Blissful Life

Category: ideas

  • The Ideal Physician AI Assistant

    When I hear “Artificial Intelligence” and “Healthcare” together in a sentence, it is usually never a pleasant thing I’m listening to. There almost always is some kind of reinvention of wheel where Google’s hardware cycles are spent in trying to solve something meaningless.

    For example, it is futile to differentiate between tuberculosis and cancer from an image of the chest where the answer may never lie in the image, but rather in the symptoms of the patient. Even if AI tells the physician that the ECG it is reading is normal (which the physician probably noticed on their own), the physician still has many reasons to refer the patient to a higher center.

    These are isolated examples. But it is the isolation that makes these good examples. AI’s role is not in isolation. AI’s role is in integration. AI (or computers) should come in and fill in where humans struggle – processing large amounts of data. (Processing data, not for the sake of figuring out patterns that humans have easily learned, but for the sake of figuring out patterns, perhaps within an individual, that a human cannot easily learn by going through information)

    AI can be a very good physician assistant. I have previously written about an intelligent EMR. The only barrier to using digital EMRs is the user interface. There are ways to optimize that interface. An intelligent combination of predictive suggestions, tapping rather than typing, reading data from text, etc will help.

    Once physicians can start using EMRs the possibilities are endless. Here is a list of things that come to the top of my mind:

    1. Intelligent to-and-fro symptom/sign/examination suggestion (that physicians can use to not miss important symptoms)
    2. Standard treatment guidelines based suggestions on medications and investigations
    3. Drug interaction checker
    4. Locally relevant and contextual antibiotic resistance patterns
    5. Patient’s past reports based insights, trends, analytics, etc
    6. Medical records exporting, highlighting important information, etc.

    If you are interested in building something like this with me, let me know.

  • Product Idea: “Explain My Prescription To Me” Service

    Many doctors have very little time to spend with the patient. So little time that sometimes they start writing prescription before even arriving at a provisional diagnosis. Imagine how then, would they explain to their patient why they have written a particular tablet for them?

    Is there a product/service idea in this vacuum of counseling that should have been provided by the doctor? Are there people who are not able to ask the right questions to the internet to find the answers?

    Perhaps there should be an app that is front-end for a prescription description service. The user uploads their prescription and also attaches a short voice note with their symptoms. This goes to a doctor/nurse/pharmacologist/pharmacist at the back-end who responds by reading out the prescription and counseling the patient about what the medication is, what it does, what side effects can be expected, etc.

    The counseling of each medicine can be recorded and reused for the next patient who is prescribed the same medicine. That way, the time required by the specialist is minimized. If a patient can afford and requires a longer, customized consultation, they can request that for a higher cost. Then it can work economically as well.

  • Healthcare With Smartphones?

    Imagine this. You are a 36 year old lady working as maid in two houses, not supported by an alcoholic husband, and mother of two school going children. Your husband had a wart on his genitals and now you have it too. It is not particularly bothersome, but you are not sure what you should do about it either.

    You think it would be a good idea to go to a hospital, but which hospital? Which doctor should you meet? Would they judge you? Do they charge too much money? Is it going to hurt? Whom do you even ask these questions?

    Enter our app.

    You will be asked a series of questions in your own language. The questions get more and more specific as you answer them. They will also be read out to you in case you can’t read. By the time you have answered about 10 questions, the app knows what your problem is.

    The app has been fed with a well designed set of protocols/algorithms that need to be followed in each situation. It has a curated collection of resources (hospitals, clinics, labs, lawyers, etc) which are guaranteed to give you quality care without judging you or making you feel uncomfortable.

    The app might suggest you to get over the counter paracetamol for a fever it thinks is not serious. But for your wart it is suggesting that you consult a dermatologist near you.

    You can book an appointment with her through the app at a time convenient to you. Your data will be passed on to her with your consent. Later at the clinic you can start from where you stopped.

    —

    Information is strength. Knowledge is power. As a doctor, I have witnessed countless situations where patients struggle because they did not know the right room number or doctor’s name. What the app does is eliminate those knowledge barriers by presenting trustworthy and relevant information in a friendly interface.

    The possibilities that this idea brings are endless.

    It can be tied together with a call center where people who do not have a smartphone can be given service.

    A subscription based service that gives discounted rates for various medical tests and consultations can be introduced.

    Micro-insurance schemes cab be brought in.

    Transgenders can be employed as distributors of the app to otherwise hard to reach strata. They can be given additional training to be able to work more or less like ASHAs in the community.

    Otherwise hard to navigate healthcare facilities can be easily navigated. (Think of how easy google maps has made walking around an unknown city)

    —

    The bottlenecks:

    Affordable healthcare providers. Whom do we have?
    Government is cheap. But what about quality and comfort?

    It might be possible to tie up with private companies under Corporate Social Responsibility to fund the charges at private hospitals. But the costs can become too high too soon.

    An intelligent mix match of services public and private using a friendly application can solve this in my imagination.

    And that is what I am working on now! Ping me if you’re interested to join forces.

  • Reimagining Kinnars (Hijras) as Health Workers for Reproductive & Sexual Health

    Supreme court struck down section 377 a couple of days back. India is moving forward in the right direction. There is still a lot of work left to do.

    We see kinnars (hijras) working on roads asking people for money. They are “strange” humans in other people’s mind and when people look at them all that comes to their mind is their gender issues and sexual connotations of those.

    This strangeness might be the reason why people are unwilling to give them jobs. Could we turn this around?

    Reproductive and sexual health is a topic that is absolutely neglected in our education system. There is plenty of embarrassment in discussing topics related to sex as well.

    What if, then, we put them both together? What if we empower kinnars by giving them training and other resources required, to go into the community and work for reproductive and sexual health? Like ASHA workers they would promote health. They can even distribute interventions like condoms.

    This is a positive change in many ways:

    • The kinnars themselves gain a better understanding of health
    • This can become an income generating opportunity for them
    • The perceptions about transgender community can slowly be changed to a positive one
    • Otherwise hard to reach strata of society can be reached

  • My Idea of a Perfect Electronic Medical Record System

    The COWs are coming to our hospital.

    No, not these. They’re getting more attention than they deserve.

    Our hospital might soon switch to an Electronic Medical Record system. And this will bring in Computer on Wheels, COW as they’re affectionately called in other hospitals.

    More like this

    While that makes me more happy about where I’m working, it also brings back a lot of ideas I’ve had during medical school. I have seen hard problems for humans that are pretty easy for computers to solve. I have seen processes that could be hastened by leaps and bounds if computers were involved even partially.

    The Perfect Electronic Medical Record System

    The perfect EMR does not just record what the physician or nurse puts in. It is an intelligent assistant that does some thinking of its own and comes back with suggestions and autofills for the physician or assistant.

    For example, when a child comes to you and her mother says she has fever, you start entering “fe…” and the EMR autofills fever. Next you can enter the duration from a dropdown menu. Also associated symptoms can be ticked “yes/no”. As you’re done and move to the next row, the computer automatically populates an entry – “Cough? Yes/No”. If you choose yes, it asks you for characterization.
    If at any point you’re in doubt or do not want to characterize a symptom, you can just delete the autosuggestion and move on to the next line.

    Once you are done with the symptoms and exhausted the negative history that the smart EMR suggested for you, you can enter the examination findings. Again the EMR will suggest for you the most important findings you should not forget to look for based on existing data on what the most common findings are for that particular set of complaints.

    Later, the computer will show you a list of provisional diagnoses based on the data you’ve entered about the patient, and the past set of data the computer has, and even the compendium of knowledge that it potentially has access to. You can reorder the diagnoses if you like.

    Accordingly the computer will suggest investigations and management plans with dosages calculated according to weight or dosage adjustments that are required for special conditions.

    The computer will always suggest and ask you for guidance when in doubt. You can always override the computer, but you can also take computers help in not missing important things.

    Such a system might not be useful for an expert clinician, but it will definitely help a new doctor in emergency rooms late night. Most importantly, the system makes sure that an exhaustive history taking and examination has been done. It also helps in making clinical decisions based on data and evidence.

    Impossible?

    Certainly not. I just discovered that Dr Lawrence Weed, MD has been saying this exact thing since years. He might even have developed such a system already. But EMR systems do not seem to have this kind of intelligence integrated yet.

    The accuracy of simple software powered by big data like Akinator is testimonial to the power of computers when it comes to problems like this.

    Clinical decision making is no holy grail and it will soon be heavily relying on, if not replaced by, artificial intelligence.

    Pictures courtesy pixabay.com