RamaOnHealthcare September 29, 2021

We talk of data as the foundation of all things healthcare, yet the debate continues between the value and use of consumer data and institutional data. No surprise that we are challenged with the extent privacy is compromised for customer experience. The regulatory framework has allowed and demanded, at times for data silos to open up. How can data play in healthcare and what are the opportunities for healthcare professionals to deliver value to consumers?

Dr. Inderpal Bhandari, Global Chief Data Officer at IBM is my interview guest today. He brings knowledge and perspective few can match as his experience is only challenged by his wisdom. In addition to IBM, over the course of his illustrious career he has also held leadership data positions at Cambia Health Solutions and Express Scripts/Medco Health Solutions, among others.

Mohan Nair, CEO of Emerge Inc.

Mohan Nair, CEO of Emerge Inc.

Inderpal Bhandari, Global Chief Data Officer for IBM

Inderpal Bhandari, Global Chief Data Officer for IBM

Welcome Inderpal to “A Mohan Nair Interview”

You have a lifetime of experience in corporate data. This interview is hopefully focused on three aspects of your journey. One, your own transformative thoughts about healthcare with a consumer lens. Two, how your global platform at IBM has and will impact or elevate our world around us and three, your personal passions that survive all things.

Mohan: Healthcare. Some say the more challenging complex combination of regulations, human condition, and societal challenges. What is your view of how data engages to raise our healthcare industry?

Inderpal: Healthcare organizations are tapping into AI to drive faster patient outcomes, organizational efficiencies while increasing both life-enhancing discoveries and life-saving predictions.

Although the use of artificial intelligence continues to grow across industries, including healthcare, a major obstacle to the widespread deployment of AI is a lack of trust. According to Morning Consult, more than half of companies cite significant barriers in getting there, including lack of skills, inflexible governance tools, biased data, and more.

Although the use of artificial intelligence continues to grow across industries, including healthcare, a major obstacle to the widespread deployment of AI is a lack of trust.

Trusted data, transparent modeling systems, and robust business processes are important to AI—especially in healthcare. To build trust in new AI-driven technologies, we must start with ensuring the right data and AI foundations are in place. Building trust begins with governance to ensure that data and AI can be trusted. Data must be accurate, accessible, governed, secure, privacy respected, and relevant. AI models need to be transparent, explainable, robust, fair, and privacy respected. Organizations recognize that it takes a holistic approach to manage and govern their AI solutions across the entire AI lifecycle.

Mohan: If I carry that thought to a family of four sitting at the dining table, how would you describe implications to them? Please note that I pick a family of four only as an example.

Inderpal: The vast amounts of data generated by the healthcare industry, related to patients, members, and citizens can be invaluable to meet individual and population health needs. Unlocking the value of data is one of the greater opportunities of our time – leading to discovering new medicines, better health for the patients, better health outcomes for the population as a whole and reduced costs. It is important to understand that AI does not replace doctors. AI generates ideas and presents options, but the eventual judgment and decision must be between a doctor and a patient, between two human beings. That will not change.

Unlocking the value of data is one of the greater opportunities of our time – leading to discovering new medicines, better health for the patients, better health outcomes for the population as a whole and reduced costs.

The healthcare industries generate and collect a lot of data. Much of it is inaccessible, unorganized, and hence untapped. The right data and AI foundation can help address this situation to generate meaningful insights and lead to actionable results.

The healthcare industries generate and collect a lot of data. Much of it is inaccessible, unorganized, and hence untapped.

Mohan: Please provide us a perspective on how we move an industry to deliver that value offered to that family through the complex ecosystem in healthcare?

Inderpal: One of the goals for implementing AI in healthcare is improving the customer (or patient) experience. Once the healthcare ecosystem has access to patient data in a safe, secure, and authorized way, the benefits will be tremendous. It improves the day-to-day life of healthcare practitioners, letting them spend more time looking after patients. Having the patient data available to hospitals, doctors, and pharmacies, for example, can improve the speed and accuracy in use of diagnostics and treatments, give practitioners faster and easier access to more knowledge, patient history, and enable remote monitoring and patient empowerment through self-care. From a routine visit to the doctor to an emergency entry in the hospital, the readiness to assist the patient will be brought to the next level.

Most importantly, the patient should understand precisely how their data is being leveraged. Transparency around what data is collected, how it will be used and stored, and who has access to it is critical to establishing trust. If we are to use AI to help make important decisions, it must be transparent and explainable.

Most importantly, the patient should understand precisely how their data is being leveraged. Transparency around what data is collected, how it will be used and stored, and who has access to it is critical to establishing trust.

Mohan: Healthcare is not one monolith. It is diverse in construct and even more diverse in customers. Hospitals can serve rural, national, or regional, for profit or not and so on. Clinics or mega-health centers. Perhaps service providers of technology? How do you view this varied ecosystem as you formulate and execute a data framework?

Inderpal: The key is to provide access to the right data just in time, regardless of where the data is stored. In the past, organizations have attempted to address data access problems either through point-to-point integration or introduction of data hubs. Neither of those are suitable when data is highly distributed and siloed.

The data fabric is an emerging architecture that aims to address the data challenges arising out of a hybrid data landscape. Its fundamental idea is to strike a balance between decentralization and globalization by acting as the virtual connective tissue between data endpoints.

With data fabric, you can hyper-automate data discovery, data governance, and data consumption in a hybrid and multi-cloud data landscape.

Mohan: As the global citizen that you are, what lessons can the US healthcare Institutions learn from this experience?

Inderpal: Trust is the major currency in any industry, in any country. Regulators need to ensure and trust that healthcare organizations and life science companies remain compliant with exponential technologies. Health organizations need to trust the infusions of AI into core processes, as it creates higher accountability. Citizens, patients, and members need to trust their healthcare providers and payers to work ethically and fairly, while protecting and accessing their data. As the healthcare industry grows increasingly digital, understanding the strengths and the weaknesses of AI predictions, as well as managing modeling risks to make processes robust, transparent, and fair, are essential requirements to consolidate and increase lifetime client value.

Health organizations need to trust the infusions of AI into core processes, as it creates higher accountability. Citizens, patients, and members need to trust their healthcare providers and payers to work ethically and fairly, while protecting and accessing their data.

Mohan: AI/ML is partly about learning from past data to predict and serve the future. But we now realize that in several cases, the data is non-diverse, and sometimes excludes underserved communities. How do you help?

Inderpal: In this case, instrumenting AI for fairness is essential. Properly calibrated, AI can assist humans in making more informed choices, process and evaluate facts faster and better, or allocate resources more fairly — allowing us to break the chain of human biases. Organizations need to be vigilant about rooting out biases in the models – whether intentional or unintentional; ensuring their data is prepped and cleansed properly; and ensuring their models are analyzing complete datasets.

We’ve been working with AI for over a decade, and we initially thought it was all about algorithms. We realize now that it is perhaps even more so about these other attributes that foster trust, like transparency, explain ability, robustness, fairness, and privacy. It was challenging for the average enterprise to deal with that, which caused us to recognize that we were missing a platform that enables enterprise clients to deal with the entire complexity of these attributes. That led us to develop the IBM Cloud Pak for Data, which offers end-to-end data and AI governance capabilities to help enterprises establish trust across the entire AI lifecycle. From data ingestion and cataloging to model development, optimization and monitoring, the unified platform leverages intelligent automation to deliver a complete view of quality data from across the enterprise and increase insight accuracy with improved model development, validation, and bias mitigation. We’ve been using this platform, internally, for 4-5 years and it’s improving and evolving every day.

Tools or recruiting systems that screen candidates have long demanded attention when it comes to issues around fairness. One major U.S. retailer was eager to tackle the problem and turned to IBM for help. It was critical for this employer to embed fairness and trust, including the ability to identify bias and explain decisions, within its AI and ML model used for hiring. Our platform was able to consistently manage AI models for accuracy and fairness and reach the corporation’s goals. Now, the company is proactively monitoring for and mitigating bias in its hiring processes.

Mohan: You are living a very important role at IBM. With it comes much work. How do you keep your focus on the more important and less on the more urgent?

Inderpal: We all face challenges when it comes to prioritization. It is important to isolate the most impactful elements of important tasks. I like to think it through a three-dimensional approach: mission, value, and readiness:

  • Mission – is it aligned with IBM’s mission and my own mission of leading IBM’s data strategy?
  • Value – is it aligned with IBM’s imperatives, bringing value to our clients and employees?
  • Readiness – how is the readiness of our organization, the data, and the project team?

Mohan: What has the pandemic taught you?

Inderpal: From a business perspective, the value of being able to provide real-time or close to real-time, highly accurate, trusted, reliable, robust data. When the pandemic hit, there was a moment where we recognized that the models in place were not going to work because it was an unprecedented time. We had to pivot our efforts to essentially pump out data as quickly as possible so that people could make the most informed decisions. For example, which suppliers were at risk, which customers would be affected, which location, etc. More than ever before, we need to make decisions quickly which requires timely and trusted data.

From a people perspective, especially with our teams working from home during the pandemic, to really focus on our team’s well-being and balance between work and life. Making sure they are supported, getting the necessary time to address their personal needs, like parenting, homeschooling, elder caring, and taking care of themselves while managing their jobs. The pandemic reminded us of how important interactions are (even though virtual) in keeping us connected and helping us to run the business.

Mohan: Thank you Inderpal.

Healthcare needs your energy and intellectual support. I have witnessed your grace in how to serve others and how you use your vast knowledge and insight for greater good. I have also known you to not settle for the obvious and drive for solutions others can only dream of. I wish you well in this journey of service and transforming the institutional healthcare system to one that consumers understand.

About Inderpal Bhandari

Inderpal Bhandari is the IBM Global Chief Data Officer, where he has leveraged his extensive experience to lead IBM’s data strategy to ensure IBM remains the number one AI and hybrid cloud provider for the enterprise. Under his leadership, the Cognitive Enterprise Blueprint was created a roadmap for Cognitive Enterprise Blueprint was created; a roadmap IBM’s clients on their own transformation journeys. Inderpal brings to IBM more than 20 years of experience in leadership roles at such leading companies as Cambia Health Solutions and Express Scripts/Medco Health Solutions.

About Mohan Nair

Mohan is CEO of Emerge Inc, about all things business transformation. He is a 3-time corporate executive, 3-time emerging business executive, 10-year Innovation Officer and 3-time author. He is Edmund Hillary Fellow for Aotearoa New Zealand and is Medical Innovator in residence for MOVAC Capital, the largest New Zealand venture fund.

 
Topics: Interview / Q&A, Trends
Q&A: Justin Dearborn, CEO, PatientBond
"Democratize Mental Healthcare Globally": Russell Glass, CEO, Headspace Health
"Importance of centering health equity in quality improvement work": Dr. Kedar Mate, President and CEO of the Institute for Healthcare Improvement (IHI)
A “Mohan Nair Interview” with Les McPhearson, President of Dean Health Plan
A “Mohan Nair Interview” with Dr. Susan L. Turney, CEO of Marshfield Clinic Health System