As seen in Forbes Insights on February 11, 2019
Anyone who has visited a hospital recently has been confronted by a gleaming array of technological marvels.
State-of-the-art medical devices such as MRIs and CAT scanners, 3D-imaging devices, proton-beam cancer therapy machines, and robotic surgery platforms have utterly transformed the clinical side of hospitals—the part that patients see.
Behind the scenes, however, things aren’t nearly as cutting edge. Back-office technology that helps hospitals perform critical, everyday tasks like compiling and storing medical records, scheduling staff, reserving operating rooms, and billing patients is, by most accounts, so antiquated that it often causes more problems than it solves.
Take electronic health records (EHRs), for example. Although an estimated 90% of physicians nationwide have adopted electronic record-keeping (up from 10% a decade ago), the benefits of EHRs have been elusive, to say the least. EHRs have been hampered by inconsistent standards, cumbersome user interfaces, and a lack of interoperability. This has contributed to burnout and information overload among physicians, and less hands-on care for patients.
In a 2018 poll for Stanford Medicine, surveyed doctors reported that more than 60% of their time spent on behalf of patients is actually devoted to interacting with EHRs. Additionally, half of office-based primary care physicians stated that using an EHR actually detracts from their clinical effectiveness. During a busy, 10-hour emergency room shift, a clinician will make an estimated 4,000 keyboard clicks merely to keep up with the demands of EHR.
Even hospitals with established EHR systems often have to transfer information on paper because of interoperability problems with other systems. Patient records are still routinely passed between hospitals by fax machine—a 1980s answer to a 21st century problem.
“On the clinical side of hospitals, there’s more cutting-edge technology than in practically any other industry,” says Stephen Filler, a director at the Boston Consulting Group who specializes in healthcare and a former hospital CIO. “But on the operational side of hospitals, the technology is so antiquated, we’re literally 20 years behind most other industries in digitizing.”
Hospitals Turn To New Technologies
But there’s a quiet revolution underway in hospital IT, as forward-looking hospitals turn to new technologies such as artificial intelligence (AI), automation, machine learning, advanced algorithms, and decision engines to modernize behind-the-scenes operations.
Hospitals are updating IT because they can’t afford not to. Squeezed by rising staff and operational costs and lower reimbursement rates, hospitals are being forced to wring new efficiencies out of expensive and time-consuming back-office operations.
“Every healthcare organization is trying to give better care at lower cost,” says Dr. Lloyd Minor, dean of the Stanford School of Medicine. “Technology plays a big role in that.”
UCHealth in Colorado is one of a growing number of healthcare systems using predictive analytics to improve its operations. In partnership with Silicon Valley company LeanTaaS, UCHealth has adopted a program called iQueue, which analyzes data about how surgeons use their operating room (OR) time and identifies the causes of delays and unexpected bottlenecks. The program uses predictive analytics and machine learning to uncover inefficient patterns of use and reallocates operating room time as needed.
“iQueue can tell us, ‘Here are the three surgeons that are maximizing the use of their OR time, and here are the three surgeons that aren’t using their OR time to maximize capability,’” says Steve Hess, chief information officer for UCHealth. “So we can talk to those latter three surgeons and maybe even take some OR time away from them and give it to the other three surgeons.”
iQueue has been implemented in about 100 of UCHealth’s operating rooms and has spread to about 50 healthcare providers nationwide, including Stanford Health Care, UCSF Health, and Memorial Sloan Kettering Cancer Center.
For UCHealth, the 4% increase in OR utilization due to iQueue translates to about $400,000 in additional revenue per operating room annually. That’s important because operating rooms are among a hospital’s most valuable real estate.
“The operating room typically generates 40% of the revenues at a hospital, and every minute of an operating room is worth about $300,” notes Mohan Giridharadas, founder and CEO of LeanTaaS. “What the iQueue platform does is become the air traffic control for the OR.”
It analyzes prior patterns and from that suggests an optimal sequencing of surgical procedures, every minute of every day.
UCHealth has also adopted Alaris IV pumps, smart intravenous pumps that connect directly with a patient and their electronic medical records. An LCD screen displays the IV fluid name, dose, and rate to confirm the right patient is getting the right medication at the right time. Before the smart pump, such information was entered manually by one or more clinicians, making the process slower and much more prone to human error.
“The smart pumps are helping us make better decisions and be less error-prone,” says Dr. Richard Zane, chief innovation officer for UCHealth.
AI To The Rescue
Hospitals are also increasingly turning to AI to improve a critical healthcare task—physician decision making. RxRevu is an AI-powered decision support tool embedded in the EHR that helps doctors prescribe the most cost-effective, clinically appropriate medication for a patient. Patient-specific cost, formulary, and benefit data are integrated into the workflow, so a doctor never has to leave the EHR to access prescription decision support.
Similarly, a decision engine called AgileMD uses predictive analysis to provide doctors with a list of recommended tests and treatments for their patients, based on a wide variety of clinical factors. Without leaving the EHR, doctors can access clinical pathways and protocols to support clinical decision making, order tests, and streamline documentation. As a result, doctors spend more time interacting with their patients and less time chained to their computers.
Hospitals Look To The Future
The quiet revolution in back-office hospital IT is also causing the traditional lines between IT and facilities departments to blur. A large hospital can easily contain 50,000 medical devices, $1 billion worth of equipment, and a substantial team of facilities staffers charged with keeping the equipment in top working order. As recent as five years ago, a facilities department often operated in a vacuum, rarely communicating directly with IT staffers. But increasingly, the two departments are working closely together.
“In a hospital, you can no longer make the distinction between a respirator and the physical environment in which that respirator exists,” says Tom Stanford, CEO of Nuvolo, a New Jersey–based firm that provides companies with cloud-based enterprise asset management platforms. “Both are responsible for patient care, and both are measured, quantified, and evaluated in terms of risk.”
And what about that bane of hospital operations, electronic health records? There are already moves afoot to streamline and improve EHR systems so that they’re more of a clinical asset than a liability. A 2018 Stanford Medicine study made a series of recommendations to improve EHR, including significantly increasing interoperability, developing systems and product updates in partnership with doctors, and eliminating the need for manual data entry by drawing on new technologies such as AI and natural language processing.
One recent pilot program between Google and Stanford, known as the Digital Scribe, gives a glimpse of a future in which doctors are entirely freed from their keyboards. In the nine-month pilot study, which concluded in mid-2018, doctors wore microphones that recorded their encounters with patients. Machine learning algorithms detected patterns from the audio recordings that were used to automatically complete a progress note, which is the primary EHR document that describes the office visit, including everything from vital signs and symptoms to a diagnosis and treatment plan.
Most of the voice recognition technologies used in the pilot study are already available in Google Assistant, Google Home, and Google Translate. The next goal will be to create a product that can be used on a broader scale in order to address the fundamental weaknesses of most EHR systems. Only then will hospital back-office technology begin to rival the high-tech medical devices that patients see.
“EHR systems are not as interoperable or as user-friendly as they should be, and they don’t provide the type of decision support we’d like to see,” says Dr. Lloyd Minor, dean of the Stanford School of Medicine. “But we’re getting to it now. We have to. Consumers are expecting it.”