It is a known fact that the adoption of technology in the health care sector has had a tremendously positive impact on medical processes. Artificial Intelligence-led predictive analytics, especially, is finding increasing applications in healthcare delivery and is largely explored to bring about positive patient outcomes.
Deeksha Senguttuvan, Head of Digital Strategy, Kauvery Hospitals, bets big on predictive analytics in healthcare. Deeksha, in this exclusive interaction with Anusha Ashwin, Consulting Editor, HCITExpert, shares her knowledge on the working of predictive analytics in hospital settings and how the technology plays a major role in tackling today’s pandemic health delivery needs.
Excerpts of the interaction:
Anusha Ashwin: What is your take on the power of predictive analytics in healthcare and how does it bridge gaps in healthcare delivery?
Deeksha Senguttuvan: Predictive analytics in healthcare has a wide array of use cases across the spectrum of care. Healthcare delivery not only includes the clinical aspect of care but the operational and patient experience aspects as well.
From a clinical perspective, predictive analysis could play a major role in clinical decision support, reducing readmissions, avoiding adverse events, chronic disease management, identifying at-risk patient cohorts, etc. The non-clinical scope can include accurate cost prediction based on patient profile, insurance approvals, managing appointments, managing supply chain, and more.
We are at varying levels of extracting meaningful insights for the various use cases highlighted above, and a lot of factors such as availability of good quality data, implementation challenges for actionable insights, etc. play a factor in the extent of impact predictive analytics has on healthcare.
Anusha Ashwin: How can physicians and patients gain from predictive analytics?
Deeksha Senguttuvan: In the current caregiving scenario, the physician’s role is limited to the care provided during his presence and at other times relying on an already stretched nursing team. Predictive analytics helps by constantly working in the backend by monitoring multiple aspects of a patients’ care and recovery journey, and triggering alarms and indicators to enable immediate action by the caregiving team, and plugging any gaps they might miss.
To explain using an example: on-set of sepsis infection while admitted at the hospital is one of the causes of deterioration in patient condition, causing patients to be shifted from the ward to ICU. We have algorithms now available to predict the onset of sepsis by monitoring a patient’s vitals, which gives a warning regarding the possibility of sepsis infection before any symptoms are visible. This enables the physician to immediately act to suppress the infection, and the early intervention leads to a better chance of recovery for the patient.
Another example: during post-discharge recovery of a patient, predictive analysis can help generate a risk score of a patient by integrating their demographic data to the symptoms they exhibit. This risk scoring helps the patient to proactively get a specialist’s consultation leading to lesser readmissions to hospitals and avoiding any post-surgical complications.
“Having a robust EMR system at the hospital would help in
enabling more use cases for prediction-based diagnosis”
Anusha Ashwin: What tools can help capture data that lead to increased prediction-based diagnosis?
Deeksha Senguttuvan: Prediction-based diagnosis is a vast area, with each type of use case requiring its own set of tools and processes to enable a successful implementation. Taking the same two use cases shown for the above question – sepsis prediction requires a low-cost continuous vitals monitoring device, whereas risk profiling post-discharge involves active input from the patient on the signs and symptoms being seen.
However, having a robust EMR system at the hospital would help in enabling more use cases for prediction-based diagnosis. Data captured in EMR can help in designing clinical triggers that help in reducing adverse events and highlight to physicians any potential complications that might arise.
Anusha Ashwin: You had mentioned that predictive analytics can play an important role in avoiding hospital readmission for patients. How is this possible and how do you think scaling up this data science for hospitals will bring mutual benefit to patients and physicians?
Deeksha Senguttuvan: I believe predictive analytics will play a great role in avoiding readmissions, however, the problem of capturing data needs to be sorted to deploy this at scale. The key to avoiding readmission is to keep a constant track of patients’ symptoms and vitals. One aspect of this equation has been addressed in recent times with the ease of availability of vitals monitoring devices that can be deployed at home. However, other symptoms also need to be factored in and a care manager from the hospital needs to follow up with the patient regularly to track their recovery journey.
This becomes a highly labor-intensive process and current solutions are being developed to enable patients to automatically capture their symptoms through mobile applications. But that still doesn’t help achieve scale in a market like India, with the complications of low internet adoption and requirement of vernacular interaction.
The mutual benefit would be obvious – reducing readmission helps reduce costs for patients, and helps improve quality of care for physicians (which might also help negotiate lower insurance premiums for large healthcare providers).
Anusha Ashwin: Predictive analytics can also have a large impact on home healthcare services and telemedicine. What are your views on this?
Deeksha Senguttuvan: Predictive analytics in teleconsultation can be used in clinical decision support by capturing vitals, signs, and symptoms from the patient and triaging them before being presented to a physician for consultation. The premise is that this enables to reduce the time for a physician in basic data collection and allows them to focus more on the patient interaction.
Predictive analytics might have a larger role to play in-home healthcare services due to the increase in the availability of home monitoring devices. However, I think predictive analysis will play a larger role in wellness management through the use of wearables, which are more consumer-oriented rather than provider-oriented. From a healthcare provider perspective, predictive analytics I think has a larger role to play within the hospital, provided the necessary infrastructure and processes are set up in place.
*About Deeksha Senguttuvan: In her current role at Kauvery, Deeksha leads technology adoption in the hospital/healthcare ecosystem, thereby improving access to affordable healthcare.
Right after her PGDM from the Indian Institute of Management, Kozhikode, Deeksha joined the prestigious TAS (Tata Administrative Services) program. In her 5-year stint, she gained varied perspectives across multiple industries (telecom, healthcare, e-commerce, utility services, manufacturing & construction) and domains (B2B Marketing, Product Management, Business Development, Project Management & Strategy). Outside of work, she is an avid Bharatanatyam dancer, who has won many state-level accolades including the coveted ‘Yuva Kala Barathi’.