I have a lot of experience - several years professionally - with mental illness and drug addiction. Twice (that I can recall), I have made potential diagnoses in my answers regarding people the OP is having problems with.

I was asked about the wisdom of doing this by a moderator, so I stopped, even though I have pretty fair confidence in my ability to diagnose given enough information (and I would not diagnose if that information was not presented in the question, either deliberately or unintentionally.)

One thing I would not do, however, is to try to diagnose the OP. For one thing, it's unprofessional. Would you want a bunch of people on the site to read that you were probably a narcissist? Even though it's only an SE [gr]avatar? Probably not. Secondly, it's tricky. It's easier to diagnose someone that a person is talking about (when they have no reason to misrepresent facts) than it is to diagnose the person who's talking (who is probably presenting themselves in a selected light).

Yet on several occasions, I've seen this done, and have flagged a moderator about it, but nothing was done, most recently, on this question: How to deal with anxiety/control issues in relationship.

In this answer, the user suggests that the OP may have Borderline Personality Disorder (I doubt he means Bipolar disorder.)

My flag was deemed helpful, but the answer remains, unchanged. (The underlying cause of BPD is an intense fear of abandonment, but there's a lot more to it. Lots of people have abandonment issues but don't have BPD.)

What do we as a community want? Do we diagnose or not? (N.B. When I diagnose, I cite references. This should be the minimum requirement.) I'm ok with not diagnosing, but not with the inconsistency.

Edited to add: I would have edited it, but, in case no one has noticed, I also answered that question, which would make it entirely inappropriate for me to edit. So the diagnosis has remained (as has the double standard.)

  • 5
    On a site like this, a bunch of hammers will find nails.
    – user5547
    Commented Oct 18, 2017 at 21:32
  • 4
    @Walfrat - I agree I should not make a diagnosis. There are no legal ramifications (individually or corporately) of making a diagnosis on SE (the lawyers had at it before Health.SE went public, precisely because the professionals on that site were concerned about it.) Commented Oct 19, 2017 at 7:36
  • 1
    We had a similar discussion over at The Workplace Meta.
    – David K
    Commented Oct 31, 2017 at 17:22

3 Answers 3


As a rule, I would lean towards not. While you may be a medical professional, and have all the surrounding education and experience, most users probably don't.

Unfortunately there's nothing stopping users from claiming professional status, training, etc. when they don't actually have it, and we don't really have a great way to verify someone's claims. (If I remember correctly, you may have already commented on one question because the OP had claimed a professional status that seemed to contradict their current question...)

It's one thing to mention that something sounds like X, for Y and Z reasons, but I would try to discourage actual diagnosis. It's not really what we're here for and there's a pretty big risk for abuse.

  • 6
    I agree with this answer, which is why I stopped. The user I called out, though, I already suspected was not what he claimed he was because of questions and answers I'd seen across a couple of sites that didn't quite seem right. He might be a student, he might be new, or he might just be misrepresenting himself. You're correct that there is no way to verify one's creds on SE. So if you're going to claim creds, you'd better damn well know what you're talking about. Otherwise the best answer is "seek treatment". Commented Oct 18, 2017 at 16:38
  • +1 from me as well. I fell into that trap in one of my posts but have refrained from doing that since then. But it's pretty hard not to when all the symptoms are there and I have tackled these issues for biggest part of my career.
    – Xander
    Commented Oct 25, 2017 at 10:51
  • @Xander - You must have an interesting career when situations like this arise. Commented Oct 26, 2017 at 15:47
  • @anongoodnurse has Catija approached you about the sexual harassment post?
    – apaul
    Commented Oct 26, 2017 at 15:51
  • @anongoodnurse She seemed to think that our answers to interpersonal.stackexchange.com/questions/5804/… fell into the category of diagnosis. Had a lengthy argument with her in chat about it. Figured it was worth giving you a heads up, so you could be prepared. –
    – apaul
    Commented Oct 26, 2017 at 16:17
  • @apaul - No, I haven't heard anything. Thanks for the heads up. Commented Oct 26, 2017 at 16:48
  • @anongoodnurse The perks of working in a hospital with the biggest psych ward in the country, never gets dull!
    – Xander
    Commented Oct 27, 2017 at 5:52

Do we diagnose or not?

There is a third possible option and this may actually be a rather safe and useful one for a user with expert knowledge like you, anongoodnurse. This is the case, when not enough information is provided by the OP and/or the OP has a suspicion:

Use your expertise to not diagnose.

What I'm referring to is How to constructively engage a female friend with possible body image problems?, where the OP saw a friend after a long time again, and noticed that she became thinner with every day. His concern was, that she may have an eating disorder.

The OP gave some information (in the question and in comments):

  • I did notice she left more than a few strands on the pillows.
  • She complained of being tired quite frequently during her visit.
  • She also has noticeably sunken cheeks, and her ribs were visible through a thin tight fitting shirt she was wearing.
  • The weekend she spent with me she ate what I considered to be a smaller than normal portion (ie for dinner she ate ~6 chicken wings and a handful of berries and yogurt for dessert after having a small salad for lunch) but she didn't seem to be avoiding food or uncomfortable at meal time.
  • I don't think she has any friends/support system in her new city and has minimal close friends in this country at all. Her family is also still in her home country;
  • She had her advisor tell her that "you can't always rely on looking so hot to get ahead" She was quite upset by that episode since it embarrassed her in front of her cohort, but that was something else I didn't know how to broach.

I am not a medical professional, but I only have personal experience with one type of eating disorder. Since this is quite a sensitive issue for me, I hesitated to answer at all and only did so after a while, when answers came from people who explicitly had no experience with eating disorders (and one of them included the advise "Tell her you do not want to be around her when she engages in this behavior because it is to hard to watch her self destruct."), just with once having been quite thin.

Your expertise may not only come in handy in enabling you to tell what it is (or not) with greater accuracy, but also to show that an array of symptomes can have different causes. E. g. I may panick and write "Bob has fever. He must suffer from flesh-eating bacteria!" and then an expert comes along and says, that it could also be something else and give helpful advise.

Or in this example, an expert may be able to use the background information (social isolation in a foreign country, stress at work) together with the physical symptoms (hair loss, tiredness, weight loss) and conclude that it may be an eating disorder, but could also have a physiological cause, be mainly related to stress or whatever.

Basically, to use your knowledge not to narrow it down to one diagnosis, but to widen it to more possible diagnoses.

  • Thanks for this insight! Commented Oct 27, 2017 at 14:17

It is fundamentally difficult and not recommended to make a definitive diagnosis based on information supplied in a written account online, especially for mental health disorders, and even an expert health care professional (whether physical or mental health) needs to make a thorough face-to-face personal evaluation before pronouncing a verdict.

Moreover we probably do not derive any practical benefit by making a 'looks like BPD' 'looks like OCD' diagnosis since the management of the condition then requires offline professional help anyway.

So it is not appropriate for a user to 'diagnose' OP or another person in their answer. In such cases it would be much better to comment that OP or someone else described in the question needs a thorough psychological evaluation by a qualified mental health professional.

If a user thinks the question as presented needs expert help and cannot be given an interpersonal solution, then the user can also vote to close it as 'off-topic because it needs professional help which is beyond the scope of this website.'

Related meta questions:

We are not in a position to make a diagnosis based on a written account submitted online by or about an unknown person who is not available for detailed direct assessment. Nor is Interpersonal Skills a website intended to pronounce professional opinions or diagnoses when people ask their questions here. If users see a diagnosis in an answer they should flag it for moderator attention. When moderators review such a post I sincerely recommend that they should edit the 'diagnosis' part out of the answer and inform the author not to roll back the edit.

  • 2
    Sorry to disagree, but it is not "fundamentally impossible" to make a definitive diagnosis with information provided by a user. The more information, the better the chances of making a correct diagnosis, but words convey a lot of information. "65 yo WF fell onto outstretched hand off sidewalk into street, immediate severe wrist pain. Will not use hand or arm." I already know that the 65 yo woman has a distal radial fracture without seeing the patient or the xray, no exam required. Treatment, however... that I need to see the Xray for. So don't underestimate the power of knowledge... Commented Oct 18, 2017 at 19:30
  • 2
    ...or words. If you don't know medicine, it seems like a much harder field than it is, but after 7 years of training and three decades of practice, surprises don't abound. At least 15% of the time, I could make a definitive diagnosis from the nurse's chart alone (just words and vital signs.) At least 60% of the time, I could guess correctly. We have a saying: You never want to be an "interesting" patient. So, please, let's be realistic, not hyperbolic. I do not claim any knowledge of programming, so I'll never make definitive statements about that. Commented Oct 18, 2017 at 19:37
  • 2
    Also, the question isn't about when to close questions. It's about answers. Commented Oct 18, 2017 at 19:38
  • 1
    Well, I was mainly talking about mental health diagnosis which is much more difficult to definitively establish without a detailed face to face assessment of the person. A written account on a SE site just does not convey sufficient information to make a definitive diagnosis. About distal radial fracture you are probably very right but we don't get too many such descriptions on IPS.SE (Note: I have interpreted your question about 'diagnosis' as your opinion that OP or another person 'needs professional help.' That is a reason to close a question if we cannot give an interpersonal solution.) Commented Oct 18, 2017 at 21:25
  • (contd.) I don't approve of any user making a definitive diagnosis based on written information supplied by an unknown person about themself or another person. Note too (as you said in comments) we cannot verify the credentials of any user who claims to be an expert. Now edited my answer to clarify that it is not appropriate to diagnose OP or another person in an answer -- if the user feels that the person needs professional help they could suggest it in a comment and vote to close the question for that reason. If users see a diagnosis in an answer they should flag it for moderator attention. Commented Oct 18, 2017 at 21:55
  • 1
    Part 3: I have modified 'fundamentally impossible' to "fundamentally difficult and not recommended (...) especially for mental health disorders" based on your objections. Also added: "when moderators review such a post (flagged as containing a diagnosis) I sincerely recommend that they should edit the 'diagnosis' part out of the answer and inform the author not to roll back the edit." Thanks for the feedback that enabled me to improve my answer @anongoodnurse. Commented Oct 18, 2017 at 22:42
  • I agree, a big complicating factor with diagnosis here, online, is that it's particularly difficult to distinguish which parts of an OP's question are completely accurate and accurately understood, which are unintentionally misleading or misrepresented, and which are outright fabrications. In a regular intake situation there are physical and body language cues that the patient is actually a 65 yo WF and not a 13 yo WM, or that maybe you should probe into whether she was pushed off that sidewalk. Those kinds of judgment are much harder based solely on an anonymous written description.
    – 1006a
    Commented Oct 19, 2017 at 15:22
  • All the more so for a psychological situation @1006a: any OP asking an interpersonal question will almost certainly tend to picture themself in a positive light, and their descriptions of others may intentionally or unintentionally distort or suppress important information to show that individual in a positive or negative light, depending on whether OP likes or dislikes that particular person. In short, anybody attempting a diagnosis on this basis is forced to depend on OP's view of the situation. Commented Oct 19, 2017 at 20:03
  • @EnglishStudent - I think the vote is in, but again, you underestimate the power of knowledge + words. If you are not a doctor, you will have no ability to generate a differential diagnosis. If you have no training in psychology, you will not recognize a classical hebephrenic schizophrenic when they befriend you. It's not an unsolvable great mystery. Most of the time, there's not enough info, but sometimes there is. Commented Oct 23, 2017 at 14:06
  • "It's not an unsolvable great mystery. Most of the time, there's not enough info, but sometimes there is." __ OK, very fair assessment. But the key statement as I see it is "I was asked about the wisdom of doing this by a moderator." What I don't understand is why your later flag on an answer containing a diagnosis did not lead to any action, @anongoodnurse. Commented Oct 23, 2017 at 23:27
  • Because I answered that question as well. An edit by me, as another poster, would have every appearance of impropriety. Which is why I haven't touched it. And am still wondering why the mods haven't either. Or anyone else. Commented Oct 23, 2017 at 23:51

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .