Dr Atish Laddad, Pediatrician and Founder Member, The Pediatric Network, gives an insight on the need to adopt Machine Learning which can become another dimension for doctors to upgrade their existing skills
Machine Learning (ML) is a fancy terminology for doctors. Most doctors think that it is for the tech nerds and data scientists hell-bent on taking a doctor’s data, analysing it and building something which is more powerful than a doctor. ML is nothing but building smart algorithms and workflows inside a tech application for a specific purpose. In our platform, we have used this tool smartly to generate prescriptions faster than pen and paper.
The biggest and smartest machine is the human brain. Technology is there to assist the human brain. No amount of technology can replace the personal touch or clinical bedside skills which a doctor offers. Even techies who are trying to replace doctors would invariably get a ‘reality check’ over a period of time.
So, how does the emerging zone of paediatrics make use of ML to assist doctors to give prescriptions? The process of ML typically involves a code built into it to study how a doctor approaches a particular ailment and suggests drugs, lab tests, vaccination and preventive measures based on doctor’s previous behaviour. This makes a doctor generate a prescription in record time. The machine learns the dosing of a particular drug making it less prone to human errors. For example, if a doctor prescribes Drug X in a twice-daily manner commonly, the system will do it for him. Or the system will show which illnesses are trending in the patient community around his locality based on previous data.
Doctors still prefer the handwritten way. Why? Foremost, because they cannot afford to spend more time on the machine then on the patient. Secondly, generating a printed prescription is not the draw. Doctors are yet to get convinced that technology can actually ease their workflow and help in patient care. Last but not the least, India needs a tighter and more stringent healthcare data norms. Lack of it or unawareness of the same makes the healthcare provider nervous. Till then, pen and paper way continues to dominate the scenario.
But, if doctors adopt technology it would add another dimension to their existing skills. A doctor works on three dimensions to come to a diagnosis: Medical history, clinical examination and third, laboratory tests. Data analytics will add the fourth dimension to their armoury which will be as powerful as the other three. For example, data will be able to predict what illnesses are trending presently in that part of the country. Data will be able to help them treat chronic ailments like asthma, thalassemia, etc. Data will help them to predict therapeutic responses and counsel patients accordingly. If done right, it will drastically get the cost of healthcare down as data analytics will supersede even lab and radiology tests to predict diagnosis and outcomes.
ML is presently the talk of the town. Most technologists are using ML in imagery data like radiology and pathology. The day is not far away where machine will help a semi-skilled person or a bedside nurse diagnose a heart defect or an ectopic pregnancy reliably and accurately. Wherever data can be standardised or reproduced, ML will have a huge impact. If ML can be used to make doctor adopt e-prescriptions, it will make a huge impact on doctor’s practice in time to come.
It’s essential and fundamental for healthcare to shift from thinking of ML to be an innovative concept to seeing it as a practical tool that can be deployed today. If ML is to have a role in healthcare, then it must be built in such a way, that it is a collaborative partner for an expert that will help recognise specific areas of focus, illuminate noise, and facilitates to lay emphasis on high possibility areas of concern.
Possible benefits of ML in times to come:
1. Prediction of epidemic outbreaks
2. Radiology: Read images and based on previous data suggest diagnosis to radiologist
3. Pathology: Read pathology slides and based on previous data suggest diagnosis to pathologist
4. Clinical trials and research
5. Precision medicine
6. Disease diagnosis and outcome predictions
By utilising ML to its maximum capabilities, there is a tremendous scope of development and major modifications can be brought about in the child healthcare field. If technology is to advance care in the upcoming years, then electronic data and material provided to doctors should be boosted by the power of analytics and ML. This is just the beginning, as these technologies mature, novel and better-quality treatments and diagnoses will save more lives and treat more viruses. The future of medicine is indisputably based on ML.