Sunday, July 19, 2015

Should Physicians Profile Our Patients?

Profiling in this country is highly discouraged and is illegal in many circumstances.  Anti-profilers decry this technique which, they argue, unfairly targets innocent individuals violating their rights.  Our beloved Transportation Security Administration, or TSA, is charged not to use profiling as a screening tool.  Personally, I object to this prohibtion.  While an individual’s rights are important, it must be weighed against the rights of the community.  Our cherished rights to free speech and assembly are not absolute.

I have flown on El Al airlines, whose personnel actively profile in an effort to keep its passengers safe.  Anyone who has been a passenger on this airline will likely agree with me that he has never felt safer on an airplane.  I have a right not be blown up, and if profiling further minimizes this risk, then call me a fan.
Last year, my mother, who appears as threatening as a school librarian, was patted down twice when she traveled to visit me in Cleveland.  I’m sure that the patter-downers were following the rules and regs, but this doesn’t seem to be a well targeted effort. 

A Non-controversial Profile

I’m familiar with the argument against law enforcement using profiling to reduce crime and protect public safety.  I understand that this can lead to abusive practices by overzealous police officers.  But I wonder if, as El Al believes, this can be a legitimate tool in law enforcement’s armementarium.  I’m open to the debate here.

I’m a physician and I certainly profile my patients.  I don’t have a single template for treating abdominal pain in all  patients.   For example, if I see a recent immigrant from China with stomach issues, a part of the world where gastric cancer is relatively common, this may affect the speed and intensity of my evaluation.   If an American born patient comes to my office with the same complaint, my response might differ.   If breast cancer is more common in Ashkenazi Jewish women, shouldn’t we factor this in when we are advising them on risk reduction?  Certain populations have different health risks.  Physicians are always trying to separate out patients who might warrant special attention. 

If you are trying to reduce a certain disease that is largely restricted to one segment of the population, doesn’t it make sense to target this segment rather than everyone?

I realize that health and law enforcement may not be analogous.  I also realize that profiling in law enforcement is a very sensitive issues, particularly for minorities who have been victimized by this technique.   But, if we abandon the procedure entirely, are we forfeiting a tool that could keep us all safer? 

El Al has a different view.  Here, in America, TSA is trying to detect evil stuff. El Al is trying to detect evil people.  Which makes more sense?


Anonymous said...

Everyone knows that if you have one particular ethnicoreligious affiliation, you'll be treated relatively well by El Al, whereas if you have another particular affiliation, you can expect to spend hours being treated with disrespect, have your data stolen, and maybe still not get on the plane. Yes, they're dealing with a genuinely high threat level, but it's also true that treating innocent people like dirt or limiting their access to travel because of where and to whom they were born creates future enemies. Americans who glorify group-based profiling (as opposed to flagging individuals based on their own personal behaviors) generally dislike the same groups that El Al's security folks dislike, and assume that the same groups would be targeted. Since in this country we have many more incidents of white right-wing violence than Islamic extremist violence, I wouldn't count on that.

So should you view your patients as group members rather than individuals? Sure, group identity can raise your level of suspicion for a specific condition, but it shouldn't affect the possibilities included in a differential diagnosis. You surely wouldn't disregard a breast lump in a non-Ashkenazi woman, while it could do equally severe harm to convince an asymptomatic Ashkenazi woman that she was a walking precancer. Studies have shown that black and working-class patients get less pain medication in ERs than middle-class white patients, because they are perceived as less sensitive to pain or as potential junkies. That is profiling too, and just like harassment at the border, it creates resentment.

EMyrt said...

just taking issue with this comment

"Since in this country we have many more incidents of white right-wing violence than Islamic extremist violence, I wouldn't count on that."

That meme seems to be making the rounds, but despite being in the NYT, it doesn't hold up to even simple methodological scrutiny.

See Megan McArdle's careful analysis here:

Anonymous said...

Well, Megan McArdle and careful analysis don't usually go together, but I read the column. Her analysis consists mostly of "September 11!" and of cherrypicking a few right-wing criminals who appear crazy, thereby dismissing the relevance of political motives to their acts, and hoping we will accept them as a synecdoche for hundreds of other violent white men. Meanwhile, we are to go on attributing all violence by Muslims to terrorism even if they are plainly barking mad (cf. underwear bombing).

Not convincing. Even if you raise doubts about an individual case here and there, it's not a meme but a fact that there have been many more violent crimes and homicides attributable to white nativist and Christian extremists in the past decade than to Muslim extremists. Then there are the endless random massacres of schoolchildren, theatergoers, etc., which, despite Megan's obsessive brooding over the Mooslim beltway snipers, are the vast majority of the time perpetrated by white males.

Now imagine treating all white males as suspected Eric Rudolphs or Eric Harrises in waiting. Of course that would be grossly unfair. But it is equally unfair to treat random brown people as suspects, given that in most contexts they are no more likely to attack you than a white person would be, i.e., extremely unlikely. White privilege simply consists of that state of affairs in which - to touch on the theme of our host's OP - you can seek medical attention without ever worrying about whether the doctor might secretly hate or fear you just for who you are, while Mohammed down the street can't always count on that.

Add this