Period-Tracking Apps Say You May Have a Disorder. What if They’re Wrong?

This is the exact reason the statistics around screening tests are counterintuitive. What's the big deal of over-recommending followups? It has to do with the difference between sensitivity and specificity. You want to catch basically all the cases (very sensitive) and not catch too many people who don't have it (very selective). Why does it matter? Obviously you don't want to tell someone they're fine when they're not, but it's also - if less obviously - bad to tell people too often they might have it when they don't. It leads to unnecessary work, which can itself be harmful at a low percentage. Multiplying the percentages (say, those who get biopsies needlessly times the risk of biopsy complication) might actually - for rare conditions - make the screening and follow-up almost as dangerous as the actual condition. For conditions that aren't that rare, or for which the tests are extremely unlikely to give a false positive, or sometimes if the risk of missing a diagnosis is way worse than the risk of a bad outcome from followup, it makes sense to screen everyone. But often, including for fairly common things like prostate cancer, screening is actually being de-emphasized. Even better, it turns out that a lot of tests work way better if they're only used on people who you have some reason to run the test, which naturally excludes lots of people who don't go to get checked out because they're fine. All of which is to say this should probably not be decided by app makers.

Important to note that these apps also sell all your data

I work with an international collaboration that developed the PCOS international guideline. The issue with PCOS is that it is a very common condition but about 70% of women that have it don't know they do. This is a real problem because it is the leading cause of infertility however, simply lifestyle changes and/or medical intervention can reduce the problem. So, it is important women know early to prevent complications. I applaud the two app companies for taking this initiative and really the worst that will happen is they will discuss their symptoms with a doctor. Not a bad outcome even if they don't have PCOS.

Digital Theranos? New leaks of HIPAA data?

Just be yourself. We are all different. The age of big data makes everyone feel they are deviant from the statistical center or from perfect or ideal. Ask yourself how good the research is. (bad). Ask yourself how good the pills are (questionable since they only have to surpass placebo by a tiny amount to get on market). Ask yourself if you are taking care of yourself before you ask someone else to do it because our own bodies are unsurpassed in healing themselves if we stop impeding them from doing so. And understand that it's not only oxycontin that is addictive to the body because it changes how it works without the medication. And what are the domino effects of taking medications (laughingly called "side effects" as if they are not even real). Right now Juul has a tobacco man as CEO and is in talks to sell itself to Phillip Morris (one of them). The first one of these apps that makes it big is going to become a subsidiary of some medical-industrial enterprise. It's not only Trump delivering disinformation and fake news. He's just riding the crest of a cultural wave and we know that don't we?

There are simple and complex potential solutions. Simple solution to issue. All "apps" need to display a disclaimer: "The conclusion reached by this app is NOT medical advice. You should, however, make an appointment to discuss the issue with your doctor." If the app displays a conclusion that a life threatening illness, disease, or problem is present, then insert "IMMEDIATE" between "an" and "appointment". If we assume that the worst the news the more likely the user of the app is to freak out, then perhaps a referral to a crisis center or resource would be appropriate (800 number), along with a mandatory "contribution" to the center's funding. Personally, I think the chance of huge problems arising from misdiagnoses are so large that such app's should be banned. However, I doubt that such would be valid, at least until the first couple of users commit suicide because of the misdiagnosis, which might reframe the issue.

"Apps" are not the answer to all of life's problems.

Like a specialist, an app will probably be most concerned about diagnoses within “its area of expertise”. Look for a holistic app.

“We err on the side of caution,” said Daniel Thomas, Clue’s head of data science. “Even if we think it’s more likely that they don’t have PCOS than having PCOS, but it’s one of these gray zone cases, we would also still ask them to see the doctor.” That sort of thinking is why overtesting and overdiagnosis is turning into a plague of our times. Telling people they might have a life-changing condition has consequences. In the UK it could take you three weeks to see a GP for a non-urgent condition, and that's if you go at all and don't just sit at home getting more and more anxious while reading up on your possible diagnosis on the Internet. In the US of course, you have to pay for your reassurance. Every one of us has some 'symptoms' that are no more than variations from the norm. The more tests we have, the more we need as every previously unknown variation spotted then has to be looked into just in case. If people are worried about their periods/acne/hair growth then they ought to see a doctor. If they aren't worried about them, having an app telling them they ought to start worrying on the off chance that something might be wrong is very unlikely to he helpful.