Hmm...ethics...

The potential for identifying putatively anonymous individuals is astonishing. And so, sanitizing records can be very challenging. As an anecdote, let me share a notable finding from Latanya Sweeney of Carnegie Mellon:

"...87% (216 million of 248 million) of the population in the United States had reported characteristics that likely made them unique based only on {5-digit ZIP, gender, date of birth}. About half of the U.S. population (132 million of 248 million or 53%) are likely to be uniquely identified by only {place, gender, date of birth}, where place is basically the city, town, or municipality in which the person resides. And even at the county level, {county, gender, date of birth} are likely to uniquely identify 18% of the U.S. population. In general, few characteristics are needed to uniquely identify a person." Latanya Sweeney, Uniqueness of Simple Demographics in the U.S. Population, Laboratory for International Data Privacy (2000), available at http://privacy.cs.cmu.edu/dataprivacy/papers/LIDAP-WP4abstract.html

For more, see her work on "k anonymity:"

http://privacy.cs.cmu.edu/dataprivacy/projects/kanonymity/index.html

describing a case (even without personal identification) on an open board

Question - a patient comes to you, asking if this blog item refers to him. He recognizes that details have been changed, but he're suspicious. What do you reply?

And does that change if he makes it clear up front that he will be extremely upset if he is the patient discussed?

A few of my patient anecdotes are based on actual individuals, most are an amalgam of patients. The individuals are usually deceased and personal friends of mine who did not have a doctor-patient relationship with me.

A better policy for cases who are not amalgams is to get explicit permission from the patient (unless they're dead, in which case the privacy consideration still exists, but differently).

If a patient is not identifiable, there should be no ethical issue.

I can see that when discussing a case with colleagues, but on an open board (especially one where you are identified) isn't there a prospect that the patient hirself might read the post and realize that s/he is being discussed?

Would it be different if the case were presented by Orac in his plastic box persona instead of his secret identity?

This gets complicated.

Case presentations are fine and as far as the ethical benefit to a patient, the loophole is that it is for education - considered an excusing pursuit.

All the case presentations I've written so far have been amalgams, but when labs came up I used real labs because otherwise the risk is things don't make sense. Patients don't know their labs (and shouldn't care about an arterial blood gas unless they're totally wacky) so risk of identification or self-identification is nil.

It's also of substantial benefit for medicine as a whole to demystify how we operate and make it clear to people the evidentiary basis for our decision making. So my feeling is, case presentations, even to a public/lay audience = good.

Patients don't know their labs (and shouldn't care about an arterial blood gas unless they're totally wacky) so risk of identification or self-identification is nil.

I dunno; I like to see my labs as much as possible, because those numbers are kinda cool in a geeky way. ;-) But I wouldn't mind if my doctor discussed my case publicly anyway, so I'm not a good person to base those kinds of ethical guidelines on. Because I know a lot of people *do* have reservations about that sort of information becoming public.

What strikes me as ironic this morning is that we're having this discussion at the same time as the push for universal electronic health records is making it all moot.

Ten years from now, anyone remotely associated with the health care system will have access to all of our health records anyway.

Dr Novella posted an argument (on his "chiropractors invade my blog") which I think fits in this case:

"First, I want to point out that I greatly appreciate when readers post contrary opinions on my blog. Agreement (while also appreciated) is boring and does not advance the discourse. If I am missing an important piece of information, or my logic is flawed, I love having it pointed out - that way I can make corrections. I want my arguments to be both valid and as complete as possible, so corrective feedback is very valuable.

I also greatly enjoy sparring with those who have used logic and evidence to come to a different conclusion from me. A constructive discourse can only be a learning experience all around. I find I understand a topic much better after having to defend my position to someone adverse to it and who is using every argument they can muster to take it down."

PalMD may have 99.9% certainty about the diagnosis and how to proceed with treatment, but adding another few more 9's is a definite benefit to the patient. He would never have learned of a possible nitric oxide connection without posting about it. Low NO can cause multiple other symptoms, now PalMD knows to watch for those in his patient. ;)

He may get notice of quack treatments and remedies that his patient might be exposed to, and if he is forewarned about them, he can offer better recommendations to his patient on why they should be avoided. Quack treatments are not something PalMD will find anyplace but on the internet and then only by being a quack magnet.

In a journal, a case presentation is (or should be) exact and precisely about the single case or cases being discussed. It is inappropriate to confabulate details to preserve anonymity. Deliberately confabulating details in a scientific presentation is scientific misconduct. It is scientific fraud. I may be seen as a purist in this, but that is my position. In a blog, confusing the details is appropriate so long as that is mentioned. A blog is a casual communication, something to be taken with a grain of salt.

To me, ongoing self-education and education of other practitioners is an essential part of the practice of medicine. Supervision and being supervised is also. As a patient, I have the expectation that if my clinician has doubts or concerns about my care that he will get more information as needed. Even without doubts, double checking and triple checking is good practice. That can only happen in the context of the clinician sharing some patient information on some level with others. To me, it is unreasonable for a patient to expect that their patient information will never be shared with other clinicians in a medical context. To expect that is to have unreasonable expectations that are inconsistent with treating the clinician as a human being who is trying to do the job they are being call upon to do to the best of their ability (which is all that can be asked of them).
If the patient can identify themselves and it makes them freak out, that patient shouldn't be used as an example by the clinician.

It is unfortunate that employers will discriminate against employees due to medical conditions they have. That is one of the major problems with how health insurance is paid for in this country, by employers, and where a low claim rate means lower premiums. Excluding people with conditions from coverage doesn't save on treatment cost, it simply shifts it to someone else.

In computer security there is the saying "obscurity is not a solution to security". Keeping the details of how things function secret doesn't make them secure. Systems are made secure when even with complete knowledge of how they function they cannot be compromised. That is how supply and delivery of health care needs to be, so that no one can profit by knowing the health status of another.