I have blogged perviously about hospitals beginning to lose money after installing an Epic EHR (see: Some of the Details Behind the Maine Medical Center/Epic EHR Meltdown; Who Says a Hospital CIO Can't Get Fired for Picking the Epic EHR; The Cost of Deploying an Epic EMR and the "Oreo Cookie" Analogy). Often the details of these financial shortfalls are sketchy because of the contractual gag clauses imposed by Epic on their hospital clients (see: Gag Clauses in EHR Contracts Documented; Concerns Raised about Patient Safety). A recent opinion piece discusses how an Epic EHR installation at MD Anderson has triggered financial losses (see: Epic Install Triggers Loss At MD Anderson):
Surprising pretty much no one, another healthcare organization has attributed adverse financial outcomes largely to its Epic installation. In this case, the complaining party is the University of Texas MD Anderson Cancer Center, which attributes its recent shortfall to both EMR costs and lower revenues. The news follows a long series of cost overruns, losses and budget crises by other healthcare providers implementing Epic of late. According to Becker’s Hospital CFO, MD Anderson reported adjusted income of $122.9 million during that period a 56.6% drop over the seven-month period ending March 31. During that period, the cancer center’s wages and salaries climbed, and Epic-related consulting costs were climbed as well. This follows a $9.9 million operating loss for the first quarter of the 2016 fiscal year, which the University of Texas attributed to higher-than-expected EMR expenses....The organization didn’t announce what it was spending on the Epic install, but we all know it doesn’t come cheap....And UT leaders seem to have been prepared for the bump, reporting that they’d planned for a material impact to revenues and expenses as a result of the Epic implementation. The system didn’t announce any staff cuts, hiring freezes or other budget-trimming moves resulting from these financial issues.Having said all this, however, no organization wants to see its income drop. So what actually happened?
...[W]hen the UT system reports that a drop in patient revenues contributed to the drop in income, what does that mean? Does this refer to scheduled drops in patient volume, planned for ahead of time, or problems billing for services? I’d be interested to know if the center managed to keep on top of revenue cycle management during the transition. Another question I have is what caused the unanticipated expenses. Did they come from contract disputes with Epic? Unexpected technical problems? Markups on consulting services? Or did the organization have to pour money into the project to meet its go-live deadline? There’s a lot of ways to generate costs, and I’d love to get some granular information on what happened. Also, I wonder what steps UT leaders will take to avoid unexpected expenses in the future. While it may have learned some lessons from the problems it’s had so far, there’s no guarantee that it won’t face of the costly problems going forward.If, perchance, and the system has figured out how to stay in the black with its Epic investment, it could sell that secret to cover its IT expenses for years. I’m betting other systems would pay good money for that information!
I have come to the conclusion that the losses experienced by hospitals when installing and running an Epic EHR are multifactorial and include the following issues:
- The Epic EHR is avidly sought, particularly by large hospitals, so the company is able to charge a premium price for it with little room for the hospitals to haggle. Licensing and installation fees can amount to hundreds of millions of dollars or even more for larger hospitals and integrated systems.
- Epic has developed an ecosystem of consultants that help to install the Epic system with tight controls over entry into the group by Epic. This leads to high-end fees for their services.
- I think that Epic, and probably many of the competing EHRs, have been designed mainly to generate patient bills and not necessarily to optimize the time and efficiency of the clinicians using the system. Hence, physician and nurse productivity tends to take a nosedive after installation (see: Ambulatory Physicians Cite EHR Design for Their Recent Loss of Productivity). Hospital executives enthusiastically cling to the notion that this drop in physician/nurse productivity is transient and will return to previous levels when these professionals become familiar with the new system. I hold the belief that a partial loss of productivity with EHRs will persist over time. Because of the gag clause, it will be difficult to learn more about this issue.
- Part of the loss of revenue experienced with a new EHR is due to bugs and start-up inefficiencies with a new financial system. Many of these problems will fade over time.
- Because of the drag on physician productivity when using an EHR, hospital personnel cost often rise after installation including the incremental cost of newly hired scribes to assist physician in keyboard input (see: The Explosive Growth of EHR Scribes; Back to the Future).
I believe that the notion of running an entire health system with a single EHR from a single vendor is simply not practical and not seen in other industries. This is a legacy of the IBM/HIS philosophy of decades ago (see: What's Really Wrong with EHRs: Beginning a Deep Dive). Here is a quote from that article:
...[H]ospital executives have been saddled with the burden of a premise that is not accepted in any other U.S. industry. It's the idea that it's possible to manage most of the important hospital computing tasks with a single, massive information system. This idea was promoted by IBM in the 70's and 80's with their so-called Hospital Information Systems (HISs) and continues to hold sway today with the electronic health record (EHR).
In time, the notion of a single, monolithic EHR from a single vendor will fade in popularity as did the IBM HIS previously. Until that time, we will just need to live with the financial drag of such systems.