Aditya Samitinjay

@Harper Thank you so much for your kind words. I shall bide my time with the NHS for now. However, the idea will take precedence in the future and I shall go where the wind blows! This is something I can devote my life to. Thanks again for your message, looking forward to interact further.

Welcome Adam. Thanks for sharing your project here. Have you done a SWOT analysis for your project? Please share them here and we can discuss as we go. Thanks Aditya

@Harry Thanks so much for your response. I'd be happy if you could give me any leads or even anyone on this forum who could work on what we're looking for - reg. food recognition models. Indeed, it is abject frustration with current healthcare practice worldwide and yes I actually think most patients are tired of bouncing around different specialties! No specific moment in particular - The first time we did it wasn't so spectacular as it was relatively minor illness. However once we started doing for chronic illness, we discovered its game changing potential. Idea had been brewing for a couple of years before we made it mainstream. Thanks, Aditya

Thanks for your encouraging words! As with most things in life - just a happy accident! Became a doctor because well it was cool to become one back then (still is perhaps, I don't know). Shortly after dad started getting sick and was diagnosed with kidney disease. That's when I started having "skin in the game". Suffered quite a bit and eventually started understanding patient problems better and one thing led to another and eventually this idea surfaced. Looking forward to interact further.

Hi my name is Aditya, I'm currently a doctor (specialist doctor in Endocrinology & Diabetes) working in the NHS. Since the last couple of years, I have noticed the principal problem with current healthcare practice globally (having experienced healthcare systems in 3 different countries as a doctor) is that it has shrunken patients into a set of numbers. Too many guidelines, too many pathways and too many referrals for numbers, completely forgetting that a human being exists behind those numbers. At the outset, it will first seem like "safe practice" but once you get into the rigmarole yourself, you will see how frustrating this can get just bouncing around different specialties with no help in sight. Make no mistake, quite a number of patients do benefit with this kind of practice, however a huge number are utterly dissatisfied with this and are pretty much helpless. Why is this relevant? While IT systems have progressed leaps and bounds - democratizing knowledge to such an extent that it is so easy for an individual to enter the market, solve problems themselves and perhaps even create value. Heck, I setup my own homelab (NAS) running Truenas Scale and my own cloud server (Nextcloud), my own home theatre (Plex Media Server) and so much more because access to knowledge on how to set these up is out in the open. It just requires the will and some commitment. Medicine and healthcare though is severely lagging behind, particularly in democratizing knowledge. While influencers have tried to cover this gap, it still does not cover for the actual knowledge gap. Influencers have only covered the vacuum left by academics making healthcare too complex for the average individual. No wonder we are thus seeing extremes in diets and exercises. Think of it, these are not tailored to benefit you as an individual but to generate more views and eventually more money for the influencer. Hyperpolarizing and extreme views generate views and capital and hence we are starting to see these in these spaces. What is my goal ? I'm a staunch, a very staunch believer of the free market. One of the key principles of a free market operative is democratizing knowledge to the individual so that they are empowered and become autonomous with their own health and make better choices. In the process, they also invariably enter the free market, get exposed to the "spontaneous order" of the free market and thus generate ideas and perhaps eventually create more value and capital. How do I plan to apply this idea to healthcare ? I see a Diabetes clinic every Wednesday and I can tell you nearly all patients with long term diabetes are extremely frustrated with the care we offer. It does not necessarily suffer from a lack of quality but from a lack of freshness and ideas. No single diabetic patient is the same and they have their own fine tuning and balances which they would like us to adjust our care to, however we often tend to give generic advise, pretty much targeting numbers and variables. I have had patients who said they were comfortable with having low blood sugars (by textbook standards you'd expect someone to have symptoms with this number) to a point they were frustrated when I asked them to reduce their insulin dose. I'm sure they may not pay heed to my advise. And rightly so, the patient knows where they feel best in themselves. They'd also like to be autonomous with their healthcare decisions and of course they'd (by virtue of "owning their disease") be accountable to their eventual outcomes. These are not unusual patterns either, this is common and I've also had patients experiencing symptoms of low blood sugar at a level which is considered above normal ranges for an average non diabetic individual! Such is the variety each patient offers. I'm also sure most doctors are aware of this wide variation and do acknowledge this. I'm sure they are also quite frustrated with being unable to offer more to their patients. The biggest bottleneck in being unable to offer more is the lack of longitudinal data - lack of follow up on what happened to patients who were okay with low blood sugars as their normal (did they die, did they get hospitalized, did they need regular correction of their blood sugars and did they have a collapse and needed support?) and what happened to those who numbers were so high but the patients themselves did not feel anything (what happened to their eye disease, their feet, their nerves, kidneys, liver, heart etc.). Basically the operating question is - what and where is their sweet spot (without putting them in distress) where they appear to have best outcomes ? While studies have attempted to answer this - too many studies, in fact all studies have pretty much homogenized patients, who become numbers. Data is usually gathered cross-sectionally in snapshots (yes, even in longitudinal studies!) and thus the individual narrative is buried in these numbers! Not just diabetes, it applies to every disease - heart failure, irregular heart rhythm, neuropathy, parkinson's and practically any chronic disease! What is my idea ? My idea is to democratize knowledge to patients. We (me and my small team) have devised an idea called PaJR (patient journey record) - a PaJR is basically a patient's own healthcare record - this is usually logged daily on WhatsApp (fully deidentified and HIPAA compliant and currently running in India) and a team of doctors, along with a patient advocate is part of the WhatsApp group. Basically we use something called 'Conversational Decision Support System (CDSS)', where in the process of helping our patients, we also educate our patients on their own health problems. While current 20 - 30 min F2F consults are cross-sectional and months apart, PaJR is daily and longitudinal and brings knowledge to the patient directly. Through PaJR we observe our patient's living daily - largely through what they do everyday through their own log (allowing for spontaneity and patient's own description rather than a set of closed ended questions) and we do this continuously sometimes over months or even years and study how their daily living is affecting their clinical outcomes. We need not do guidelines or SOPs, instead we are guided by patient centered outcomes and all our medical decisions pivot on achieveing this goal. This makes the individual the centre of his/her own healthcare and also learn with us (the doctors themselves are learning how to logitudinally observe patients everyday and patients are learning with doctors on what governs their medical decision making). I see no other way than this to acheive patient centered outcomes, give more autonomy to patients and ultimately democratize knowledge to them. Here's an example - We had a 56/F with 2 year history of dry cough, particularly more in the morning hours. While most doctors are able to make a diagnosis of Bronchial Asthma (through history and appropriate tests such as Spirometry/PFT), getting to the bottom of the problem remains elusive to them. We certainly have made advances in identifying what is the allergen causing the asthma in this case and apply a blanketed template of tests. However, what we did was to enable the patient to log her daily activities to such a precision that she noticed that her symptoms are worse when in the kitchen but not as much when she is in her bedroom or in the living room. Clearly something was going on there! A few more days of logging enabled us to identify that dust from nearby construction sites was entering her home through her kitchen window and causing the allergy. When she was in her bedroom which is away from this dust, she had minimal symptoms. We asked her to buy an air purifier and voila her symptoms vanished overnight! Do you see the precision and accuracy of this form of healthcare here? No life long inhalers or steroid use (and with it a million adverse effects to be dealt with!) Take away - Doctors learned that Asthma can be situational (and not necessarily occupational) and the patient learned what is influencing her outcomes. She also went on to write to the local politician in asking for measures to tackle this problem. This is the absolute beauty of democratizing knowledge to patients (people)! They become self-governing and eventually demand better politics and shape better socieites! What is needed for this ? Patients are currently burdened by data collection. We are pitching for a sensor to ease their burden a bit. While text based logging allows for spontaneity, logging their food plates and daily physical activity is a pain in the neck! As a first step - we are hoping to devise a sensor/app which, when pointed at a foodplate can analyse the foodplate and tell us the amount of calories it has, the macro and micronutrients in that food plate. We are currently burdened with nearly a million images of foodplates from 5200 patients (all on WhatsApp! Sorry Meta!). We study those foodplates and eventually marry those foodplates to their outcomes. While patients are dilligently posting their foodplates (among other things) and are able to understand how their diets are influencing their outcomes (literally in the flesh!), the burden of analysis is on the volunteers (me and my small team). As a first step - an app or a sensor which can scan a plate and list out the contents and the micro and macronutrients would be awesome and this is what I'm looking for and wanting to collaborate on. After this we are hoping to achieve higher order data logging - for which I'll leave a badly handwritten document for you to read up. I've rambled on too long now. But thanks everyone for allowing me to post and to pitch for my idea. Looking forward to interact further. P.S. - Here's what one patient can teach you - 3 Year Old Child, Type 1Diabetes, Insulin Hypoglycemia, Intermittent Bloating1 Month - Look at the amount of data generated from a 3y old child with Type 1 Diabetes. Every doctor should be learning with their patients! Aditya Samitinjay PaJR_250209_195644.pdf