Hacker News Reader

Comment by makehistory | Hacker News Reader

It takes an average of 4-5 years to initially diagnose autoimmune diseases. https://nationalhealthcouncil.org/blog/a-major-health-crisis...

When my first symptoms appeared, I initially thought it was just aging, despite being a healthy 28-year-old man who enjoyed exercise and maintained a proper diet. I noticed health issues arising even though I took better care of my exercise, diet, and sleep compared to my friends.

Hospitals provide fragmented information across departments, which can extend the diagnosis period for rare conditions.

I believe medical services can be delivered through personal medical data + medical domain knowledge + intelligence (excluding physical intervention). AI can process more data than doctors have time to review, contains more knowledge than one person can study, and its intelligence will continue to improve.

This gives me hope, and I want to help others who could benefit from this hope.

View on HN

>AI can process more data than doctors have time to review, contains more knowledge than one person can study, and its intelligence will continue to improve.

The key part of that statement is "continue to improve" we have no way of knowing where we are on the sigmoid curve of improvement.

But right now. Today. Tools like this that spit out "you might have an autoimmune disease" aren't helping. Until they improve dramatically, the most likely outcome is that doctors get even more overwhelmed with people who think they have an auto immune disorder, but don't.

I think a lot of people have lost trust in doctors. You need to do your own research if you want to get the best treatment (especially when dealing with dentists). I don't think it's unreasonable to use AI as part of that research.

People are really bad at doing their own research and diagnosing themselves.

I'm much more medically knowledgeable than the average person. I'm reasonably smart, and interested in the subject. And then I married a doctor.

But even given all of those things, I'm very bad a diagnosing myself. I once swore that I had Giardia. I read dozens of papers and case studies. A close family member of mine was also having some similar GI symptoms at the time, and I even calculated the probability that our symptoms were independent.

After that was ruled out from testing and I kept having stomach pain, I did more research. I was pretty sure I had stomach cancer.

Eventually I got scoped--gastritis. Took a PPI for a little bit and it cleared right up. Gastritis wasn't even something that I had considered. It showed up in my search results, but because it wasn't a serious problem, I didn't focus on it.

My sister had an experience where she went to the doctor multiple times because she had felt something “off” about her health, fatigue, small aches in her chest.

Multiple times she was turned down for breast cancer screenings, she was told that she was too young (33 at the time). It was not until she had requested that the doctor write a note of their denial, along with reasoning, that they finally gave her some screenings.

It was stage 2(3?), it had affected several lymph-nodes, but she was lucky that she persisted and did not just say “oh ya it’s fatigue”. When I ask her about how she knew, it’s always the same response, that it was just a feeling, that she knew something was not normal.

Yeah it's definitely possible. And it's terrible if you or someone you love is in that situation. But for the vast majority of people feeling off, tired, and some small aches is going to be nothing.

There are costs to testing, financial and otherwise. For example you take someones family history, and other risk factors into account. There is a 1/100,000 chance that they have cancer. Then you add in the symptoms, fatigue, and minor aches and pains. That brings it to a 1/50,00 chance.

We don't have the money, technicians, or machines to run an MRI on everyone who shows up to the doctor with a problem that has a 1/50,000 chance of being cancer.

Then for many tests, with such a low prior probability, a positive test has a higher chance of being a false positive. So now you patients with massive anxiety out over a positive that has high chance of being a false positive. And you need to send them or further more invasive anymore expensive tests, that have real risks of harm.

> We don't have the money, technicians, or machines to run an MRI on everyone who shows up to the doctor with a problem that has a 1/50,000 chance of being cancer.

A guy launched a car into space; I think we have all this stuff but we — as a society — just don’t find the business case for “poor people being alive” very compelling.

There was a period where doctors would run every test on the list.. They stopped this because even if everything were free, each test has a false positive rate with a psychological cost to the patient. If anything the US has a problem with not accepting that everyone is going to die and asking them to go on to a road makes this more likely, (thus costing more than the value of the appointment.)

Being told that a doctor won’t do a test because it’s not worth it has a psychological cost that is higher than being told that the screening test came back positive and they need to run some more tests.

You are incorrectly calculating the “value of the appointment” from precisely the wrong viewpoint. Remember that healthcare is about the patient’s health, not the bill.

No, you're missing the point. Over treatment is a serious problem in healthcare. There is always some risk of iatrogenic harm, and for many tests the risks outweigh the benefits. Many tests don't even have a simple positive or negative result: often the outcome is inconclusive.

https://www.bloomsbury.com/us/price-we-pay-9781635575910/

This is such a wild non-sequitur! A single person doing a single wildly expensive thing is not comparable to hundreds of thousands or millions of people doing a far less expensive thing. The orders of magnitude don't cancel out; the 5-6 orders of magnitude in number of "people doing expensive thing" outweighs the 3-4 orders of magnitude difference between the costs of the things you are comparing.

It's also just so wildly irrelevant to the topic at hand

This is why proper differential diagnosis skills is important - something a layman would not understand. You must consider ALL possible conditions that your symptoms match, and then work your way from the most common to the most unlikely. AI is remarkably pretty decent at doing DDX.

There have been clinical decision support systems that were pretty decent at doing DDX for decades, well before the current crop of AI LLMs. The issue is getting relevant data about patient history, symptoms, signs, etc into the system and then convincing doctors to actually use it. For most simple cases, using any form of clinical decision support (whether it's based on AI or simpler deterministic algorithms) just slows them down and they don't have time to waste.

A big challenge is that most doctors can't operate outside of the most common conditions and differential diagnoses of their specialty and have to punt patients.

Of course, to the doctor's defense, patients are not an easy population to work with, "Why can't I eat candy and cake 3 meals day and my diabetes is getting worse?"

As someone also in a similar position to you, I think the challenge for the non-medical community is that medicine/health frames things in terms of some entity, instead of failures in a systemic set of interconnected processes.

"I HAVE cancer. I HAVE this autoimmune condition. I GOT a cold."

So people are in search of a label and the "thing" to get rid or fight, but imagine if all technology problems were labeled similarly. Rendering-itis. Memory-itis. PCI-Bus-itis. Imagine someone coming to a person going, "I think I have keyboard-itis, because when I type a key into the keyboard, it takes a couple seconds for the character to show up on the screen." When in reality, as the engineer, your intuition is that there's something amiss in the auto-complete microservice. Perhaps that service needs to be scaled up or something else is DOS'ing the service, Perhaps the autocomplete service is running across the country instead of the same datacenter, so you know which tests to run to isolate and solve the source of the problem. Medicine is basically that now.

Medicine has developed to the point where they have several hammers where if bloodwork == X || symptoms == Y, then A or B or C drug/procedure. So doctors are mostly the level 1 engineer after the level 1 tech support rep on a script.

----

Most medicine is merely a collections of latin labels, and unless the conditions map to well-known treatments, most doctors are in the dark. You were lucky that you responded to PPIs.

Gastritis is "inflammation of the GI lining" and the use of a PPI, is similar to a common second-line option like "reset the software settings" after the first suggestion of "reboot your computer." They scoped you, because they had no clue what was up. Scoping is basically like lighting up the debugger. They go in with a camera, and basically found nothing but inflammation, thus came the PPIs. They were mostly ruling out cancer or growths or other weird things like outside items. (ie ate a non-digestible item).

Basically, doctors are relying on pharmaceuticals to help out for a couple days/weeks, while the body addresses things on their own such that they don't become necessary anymore. That's probably most medicine. As most pharmaceuticals are symptom management, doctors really only have tools to help someone manage symptoms or alleviate the pain. They are just practiced in that troubleshooting process, the way tech folks get deep into the stack of the architecture of computing, networks, etc and can troubleshoot issues in the tech stack.

----

The way that there's only a handful of the most brilliant engineers who deeply understand the stack top to bottom in such a way that they can thread their way to "that's a problem in this specific subsystem and the latest change now makes the datastructures exceed the L3 cache size of that node and that's causing an additional 10 millis degradation." There's really only a handful of doctors who are intimate enough with the all the subsystems of the body to be able to drill into what's going on with a person, otherwise most operate on rules and mentally null pointer when they run into something that doesn't fit into their flow chart.

Outside of very specific genetic conditions, incredible amounts of sickness out there could be alleviated by supporting the biochemistry (limiting reagents) of the body, so its processes can keep it in tip-top shape.

IMHO, the REAL problem is not the individual doctors, it's the system we seem to have implemented to dole out "health care" as if it's a product that has to be rationed according to some metric.

I've met doctors that are passionate, but are hamstrung by "the process". The hospital has policies, the insurance company has policies, local health departmen thas policies, whatever it is. I hate that "some process" needs to be in place to keep things in check, lest we end up with patients abusing things, or doctors abusing things (SURGERY ALL THE THINGS, for example).

I think it's "the process" we have lost trust in. Individual people really do want to do good things for each other. But "the process" sadly has an awful lot of inertia, and doesn't make it easy to change.

Precisely right. The corporate takeover of medicine is dangerous. Private equity firms are buying up hospitals and then illegally shutting them down to cash out. (Yes, I’ve seen this happen first hand. I’ve also read about other cases.) Another example of small local governments getting railroaded by corporations that have more money and lawyers.

How does shutting down a hospital allow a PE firm to "cash out"? Let's see some actual numbers on that, not vague claims about tax write-offs. Sometimes PE firms take on too much debt as part of acquiring a hospital and paying out a special dividend to shareholders. But if the hospital later fails due to being unable to make debt payments then the PE firm takes a loss on their remaining equity stake; they would always prefer to keep it running as a going concern so that they can sell it for an additional profit later.

There is a legitimate issue with concentrated ownership of provider organizations in some regions. More antitrust enforcement is needed to protect consumers but unfortunately that seems unlikely.

I would say in your toy example the hospital has been shutdown due to the PE firm "cashing out" excessively

> How does shutting down a hospital allow a PE firm to "cash out"?

https://www.theatlantic.com/ideas/archive/2023/05/private-eq...

https://www.ineteconomics.org/perspectives/blog/private-equi...

What would you propose as an alternative? All healthcare systems have to ration care because resources are limited and demand is effectively infinite. We can discuss the process used for rationing but it has to be done on some reasonably objective basis.

I had a conversation with my friend. "Look at this new, incredible CPU cooler tech! It's sooo good at removing heat."

"What if you didn't generate the heat in the first place?"

Personally, I have a hypothesis that there are systemic areas we need to address in our food supply to lower healthcare demand. This will sound kooky, but if we can get more sulfur and trace minerals into the food supply, we could improve the general well-being of massive amounts of the population. We really need to look at doing more research into additionally fortifying our foods.

Here's a nature article on the reduction of sulfur from the food supply: https://www.nature.com/articles/s43247-021-00172-0

NPK farming has ensured that the food supply doesn't have a rich set of additional minerals outside of (N)itrogen, (P)hosphorous and (K)potassium.

Then when looking at sulfur-containing molecules, the research on them tends to herald them as miracle-health molecules: MSM Taurine, B1, NAC. All bring sulfur into the body.

Then there's companies like Buoy drops that offer 87 different trace minerals and folks are swearing by the effects of them. Asian countries consume things like seaweed and get trace minerals like this from the the sea water, but a the standard american diet is really lacking in a lot of additional nutrition that would be productive to health.

The majority of healthcare spending goes to chronic diseases primarily caused by lifestyle issues. Collectively we are a nation of obese, sedentary substance abusers. If we could do better on prevention and delay the onset of chronic diseases like type-2 diabetes, hypertension, and heart failure then that would largely fix the demand problem and leave a lot more resources available for patients who randomly get cancer or multiple sclerosis or whatever. But this is a policy issue that goes way beyond the healthcare system and no one has a practical solution. Everyone loves to give health and nutrition advice to others, but we're all undisciplined and do unhealthy stuff even when we know we shouldn't.

https://peterattiamd.com/outlive/

As for the food supply, I have no doubt that many people are impacted by subclinical micronutrient deficiencies, whether it's sulphur or zinc or D3 or something else. But the evidence indicates that's less important than excess macronutrients for most people.

I'm a big fan of Attia and familiar with his work. That's fair that excess calories is among the major contributors to Healthcare and addressing that would probably be among the largest impact items we ought to tackle.

I wonder though if with the advent of GLP1 drugs, if obesity may already be a solved problem, but merely a matter of time?

I imagine, as with anything, perfect doesn't need to be the enemy of good enough. A huge multifaceted issue like obesity could be addressed from many angles.

What's actually interesting is that there exists a model for solving obesity. Mostly it's a matter of having the will and prioritization to solve such a problem.

Japan has an obesity rate of 4%. By law they require every school to employ a professional nutritionist, which requires 3 years of additional training on top of their education degrees. In a particular example of 1000 kids, not a single child is overweight. Every nutritionist is required to design the school's food curriculum such that all meals are made from fresh foods and to be balanced according to all the food groups, and they forbid even the use of processed pastes. All meals are made from fresh foods on-site.

Then the kids are then taught how to eat well and to fully cover their nutritional needs from 6-18. This would be the non-ozempic way to influence a nation to eat well.

Americans already consume more healthcare than anyone in human history by a huge margin. A large proportion of that healthcare isn't really healthcare at all, because it does more harm than good.

If you're paying out of pocket, you have the right to demand as many tests and as many treatments as you can afford. If you're relying on an insurer to foot the bill, then that insurer has two choices - refuse to pay for some treatments, or watch costs (and premiums) trend towards infinity.

A lot of it is just that:

a) The knowledge is democratized on the internet (albeit often without proper context, and with somewhat limited accuracy)

b) You just care a lot more about your own, or a loved one's health/life and are willing to put a lot more effort than a doctor can or an insurance company is willing to pay for.

You may not have a medical degree, but if you're willing to put in 40 hours doing research and your condition is even somewhat unusual, you can probably exceed the knowledge of your doctor about your specific condition.

Regarding b) there's a term "advocating for one's health" and yes, doctors are fully aware of it.

>> you can probably exceed the knowledge of your doctor about your specific condition.

It depends on the doctor. That's why in the US patients first see a general practitioner, then are referred to a specialist. I've met some less knowledgeable doctors for my condition, but I don't discredit them for it. It's a difficult profession.

Plus "doing research on your own" is not a bad idea, but if one deep dives and starts believing ivermectin cures everything, well, I'd say they did poor research, and poor research can be worse than listening to a doctor.

"Do your own research" has turned into codeword for "Open yourself up to quackery." Nobody uses it to mean do actual research. It really means: Go onto YouTube or search and find all the various Herbal-this and Natural-that and Abc-Oils and Xyz-Supplements that confirm your suspicions that doctors are lying, and that managed to SEO their way to the top. Or, more recently, go onto your favorite political echo chamber and trust medical advice from pundits.

When I need my car fixed, I don't go to the local psychic or consult Fox News. I go to a mechanic.

This has a simple benefit: There are cases where better quality medical services can be received with less cost (time and money).

Why would this be downvoted? This is 100% true.

The medical shills are working hard on this thread.

>The key part of that statement is "continue to improve" we have no way of knowing where we are on the sigmoid curve of improvement.

Such a tired take. What we do know is we are not on the flattening part and literally every company working on these technologies is saying there is no end in sight.

>literally every company working on these technologies is saying there is no end in sight.

Literally every company making money selling Foobars is saying there's no end in sight selling Foobars?

Indeed, it’s tiring to have to keep correcting uneducated AI bros. Such is the consequences of unchecked capitalism finding a new shiny toy that it wants to suck the value out of.

The people currently enjoying a bajillion dollars of investor funding say the tool making them a bajillion dollars has no end in sight, you don't say...

If you've got evidence the curves are flattening lets all see it. Otherwise it actually does make sense to update when everyone at these companies tells you they aren't flattening and might even be bending upward.

> Tools like this that spit out

WebMD has been like this well before current iterations of LLMs. Pretty much any symptom you looked up would very meme-like say "it could be cancer". Only, it wasn't trying to be a meme. I always saw it more of a CYA on webMD's part though, without trying to sound authoritative like LLMs do.

You might enjoy the opening of the comic novel Three Men In A Boat, where the narrator reads a medical dictionary and finds he is suffering from every malady listed apart from- much to his annoyance- housemaid’s knee.

https://gutenberg.org/files/308/308-h/308-h.htm

Published 1889

The doctor's prescription is probably applicable even to this day.

Food, bitter beer, exercise, early bedtimes and "stop stuffing your head with things you don't understand".

> I followed the directions, with the happy result—speaking for myself—that my life was preserved, and is still going on.

I read the book as part of my English - II in school, but no part stuck to me as much as the Liver cure for children.

I've witnessed 4 medical emergencies, with 3 being life-threatening. Each case at different hospitals, and each incorrectly diagnosed.

The most dramatic was in 2009. My wife was admitted to emergency at 22 weeks pregnant with severe cramping, nausea, and severe IBS like symptoms. She couldn't keep food down, leading to daily stomach pumping and eventually IV nutrition.

The episodes started every few days but grew more intense and frequent. After 8 days, the first doctor seemed lost. When I asked about a second opinion, he lectured me that it was "his call to make, not mine." That killed any trust - I asked him not to return.

The next doctor, an eccentric but brilliant GI specialist, ordered new tests. I obsessively googled every result (partly to cope, partly to help). When he suggested a motility drug, my research found a case of spontaneous abortion linked to it. We declined.

Three weeks in, I noticed something odd in her lab work - different blood counts from different locations. More late-night googling led me to placental percreta, where the placenta grows through the uterus and attaches to other organs. Several ultrasounds proved inconclusive.

With daily episodes now and my wife deteriorating, we insisted on an emergency C-section despite warnings about our baby's lung development. During the procedure, even with an epidural, she went into extreme pain. Her eyes rolled back, the surgical team's demeanor changed instantly, and I was told to leave the OR.

Waiting outside for who knows how long. I knew the risks. Some noise down the hallway, a few nurses called me over. They were wheeling a humidicrib down the hall to the ICU. Little champs whole leg was no longer than my pinky finger. They took him into the ICU and sent me to the hospital room to wait.

The OB arrived an hour or so later and explained she was alive and they had to remove something. He took out a photo and and showed how the placenta had indeed broken through her uterus and attached to the intestine. He then pointed to a particular point and explained how it would have ruptured at any moment. The decision to push for the C-section likely saved both lives. Our son, never developed lung issues. Lots of luck.

This experience showed me how even amazing doctors, overwhelmed with cases and data, can miss crucial connections. In 2009, they didn't have the tools to correlate data the way I could with google search-fu.

AI diagnosis tools will save lives - they can make these connections faster and more reliably than either doctors or desperate husbands with search engines. And it will just improve over time.

Terrifying story and example. Im glad it worked out and I think there are a number of factors that this illustrates. One is the time and incentive differential between doctors and the patient. Im guessing you and your wife spent multiple magnitudes of order more time trying to get to find a cause. You have one wife and one unborn child, while doctors have dozens of patients a day. A doctor can never devote as much time and emotion as a patient. This isn't a condemnation, but a statement about reality.

Medical training and experience can reduce the impact, but not overcome it. It is a reason why someone can't entirely outsource their heath to the medical system, and something that I think people miss when they set medical expectations.

From the perspective of the medical system, it is a numbers game, where the optimal number of deaths is non-zero. More time and effort spent in one area means less in another. Saving one life might cost two others.

Because of this, I think it is most useful to discuss ways to increase medical efficiency. Better data transfer, better collation, better processes. AI/big data review has a place in this, but we should be aware of it's limitations, particularly on the data it doesn't capture.

Agree with you. Time is a huge factor. Another is determination.

I'm sorry you went through this experience, and I'm glad your son is alive and (hopefully) well.

I don't have the ins and outs of your case. I'm a student, not an obsterician, or neonatologist. That said, I do know delivering a 25 weeker is insanely risky even if perfect care is provided, for both his immediate outcomes and avoiding long-term disability. (80% of babies born before 26w have some form of disability e.g low IQ, blindness, etc [1]).

No sane obstetrician is going to deliver based on a family opinion like that - it is a true last resort. If you were wrong, or less lucky, as horrible as it is to say, your son would not be with us and the surgeon would be out of a job with the "what if" on their conscious (which trust me, all obstericians have anyway).

When your son needed to be delivered, it sounded like he was delivered. Frankly no neonatal respiratory distress at 22 weeks sounds insanely lucky to me; normally before 35 weeks it's a concern.

I can't say if this was a failure of the medical professionals in charge of your care or not, but from the information you've given us it doesn't sound like it. Tens, maybe hundreds of patients will have presented identically to your partner with a far less severe diagnosis. Getting their babies out early could have left them permanently disabled or dead, on the offchance it was percreta/accreta/increta: especially in the absence of definitive imaging, which they did attempt to perform. Medicine is not perfect by a long shot. Managing uncertainty is one of the many skills of senior doctors.

Medicine isn't as simple as a lot of people in this thread think it to be.

DOI: Medical Student

[1]: https://www.tommys.org/pregnancy-information/premature-birth...

No need for sorry, but I appreciate it. I had no intention of implying medicine is simple. Yet the point remains. Doctors cannot read all cases nor memorize all those they do. Unless they have skilled researchers standing by, relevant cases will be missed. If AI (or a human) can analysis that data then its perfectly reasonable to take it into consideration.

In our cases, her condition was deteriorating at a frightening rate. She could barely speak. Under such stress one must also consider the possibility her body giving up and aborting the pregnancy. Doctors had put cameras up and down and ultra sound etc. They were trying to keep her going as long as possible to give baby the best chance. I definitely would not advocate for anyone to make rash decisions in such delicate situations, and of course, my post omitted myriad details for brevity.

The cause was later determined to be the result of a D&C for retained placenta that was performed following the birth of our first child a few years earlier.

I don't know about autoimmune diseases, but I had someone I care about that had a rare (1/50,000) bone disease recently that Claude was able to successfully diagnose, suggest appropriate labwork, and give treatment and prognosis information for. I was seriously impressed. The doctors eventually arrived at the same conclusion, but it was multiple specialists and it took them weeks.

Do you have any data for how much time is spend running tests and exams on people without autoimmune conditions? That might impact autoimmune diagnostic timelines.

Anecdotally, I have heard that hypochondria is a problem in this space. So I would want to know the ratio of incorrect self-diagnosis to confirmed diagnosis, and what percentage of the system's totally human and technical resources are spent recognizing false positives.

I think that this sort of effort is probably a positive one. Patients should have access to their records are are more likely to be able to centralize data for analysis by a 3rd party - with all the attendant risks - more effectively than any other party. Secure systems without conflicts of interest - ie, insurance companies - to aid multidisciplinary diagnoses over longer time frames is a good thing.

But the risk of AI hallucinations and doctor/diagnosis shopping are real. This post was light on details on the diagnosis, how many rheumatologists they saw to get it, what tests confirmed it, and if the treatment has been effective in a way that cannot be skewed (https://pubmed.ncbi.nlm.nih.gov/30139289/)

That depends on how you define confirmed diagnosis and false positives. For a lot of autoimmune conditions there's no real solid diagnostic criteria. Like the patient has a variety of symptoms which could indicate a few different conditions. The doctor suspects it might be lupus so treats for that and the patient's symptoms seem to improve. Does that mean the patient should now have a confirmed lupus diagnosis? Hard to say.

I don't think hypochondria is the problem as much as there is a degradation in the overall health of the population and they do feel unwell, but the system is unable to diagnose them because they aren't "bad" by any quantitative standards medicine operates on (ie bloodwork/weight/etc being off).

It's like how the tech world did not notice or address "the website is slow" in the 90s until folks started writing about how the 95th, 98th, 99th, 99.9th percentile times affected user experience. Once we could link 95th percentile page latency to revenue/profit, all of a sudden, investing in performance became a priority.

I'm not sure that we have managed to link and make a case of linking GDP (or some other metric) to overall health of the population at the fringes.

I agree with you. And I have also been having health issues for years and doctors cannot really find what's going on. You know what's worse? They don't even care to find what's going on, they pay very little attention to me and just send me out with the same medication they give to everybody because they get 5 dollars for each patient that buys it. I cannot wait for many of them to be replaced by AI, it will be an improvement over the status quo.

Do you know how many times I've got prescribed a drug that would harm me more than do good? Do you know how many times I've got prescribed an antibiotic I'm allergic to, even after telling them I'm allergic to it, even after three different people wrote that down in many different places? Most modern doctors are criminals and I will never be ashamed to say that, because it's true.

Please keep doing what you're doing and ignore the naysayers, they're usually on some pharma's company payroll, sad.

Curious if you've tried any harder core dietary interventions?

I eat pretty much everything, in a healthy way.

But it could be something I eat, definitely, what I have feels like a sudden allergy to something that I haven't been able to single out from the environment.

>because they get 5 dollars for each patient that buys it

You probably won't believe it, but that's not a thing. My wife is a doctor and I can tell you exactly how she gets paid. The only real misaligned incentive I've been able to find is that she could pad her paycheck by spending more ordering tests.

But she her pet peeve is doctors ordering tests that aren't going to change the course of treatment, so she's definitely not going to do that.

I may have missed it, but what autoimmune disease were you diagnosed with? It seems rather odd that you wouldn't mention it.

Reply on news.ycombinator.com