What a sad and terrible loss, my heartfelt condolences to Sheryl and her family..
I wanted to add my thoughts about false positives and executive health checkups, which fall under the larger umbrella of screening. I conduct executive checkups fairly regularly, but i’m not promoting anything here. I just want to share a personal philosophy that you, as a patient, could adopt to reduce risk.
Firstly, it should be stated that the utility and performance of any diagnostic test that has a binary outcome (eg. diseased/healthy) depends on 1) its intrinsic test characteristics 2) the prevalence of the disease in the population being tested[1], 3) the available diagnostic/treatment options for positive people, and 4) the threshold chosen to call something ‘positive’ since you are virtually always measuring a continuous variable.
People talk about false positives often, but you should know that there is some room for judgement here. The typical threshold for PSA has historically been 4.0 ng/mL. Above this is considered a positive screen, and you are recommended further testing to confirm prostate cancer, which can be invasive and lead to complications (bleeding, nerve damage, etc). This kind of two-stage testing is a common theme. The issue is that PSA values in people with cancer and people without cancer are normally distributed and have some overlap. A value of 4.0 is in the overlap zone, and the vast majority of people with 4.0 will not have cancer. However, lets say hypothetically that guidelines changed and we now use 20. At this level, most people who test positive probably have something bad, and the risk:benefit ratio is in favor of confirmatory testing and probably in favor of treatment too. I won’t bother going into it, but there’s a fair amount of liability, psychology and economics at play when choosing thresholds. Biochemical tests like PSA do have some variance, but are mostly reliable. It’s more complicated with imaging and amorphous things like mammograms that have some operator variability in judgement (and hence more liability).
This was a long-winded way to say that information is not necessarily bad, but it’s what you and your doctor choose to do with it. There’s room for discussing probabilities, and the various available pathways. Trust is huge. Your doctor will act in your best interests when he feels safe from lawsuits, otherwise he'll be focused on not missing anything regardless of likelihood or cost. The financial misalignment of our system doesn’t help in building trust either. Most first-stage screening tests have very minimal risk involved. Things like checking your blood pressure, cholesterol, glucose, doing a treadmill stress test, reviewing your meds, vaccinations, etc which are part of these annual checkup programs are, on balance, usually good ideas.
There should also be some discussion about why the US uses more prescription narcotics than any other country. Some interesting data and charts here on global consumption and how it's changed over time.
Patients who know their conditions to the extent that they're able to have a well informed discussion with you at the visit are still quite uncommon. Most people google their diagnoses and glance at the top 2-3 hits, but very very few really dig deep into the guidelines, physiology or publications to see if what is being recommended makes sense. I think there's still a lot of faith being placed in experts, which in my opinion is appropriate because the likelihood of someone misunderstanding what they've read still exceeds the chance that they've uncovered ways to further optimize.
That being said, learning is always good. I always encourage people to discuss anything they've read with me. I print out guidelines, trial data, etc in the interest of leveling the playing field and complete transparency. There's really no point in trying to maintain an information gradient.
Now the article is not about understanding your disease, but rather, having complete access to your medical records- which is a related but different issue. Unfortunately, 99% of people have no idea what people are writing about them and I find that really troubling. Not sure why, but patients still seem to feel that it's somehow antagonist to ask to have a look. Like it's some threat or precursor to a lawsuit. It's not. When I turn the screen and ask the family to gather around so I can explain stuff, there's often hesitation or even surprise - "What? Can I really look at that? Is that ok?". Yes You Can folks.
Great advice, thank you- will make a note of these things for when we start deploying. At the moment, I think we will only need one EC2 instance attached to an encrypted EBS volume with the database on it. We're not using RDS. When you say encrypt PHI traffic between servers, you mean like EC2<-->S3?
HIPAA is surprisingly vague about the minimum standards for compliance, calling encryption "addressable" instead of "required"[1]. I believe this goes for both data in transit(HTTPS), and data at rest. I've seen some say that HTTPS is required but I can't find this on the gov site. My understanding is that if you choose not to encrypt, the burden of proof is on you should anything bad happen, to prove that implementing it was not "reasonable and appropriate". Since I can't think of any circumstances where sending plain text patient info over the internet is reasonable, i'll choose to encrypt. The other things for HIPAA compliance are complete audit logging so you can see who has accessed anything, and training of staff who have access to protected health info.
Most HIPAA recommendations seem to be a good idea to do anyway.
My wife is co-authoring a review paper on this topic with the group from Zhongshan. The rise in myopia prevalence is really quite staggering, especially among asians.
It's a long article, but the essential point is that insufficient time spent in bright light >10k lux (roughly corresponding to being in shade on a bright day) during early childhood increases the risk of developing myopia.
Normal eye growth requires dopamine which is released in a circadian fashion, which requires exposure to bright/daylight. Absence of this dopamine cycle may cause the eyeball to be more elongated, leading to refractive errors.
The precise number of hours needed to prevent myopia is difficult to say, but studies in schools with even 40-mins extra have shown benefit, and there's some evidence supporting 3+ hours/day.
Importantly, it doesn't seem to matter how much close-focal work kids do, such as computer use/studying, as long as they spend sufficient time in bright light. I'd consider setting up a study desk in front of a large window, and taking breaks at school outdoors instead of inside.
Yes, if you purchase the item using the card. It's an extended warranty. Quote from benefits section of the gold card:
Extended Warranty can extend the terms of the original manufacturer's warranty on warranties of five years or less. We match the length of the original warranty if the original manufacturer’s warranty is less than one year, and we provide one additional year if the original manufacturer’s warranty is between one year and five years. Shipping and handling costs for the purchase will not be refunded. Certain purchases are not covered, e.g. items with physical damage, damage as a result of natural disaster, software, motorized devices and vehicles and their parts. Other important exclusions apply. You will only be covered up to the amount charged on your eligible Card; coverage is limited up to a maximum of USD $10,000 per occurrence; not to exceed USD $50,000 per Card Member account per calendar year. You may be required to send the purchased item to us, in which case we will reimburse you for shipping cost. This product is in excess of other warranty plans that you have in place for the eligible item.
Generally he means addressing the problems "heavy car usage, not enough walking, and no safety for humans". Optimizing city design for people, and not for vehicles. This is generally done by increasing density, and requiring mixed-use buildings with commerce on ground level, and more public spaces and wider side walks, and dedicated lanes for bikes and public transportation.
Cars aren't going anywhere anytime soon in the US because sadly you can't function without one in the vast majority of places. It's far easier to implement this design philosophy for a growing city[1], but really hard to change the layout of a mature one, though it can be done thoughtfully [2].
It would be great to have the note kept in sync with the comments section on a file/folder when you Get Info on it so it's easily passed along to others. But i'm not sure if that supports formatting/bullets, etc.
Currently if a file is moved around, does it keep the note? Can you add a note to any URL, or only ones you've bookmarked?
I'm by no means a physicist but find topics on the nature of universe deeply fascinating. One thing i've always found frustratingly difficult to grasp though, is the expansion of the universe - but just found this article to be a total eye opener: http://en.wikipedia.org/wiki/Metric_expansion_of_space
It was counterintuitive to me how we can observe light from objects 12 billion years old, when supposedly things are flying apart faster than the speed of light. Is there an edge of the universe, and what is outside that we are 'expanding into'. If everything is moving apart, why are we going to crash into Andromeda? Why haven't black holes consumed all the mass in the universe by now? If the cosmological principle is true and the universe looks the same in all directions, doesn't that imply that we're either at the center- or that it wraps around and we just can't observe it? Saying it's infinitely large seems like a convenient explanation.
The idea of metric expansion of space between gravitationally unbound matter, local groups, and Hubble flow explain a lot.
I do still wonder though how accurate our understanding of the universe really is, or if there's some fundamental thing we've missed and have generated theories to fit the observations we're currently capable of making.
My thoughts exactly. Star Trek, and Spock form an indelible part of my childhood that's shaped my early personality more than I probably realize. Truly a sad day. Rest in peace, friend.
I would be very interested in seeing a list of diagnoses from cases that were solved correctly. It seems to me that once someone with a difficult to characterize condition has gone through 5-7 physicians, any 'diagnosis' that comes out at the end is either wrong or a placeholder until technology advances.
Hi there, great images! I'm curious about how West Kendall Baptist Hospital was chosen for the clinical trial? Is this a single-center study? Lots of angiography in the sample images, so is the trial comparing this to cath?
I've been giving your idea some thought, and would be happy to discuss further over email if you like (in my profile).
I don't think HIPAA is the main problem you have. The biggest obstacle is whether any healthcare provider would be willing to participate. People in this area have little time, too much work/obligations, are generally not tech savvy, have fixed workflows that they're unwilling to deviate from, and even the smallest amount of friction in your product will make them give up.
If you're going to do this (and you probably should if you're passionate about it), then you have to think of a way to integrate seamlessly into existing workflows and minimize their time commitment. Your goal should be 5-10 mins per day, max, regardless of the number of people being "monitored". Think about the incentives to get providers to participate. How much turn-around time is realistic? I don't see how this will be an "imminent suicide" prevention system and there are already lots of hotlines for that, but it could be great to monitor mood over days/weeks.
Perhaps have a single daily aggregated and anonymized email to a provider's nurse at 7am (when most check their patient data) that updates them on all patients in the practice who are potentially in trouble. They click a link, enter a password, and see the non-anonymized data. Then the nurse can get on the phone and start calling people.
Don't doubt yourself so much. While talent is undeniably an advantage, in the long term, someone who is persistent, overcomes obstacles, and has real interest in the work will win over raw talent 9/10 times [grit,1]. The fact that you were accepted to a good school is justification that you have more than enough intrinsic ability. Just find some area that you're interested in and you feel is important and challenging, and stick with it day after day and you'll be rockin it soon.
"All these logos are sized to fit in 16 x 16 pixels." - I was surprised how crisp these look for 16 pixels, then noticed the file was actually 32 x 32 and shows up as such on a retina display. It's a fair point that only the first four logos are recognizable when 16 pixels though.
Always enjoy reading your posts 7Figures. I'm in this exact situation, and decided to go the agency route for the same reasons Alex did for Groove. It's hard in some occupations to have technical friends in your circle, creating some distance between the people who have the domain expertise and the people with the skills to execute on ideas. It's nice to see alternatives. It's worth mentioning also that developing Groove with the agency costed $300k.
I wanted to add my thoughts about false positives and executive health checkups, which fall under the larger umbrella of screening. I conduct executive checkups fairly regularly, but i’m not promoting anything here. I just want to share a personal philosophy that you, as a patient, could adopt to reduce risk.
Firstly, it should be stated that the utility and performance of any diagnostic test that has a binary outcome (eg. diseased/healthy) depends on 1) its intrinsic test characteristics 2) the prevalence of the disease in the population being tested[1], 3) the available diagnostic/treatment options for positive people, and 4) the threshold chosen to call something ‘positive’ since you are virtually always measuring a continuous variable.
People talk about false positives often, but you should know that there is some room for judgement here. The typical threshold for PSA has historically been 4.0 ng/mL. Above this is considered a positive screen, and you are recommended further testing to confirm prostate cancer, which can be invasive and lead to complications (bleeding, nerve damage, etc). This kind of two-stage testing is a common theme. The issue is that PSA values in people with cancer and people without cancer are normally distributed and have some overlap. A value of 4.0 is in the overlap zone, and the vast majority of people with 4.0 will not have cancer. However, lets say hypothetically that guidelines changed and we now use 20. At this level, most people who test positive probably have something bad, and the risk:benefit ratio is in favor of confirmatory testing and probably in favor of treatment too. I won’t bother going into it, but there’s a fair amount of liability, psychology and economics at play when choosing thresholds. Biochemical tests like PSA do have some variance, but are mostly reliable. It’s more complicated with imaging and amorphous things like mammograms that have some operator variability in judgement (and hence more liability).
This was a long-winded way to say that information is not necessarily bad, but it’s what you and your doctor choose to do with it. There’s room for discussing probabilities, and the various available pathways. Trust is huge. Your doctor will act in your best interests when he feels safe from lawsuits, otherwise he'll be focused on not missing anything regardless of likelihood or cost. The financial misalignment of our system doesn’t help in building trust either. Most first-stage screening tests have very minimal risk involved. Things like checking your blood pressure, cholesterol, glucose, doing a treadmill stress test, reviewing your meds, vaccinations, etc which are part of these annual checkup programs are, on balance, usually good ideas.
[1]http://en.wikipedia.org/wiki/Sensitivity_and_specificity