Mudit Garg, MSEE, MBA is co-founder and CEO of Qventus of Mountain View, CA.
Tell me about yourself and the company.
Qventus is an AI platform. We work with hospitals and health systems to help them manage their day-to-day operations. I’m one of the founders and CEO of the company. My background for the last 10-11 years has been in healthcare operations, specifically in lean process improvement. I’m proud of that. I started a few technology companies and spent time at McKinsey & Company’s healthcare practice. I’m an engineer by background.
How can data and dashboards be made useful to frontline people as they are making operational decisions?
That was one of the biggest prompts to start the company. The first time I walked into a hospital, I was struck by two things. One, how the managers and caregivers did whatever it took to provide care. Second, maybe because they cared so much and because the system was dependent on them doing these heroic acts day after day, the system itself never developed. My perhaps biased view in the beginning was that data could help folks be more prepared and to require fewer of these heroic acts.
We are comfortable in the conference room thinking all we want about dashboards and information, but in the moment when people are busy, nobody has any time to do something about it. Nobody logs into a dashboard. Nobody has time to read through a graph or a report and understand it. That was the earliest insight into the way of using data that we learned from.
We started talking a lot about what was needed in 2013. We said, what makes it so hard in healthcare operations? Typically the answer came back as, half my patients are unscheduled, we don’t know how long they will stay in the hospital, and the resources they will need is unknown. Predicting would be great.
It’s an often-used buzzword, but we started using machine learning tools back in 2013. My co-founder and I both had a background in it. We started predicting.
But we learned that predictions by themselves are sometimes counterproductive. While an average manager doesn’t have time to stare at a dashboard, they also don’t have time to interpret a prediction. A nurse we worked with at that time said, I don’t have time to figure out 30 percent chance do this, 40 percent chance do that. If my GPS said it’s a 30 percent chance to take a left, 40 percent chance take a right, I would toss it out the window. I have more load than I do while I’m driving. Just make it simple.
The goal of the product is not to expose more data, but to take those things that a really good manager would do. A really good manager in an emergency department anticipates. They say, things are getting really bad, I had better have my lab manager start doing X or start prioritizing these things. I had better tell the house supervisor to prioritize some beds. By doing those things two, three, or four hours in advance, they can get ahead of the situation. But that only happens when they have a calm environment where they have the time and capacity to look ahead and solve those problems.
Our product’s goal is to take away that mental load — the data processing, the evaluation of options — and to offer a suggestion in the moment as a message, discussion, or into the workflow in some way.
Hospitals usually have some internal expert they call in when they have a problem, but are lost when that person isn’t available. It would seem that once a hospital has formalized the decision-making process, it would be easier to then enhance it.
Absolutely. An excellent manager has to look at data and make sense of it. That depends on that manager’s time. What judgment they apply depends on that manager’s experience. All those things create inconsistencies.
But in that ED example I gave, the system would be saying, it’s Monday after Thanksgiving. The patients in the waiting room are much sicker. Dr. Smith is working and he tends to he tends to order more labs, but our lab is really slow right now. Based on all of this, we will run out of capacity in the next three hours.
Then the hospital can connect those subject matter experts. Gather the lab manager, house supervisor, and charge nurse and say, “Here is something that we see. We suggest you do this.” Let them have a workforce huddle on that discussion topic and do something about it well before the problem becomes bad.
Who would typically serve as the internal champion of that kind of real-time monitoring?
The executive sponsor often ends up being someone like a chief operating officer or a chief nursing officer. But the internal champion often comes from the lean groups in the hospital. They are the ones who have seen the day-to-day problems, are trying to improve them, are trying to build a system around them, and are connected enough to the day-to-day problems. They can be good champions. Oftentimes department heads will see these challenges, such as the medical or nursing director of the ED.
Those are the internal champions who want this to become a part of the system. The executive sponsors typically are the chief operating officer or the chief nursing officer, who are day-to-day focused on these problems and who jump in to help when things don’t go well.
What is the physical and operational manifestation of how your product gets used in a average hospital?
The ideal end state of the product is that there is no physical manifestation. The ideal end state is that it is invisible, like a really good assistant or someone who is helping you have the insight. It just disappears into the background and brings in the right information at the right time. That’s why it is like virtual air traffic control.
The product has three parts. The most important one brings the insight into the moment. It tells you, this patient in room 434 is likely to get admitted. We don’t have an admit order. We probably won’t have one for the next three hours, but let’s start preparing the bed. Or, this patient is likely to leave without being seen, or that we’re going to have a bad situation with this patient. It’s processing these insights in the background and delivering them in the moment — on a Vocera device, on a secure messaging device, or whatever the right mechanism might be.
Our system provides situational awareness, a sort of mission control. It can be in the break rooms or the huddle rooms, where people can have meaningful information displayed to help them understand the situation. Some of these nudges can be shown at that same place.
The last part is being able to understand the data to see where changes need to be made. An average department will get insight in the moment when they need to do something.
As hospitals centralize and and have larger deployments, there is an interesting role to play for a centralized place. In General Stanley A. McChrystal’s book “Team of Teams,” he talks about how the traditional image of command-and-control came to fail. The military started it, but in the most recent war, we struggled with that approach. They had to rebuild it and dismantle the command-and-control approach. He talks about the importance of spreading shared consciousness throughout the frontline people who are experiencing the situation and who have the most knowledge in the moment.
Our job is to spread the context, consciousness, and best knowledge to the people in the moment who are about to make that decision. While there’s a role to play for the central manifestation in escalation and awareness, the ideal situation is one where the information and the shared consciousness is going to the front lines. That’s how our product works.
Your site allows looking up any hospital’s efficiency index as calculated from publicly available information. What metrics might improve in using your system?
Our product is in 60 or 65 hospitals. Patient flow is a big use case — in the ED, inpatient, and OR. Length of stay, as you can imagine, is a really important metric, because it’s one of the most important measures of affordability and survivability for an organization to be profitable on Medicare patients. Length of stay is a big one on the inpatient side.
Length of stay is important in the ED, but so is patient satisfaction. The number of patients who are leaving without being seen is important.
On the operating room side, they look at efficiency — how much time it takes to turn a room, how many of the rooms are being used, whether supplies are being used appropriately, and how well patients are being informed throughout.
Then we have use cases for pharmacy and outpatient clinic access. In pharmacy, how to manage the drug spend. In outpatient access, how can the health system, with the resources it has, provide patients with quick access to care?
These metrics are beneficial regardless of the payment mechanism or the healthcare system’s economic model. As an example, one hospital freed up about a million minutes of patient wait time in their ED when they deployed the system. That helps them provide care to more patients in the community with the same resources. That lowers the cost, helps the hospital, and helps the patients. Regardless of the economic model, it helps both the health system and the patient.
Where do you see the company’s future being?
I grew up in India. We have in the US healthcare system the best clinicians, some of the best equipment, some of the best therapies. What’s holding back the potential of our system is oftentimes is the ability to execute on the processes day-to-day consistently and reliably, without placing an excessive burden on the people who provide it. If we can do that, if we can create a mechanism where it doesn’t take the heroic effort to provide that consistency and reliability, we can do that across every aspect of delivery of care. Whether it’s your experience in the unit, how well informed you are, your billing, or your staffing. Whether its in the emergency centers, in urgent cares, or in outpatient clinics.
My hope is that we can provide the infrastructure to allow for consistent, reliable execution of the clinical practices we know. Managing the logistics around delivery of care so that the human connection, and the calm that we can provide to people while delivering the care, is feasible. That’s my hope. I’m hopeful that we’ll be able to play a meaningful role in bringing that about.
Comments Off on HIStalk Interviews Mudit Garg, CEO, Qventus
I had an old physician colleague whose favorite hobby was bitching about EHRs, and one day told a story about…