Ask HN: New to US, puzzled why tech hasn't simplified health insurance

All aspects of health care and health insurance in the US is inefficient by design to employ more people. The solution to that inefficiency is to eliminate many jobs. This is commonly known.

The problem though is that health care, while very expensive and not universally available in the US, is very good. It is not clear if destroying the large number of jobs to sufficiently make the health industry more cost effective will impact the quality of care and it will be economically disruptive. The primary reason why people are sheepish to solve that problem is that it’s politically toxic with no immediate profit incentive.

> we can't figure out how to automate an insurance call

You obviously have not encountered an AVR (Automated Voice Response) tree yet. When you do, you'll be glad they don't 'automate' the call, as it would be worse than being on hold waiting to talk to an actual person.

> Is there a technical reason why so much insurance stuff has to be done over the phone?

Unsure. One guess is this falls out as a HIPPA requirement for "security" reasons. There are a lot of laws/rules in the US that were written years ago back in the day when the phone network was a single monopoly company and an isolated network such that "phone calls" were "secure" against eavesdropping. The phone network has changed such that this is no longer the case, but none of the laws/rules have been updated in the interim to account for the fact that "phone call" is no longer the same as it was when AT&T was the highly regulated monopoly provider of all phone service throughout the US.

> Why is it so hard to get a straight answer on how much a doctor's visit will cost?

Because the actual cost depends upon at least two things:

1) a secret contract agreement for pricing between the doctor (or the 'group' the doctor works for) and the specific insurance plan; and

2) how the doctor "codes" the visit for billing purposes (with the code referencing into that secret contract agreement).

> Are there regulatory challenges that make this all so complicated? I'm wondering if there are laws or compliance issues that make it hard to simplify things.

Yes, much of the mess that is medical insurance in the US is the direct result of many regulations/laws enacted over many years, that all sum up to the ugly monster that is the system as it exists now. And because it is profitable for the incumbents, they purchase politicians in order to keep the mess messy to maintain their legacy profitability.

They don't want you to claim, anything that makes it easier to claim is going to be opposed for that reason alone.

You cannot improve this system, its not supposed to be helpful.

a. Claims calls are often used to better detect fraud (tone/pauses etc) and ask probing questions to deny claims. The friction of having to make a call also helps to reduce claim volumes.*(not all insurers)

b. Not sure, the insurer should negotiate competitive rates and know what they are. This could be on a sliding scale or otherwise vary dynamically based on claims and performance criteria.

c. Insurance is state-regulated in US. There may be specific requirements relating to this.

I spent a decade working in insurance and reinsurance in the UK. The US healthcare system is a different beast but there are some similarities.

Insurers are using tech in lots of interesting (and... unethical?) ways. Some reasonable ways would be linking to wearable or IoT devices for real-time data and offer incentives to care better for your health and home.

Others, like pricing and claims optimization strategies are a bit more questionable. This is where they'll vary prices and claims to maximize profitability. With claims optimization, the insurer will use data about the person's financial situation to determine how low a pay-out they can make e.g. £1000 today or new car next week.

With AI, they can analyze more datasets e.g. social media activity to analyze behaviours and likelihood of claims in order to adapt pricing... they've done this forever. 20 years ago they'd vary pricing based on your email provider.

There are new models like UBI (usage based insurance) and on-demand insurance, and some new entrants are using tech to streamline operations including cutting the contact centres - so it's being done by some.

I think there are more fundamental opportunities to disrupt and build more accessible and profitable insurance models with a longer-term view though.

The problems with health care in the US are not technological ones, so technological solutions can't really address them.

A lot of people are focusing on why insurance is complicated in the United States, but I'll do my best to answer why tech has not made this more simple.

  Why tech hasn't simplified health insurance?

Beyond that, yes there are regulations that make things complicated. Chiefly HIPAA, but also the ACA. The real challenge is a solution that makes everything more simple for the patient while appealing to providers and insurance companies that have conflicting incentives. Is there a tech solution for this?

Some food for thought: More than half of all medical costs are accrued in the last 2 years of someone's life. That means the vast majority of the medical system, including billing and insurance is geared around elderly people with extremely complex medical bills. Young, healthy people who occasionally need to visit the doctor is an afterthought. Which is why you still get a person over a phone.

That said - yes, the system is incredibly dumb. Not only is it insanely regulated, but there is a constant 3-way game being played - not just between insurance and hospitals, but also between hospitals and Medicare/Medicaid. Medicare and Medicaid only pay out a percentage of the list price for any procedure, but unlike private insurance, the payout is non-negotiable. So providers play around with billing codes to cover their costs. They are also cagey about publishing prices because that's ammunition for insurers to negotiate harder.

Some providers try to offer transparent/straightforward "cash" prices (shout out to Zoomcare in the PNW). But this gets them in trouble with insurance providers for some reason.

Kaiser Permanente is an insurer that also provides health care and solves all of the aforementioned problems.

Like almost all problems, it's a social/political mess, not something you can solve with better technology.

Most insurers have a website where you can estimate costs and find providers. If you want to know about costs, ask the provider to estimate the cost as they will know the codes to bill it as.

The main problems are the incentives. Providers and insurance are fighting each other and the customers to cut their own costs and retain/recover as much money as possible. But yes, if you could overcome these fights while improving the billing and reimbursement efficiency then it would save a ton of money.

A typical business wants customers that call in to get a quick and efficient response, so they will potentially buy more things in the future.

Your health insurer, on the other hand, already HAS your money. Their goal is to do exactly as much customer service as will keep you (or more likely, your employer) from switching companies. The more efficiently they let you make claims, the more you cost them.

I increasingly feel the route to simplifying health insurance and medical care in the US is to bypass the existing legacy players entirely.

Closest example I can think of that already exists - concierge healthcare services. But maybe that is only possible because it's a small scale boutique service. Can it be scaled up and survive? I'm not entirely sure.

Because it makes money for Big Pharma and insurance.

And no one else. That's it.

If we switched to Swedish-style healthcare, even accounting for the population difference, the US government ALONE -- ALONE -- would spend about 1/4 as much on healthcare (medicare+medicaid) as it currently does. The public would spend less. And businesses would probably spend less, due to not needing to pay such high premiums. The system would be simpler and it would be easier to catch fraud.

It's a combination of polarization, inertia, and effectively bribery by big companies that we haven't switched yet. That would probably help with a lot of your questions.

Our healthcare system is just crap. Most people don't like it, but...yeah. (By the way, if I had to guess, patient privacy rules might interfere with automation. Not sure/at all confident, though)

Good luck in figuring out our healthcare system. And I hope you like the US! We're not as idiotic as our politicians make it seem. :)

Possibly because, as some have claimed, society is effectively not run for the benefit of all people but for the benefit of a few people who are making lots of money.

What healthcare plan do you have? Why are you dealing with the insurance and not the healthcare provider?

"For profit" and "healthcare" are just two things that shouldn't mix, but here we are. Everything in capitalism has a profit motive. Everything in life does not.

Upton Sinclair: "It is difficult to get a man to understand something, when his salary depends upon his not understanding it!"

What you might see as "simplification", others might see as "threatening our jobs." Technical excellence (or lack thereof) aint the issue.

There is no technical solution. The problem in the US is that we are a combination of a free market and government controlled single payer system. To improve things the government would have to pick one which is impossible.

a. Insurers don't really innovate. b. Because the doctor doesn't know what price they negotiated with your insurer. c. There are a ton of regulatory constraints.

It is because the insurers and interests [provider groups] represneting the providers want it to be a convoluted mess.

a. No. Some of it is handled by different apps now including Healow, MyChart, etc.

b. The negotiated rates change depending upon what insurance you have. It is coked-up monkey with a random number generator that is making the actual billed prices.

c. Yes. Hippa, the ACA to some extent, different state political interests too.

TLDR -- Healthcare is a hot mess in the USA for all but the very wealthy.

The problem space you describe is actually infinitely more complex than you make it out to be, but to truly understand it you have to understand how centralized systems fail. The vast majority of people out there, even the so-called specialists haven't actually solved the underlying problems because in some ways they cannot be solved.

If you would like to educate yourself to better understand the world, and life in general, when it comes to centralized systems; I would recommend Mises on Socialism. It was written in the 30s, but describes systems based on structure, maps failure domains within those systems, and coins problems that remain un-refuted today following rational principles. It can be a difficult read as it was written at a time when hyper-rationalism was the lingua franca, and each word had one specific non-contradictory meaning with little ambiguity.

Almost everything discussed in his book impacts and applies to every centralized system, this includes bureaucracy, government, education, and any heavily regulated industry, or inflationary economy (where money gets printed).

The latter most proof is a bit indirect, but basically Producers and Consumers have requirements that must be met in terms of purchasing power. Ponzi's purchasing power collapses in the final stage where outflows exceed inflows, and both those requirements for exchange then fail leaving a defacto state of socialism, or annihilation. The producers stop producing when they cannot make a profit, leaving only state-run entities for survival.

Socialism has 6 main problems, and depending on the structure potentially many others; the economic calculation problem is the impossible to solve problem. It is a progressively mathematically chaotic n-body limited visibility problem with arbitrary variables instead of constants. Inevitably shortage occurs, then self sustains, order wanes, food production ceases its current levels, and chaos reigns. Most of humanity or all dies out from ecological overshoot reversion or its complications (MAD) (i.e. based on Malthus law of population growth/Malthusian Trap).

Cascade failures are some of the most complicated and difficult problems any system's engineer deals with, and resilient systems design are needed, but centralized systems create brittle designs with single points of failure and front-of-line blocking.

Anytime you have government regulation, you must necessarily have a good understanding of the failures of centralized systems, and the misleading lies and false solutions that often accompany those systems when people are involved.

Atop the structural changes, the psychology of people involved in centralized systems changes dramatically when compared to people in systems that have a hard loss function which they are measured against (i.e. where they get fired if they can't perform).

To answer your questions:

a. Government Regulation/Privacy, Lack of Free Market (inflationary economy) b. Monopoly/Collusion c. Anytime government imposes ambiguous restrictions, or impossible to fulfill requirements you have a sieving filter applied to the business sector. All other business then dies off creating a moat, so no new business can enter the market.

As for why we cannot automate 'insurance calls' that is more simple than everything else.

When humans communicate with other humans, there is a component of communication called reflected appraisal. This happens beneath perception, and when there is an inconsistency, it induces irrational responses and actions. This also is one method that real-world torture aims to exploit for either information gathering or thought reform (breaking someone permanently, which is used often in adtech/marketing; a topic for another time).

Humans are capable of handling multiple contexts at the same time and choosing the correct one where other humans will share a common single shared meaning utilizing the words, and other indicators as a medium. This is basic Uni-level Introduction to Communications coursework.

Computers are unable to handle multiple contexts given the same inputs. It breaks determinism which is a required property for computation to do work (on von-neumann architecture), and the best it can manage is approximation using weights, which will never be able to handle and perform consistently with the same words that have contradictory meanings depending on context. Large companies are also incentivized to not address their customers needs and instead torture them through CSR doom loops (struggle sessions). You can learn more about torture by reading Robert Lifton, or Joost Meerloo. The systems in place at these companies are designed with this in mind (intentionally), to cut their biggest cost, labor.

Data breaches of medical data are also far more harmful than others. Imagine someone finds out you use a specific medicine regularly and then introduces a common chemical that interacts with it into other products you use. While this was once the stuff of fiction, with big-data having no liability for security, and fines that usually just further concentrate an existing market sector, anyone with sufficient money can know this; or even target this if you make yourself a target, or to cause manipulation towards products where they make a profit. Corporate Espionage with megalithic corporations is a real thing.

[dead]

[flagged]