Brian Robertson, MHSA is founder, chairman, and CEO of VisiQuate of Santa Rosa, CA.
Tell me about yourself and the company.
I’m 30 years in and a data geek at the core. I started helping the provider industry achieve yield improvement from the revenue cycle 30 years ago, initially at a consultant that had a boutique. That turned into a company called MedeAnalytics. It was taking the application of what you learned in consulting walking the halls of the hospital. People were interested in visualizing what was going on in the enterprise. MedeAnalytics was focused on that and still is.
I departed in 2009 and started VisiQuate, fundamentally doing the same thing, although my passion grew from targeted point solutions to a broader data platform for the revenue cycle as opposed to having too many point solutions. We deliver that as a service-enabled technology, because we are doing the data aggregation and processing.
I am trying to the help CFO, the VP of revenue cycle, and their staff do two things. Drive yield improvement, but also those data signals can also tell you where there’s redundancy and where automation — more buzzwords, such as AI, machine learning, RPA, all of those things – create a tremendous opportunity to take waste and process inefficiency out of the revenue cycle. My passion remains. The bots have arrived and we’re helping clients get things done through intelligent bots.
What has changed the most in analytics and technology since you started the company?
There’s a nice tipping point, in my view. Let’s just go back five years. We started our initiative for AI and built a chatbot, essentially Alexa or Siri for the revenue cycle. You could say, “Ana, what’s going on with Medicare bad debt, or what’s going on with Aetna, payer code 1234?” That then evolved to looking at deep data signals on where there is redundancy. Clients would say, “We have to start automating things. We are growing to so many people growing through acquisition. I have 1,200 revenue cycle FTEs, the company is growing, and the CFO wants us to maintain 1,200.”
We started to lean into that. For those years it was pilot projects, proof of concepts, interest. Everybody knew it was something you should do. COVID arrives, we’re all working from home, and you look at getting people at the home. Then some labor shortages, some problems with the sheer volume of accounts that need to be processed every day. The conversation went from “nice to have, we should do that” to “what can we automate?” We are in COVID hangover, but many folks are still at home. Many of our clients tell us, and we see it in the data, that they are having problems. Largely in the back office, where it’s hard to find FTEs that are doing account follow-up call center type activities. The line you hear is, “Target pays more than I can pay the most senior collector.”
We are addressing that shortage by using one lens of insights that is driving yield improvement. Who’s paying and who’s not paying? Where are there under payments? Where are their denials? Those types of things. But now we are training the data signals to also look for redundancy, where any kind of revenue cycle FTE is doing the same thing. Filling out a payer downgrade appeal form, and they do it 15 times in their day. And you say, we have all the data, what if we automate that? Oh yeah, I could work more accounts.
Our approach, instead of pure robotic process automation, is what Gartner and others call intelligent or cognitive process automation. Because we are letting Ana, which is our AI analyst, first go do the discovery, companionate smart people to say, we have a lot of redundancy here, here, and there. They say that qualitatively. Then we say, let’s go look at the data, let’s look for that redundancy, and then let’s do one bot at a time.
We are trying to focus on smart bots, leverage the Pareto principle, get people excited about automation, get them familiar with it, do one and go to the next, and make sure you maintain them. You hear sentiment like “Bots break. Bots are brittle.” Yes, they are. But so is the contract management system, where you have to update tables, profiles, and dictionaries. It just has to be built into the service model.
We are advanced analytics versus BI and reporting. Insights can focus on where to automate. That’s where we are passionate and getting some nice traction.
How does a health system that has revenue cycle opportunities decide whether to bring in outside help, outsource, or invest in technology?
When we are talking to clients, we can walk in and say, “Here’s what we can do, You push the data to us, we’ll take care of all the heavy lifting. You’ll be on the assembly line.” Many clients have already invested in bots and RPA or they are about to, or they’ve got a consultant. We try to be compatible and complimentary in all the things that we do. I hate to use Lego blocks as a metaphor, but I don’t have e better one. Whether it’s APIs or just containers, all the techy stuff, we try to make all of our offerings plug and play. Because half the time it’s fully outsourced to us, and half the time we’re working with a combination of the VP of revenue cycle and CIO and should complement their initiative.
For example, if they have bought or made an investment in UiPath, Automation Anywhere, or tools like that, they have existing licenses. You say, great, let’s leverage them, Our cognitive bot Ana is benefiting from crowdsourced data across many, many clients. That’s the cognitive brain that lets us do that part. For the carpentry of building the bot, if you have programmers and you want to do that, it’s like we’re doing the architecture. We can do carpentry or you can do carpentry. We try to be plug and play friendly, because if you don’t, then you are leaving market opportunity on the table.
How has hospital consolidation into larger health systems impacted the capability for revenue cycle management to scale?
I’m famous for saying that you can end every sentence in healthcare with dot, dot, dot. It depends. We have seen all of the above. Some grow through acquisition, and maybe there’s two-thirds on one platform like Epic. They have robust, capitalized product development dollars. IT shops that have actual software developers, an architect, a true world class DBA are the shops we tend to be complementary to. They have some existing investment. Other shops are resource constrained and are just keeping the lights on in many cases.
People will say things like, they’re an Epic shop, or they’re a Cerner shop. I would say that they have chosen Epic to be their vendor of choice for the HIS and system of record, but they are on Epic, Cerner, Meditech, and Allscripts and they are moving across a five-year journey to centralize Epic. Many times, clients think that we are the bridge. We are still giving a consolidated view of the enterprise, because we take feeds from all those systems and give the CFO, the VP of revenue cycle, and the case manager their dashboards. Everybody gets the intelligence that they need and we normalize that data.
A lot of our advanced analytics is leveraging embedded wisdom across a lot of years. That’s the part we’re always making sure that they take advantage of. I can also sit down with folks if they’re intellectually honest and say, “I know you have invested in licenses and all that, I can show you a TCO calculator, and it would be hard to compete with our benefit of scale. Because we have a massive cloud store, our return on terabyte is going to probably have a benefit of scale that you can’t compete with. We have over 400 hospitals, and we do this every day. Whether it’s the private cloud or it’s running daily ETL, personalizing dashboards — because in healthcare, it’s hard to be a cookie cutter solution, things change too much – we are very malleable in all of our solutions, and to be malleable, you’re supporting them every day.” We tell clients, if you want to do this portion, let’s make sure we’ll consult with you, what’s the maintenance, what’s your maintenance plan. Because these are the hours it takes to keep all the lights on. Data action versus reporting is not for the faint of heart. You have to be a little bit crazy and you have to have done your 10,000 hours.
What do health systems gain from using workforce performance analytics?
One of the most exciting things that I’m passionate about came from a client. The hospital side of the house and the physician side of the house were very different in how they did incentive compensation. They were using tools out of PeopleSoft and traditional systems like Kronos to not only get time and attendance, but try to have quality performance scores. They would take a random sample of transactions. For example, if you were a patient access clerk, there was a threshold of errors. You can’t miss the Social Security number, the subscriber ID, the really important information required to get a claim paid. It was saying that you have to be at this threshold, and if you’re at this threshold, you’re eligible for points in monthly giveaways in the fishbowl, or you’re eligible in some shops for incentive comp.
It started there, and we started getting deeper into the quality scores, because that particular client got us excited about the notion of gamification. Shops that can’t do incentive pay can do point systems and badges, like a Fitbit. People love bragging rights, to go in the lunchroom and announce, “Hey, did you see I got the badge? I improved my patient accounting collections 3% over last week.” The attaboys and attagirls and things that come from gamification really started to move the needle.
We took time and attendance and added measures of quality that people would usually do through an audit. We automated that audit, so instead of looking at 15 accounts, do a score, and see which threshold they are at, we took all their Boolean logic and automated it. They ended up with something that is similar to RVUs. We called it a PVU, Performance Value Unit. It’s multi-variable calculus on, what does their time and attendance record look like? What does their quality score look like? Some people do things like training, so what is their ongoing training and CEUs? It’s more holistic than grading somebody’s paper on pure time and attendance and leveraging the Hawthorne effect in a positive way.
Where do you see the company going in the next few years?
I expect, and we see this in the market, that our revenue mix will shift from 80% insights business and advanced analytics to 80-20 or maybe 60-40, where the 40% will be intelligent process automation. It will be us tackling administrative waste in the revenue cycle in a way that’s compelling and delivers an ROI. Right now we deliver an ROI by improving cash flow, bad debt, and underpayments and the like. I think that because the need is so great, our ROI will now be a combination of analytics and the results of automation, taking out the waste and also upskilling the revenue cycle folks to be directionally headed to being knowledge workers.
I share your concerns about a lot of similar projects, but I think this one's an outcomes comparison thing, not…