By Kanika Gulati
If Jarvis, Ironman’s artificially intelligent assistant, worked in a ward in a hospital, we might hear it say: “Brother Vinay, Mrs. Sivalingam, the patient in bed 3, is trying to get out of bed but has a high risk of falling.” “Doctor Swati, the labs you ordered for your patient in Ward Unit 2, Bed 4 have come back; he is much improved. The patient is indicating his pain is gone, and his temperature is now normal. Would you like to consider discharging the patient?”
Adequate staffing and the quality of care a patient receives in the hospital will directly affect the outcomes and length of stay of the patient. While certain medical staff: patient ratios are generally considered as standards (e.g. 1 Critical Care Physician: 8 ICU patients1), they are challenging and costly in practice to achieve. Other than the numbers, the training and skill level gaps also make it difficult, if not impossible, to bridge the gap for the entire population.
Even when adequate staffing is available, monitoring patients can be both exhausting and monotonous. Doctors would be much more effective if they concentrated on highly cognitive tasks such as diagnosis and planning of the treatment as well as performing highly skilled interventions and procedures. However, they spend a significant portion of their time reviewing the information available in various systems or paper files, as well as documenting their own observations and actions.
Telemedicine offers a cost-effective approach to improve access to highly qualified medical professionals who can support the team on the ground for more critical or complex cases. Remote monitoring is often part of the solution, where nurses monitor the patient directly through video as well as through data received in the form of vital parameters, lab reports, scans etc. This solution can help to both identify cases of concern which require active intervention as well as guide the team on carrying out the intervention itself.
However, replacing medical staff on the ground with clinicians connected virtually would be less effective. It would also add layers of complexity and duplication that feel like more work to both teams than before. When technology automates repetitive tasks and provides insights to clinicians that they might otherwise miss, the entire system begins to work at a much higher level of efficiency and effectiveness. Healthcare can take a leaf from Aviation, where technology augments highly skilled humans to improve safety and efficiency.
Back in the 1950s and 1960s, doctors were already experimenting with models of tele-medicine to provide access to healthcare, especially Emergency Care, quickly when physically doctors would be far away. Today, we have patients or doctors sending images or documents over their mobile phones for review. But the true power of technology is not only in bridging the gap but also in adding intelligence along the way.
The field of diagnostics has already adopted AI in a big way, with Artificial Intelligence not replacing doctors but augmenting them, thereby increasing their capacity to do more and with better accuracy. One great example of AI augmenting the clinical capacity was shown by Google AI through LYNA, or Lymph Node Assistant in 2017, which was shown to help pathologists by highlighting the areas of concern for pathologists to review and determine the final diagnosis, based on a much higher detection rate than what had been manually recorded1. Not only was the AI based tool shown to be more accurate, it also helped make the laborious work of pathologists easier and halved average slide review time. However, on its own, LYNA can neither make a final diagnosis nor advise the course of treatment.
Using AI as an augmentation tool offers many exciting opportunities in clinical care as well. For example, by pre-selecting and alerting physicians as to the developing conditions or deterioration of patients, AI algorithms can help improve physician response to the most critical patients first. Similarly, based on multiple different criteria, including non-medical ones such as patient preferences or religious beliefs, AI could help personalize medical as well as diet prescriptions.
Smart-ICU solutions reliant on AI based digitization of patient vital parameters as well as lab reports without the need to custom-build software integrations with heterogeneous systems and medical devices found in Indian hospitals. Another AI assistant helps collate available information about the patient to reduce the note writing time of the consultant and allow the consultant to focus on new information about care planning.
As we continue on this journey of digital transformation, there is no doubt that leveraging the power of Artificial Intelligence can help healthcare practitioners respond better and faster to their patients, while taking away laborious and repetitive tasks, thereby leaving them with more time to spend with their patients than with data and documents. With patient safety and good outcomes being the primary objectives, this would augment, and not replace, human intelligence and decision-making.
(The author is Director of Product Management, Cloudphysician. Views expressed are personal and do not reflect the official position or policy of the FinancialExpress.com.)