What I'm getting from this is
a) It's mostly conjecture
b) The author thinks it's likely that they will be creating something akin to "Kaiser Permanente lite"
Yea, sounds like Warren Buffet isn't really behind this whole thing (entirely my observation). This morning on CNBC he was asked about the deal and only mentioned about how it was really Todd Combs that orchestrated this thing. Buffett doesn't even commit to forming a company for this new insurance idea (also of note, insurance businesses are one of Buffett's largest sectors of ownership, why note just instruct them to fix this issue???)
Yes it's almost completely different. A few of the actuarial concepts are similar but almost nothing else carries over. In fact, most health "insurers" have shifted away from selling insurance policies and toward becoming third-party claims administrators for self-insured group buyers.
Nah. Buffet wants in, but he wants his niche which is reinsurance; covering pools who get hit with cancer patients and car wrecks so they can expense risk as a line item and keep much lower cash reserves.
One very simple thing they could do that might be helpful is provide an easy to understand guidebook for understanding the quality of various healthcare providers. Sort of a Michelin guidebook for healthcare
Currently, patients don't have the resources or information to assess the quality of their providers. Hospitals with fancy marble lobbies are more likely to be rated as high quality than hospitals with better quality scores but dingier interiors. People think that if care costs more, it must be better
Insurance companies have tried to build narrow networks that exclude high cost hospitals that nonetheless are publicly perceived as high quality, but public backlash, however misinformed, can quash these efforts. One example is UCLA vs BSCA a few years ago
A few San Francisco companies seem to be doing this already: Grand Rounds, Amino, and to a much lesser extent Castlight. They sell through the employer-payer; not sure if they are competition for AmaMorganShire or potential vendors to them.
>One very simple thing they could do that might be helpful is provide an easy to understand guidebook for understanding the quality of various healthcare providers. Sort of a Michelin guidebook for healthcare
there are plenty of reasonably good definitions of quality and value-based frameworks out there. medicare defines and enforces several of these, and the amazon JV could potentially play a similar role in the commercial payer world
> One very simple thing they could do that might be helpful is provide an easy to understand guidebook for understanding the quality of various healthcare providers.
That's not a simple thing to do, especially well, and you kind of implicitly note with your reference to existing quality ratings that lots of people have tried to do it with mediocre results.
I think there is an opportunity for the amazon / BH / JPM JV to use its scale, visibility and "outsider" status to create a standard that is simple and easy to understand by consumers, and actively market that standard in a way that medicare, startups or industry coalitions have not been able to
most quality standards are not even well known or standardized within the healthcare industry, and there are so many competing interests that it is hard for providers, payers, other stakeholders to align on a standard, useful and enforceable definition of quality. the JV is already insanely visible / high profile without anyone knowing what it does; if they create a quality rating system it will certainly get more attention than any other one created to date
also, amazon / bh / jpm in theory is better aligned with society in terms of what they want out of healthcare: 1) lower cost and 2) customer satisfaction. that group cares about making its employees happy, whereas other stakeholders dont have the same direct accountability to employees / patients: payers are only accountable to patients through the employers they serve, and providers are in theory accountable to patients but in reality more influenced by the health system that owns them or the payer that pays them
The major quality standards are published by CMS (Medicare) and they're very well known because reimbursement is now tied to hitting those metrics.
But in a broader sense it's really impossible to define objective quality metrics for individual healthcare providers. If a heart surgeon has a high patient death rate is it because she's incompetent or because she takes on the most fragile patients that no one else is willing to operate on? You might think that those differences could be controlled for but in reality there are too many confounding factors and research hasn't even been done to accurately quantify many of them.
It is hard, and once you've defined quality metrics, it's even harder to write a contract that captures them and get parties to agree on them
The latter issues might be resolved somewhat by the Amazon group; the former is still a concern, but I think there are opportunities to do meaningfully better things
This is just an idea, and I've not done much research on it, so take with a grain of salt: I think patient focused quality measures might be better. For example, look at cohorts of patients with similar characteristics (like in case control studies) who are experiencing a similar health event (knee surgery) and compare providers based on the outcomes these types of patients have. This is something the JV could do, is patient focused, and avoids all the challenges you mention with metrics measuring / managing individual providers (which isn't just hard to measure but demoralizing to docs)
That's exactly the problem. There's no reliable way to construct a cohort of "similar" patients. We don't even know all the variables other than provider quality which impact the outcome of knee surgery. Do we have to consider age? Genetics? Sex? Fitness level? Co-morbid conditions? Blood Type? Personality? Something else? That data simply doesn't exist, and won't exist in our lifetimes.
With large enough populations some of those confounding variables even out. But that doesn't help when trying to evaluate individual providers who treat relatively small numbers of patients per year.
Of course in any area all of the local doctors know from subjective experience and word-of-mouth which other doctors are skilled and which are quacks. But none of them will say so publicly, and you can't quantify those opinions.
People do case control studies all the time, the factors used depend based on disease but there is plenty of literature to support selection of matching factors
I think it would make sense to start at the hospital level; provider level probably isn't feasible as you say. Hospitals see enough patients where the sample sizes are pretty big, and they are the main cost driver in healthcare, so rating hospitals by cost / quality could work
They could provide a platform for crowd-sourcing the gathering of these metrics. Of course controlling for bias could be a huge problem - but then again, healthcare providers could simply correct biased metrics if they so chose.
The worst thing about hospitals is how they are incentivized to keep quiet about extremely important metrics such as sepsis/staph infection rates, effectiveness at treating certain conditions, etc.
Show us all what leverage is worth, perhaps? For my own small biz, $2500/month for one employee+family bought parity for solely health. I was far behind for dental/vision/etc. As a now employee of a large company, $500/month gets better returns than $2500/month as a nobody did.
yeah, would really like to know what the actual cost the company pays is, vs the employee share. might still be lower than $2500/month (probably) but I can't imagine a family plan in any major state costing $500/month, even with a huge deductible (I sorely wish it did tho!)
1) UX for doctors. Make it easy to log records of visits without secretarial support.
2) Pharmacy at the Amazon warehouse for 90% of most used drugs.
3) Diabetes/endocrine specialist by Amazon as a kiosk in Whole Foods. Most preventable illness in the US is because of sugar over consumption, the rest we can't prevent.
22 comments
[ 3.1 ms ] story [ 55.9 ms ] threadIs there something else I'm missing?
[1] https://www.cnbc.com/2018/02/26/full-transcript-billionaire-... (Search for Todd Combs, it'll bring you to the first paragraph talking about the healthcare insurance venture thingy).
Currently, patients don't have the resources or information to assess the quality of their providers. Hospitals with fancy marble lobbies are more likely to be rated as high quality than hospitals with better quality scores but dingier interiors. People think that if care costs more, it must be better
Insurance companies have tried to build narrow networks that exclude high cost hospitals that nonetheless are publicly perceived as high quality, but public backlash, however misinformed, can quash these efforts. One example is UCLA vs BSCA a few years ago
Define quality.
That's not a simple thing to do, especially well, and you kind of implicitly note with your reference to existing quality ratings that lots of people have tried to do it with mediocre results.
most quality standards are not even well known or standardized within the healthcare industry, and there are so many competing interests that it is hard for providers, payers, other stakeholders to align on a standard, useful and enforceable definition of quality. the JV is already insanely visible / high profile without anyone knowing what it does; if they create a quality rating system it will certainly get more attention than any other one created to date
also, amazon / bh / jpm in theory is better aligned with society in terms of what they want out of healthcare: 1) lower cost and 2) customer satisfaction. that group cares about making its employees happy, whereas other stakeholders dont have the same direct accountability to employees / patients: payers are only accountable to patients through the employers they serve, and providers are in theory accountable to patients but in reality more influenced by the health system that owns them or the payer that pays them
But in a broader sense it's really impossible to define objective quality metrics for individual healthcare providers. If a heart surgeon has a high patient death rate is it because she's incompetent or because she takes on the most fragile patients that no one else is willing to operate on? You might think that those differences could be controlled for but in reality there are too many confounding factors and research hasn't even been done to accurately quantify many of them.
The latter issues might be resolved somewhat by the Amazon group; the former is still a concern, but I think there are opportunities to do meaningfully better things
This is just an idea, and I've not done much research on it, so take with a grain of salt: I think patient focused quality measures might be better. For example, look at cohorts of patients with similar characteristics (like in case control studies) who are experiencing a similar health event (knee surgery) and compare providers based on the outcomes these types of patients have. This is something the JV could do, is patient focused, and avoids all the challenges you mention with metrics measuring / managing individual providers (which isn't just hard to measure but demoralizing to docs)
With large enough populations some of those confounding variables even out. But that doesn't help when trying to evaluate individual providers who treat relatively small numbers of patients per year.
Of course in any area all of the local doctors know from subjective experience and word-of-mouth which other doctors are skilled and which are quacks. But none of them will say so publicly, and you can't quantify those opinions.
I think it would make sense to start at the hospital level; provider level probably isn't feasible as you say. Hospitals see enough patients where the sample sizes are pretty big, and they are the main cost driver in healthcare, so rating hospitals by cost / quality could work
The worst thing about hospitals is how they are incentivized to keep quiet about extremely important metrics such as sepsis/staph infection rates, effectiveness at treating certain conditions, etc.
Fixing healthcare in the US will require new vertical integrations, not just tools to navigate the existing system
2) Pharmacy at the Amazon warehouse for 90% of most used drugs.
3) Diabetes/endocrine specialist by Amazon as a kiosk in Whole Foods. Most preventable illness in the US is because of sugar over consumption, the rest we can't prevent.
Boom. They just shaved 20% off costs.