular individual culled by the system is not significantly more likely to be guilty than any randomly selected person.