In this episode, we are joined by Phillip Tseng, author of the study “Administrative Cost Associated with Physician Billing & Insurance-Related Activities at an Academic Healthcare System” to discuss his study’s findings and their correlation to EHRs
Follow Phil on Twitter at @_PhilMD
Highlights of this episode include:
- Overview of Phillip Tseng’s study and the methods used for his research
- Review of the five different types of patient encounters that were common at the studied health system
- Analysis of EHRs performance in the study
- What are the future implications for health systems?
Mike Passanante: Hi, this is Mike Passanante. And welcome back to the Hospital Finance Podcast.
EHRs have long held the promise of reducing billing cost and increasing efficiencies within health systems. But that may not necessarily be what is happening. To shed some light on this topic, I’m joined today by Phil Tseng, an investigator on a study published in the Journal of the American Medical Association that looked at this correlation.
Now, Phil just graduated a few days ago from Duke University with both MD and MBA degrees. He has a special interest in helping to bridge the gap between the world’s medicine in business which had been, at least historically, rather divided. Phil believes that the bottleneck of healthcare in the coming decades will be in not the science of medicine, but in the financial sustainability and equity of care.
His decision to pursue the MBA is largely informed by his previous experience serving in the Teach for America Corps in Phoenix, Arizona.
Phil will soon be practicing emergency medicine physician and hopes to spend increasing amounts of time in health system administration as he moves forward in his career.
Phil, welcome to the show!
Phil Tseng: Thanks, Mike. It’s great to be here.
Mike: Phil, the study we’ll be discussing today is entitled Administrative Cost Associated with Physician Billing & Insurance-Related Activities at an Academic Healthcare System.
Can you explain why the study was initiated and what the goals were?
Phil: Absolutely, Mike. I’m very happy to. So as your listeners are undoubtedly aware, the cost of healthcare in the US has risen dramatically over the past several decades from about 5% GDP in the 1960s to double that, 10%, in the 1980s which, at the time, everyone thought was crazy, but now we’ve managed to double that figure yet again with the latest figures showing that we spend a fifth of our economy in healthcare last year in 2017.
We’re a clear outlier when compared with the rest of the globe. America is definitely number one in terms of per capita health expenditures both in dollar amounts as well as in proportion of GDP. And of course, there have also been plenty of other studies—perhaps most notably one released by the Commonwealth Fund — sharing that we don’t exactly get what we pay for in terms of patient value in healthcare and that we’re frankly laggards among multiple developed nations across multiple factors, not the least of which include efficiency, equity and cost.
So, why did we look at administrative cost? Well, it turns out that only about two-thirds to three-quarters of every dollar spent in healthcare actually goes toward the clinical care of the patient that is received in any tangible way by a patient. And at the end of the day, we’re all patients. In some way, shape or form, we all have a vested interest on optimizing this metric, I would argue.
The remaining fraction really contributes no material value to what a patient sees in her/his care, cost spends on administration. And if you look at how these administrative costs are broken down—there was a great paper written by Jiwani, Himmelstein, Woolhandler and Kahn a few years ago. Two-thirds of these administrative costs, the bulk of these costs, are attributable to billing or insurance-related activities or what came to be known as BIR costs in the academic literature for short.
So, given that these administrative costs constitutes such a large part of US healthcare costs, my team wanted to take a deeper dive and look at what makes these billing or these BIR costs so expensive and whether we could trace these costs to any specific activities that are performed within the health systems that generate these costs.
So, to date, including the Jiwani study that I just cited, most studies have looked at these costs through top-down aggregate approaches. We sought to see if we could arrive at the same results through bottom-up micro-costing approach, and in so doing, also shed more light on these specific activities that are generating and contributing so much costs to the overall cost of care in the States.
Mike: That’s a great set-up, Phil. Would you mind going over your methods briefly for the audience?
Phil: I’m happy to. It was essentially a two-step process though I have to say, in practice, it was much uglier and more complicated. Our core methodology is called time-driven activity-based costing or TDABC for short. It’s a variant of ABC or activity-based costing in the accounting world, and it was developed by Bob Kaplan at the Harvard Business School years ago and had been lauded by the academics and practitioners alike.
TDABC is essentially the marriage of two techniques: the first of which is process mapping from industrial engineering; and the second of which is ABC or activity-based costing from accounting.
I mentioned it was a 2-step process. The first step is to develop that full process map for what we started calling “the life of a bill,” meaning a full list of activities that needed to happen for an insurance claim to be paid for from the health system’s perspective, starting all the way from the very beginning at appointment scheduling to the very end when payment is finally posted to the health system ledger.
Once we had this life of a bill and it was effectively an activity flow sheet, we moved on to step two in which we wanted to know five things about each of those activities. What was the activity that was being done? Who’s doing that activity? How much is this person paid? How much time does it take this person for each instance of the activity? And what economic resources are consumed by this person within the organization perhaps indirectly outside of their wages and benefits?
The consolidation of all this information yielded the results of our study.
Mike: And let’s go through those results. Could you walk through those for us?
Phil: Absolutely! So, we looked at five different types of patient encounters that we saw as common within our health system—primary care visits, ER visits, general inpatient stays, ambulatory surgeries, and inpatient surgeries.
For each different type of encounter, we followed the activities that needed to be done in order to generate a bill and we calculated costs for each of those activities. Our final billing costs range from $20 for a primary care visit, which represents 15% of average revenue within the health system we studied, to $215 for an inpatient surgical procedure. These costs are solely attributable to billing and insurance-related costs or paper-pushing costs, if you will, and are entirely in excess of the cost of actual patient care.
From the primary care standpoint, this cost represents $100,000 per primary care physician per year which corroborates with previous studies published years ago even before large-scale EHR implementation. So it’s no wonder that private PCP practices have found it increasingly financial unsustainable to operate independently over the past several years when they have these costs to worry about on top of massive software licensing costs with EHR implementation—
Not even to mention we conducted this study in a very large health system that enjoys some significant economies of scale. Imagine what a tiny 3-physician practice would need to spend in order to avoid denials.
Mike: Indeed! So I just wanted to drill in on one point with you. So the EHR adoption did not seem to reduce billing cost. I think that was really the bottom-line. But your team did not place the blame on EHRs, per se, but rather attributed the issue to other factors. Can you explain that for us?
Phil: That’s a great question, Mike. I mean there’s no doubt that EHRs can be better. It’s no secret, for example, that the largest and most fully featured EHR system relies on computing technology that originates from the 1960s.
The core problem we saw though as a result of our study was that there was too much variation in the billing process for it to be fully automated. Health systems manage hundreds of different health plans, each with different billing requirements and nuances of what is covered and what is not.
These requirements change multiple times per week… literally. And they’re frankly a part of a current business model of payers. It’s a way for them to avoid having to pay for these astronomical healthcare prices which is a whole other story in and of itself.
EHRs are not currently equipped to handle these variations. And even specialized Clearinghouse Bolt-On Software that health systems most often use are limited in what they can do before it requires human input.
So there’s also no doubt that a somewhat significant addition of AI into these workflows could possibly improve the cost profile of the billing activities we saw. Though of course that I’d say that guardingly, the cynical part of me thinks that this AI would just be priced marginally less expensive than current workflows. I could be wrong.
The second caveat would be that this change would likely be met by equally effective responses from the payer’s side, taking a page out of game theory.
So anyway, AI might help.
The fact remains that there is today too much variation—and the emphasis is on variation—that necessitates manual human input. And this variation doesn’t exist in countries that are on a single payer system. It’s also minimized in countries where health plans are much more standardized.
Switzerland, for example—which my mentor loves to talk about—like the US, is a private multi-payer system, but it has overall administrative costs from the order of 4%. And BIR costs are only a fraction embedded within that 4% of those already low costs.
I’m kind of giving a very long-winded response here. But to borrow the words and paraphrase Vivian Lee who is one of the authors on the editorial on our paper, the political viability of a single payer system where none of this variation would be there, in this country, I think currently is very questionable. What we propose in our paper is a middle ground approach, though it would also be likely met with significant resistance. It would be to have a set of standardized contracts to which health plans would need to align which would reduce this type of variation, thereby decreasing the paper-pushing costs that we see today.
Mike: And to that end, what do you think the implications will be for health systems going forward?
Phil: I think changes are coming one way or another. The fact that health systems need to spend 15% of revenue collecting fees for simple outpatient visits to me is astounding.
By comparison, credit card transaction costs are in the order of 10 times less.
Now, I think it’s an unsustainable status quo right now, and especially so as we’re in an age where we’re pushing for more and more care to be delivered outside the hospital in an outpatient setting. So if we extrapolate our study findings for outpatient encounters, this trend would only be associated with further increased administrative costs. And I think once there’s enough pain—though hopefully we won’t get that far—something big is going to have to happen. Maybe it’ll be AI. Maybe it’ll be a massive policy overhaul.
And this is probably going to sound trite and overused. But health systems need to be ready for coming change—it could mean for thousands of employees that currently support their administrative functions. I mean, what will health systems do to support them through that change? I think that is going to be a very important question for the years to come.
Mike: Phil, thanks for adding the study to this very important discussion that we’re having in the United States around healthcare costs. And thank you for joining us today on the Hospital Finance Podcast.
Phil: Thanks for having me, Mike. It was great to be here.