The Black Liszt

Medical Services Before, During and After an Event

 One good way to get an understanding of the vast array of medical services and treatments is to organize them into a spectrum based on some parameter. I earlier looked at medical innovations based on how “fancy” they are, from AI to cleanliness.

https://www.blackliszt.com/2016/09/healthcare-innovation-from-washing-hands-to-ai.html

Another useful spectrum is the time relationship between a medical event that requires attention and the treatment of that event. At one end of the spectrum is treatment to prevent the medical event from taking place. The middle is treatment at nearly the same time as the event. The other end is an event in the past that now needs treatment.

We can understand a great deal about modern changes in healthcare practice and cost by looking at treatments in terms of this spectrum.

First, let’s understand the spectrum a bit more.

Treatment before the event

This a normally called preventative medicine. It is the subject of wellness visits to the doctor, and advice given by experts to people about how they should eat and otherwise behave. Medications and even surgery can be involved. If the treatment works, the event doesn’t happen.

Treatment to detect the event early

This is often called disease screening, a procedure to detect the early presence of a disease that the human doesn’t notice, but could grow into something serious. It could be performed personally or by medical services. The idea is that a problem caught earlier can be treated with greater success than one caught later.

Treatment around the time of the event

This is classic “I know something’s wrong, I need medical help.” It ranges from a bad fever to being hurt in an accident. You don’t make an appointment, you go or get taken to an Emergency Room or Urgent Care.

Treatment long after the event

These are treatments for conditions that are ignored or develop slowly, conditions that might involve some pain or inconvenience. As the condition worsens, medical advice and treatment may be called for. Examples are joint replacements (knee, hip)

Evolution of the spectrum

We all recognize the medical spectrum today. However, it wasn’t always the way it is today. Pre-event treatments were rare prior to the modern world. Most of these had to do with common and personal hygiene.

Prevent the Problem and detect it early

Discovery and implementation of disease prevention methods started in the 1700’s. It’s estimated that between the years 1500 and 1800 scurvy killed about two million sailors. Only in 1747 did a Scottish Naval surgeon publish a book with a trial that proved a cure. Another Scot, Joseph Lister, discovered the benefits of antiseptic surgery.

We’ve come a long way since the acceptance of the long-resisted ways to prevent scurvy and minimize infection. Now we are told to visit the doctor for regular “wellness” visits, during which we may be prescribed medicine to prevent heart failures (blood pressure, cholesterol) and encouraged to get procedures that will detect things like breast cancer and colon cancer early so they can be treated more effectively.

Sounds like a good idea, right? Most folks you talk with agree. There’s just one little problem: the wellness visits are basically sales sessions during which the doctor tries to convince you with tests, numbers and serious words to spend more healthcare dollars taking pills and injections and undergoing procedures that … don’t make you healthier and don’t extend your life.

Here are details about wellness visits, heart health visits, and health insurance company promotions. The links contain details about how and why both prevention and early detection do more harm than good, and in particular don’t lead to longer, healthier lives. Of course, there may be exceptions. But they must be proven, unlike most current practices.

A major barrier to improvement is the wall of secrecy around medical events and tests. There is a proven path to fix this problem in the software world: open source. It will lead to a positive revolution in the medical world as it has in the software world. Here is further detail.

The rest of the spectrum

Emergency care has advanced a fair amount in recent decades. But the changes are more evolutionary and gradual. For patients treated by a system of which they have long been a member, having on-line access to the patient’s medical records has certainly been a plus.

Post event treatments have grown a fair amount in recent decades. Organ transplants can now be achieved with unprecedentedly positive outcomes. Joint and limb replacement have now, in some cases, become routine, which is a big improvement in quality of life for those involved.

Conclusion

The spectrum of the relationship between the time of an event and the response to that event is important for understanding medical care, both quality and cost. While there have been substantial advances in the later parts of the spectrum, it’s clear that the early part of the spectrum has grown by far the most dramatically, mostly from the 1950’s on. It’s also sadly clear that the early part of the spectrum’s contributions to health and longevity are largely negative. If most early-spectrum tests and treatments were dropped (subject to open-source evaluation and testing), costs would be reduced and health improved.

How to Reduce Medical Costs and Improve Health

The US has the highest per-capita cost of medical services of any nation. At the same time, average life expectancy has been going down. What can be done to reduce costs while at the same time improving our health?

Change won’t be easy. There is a monopoly of providers, health institutions, insurers, pharma, regulating government agencies; all of them want to increase their power and business. The medical-industrial complex is a growth engine.  Worse, most people accept the medical doctrines (marketed as science) that underpins this power. While much can be done to improve costs and outcomes without challenge to medical doctrine, the best results can only be achieved by taking a fresh look at medical givens.

When you want to make big changes to something as large, powerful and complex as this, incremental change is nearly impossible. It’s like the world of business in general: innovations are primarily made by small groups of motivated people who give birth to the new approach. This pattern has lasted for many decades. It's even tougher here because of the interlocking monopolies.

At a high level, the overall ways to change are these.

Zero-based re-engineering

This is a fancy way of saying, imagine you were starting from scratch: no assumptions, no nothing -- what would you do? Even in a supposedly fast-moving field like software, there is at least a 10X difference between doing things the “standard” way that professional managers require and the optimal way. The gains in medicine and health could be at least as large.

https://www.blackliszt.com/2015/10/secrets-of-software-super-developers.html

Here's a specific example of zero-based re-engineering for 10X gains.

https://www.blackliszt.com/2019/10/software-professionals-would-rather-be-fashionable-than-achieve-10x-productivity-gains.html

Automation

Automation is the proven path to quality improvement and cost reduction. The principles have been established for centuries. Automation is why food production took 90% of the population at the time of the founding of the US, and less than 1% today. A key part of automation is that work organization, job requirements and skills are re-thought from scratch. Otherwise, all you're doing is "paving the cow paths."

https://www.blackliszt.com/2020/01/the-fundamentals-of-computer-automation.html

Automation of clinical decision-making

This is an essential aspect of automation that is strongly resisted. It has been a central part of general automation ever since Jacquard looms were driven by cards centuries ago. It directly addresses the always-incomplete knowledge in doctors’ heads and enables vast expansion of personalization and accuracy.

https://www.blackliszt.com/2025/01/ai-can-automate-what-doctors-do.html

https://www.blackliszt.com/2016/12/what-can-cats-teach-us-about-healthcare.html

https://www.blackliszt.com/2020/01/luddites.html

In-old-vation

The medical world claims to support “innovation.” This is propaganda. The fact is that the vast majority of effective innovation has been proven in other domains for a long time, and “just” needs to be adapted. I tell the story of how an innovation that transformed the management of oil refineries in the 1960’s began to transform the management of medical operating rooms and infusion centers more than fifty years later.

https://www.blackliszt.com/2019/08/the-slow-spread-of-linear-programming-illustrates-how-in-old-vation-in-software-evolution-works.html

Data-driven evaluation of treatments

The medical system brags about “evidence-based medicine,” random controlled trials and the rest. But the fact is that the vast majority of spending and treatments are not backed by solid data, and huge parts of standard practice are proven ineffective and/or harmful by data. Anything that is done should be tracked by a continuous feedback, closed-loop system that includes patient input.

https://www.blackliszt.com/medical-science-research/

Regulation

While often well-intentioned, government regulation drives up costs and stifles innovation without improving quality. The FDA has a staff of lawyers and bureaucrats writing mountains of regulations that spell out in voluminous detail how software should be written, for example. This stifles innovation and drives up costs dramatically for medical devices.

https://www.blackliszt.com/2016/12/regulations-that-enable-innovation.html

https://www.blackliszt.com/2020/05/heres-how-the-fda-can-reduce-medical-device-costs-while-improving-healthcare.html

https://www.blackliszt.com/2023/01/how-to-reduce-the-cost-of-medical-imaging-and-pacs.html

A similar concept applies to drug development.

https://www.blackliszt.com/2017/01/using-software-methods-to-speed-drug-discovery.html

Education and credentialing

Practice licenses and requirements sound reasonable, but in practice they are a way to enrich the relevant organizations and restrict supply while keeping prices high. Example: why should a person who performs a specific kind of surgery require an MD and 5 or more years of further training? Why isn’t performing a breast lumpectomy like being a phlebotomist – an important job that requires physical skills and must be done correctly, but only requires weeks of training and then apprenticeship? This is a huge subject.

Optimize the system for maximizing patient health

You'd think would be obvious, but it's not. Today, the growth and profitability of the participants in the medical-industrial complex are optimized, in spite of many individual providers sincerely trying to optimize patient health.  For example, the essential data in the EMR is owned and operated by the medical system. Yes, they sometimes provide interfaces for patient access, but it's clearly theirs.

https://www.blackliszt.com/2016/01/i-want-an-emr-app.html

Conclusion

The medical system has achieved some amazing things -- just take knee replacements which are now routine but unheard of by earlier generations. The trouble is it has grown into a multi-faceted monopoly that grows endlessly while the patients take ever-more drugs and submit to ever-more procedures that have the net effect of making us pay more to live shorter, less healthy lives. While the defenses of this castle are tall and thick, the peasant/patients who serve the masters inside the castle walls have got to find a way to create a revolution, helped by the growing number of disaffected medical professionals. The nightmare anti-scientific response to covid has sparked doubts in the minds of many. When are we going to storm the Bastille?

PS: Here is a complementary earlier approach to these ideas.

https://www.blackliszt.com/2015/12/healthcare-problems-and-opportunities.html

How to Learn to be a Computer Programmer

Lots of people will tell you how to become a computer programmer. Go to college and major in Computer Science is the most wide-spread advice. In the last few years things called “coding academies" have emerged as quicker, more affordable paths to that goal. There is a faster, better way to acquire the skill of computer programming that is rarely discussed. While this post is about computer programming, the educational concepts discussed are broadly applicable.

How Skills are Acquired

Stepping back from programming, let’s think about how skills are acquired. Physical skills like playing various sports is a good place to start. How much classroom time is required? How about text books? The basic way you learn sports or other physical skills is watching someone play them. With any sport involving a ball, you pick one up and try to throw it. Then catch it when thrown to you. Someone may give advice, but basically you try a lot, gradually learn from your mistakes and get better. It’s important to note that you see the results of your effort. You see if the ball went where you intended it to go, for example. This can continue for years.

How about a more intellectual skill? We all learn how to talk and listen with understanding. The way we learn is similar to learning a sport – by watching, listening and then emulating. The world is incredibly complex and varied, with words associated with a huge number of things and actions. You start with a few words, and spend years adding many thousands more to them. Yes, parents give lots of feedback – not unlike the feedback you get when you see whether the ball you threw went where you wanted it to, whether the receiver caught it. Did you catch what I was trying to say there? Yup, like that.

The language of software

By contrast to any human language, a computer language is amazingly simple – partly because the “world” in which it “lives” is incredibly narrow and abstract. The nouns are all data of a couple different types, basically numbers and letters. You give each place that can hold data a name. The verbs are just a handful of simple actions that just grab the data that’s in a named place on your “desk,” do something with it, and put the result back or into another named place.

A great deal of the power of software comes from what’s called the subroutine library, which is basically an attentive gang of little robots waiting to do your bidding. If you want to write something on a screen, you “call” the relevant robot, give it your instructions, and send it off. There are robot librarians and file clerks that are excellent with card catalogs and endless file cabinets, fetching what you ask for and putting away what you’re done with. There are robot butlers at each “door” (keyboard, mouse, etc.) that politely inform you when something has arrived and give you the package. Another butler will also send data where you want it to go. While central to programming is learning how to “talk,” learning about the team of robots at your disposal and what each can do for you is also important.

Yes, there are lots of different software languages. But they’re really like local dialects. They all conform to the basic description I just gave. The gang of robots available in a given language typically vary depending on what the language is mostly used for. There are business-oriented languages that pay special attention to things like financial data, account data collections like accounts, and others that have robots that are really good at fancy math. But the way of thinking about them and writing programs is remarkably similar.

This should make you wonder exactly why you need years of courses “taught” by fancy professors, when programming is like doing things in an amazingly simple world of data using a small set of instructions that are easy to visualize and understand. The answer is that you can do it on your own, given access to a computer, some widely available tools and a bunch of organized exercises that start super-simple and get gradually more elaborate. That’s how I learned!

How I learned to program in ancient times

I went to high school at the newly-opened Morris Knolls HS in Rockaway, NJ. It took students from a couple of towns and had courses that were explicitly college and non-college. There were vocational tracks such as HVAC, auto repair and secretarial. The towns weren’t poor but definitely weren’t elite, with few professionals.

My introduction to programming and computers was a course offered in my junior year of high school, the academic year 1966-7. One of the high school teachers wanted to move from teaching to software and had no training or other way to get there. He somehow persuaded the administration to let him teach a course on programming. His break-through was arranging for the class to have computer time at a nearby company, Reaction Motors.

The teacher found a FORTRAN textbook to serve as the basis of the class. While he was the “teacher,” it was clear that his teaching amounted to trying to be a couple chapters ahead of the class in the book. But it didn’t matter. The important thing was having a book that laid things out, along with exercises and answers. And above all, the Saturday time slot when we had access to the computer.

Programming involved writing the program on paper, and then keying the program onto cards using one of the punch card machines that were in a room adjacent to the main machine room. We got to operate the room-sized computer, something which in the normal corporate environment was an absolute no-no, as I later learned. It was terrific fun.

I got lucky the summer after my senior year, and got a job at a local company, EMSI, Esso Mathematics and Systems Inc. The company applied computers and math to a wide variety of issues running the oil company. I started by fixing bugs and making changes to their math optimization FORTRAN program that helped them run their oil refineries better. Did I “know” how to do this? In each case I dove in and figured it out.

Then they needed some test programs written a new language, PL/1. No one there knew the language, so I dove into the manual and wrote the programs. Then they needed some accounting programs to help run their facilities in Asia. They wanted them written in COBOL, which made sense for financial programs, so I learned the language as I went along. I quickly fell into the pattern of getting some data from my boss, writing code that did what he wanted with it, showing him the output, and cycling for more. Each cycle required less self-education in COBOL, so I got faster. I tell more of this story here.

What I did wasn’t that unusual at the time. Most of the early programmers basically figured things out as they went along. They needed to get stuff done. They studied programs written by other people and learned how to write their own. Compared to learning to read and write the German language, I found FORTRAN to be amazingly simple by comparison; in fact, it was a relief compared to the endless complications and words of any human language, not to mention idioms.

How you can learn in modern times

You won’t have to deal with card and card punch machines. You won’t have to go to a place with a room-sized computer. You have the internet to give you access to all the information about any language you want, and tools you can access to enable you to write code in that language, try to run it and see how it goes. To start you can use formal exercises with programmed solutions you can study. You can learn and advance at your own pace.

If you want, you can poke around for a local or remote job. It’s good to start with some narrow domain of problems and tools that you’re interested in and have gained skill in. The first job doesn't even have to be programming -- it could be testing, for example. Just get in somewhere and start producing. Once you prove value, you’ll move up from there, because people who can produce results aren’t that easy to find, and a proven producer is a keeper. The important thing is to keep learning and move up the skills hierarchy.

Finally, count your blessings that the ridiculous world of job certification hasn’t yet latched its evil claws into programming very much. Yes, there are ignorant HR people who insist applicants have a degree in Computer Science, but the best programmers with degrees are good in spite of their mis-education in college.

AI can automate what doctors do

There has been a decades-long evolution towards creating an effective clinical diagnosis and treatment AI system, essentially automating the mental part of what doctors do. A solid basis for the content of the system has already been built in the form of medical text books, procedures, published databases, studies and clinical standards such as HEDIS.

The major elements of a fully automated system have been built and put into practice in a variety of medical practices. When a comprehensive system will be built and deployed is impossible to predict. No fundamentally new tech needs to be invented for this to be created; no “break-throughs” in AI! It “just” needs to be applied.

While having an AI-driven medical diagnosis and treatment system would be amazing, much more important than the AI aspect of it would be the fact that it would be data-driven instead of human-created-policy-driven. This means that the system would, over time, determine what actually works based on the data and results, rather than what human “experts” and self-interested institutions say works. In other words, it would support true evidence-based medicine, replacing the too-often corrupt practice of studies published in medical journals. This is a huge subject.

What do doctors do?

They start with the patient’s complaint, why they’re seeking help.

They then get from the patient and/or medical records a time sequence of conditions (like a cough), tests, observations, events (like falling down), related personal things (age, heritage), and finally diagnoses, treatments and outcomes.

Based on this, they make further observations, tests and measurements. The tests may involve other people and equipment, for example a CAT scan. Depending on the expense and trouble of the test and the chances it will affect the outcome, further tests may be performed.

The result is that the doctor recommends and/or performs treatments that will resolve the issue. The treatments can include drugs and operations. The results of all of this are stored in the patient’s EMR, partly coded data and partly written clinical notes.

In order to do the above, doctors receive a great deal of training, both general and clinical. While in practice, they are guided by their knowledge and experience, and also by clinical guidelines and protocols, which evolve over time.

Doctors are limited by a couple of things. First, missing information: they may not have access to and probably don’t have time to read all the patient’s medical history. Second, missing knowledge: there is a huge and ever-growing body of medical knowledge and treatments. It’s amazing that doctors have as much of this in their heads as they do, and not surprising that they sometimes forget or haven’t had time to read and absorb information that is new to them.

Is all the technology required really available?

The pattern of an innovation being proven and waiting sometimes for decades has been demonstrated many times. For example, an algorithm applied in production more than 50 years ago (!) for optimizing oil refinery operations has only recently been applied to optimizing some aspects of health care scheduling. Here’s a detailed example.

No new math or fancy algorithms are needed. The fancy new AI LLM’s (large language models) that are getting attention these days don’t apply to this problem. The vast majority of the effort is in centralizing, codifying and standardizing data that is entered into medical EMR’s, which has already been done and is being refined. Even the tricky work of extracting value from doctor-written clinical notes is largely automated. Large databases of this kind are in use today by pharma companies to help them discover and refine targets for drugs.

The path to automation

The word “computer” was originally applied to people, mostly women, who spent hours and days bent over desks, often with calculators, computing the result of various mathematical formulas. For example:

Barbara “Barby” Canright joined California’s Jet Propulsion Laboratory in 1939. As the first female “human computer,” her job was to calculate anything from how many rockets were needed to make a plane airborne to what kind of rocket propellants were needed to propel a spacecraft. These calculations were done by hand, with pencil and graph paper, often taking more than a week to complete and filling up six to eight notebooks with data and formulas.

While not as precise, doctors are also human computers, in the sense that they confront a new case (problem), get inputs from the patient and the database of the patient’s history, make observations (like calling a data-gathering subroutine), search their memory for a standard to see what to do next (if X and Y, then do a blood test to see if Z). Depending on the results  of that test, there may be further branches (if-then-else) to see what other tests and procedures may be required. Finally you reach a diagnosis and a treatment plan. The results of everything including the diagnosis and plan are recorded in the EMR for the patient to form the basis of future medical interactions.

All of these things are in medical text books, treatment protocols, check lists, medical databases and academic papers. They are all pounded into doctors’ heads by clinical training and apprenticeships. Doctors are expected to remember everything.

The path to automation isn’t fancy. It basically amounts to getting a computer to do what a doctor does: interacting with patient (taking input and providing information), organizing and enhancing the records about the patient, standardizing and digitizing all the existing protocols, and creating digital channels to orders for tests, procedures and drugs. Most of which are already a feature of EMR’s.

Most of the elements of this automation are already in place! WebMD.com, for example, has a huge amount of information about symptoms, diseases and treatments online. It’s medically reviewed, and organized for access by patients. Major hospital systems have similar websites. The websites are just the visible part of the iceberg, with vast underpinnings.

The most obvious missing elements is the ability to request tests and procedures – for that you have to go to a human. But the ability to input requests for such things is already a feature of the EMR’s used by most doctors. Making the connection from the EMR to software instead of a human is a minor task.

Automating doctor decision-making is the heart of the job. It’s essential that this be done using an editable, extensible decision tree. This can be enhanced with probabilities and ever-increasing amounts of personalization. This should not be created by training of any kind; it must be human editable and fully transparent, so that you always can know exactly how and on what basis every decision was made.

Among the biggest missing elements are things that doctors learn during their clinical training and personalization.

Once all these elements are put together and working, you would enter a parallel production phase, in which the computer would get the same inputs a human doctor would and propose what to do next. This would be recorded and compared to what the human doctor did in classic champion/challenger fashion. The system wouldn’t have to be 100% complete to be put into live operation, so long as a good system for bailing out of the computer and shifting to a human doctor was in place. But since such a large number of patient visits are routine, the computer is likely to be able to handle a large fraction of cases from early on.

There is a huge amount more detail in the building of such a system. However, surprisingly little needs to be “invented” to make it work, given that large elements are already built and in production in limited ways.

Related posts

Doctors too often get the wrong answer. This is the kind of thing that makes some people hope that automation could do a better job:

https://www.blackliszt.com/2016/12/what-can-cats-teach-us-about-healthcare.html

Massive spending has gone into "cognitive computing" and healthcare, with nothing to show for it.

https://www.blackliszt.com/2015/07/cognitive-computing-and-healthcare.html

You don’t need AI or cognitive computing to discover or promulgate the new discoveries that humans make.

https://www.blackliszt.com/2015/08/human-implemented-cognitive-computing-healthcare.html

Health systems have trouble just making computers work. When they try to do something "fancy," the results are usually poor. But there are promising exceptions.

https://www.blackliszt.com/2016/05/healthcare-innovation-can-big-data-and-cognitive-computing-deliver-it.html

Healthcare systems spend huge amounts of money on things related to AI, but they don't know what they're doing and neglect to spend on simple things that could make an immediate difference.

https://www.blackliszt.com/2016/09/healthcare-innovation-from-washing-hands-to-ai.html

Avoiding error is hugely important.

https://www.blackliszt.com/2017/06/how-to-avoid-cutting-off-breasts-by-mistake.html

A major lesson from the above posts is this: while AI can certainly automate what doctors do, having the usual major corporations and medical systems be in charge of the effort guarantees failure -- which billions in wasted spending to date demonstrates.

The benefits of medical automation

The potential benefits of automation are huge.

Cost of medical care: As medical workers are replaced by software, costs will go down. Not just salaries, but also office space, etc.

Medical care waiting times: The software doctor is available 24 by 7, no scheduling required.

Accuracy of care: Medical people can’t be as consistent or up to date as data-driven software. Elaborate measures such as HEDIS for judging medical care after the fact will be applied as the care is delivered, assuring its accuracy.

Transformation of care: Dramatically better health and lower costs will result once the system is in place and real-world evidence from it supplements, personalizes and replaces existing care practices.

Automation of medical care isn’t without problems. The institutional obstacles are huge. Mountains of regulations and standard practices would have to be changed, with entrenched forces fighting every step of the way. The people whose jobs are threatened will resist. A large number of patients value interacting with a human doctor. Corporate forces will fight to have their interests supported in the rules and data of the automation. There will have to be a way to provide alternatives and avoid centralized government control, which will be a major struggle, and a danger I fear.

Conclusion

           Automation of medical care has been underway for decades. All the technical elements to enable it are available. The benefits of automation are large, but so are the obstacles to implementation. Centralized control of medical diagnosis and practice is already strong, and automation would make it stronger and less visible. The path forward is  likely to remain slow. While there are substantial potential benefits in terms of cost reduction, better time and accuracy, the largest potential benefits of huge cost reduction and improved patient health are threatened by a version of the centralized control embedded in the current partly-mechanized system being translated to the automated one.