[25/11, 12:27] Patient Advocate 29M Quantified Self: Various possibility for this rare event, including possible mannual error (but not only that).
[25/11, 12:34 rb: Need more details of the procedure including OT notes to comment
Lots! Transparent and accountable healthcare still sounds like a dream promise of health IT but the above case scenario is the reason we need more health IT driven transparency and accountability in data driven healthcare
[25/11, 12:48] AC: First... I think the doctor is a thorough professional with very high integrity. Lithotripsy is not a "guaranteed" treatment
Second... From medical devices perspective - the brand of the machine could be at fault too and not so sure about the professionalism of the technician.
Is the planning and targeting correlation demonstrated for the device. Ultrasound penetration is heavily dependent on tissue type. Improper targeting "can" cause pain during lithotripsy. Does it? Will need to check literature
[25/11, 13:28] RKB AIH: When an engineer makes a faulty bridge or financial error with significant losses he literally pays with his job. It becomes his responsibility primarily, not the bull dozer or the computer on which he did his calculations. And yes, if there are inconsistencies those external factors are also punished.
[25/11, 15:34] RKB AIH: The pilot is first suspended & put on probation until pending investigation of Blackbox contents before expert panel & TSA reps
Clinical science doesn't have checks and balances for high accuracy, neither funding to brute force diagnostics, nor safety profile to brute force treatment plan. So we keep steering decision (approach) for improving it according to data gathered while discourage overdiagnosis and overtreatment.
Airlines and bridges are not operated in uncertain / risky environment beyond a small limit. Thanks to Technological advancements.
Ex. - failure rate of lithotripsy (kidney stone breaking procedure) is 10-30%
The reason I mentioned the pilot example is from the perspective of safety critical systems.
Medical systems are some of the most common case studies in safety critical systems engineering
[25/11, 16:35] rb: Yes and the other point is that medicine has been hyped so much over the last few decades by unscrupulous vendors wanting to profit that the public feels it's as good as a plane ride most of the time while the truth is it's much more complex than that!
Perhaps one could have an interesting model interface based on a commercial flight analogy where one may enter the predictive variables of any patient (aka plane traveller) entering the hospital and based on his her comorbidities try to predict the likelihood of his her reaching destination health safely and very importantly the probable time taken to reach there! @AC @RKB AIH
[25/11, 16:40] RKB AIH: I respectfully disagree…every faculty has risks & guardrails. Aviation risk factors along with flight parameter risks coupled with environmental factors are one of the most dynamic risk factors in the world.
[25/11, 16:43] Patient Advocate 29M Quantified Self: Medical science is lagging far behind in terms of technological advancement and it's access.
In an aeroplane you have a normal person without any illness boarding the plane and getting down at the destination
In an OT you have a person with pre existing illnesses getting operated.
Now his chances of surviving the operation will depend on his pre existing illnesses vs no pre existing illnesses.
Next would come the efficacy of the intervention in solving the problem (here the plane analogy may not hold good unless dragged into a further train or car journey till they reach home)
[25/11, 16:49] Patient Advocate 29M Quantified Self: *Lagging far behind in terms of promising high certainity for individual case.
Surely huge advances for humanity.
[25/11, 16:50] RKB AIH: In case of a flight there are thunderstorms, mechanical failures/bugs (remember boeing 737 rudder issues on fly by wire) etc etc, shortage & overworked personnel, wars & diversions to unknown terrains…yet the casualties r minuscule
[25/11, 17:00] Patient Advocate 29M Quantified Self: Atul gawande pioneered checklist use 20+ years ago, still same example is repeated in these papers..they are useful. Even 1% error correction is good and it did more..but still very very tiny thing if we consider over all clinical practice.
[25/11, 17:02] Patient Advocate 29M Quantified Self: There are many more strategies Implemented and practiced and then the current quality standards are met which is still too low compared to airlines accuracy.
[25/11, 17:02] RKB AIH: So nothing changed to correct course in 20 years ? Maybe its the reticence that need to be addressed first
Because everything is controlled factors like random pockets of turbulence are avoided. Passenger have died owing to turbulence in the past few weeks
25/11, 18:50] rb: This would be analogous to anesthesia related deaths and are fortunately rare
[25/11, 18:23] AC: This video will most likely change the perception people have of flight safety.
Healthcare is a journey and has a certain course and every illness as well as patient journey is different because of the greater amount of known and unknown variables
An illness as mentioned earlier can be likened to a plane ride and there are possibly multiple interventions involved including medications, operations and physiotherapy (each of these interventions are different vehicles involved in that plane ride including the vehicles one uses to reach the departing and arrival airports).
As mentioned earlier the outcome of that travel ending safely and quickly can be predicted to a certain extent by the comorbidities the patient traveller already brings as a baggage.
Now in this flight analogy while the plane offers a lot of significant contribution in the journey, not all operative interventions may offer the same support in terms of solving the patient's main problem other than guaranteeing a smooth passage into deeper planes of anesthesia with some statistical guarantee of emerging out from it.
The main problem/s of the patient could fester and persist prolonging the journey (most of the times it's between the airport and home) and eventually time emerges as the greatest healer. Inspite of all the hype when it comes to healing efficacy no single intervention has been able to compete with time in terms of healing (which also means it's the internal medicine intelligence of our 30 trillion celled bodies that eventually decides their own outcome).
Human reticence to understand and overcome these challenges through data driven healthcare stems from their reticence to capture and share data in a transparent and accountable manner
[25/11, 19:18] AC: Nice points. One major difference - airlines usually begin the process with healthy people and healthy machines.
In healthcare, again usually - maybe hopefully, the machines are healthy, but invariably 50% of the equation are not. They are called patients!
Or, flying a sick plane.
[25/11, 19:35] S Ed: You know the structure and mechanics of the plane because humans designed it. With humans, the "normal" is also not fixed. Your heart may not be in the place you expect and you may be fine.
[25/11, 19:43] RKB AIH: What stops us from perfecting the understanding ? This is why rb’s last paragraph is very important
[25/11, 19:45] RKB AIH: We do have sick planes !! Pilots deftly maneuver around those drawbacks…u will dread flying and I will guarantee u will have heart attack 😀
[25/11, 19:48] RKB AIH: Believe me I pray to thousand gods before setting foot on a plane…ignorance is truly bliss in this regard
[25/11, 19:48] AC: So we need people designed by humans.
[25/11, 20:02] rb: Problem is treating digital twins will not cure their physical brethren!
[25/11, 20:03] rb: Robots yes but most of them will likely have their own healing systems embedded in their design very much similar to how evolution created our trillion celled collaboration over ages
[25/11, 20:06] AC: Yes. Synthetic diseased beings get recycled without a thought.
Biological ones get recycled too... But the emotional loading is high.
From a disease loading perspective, make designer babies with all attributes one likes.
[27/11, 10:18] rb: "3 men were killed after their car drove off an incomplete bridge and plunged onto a riverbed while they were navigating with Google Maps in Uttar Pradesh, India.
A local police officer reportedly said the front portion of the bridge collapsed into the river during floods earlier in 2024, but this change was not reflected on the navigation app.
He added that there were no safety barriers or warning signs on the bridge.
A Google spokesperson said: “Our deepest sympathies go out to the families. We’re working closely with the authorities and providing our support to investigate the issue.”
[27/11, 10:21] RKB AIH: So is Google liable for their deaths ?
[27/11, 10:22] rb: Very interesting question!
[27/11, 10:23] JG: No, read terms of service disclaimer. Liability is with local authorities for lack of road safety signs.
[27/11, 10:23 rb: Let's reframe the question and answer in the light of aviation and healthcare analogies and let's try to spot the transportation and healthcare analogy here
[27/11, 10:27] rb: We too practice this kind of healthcare in what we have started calling "global learning toward hyperlocal caring" and we often with all our texts to the patient ask them to get our suggestions verified with their local doctors in the same style as Google! But can they? Obviously not and that's the reason they are reaching out to us distantly with their hyperlocal needs!
So this is a current problem in E healthcare that we have to contend with?
[27/11, 10:28] rb: It's a collaborative endeavour and everyone including the drivers (tech end user) is liable?
[27/11, 10:29] JG : The user has to first accept terms of service then only can use it. So Google most probably won’t be held liable. They can challenge it in courts.
[27/11, 10:31] rb: Yes and accepting it is as simple as clicking on "i agree" at the time of downloading it from Google playstore? What if the app was pre-downloaded by the mobile phone vendor?
[27/11, 10:31] RKB AIH: Exactly!! Google can’t look for you & you can’t put signs everywhere. It’s a collective responsibility, yet when an incident happens Nobody is automatically absolved, hence Google had to reach out and intervene despite its job being accomplished
[27/11, 10:34] RKB AIH: I dont know how it works in medical field..can someone extend this to a medical world…what happens when it goes south with a patient?
[27/11, 10:34] JG: If ordinary folk get large compensation from negligence of administrators, we would have a different scenario. Holding tech companies liable is an easy way out.
[27/11, 10:35] JG: Same for using AI in medical field. No liability
[27/11, 10:35] rb: We don't have a Google map of healthcare!
[27/11, 10:37] RKB AIH: Isn’t that what medical education is supposed to do ? Ur doctors are ur gps to maneuver ur health in right path ?
[27/11, 10:39] rb: Yes but our map is not as great as the current google map!
Healthcare is still largely 99.999% uncharted territory! Although the percentage perspective will vary with your years in healthcare. For the first year medical student its 99.99% charted territory
[27/11, 10:43] RKB AIH: That’s what boggles mind. Pardon my medical ignorance. If one company could map every single road in the world in a span of 2 decades, why can’t a conglomerate of medical professionals across the entire globe map a human body correctly with it’s characteristics
[27/11, 10:44] rb: The human body is analogous to the universe and the mapped surface of thd Earth is just one of it's trillion cells!
[27/11, 10:48] RKB AIH: Agreed. But most treatments are at bone/muscle/nerve level wouldn’t that make it easier to atleast partially cover major ailments?
[27/11, 10:50] rb: Yes that's covered but currently that is way inadequate to match up to the travellers requirements and expectations
[27/11, 10:54] rb: I think i shall offer that as a hook to the audience tomorrow to break the ice! Can we engineers and doctors think of building a google map of healthcare together to help navigate billions of healthcare journeys!
That will be our ask!
[27/11, 11:23] AC: It needs to be formed like a crypto
[27/11, 11:27] AC: Because google maps is crowd sourced
It has been done on humans
1. HeLa
2. The subjects of visible human project
3. Human genome project
4. Human Brain project (on going)
Just like physicians are the only "doctors" folks encounter, gmaps is the only mapping folks encounter.
For instance, not many here would be aware of open cell id project
[27/11, 11:53] AC: This was sunset last year
[27/11, 11:55] rb: Yes but here the analogy is to gmaps and individual journey navigation.
The projects you have listed may not compare?
We need to compare it with individual patient journeys!
Let's say for example an 80 year old man has a fall and other than a black eye also notices left upper limb weakness and visits a neurosurgeon who advices an operative intervention to decompress what appears to be a spinal cord compression on MRI in the hope that this will give him better mobility and then the journey begins with the current google map of healthcare!
[27/11, 11:59] AC: Think again!
The map is one, how it's used presented is unique
The same happens with genomic medicine
Should begin to happen with neuro
.....
The challenge lies in not being able to crowd source the map of each individual.
Even gmaps fails outside cities.
....
The only crowd sourcing individuals can do is my the crowd of themselves... Their longitudinal EHR
....
Don't disagree with your view on personal patient journeys, but maybe the analysis you applied needs a refactoring
[27/11, 12:02] RKB AIH: Map is only the guideline, how you use it for individual patient care should be ascertained
I draw analogy as map is like foundational model & you transfer learn for downstream tasks
[27/11, 12:02] AC: While you type let me try chatPakoda... Whether it gets the nuances
[27/11, 12:03] RKB AIH: Like Dr. Hinton eluded trying g to make sense of entire brain in one shot may be too much…localized understanding may be the key
[27/11, 12:06] AC: Your and Dr. rb's approaches differ in focus and perspective, stemming from your professional backgrounds:
1. Approach to the Problem
Engineer, Reductionist perspective:
Focuses on the system-level architecture and foundational principles (e.g., maps as a unified dataset, EHR as the data backbone).
Reductionist in that you break the problem into components (data, crowdsourcing, maps) and analyze scalability, reliability, and use cases.
Emphasizes individual data aggregation (longitudinal EHR) as the key to personalizing the map.
You view the map as a system shared by all, where the uniqueness arises in interpretation and application.
Clinician, Patient-Centric perspective:
Views healthcare from the lens of individual patient journeys.
Focuses on the experiential aspects of the journey, such as decisions, interventions, and their outcomes.
Analogizes the map to the navigation of individual clinical situations, where personalized deviations from the norm are key.
Highlights real-time problem-solving, especially in complex, patient-specific contexts.
2. Underlying Differences in Professional Backgrounds
Engineer:
Deals with complex systems, requiring abstraction, scalability, and process optimization.
Prioritizes systems integration, repeatability, and leveraging data infrastructure for consistency and personalization.
Often reductionist to analyze individual components of a system, but capable of understanding emergent behaviors in complex systems.
Clinician:
Focuses on human-centric, real-time decision-making, rooted in individual variability and holistic outcomes.
Experience-driven, often anecdotal, yet attuned to system-wide limitations impacting patient care.
Less focused on systems engineering but deeply invested in navigating complex biological systems.
3. How This Shapes the Analysis
Engineer:
Emphasize creating a reliable, universal framework (the map) where personal journeys derive from individual data integration.
Advocate for longitudinal health data to serve as the patient's own "crowdsource" resource.
View variability as a result of system limitations in data aggregation or infrastructure, solvable via better tools (EHRs, genomic databases).
Clinician:
Focuses on the individual's immediate experience, prioritizing the journey itself over the infrastructure.
Sees variability as intrinsic to patient care, emphasizing adaptive decision-making over universal frameworks.
Conclusion
Your differences likely stem from your distinct professional paradigms:
The engineer sees systems; the clinician sees stories.
Both approaches are complementary: the system you envision supports the individualized journeys he focuses on. The challenge is bridging your system's abstraction with his patient-specific adaptability.
[27/11, 12:07] AC: I would say this is an "artistic" reply from the AI
[27/11, 12:07] SA: Very very nicely put
[27/11, 12:10] RKB AIH: Well summarized. I agree, this is why medical AI help as they r complementary
[27/11, 12:14] AC: I liked the part where it says "engineers see systems, doctors see stories"
[27/11, 12:21] rb : There is a story inside every system!
[27/11, 12:22] rb: Can no longer call it chat pakoda! It's proving to be anandamide!
[27/11, 14:16] JG: All the best. If you can map all the trillions of neural pathways in the brain of a single individual, that alone will be similar to mapping the universe.
[27/11, 15:55] AC: I guess I know why... We can't even lend a simple pen to someone without worrying that it's not coming back to us!
What a population! Our lack of integrity is the biggest proof that modern humans originated here.
Originated here and conned all the other hominid species out of existence.
[27/11, 23:32] SN: The road department got away with the death of the Tata executive where 3 lanes suddenly became two with a walk in front. Regular people are immaterial for our killer roads.
[27/11, 23:35] SN: They could have had the same accident even without Google. Fault lies with whoever is responsible for maintaining the road. They should put barriers and signs to close the road.
[28/11, 00:41] AC: The robots are not the challenge, the network latency is.
One off demo can be done, but deployment at scale is not feasible
[28/11, 00:43] AC: Ref CNS for the planet
It should now be obvious why a global round trip for packets needs to happen faster than our NCV to the digits
[28/11, 01:57] RNG HIT: And that is under major threat from uncontrolled growth of AI 🤔. End of Kalyug?
[28/11, 06:20] AC: Rather beginning of its patal phase...
Patal is true singularity... Before that there will be the various other layers.
[28/11, 06:24] AC: Anyway, pataal, rakshasas etc are not what mass media of the oceanic type (read in English) have made it out to be.
I tried to use AI to generate images of Pataal which is an utopian abode with Rakshasas who are the most attractive beings. Will share if I find it.
[28/11, 06:35] AC: Found this. Most of the description I remember were captured by AI.
1. No sunlight, lots of "mani"
2. Even nights are glittering bright
3. Lots of magnificent contraptions
4. People who live there use artificial means to look very attractive
5. Diseases are almost non existent
6. Abode of other life forms like Nagas who are super intelligent
7. Buildings touch the sky
[28/11, 06:38] AC: Anyway, this deserves a little explanation of how it's digital health related.
Patal lok has a model of UHC. Universal Health Coverage
Those were the thought experiments of our Rishis on how health for all could be achieved
Patanjali was claimed to be a Naga.
[28/11, 06:42] AC: This article touches upon how the purported Pataal entry remained covid free.
(Though I disagree with it literal interpretation)
[28/11, 07:06] rb: Sounds like internal medicine as in getting into our internal universe of trillion cells and their interstitial SPACE!
[28/11, 07:17] AC: Ayurveda and Yoga Sutra do link these to the body
[28/11, 07:18] AC: Matches our maps of healthcare and geo maps discussion yesterday
The anecdotal stories have survived... Which is good
However, most of the anecdotal interpretation have been lost
[28/11, 07:21] rb: Yes so in this collaborative Google map of healthcare that we hope to collaborate on here, there's an external road map of external patient life events and there's a parallel road map for internal medicine life events that happen inside the patient's body and these maps are connected but we still need to make the connection explicit
[28/11, 07:22] rb: Essentially external and internal data points surrounding an individual's car journey through birth and death
[28/11, 08:59] AC: Medical informatics needs robustness like financial informatics
The amount of modelling that exists for Fintech is exponentially higher than medtech (for the clinical encounter, not high end devices)
Here's a summary, key words, and thematic analysis of the conversational transcripts:
Summary
The conversation revolves around the analogy between aviation and healthcare, exploring the complexities of healthcare and the potential for creating a "Google Map" of healthcare. The discussants delve into the challenges of mapping individual patient journeys, the role of technology and data, and the differences between engineering and clinical perspectives.
Key Words
1. Aviation and Healthcare Analogy
2. Google Map of Healthcare
3. Individual Patient Journeys
4. Technology and Data in Healthcare
5. Engineering and Clinical Perspectives
6. Medical Informatics
7. Robustness and Reliability
Thematic Analysis
_Codes_
1. Aviation and Healthcare Analogy
2. Complexity of Healthcare
3. Technology and Data
4. Individual Patient Journeys
5. Engineering and Clinical Perspectives
6. Medical Informatics
7. Robustness and Reliability
_Themes_
1. _Aviation and Healthcare Analogy_: The conversation starts with the analogy between aviation and healthcare, highlighting the complexities of healthcare and the potential for creating a "Google Map" of healthcare.
2. _Complexity of Healthcare_: The discussants emphasize the complexity of healthcare, with multiple variables and uncertainties, making it challenging to create a comprehensive map.
3. _Technology and Data_: The conversation explores the role of technology and data in healthcare, including the potential for AI, machine learning, and data analytics to improve patient outcomes.
4. _Individual Patient Journeys_: The discussants highlight the importance of individual patient journeys, emphasizing the need for personalized care and the challenges of mapping these journeys.
5. _Engineering and Clinical Perspectives_: The conversation reveals the differences between engineering and clinical perspectives, with engineers focusing on systems and clinicians focusing on individual patient stories.
6. _Medical Informatics_: The discussants touch on the importance of medical informatics, highlighting the need for robust and reliable systems to support clinical decision-making.
7. _Robustness and Reliability_: The conversation emphasizes the importance of robustness and reliability in healthcare systems, drawing parallels with financial informatics.
Learning Insights
1. _Interdisciplinary Collaboration_: The conversation highlights the need for interdisciplinary collaboration between engineers, clinicians, and other stakeholders to create a comprehensive map of healthcare.
2. _Personalized Care_: The discussants emphasize the importance of personalized care, highlighting the need for individual patient journeys to be mapped and understood.
3. _Technology and Data_: The conversation reveals the potential for technology and data to improve patient outcomes, but also highlights the challenges of implementing these solutions in complex healthcare systems.
4. _Medical Informatics_: The discussants highlight the importance of medical informatics, emphasizing the need for robust and reliable systems to support clinical decision-making.
5. _Complexity of Healthcare_: The conversation emphasizes the complexity of healthcare, highlighting the need for nuanced and multifaceted solutions to improve patient outcomes.
CC licence for the above image: https://commons.m.wikimedia.org/wiki/File:980310-N-7355H-003_Simulator_Training.jpg#mw-jump-to-license
