Parallels in Factory Mechanics & Healthcare Administration
The U.S. healthcare process in a nutshell:
- Step 1: get care
- Step 2: ?
- Step 3: insurance company pays and/or you get a bill
While all 3 steps are the subject of myriad debates, coffee shop talks and speeches, step 2 is the least understood by many who don’t work in healthcare. Even clinicians with little familiarity of how the “business” side of a health system works find step 2 to be a black box. Let’s demystify this a little. There’s too much ground to cover in one article on how what’s been coined as the “revenue cycle” of healthcare works in its entirety. So, we’ll focus on a few observed parallels between how factory floors work and the nature of working in healthcare administration.
1. After you’re discharged or your outpatient procedure is complete, a claim is created and handed off in an almost linear sequence between a bunch of different teams before it’s sent to the payer.
This is the official “Healthcare Financial Management Association® (HFMA®)” patient-centric revenue cycle infographic:
This graphic shows every step in the healthcare administration lifecycle, but the linear hand-off between each team is evident. There are tons of teams that touch your claim post-”patient care delivery”. Like a factory, each team is highly specialized to perform some specific function on getting your claim paid for the health system. Unlike most modern factories, most of this work is done by human beings interpreting data across an army of systems and annotating either the claim itself or any follow-up work they have done on your “encounter” with the health system.
2. The system incentivizes teams to get as many claims out of their backlog and in the next team’s backlog as possible.
Of the few metrics used to analyze team performance, the dominant ones examine things like how many times a healthcare administrative team member “touches” a claim before it is paid and how many “good” touches they make in a given day. The goal is to try to work as many claims (aka “widgets”) as possible and fill up the next team’s bag of things to “add” to the claim as possible. It’s an assembly line where staff are incentivized and praised for performing their function in the value chain as fast as possible so the next team can hurry up and get their function done with the overarching goal of getting a claim manufactured and paid as fast as possible.
3. Defects are measured by whether or not a claim is paid or not.
As you can imagine, the possibility for introducing a defect in such a human-driven factory assembly line is high. Systems called “claim scrubbers” do exist that run automated rules to check if a claim is in proper order before it is sent to the payer, but payers frequently change their terms of how they pay and it’s a daily struggle for providers to stay on top of such changes. In the end, most providers measure how well their process is running based on whether they get paid or not. Ultimately, this means a ton of teams have to touch a claim before a defect rate can be measured (because the claim has to be sent to the payer and the payer needs to respond), so root cause analysis of where the defect was introduced is very retroactive and delayed. Other management KPIs exist that mitigate this problem, but it’s still really big. $40.6B is expected to be spent on administrative complexity, including working denials on top of upwards of 3.3% in revenue loss per hospital each year due to denials — a staggeringly high number when you think about the $3.6T spending size (using as a proxy for total revenue) of the U.S. healthcare system.
4. Automation is the obsession du jour
Given the above, figuring out ways to get claims paid faster, reduce defects and reduce the overhead associated with administering health systems has been the rallying cry of industry denizens. Over the past 5 years, the following sequence of buzzwords were really just proxies for people who wanted solutions that increased revenue and/or decrease cost: machine learning, block chain, chatbots, robotic process automation, intelligent process automation, AI. In short, automation and systems consolidation. Or, introducing robots into the factory floor ecosystem.
There are a number of industry initiatives underway attempting to break this mold. You’ve probably heard about population health, measuring social determinants of health and maybe even the moniker, “value-based care.” All of these seek to implement capitated/bundled-payment style remuneration schemas to replace “fee-for-service.” In short, they aim to pay providers for outcomes and not for each of the individual procedures/charge codes they so meticulously groom on outgoing claims (that’s not sarcasm; they have to do this to avoid costly denials of payment by payers). But ensuring outcomes is no easy feat given the relative complexity of the human body and our still nascent understanding of all the ways in which the “practice” of healthcare can go wrong (and all the ensuing implications of when it does).
So, the factory floor is here to stay — at least for a while longer. Building product in this space necessitates an intimate understanding of these dynamics if you’re actually going to solve, or at least mitigate, problems.
