Doctors have always looked at artificial intelligence with some skepticism. Most believe that AI and machine learning technologies are overblown and cannot solve real-life clinical problems.
Doctors tend to dislike their decision-making machines. They prefer to rely on their clinical acumen and judgment to make clinical diagnoses and decisions.
But in today’s changing care delivery landscape and consumers demanding better care engagement and experiences, doctors are rethinking how they can improve service delivery. take care of.
Empower decision making, improve health outcomes kết
The question has never been about AI versus physician clinical decision-making. As Atul Gawande said in his best-selling book, Complications, “No matter what measures are taken, doctors will sometimes falter and be unreasonable in asking us to achieve perfection. . It makes sense to ask us to never stop aiming for it.”
Gawande, in this book, provides real-life anecdotes of the mistakes surgeons and doctors have made.
When supported by a layer of AI and ML-enabled assistants, sifting through historical data and extracting similarities and relevant insights about a case, decision-making can go a long way. goes a long way toward accelerating diagnostic and decision-making processes.
Consider the case of a diagnosis of sepsis. AI algorithms are widely used in critical care units to diagnose sepsis. A sepsis-sniffing algorithm alerts doctors at least 3 to 4 hours before an escalating event leads to severe sepsis. This can reduce mortality.
Initial readings are given by an algorithm operating in the background that collects all the data generated from the patient’s bed, lab, and generates intermittent results to alert the physician to a medical condition. impending crisis.
Hospitals have seen a average 39.5% reduction in mortality on admission, 32.3% reduction in length of hospital stay and 22.7% reduction in 30-day follow-up visits for patients hospitalized with sepsis.
AI is one of the levers that can be used as a second opinion to make diagnoses in complex cases and towards perfection.
Patient engagement and care experience
In today’s scenario of integrated virtual care with on-site care, AI-powered doctors can delegate routine and routine tasks like sending educational materials, ordering prescriptions, and paying Answer patient queries with the help of AI algorithms.
In larger facilities, by using AI-powered tools like symptom checkers to triage patients, doctors can further optimize the functionality of their clinic or department. The use of AI-powered chatbots to answer common questions, book appointments are other uses of AI to improve patient experience.
AI algorithms are helping identify patients with chronic diseases, sending them medication reminders, educational materials, and alerting doctors to any changes in their test tubes or labs when used connected devices. Overall, it leads to patients being more engaged and responsible for their health.
Choose use cases where AI can be successfully deployed
It is important to select use cases where AI algorithms can make a measurable impact in clinical areas. Some of the areas where AI has been successfully deployed are radiology, internal medicine, neurology, and cardiology.
In all of these areas, algorithms work silently in the background and help doctors make a difference, sometimes by providing a second opinion or simply warning of any What crisis is coming? Nowhere is the presence of a doctor overshadowing the AI.
Patients always love hearing their diagnosis from their doctor. In images, today, AI models are helping to automate the contouring of healthy tissue and organs from tumors, develop dosages and adaptive treatment plans for radiation therapy, diagnose cancer at an early stage. In the early stages, diagnose large vessel occlusion in stroke and identify disease patterns for imaging. This is then reviewed by physicians and radiologists who know the patient’s overall clinical, social, and psychological picture.
Machine learning has algorithmic bias and will always be added to the referral line or disclaimer: “clinically necessary correlation”. However, AI replaced by clinical intervention from a specialist who is aware of the human aspects of the patient is a good solution for an integrated human-machine model of care.
Many other use cases have been implemented or are being developed to help with bedside diagnosis. In recent times, technologies such as natural language programming to read unstructured information in doctor’s notes and voice-assisted assistants to predict emotional and behavioral characteristics are being explored. research.
Artificial intelligence has made remarkable strides in the administrative and operational areas of the healthcare industry and is making a measurable mark in the revenue growth of major health systems.
But AI has also encountered a series of failures in clinical areas, leading to a lack of real-world machine learning algorithms in mainstream clinical practice. IBM Watson’s failure to diagnose and treat cancer, and Google’s failure to detect diabetic retinopathy using deep learning models from images of patients’ eyes are recent examples.
So far, the potential of AI in healthcare has not been realized. There are limited reports of clinical benefits and costs arising from the use of real-world AI algorithms in clinical practice.
Although slow, AI in clinical fields is slowly evolving, but the promise of making a difference at the point of care needs to be fulfilled.
As healthcare systems and hospitals digitally transform to improve care delivery and patient experience, doctors cannot be left behind. They must also change and contribute to making this transition a more positive experience for themselves and their patients.
Dr. Joyoti Goswami is a Lead Consultant at Damo Consulting.