Healthcare organizations begin to embrace digital disruption

Register now

Information technology is transforming healthcare, increasingly being used to radically improve care and change many of the ways in which organizations have traditionally practiced medicine.

Digital approaches are changing how physicians and healthcare systems diagnose diseases, treat patients and monitor their conditions on an ongoing basis. These new iterations are coming rapidly, as technology enables care to become virtual and patient-centric.

Various technologies such as artificial intelligence, natural language processing, medical devices connected via the Internet of Things, smartphone-based apps and more are giving doctors myriad options for revamping patient care.

This is disrupting common notions of healthcare and, at the same time, counteracting negative perceptions some clinicians may have had about information technology to date, says Lyle Berkowitz, MD, a medical informaticist and IT entrepreneur based in Chicago.

Contentions that healthcare IT has unduly burdened physicians are “silly,” he says, because most problems are a result of the way technology implementations have been done, as well as the relatively basic purposes behind healthcare IT.

Emerging solutions that reconceptualize care delivery are catching on because they cut straight to the need to use IT to truly improve care—using technology’s strengths to eliminate plodding, labor-intensive processes that make healthcare expensive and unnecessarily burdensome.

“Too often, innovation starts with cool things we think we can do, as opposed to [starting with] the problems we need to solve,” says Andy Slavitt, former administrator of the Center for Medicare and Medicare Services who has become a noted health reform advocate.

Slavitt now sees technology being used to improve care and support clinicians in their work—and he believes that trend will continue. “We need to innovate in ways that give physicians and clinicians and patients more time to get the better result,” he adds.

Increasingly, those innovations are coming to market today, disrupting norms in healthcare delivery and bringing new efficiency and capabilities to provider organizations. The use of these technologies holds promise for assisting clinicians in making care decisions and improving patient health.

Artificial intelligence

Clinicians and hospital administrators are adopting computing technology to process data and make decisions based on the insights derived from it.

The growing acceptance of smart technology—artificial intelligence, machine learning, predictive analytics and deep learning—involves making use of some of the best capabilities of computing power. These advanced technologies can organize data and derive insights from that information, thus helping support clinicians as they make crucial care decisions.

AI is increasingly being incorporated into imaging products to assist radiologists by improving the speed and accuracy of their diagnostic work.

For example, Royal Philips has developed IntelliSpace Portal 9.0, the latest edition of its comprehensive, advanced visual analysis and quantification platform. The product helps radiologists detect, diagnose and follow up on treatment of diseases, while using new machine learning capabilities to support the physician. Another of its products, Illumeo, is an imaging and informatics technology that uses adaptive intelligence to redefine and enhance how radiologists work with medical images.

Radiology professionals are increasingly seeing new IT capabilities applied to image assessment. Early tests have suggested that these approaches are at least as accurate as clinicians. For example, a research team led by Case Western Reserve University studied whether a deep-learning network approach could identify invasive forms of breast cancer. The network was “trained” by downloading 400 biopsy images from multiple hospitals; then, the network was asked to analyze 200 images from The Cancer Genome Atlas and University Hospitals Cleveland Medical Center. According to Anant Madabhushi, professor of biomedical engineering at Case Western Reserve and co-author of the study, the network scored 100 percent accuracy in determining the presence or absence of cancer on whole slides.

AI, machine learning and other forms of applying computing to intelligence are being combined with other technologies and analytics to help providers anticipate patient problems and then head them off.

For example, Hamilton Health Sciences, based in Hamilton, Ontario, is using technology from Toronto-based ThoughtWire, which combines data streaming from Internet of Things devices with AI for a variety of purposes. In the realm of patient care, the technology is able to forewarn staff of potential Code Blue calls, using a grading scale to launch preemptive interventions before patients are even in danger. Code calls have been reduced, and the organization has set a goal of eventually eliminating them altogether. As a result, patient survival is enhanced by intervening before heart arrests require heroic measures, says Mark Farrow, vice president and chief information officer at Hamilton Health.

The ThoughtWire technology also uses AI to assist administrators. For example, it is able to calculate projected staffing needs, assisting Hamilton Health in scheduling nurses to specific units, Farrow notes.

Other AI targets

Thus, artificial intelligence can contribute on several levels to disrupt current challenges in healthcare. But it can make an immediate impact,by first rescuing clinicians from overwork, says Santosh Mohan, who chairs the HIMSS Innovation Committee and heads an initiative at athenahealth called “More Disruption Please” Labs. Mounting administrative tasks “are not the best use of a doctor’s time, and [physicians] aren’t great at doing assembly-line check lists anyway,” says Mohan. As a result, he sees great value in using technology to computerize and automate some of this “low-level work.”

For example, Berkowitz is co-founder and chief medical officer of healthfinch, a new company that seeks to squeeze time and cost out of some of the most ordinary tasks in clinical practice.

“Our focus is not on the most difficult, complex 5 percent of patients,” says Berkowitz, who’s also CEO of FutureHealth, a precision medicine-based primary care practice and director of the Szollosi Healthcare Innovation Program at Northwestern Medicine. “Our focus is on the 80 to 90 percent of patients who just need routine care. And if we can automate and delegate a lot of that care, we can give doctors back a lot of time to be able to focus their attention on the more complex cases.”

The objective is to clear away routine work—such as refill requests, emails about incoming patients and lab test orders—that pile up in physicians’ electronic inboxes says, Jonathan Baran, CEO of healthfinch.

Mohan says AI can make an immediate impact in performing “intelligent computing” in areas such as scheduling, predicting no-shows and cancellations, determining appropriate appointment lengths, and tackling the drudgery of insurance preauthorizations.

Remote monitoring

Advanced IT also is making inroads in providing care to patients who are not in traditional healthcare settings. As providers deal with shrinking reimbursements, they’re seeking to treat patients in the least expensive care setting possible.

New technologies are enabling organizations to monitor these patients offsite. Such capabilities are crucial for organizations that want to better manage their care continuum and handle the financial risk associated with population health and value-based care contracts.

In home healthcare, for example, subjecting vital sign and other monitored data to a “lightweight machine learning algorithm,” AI can “predict the onset of an adverse event and perhaps avoid an ED visit or some other bad outcome,” says Michael Joseph, CEO of Prime Dimensions, a healthcare consulting firm.

This class of technology also can “microsegment” patients along lines of what works for a very specific health history and medical situation, by mapping back to records of similar cases very quickly and applying a care pathway accordingly, something that couldn’t be done even five years ago, says Joseph.

Another problem with delivering care at patient homes is that a nurse can’t be present to constantly observe patient compliance with medication administration or other treatment routines. And if a patient lapses into a negative outcome while on a treatment, it’s not clear from self-reported data whether the problem was nonadherence to treatment or the treatment itself didn’t work.

By using a front-facing camera function of a smartphone in combination with sophisticated algorithms, a platform created by AiCure simulates the interaction of nurses. It identifies a specific med held up to the phone, “observes” the med going into a patient’s mouth and confirms it was ingested, even to the extent of alerting to possible “cheating” if the patient’s uniquely recognized face drops out of the field of vision at any point, says Laura Shafner, AiCure’s chief strategy officer.

Some monitoring approaches are geared to the hospital setting, enabling clinicians to identify changes in patient conditions in real time and make adjustments in treatment approaches or interventions to head off deterioration in patient conditions.

For example, Medtronic is offering a combination of technologies that enables patient monitoring software with wireless monitoring devices and customizable clinical decision support mobile applications. The company contends that its Vital Sync monitoring and CDS solution can simplify time-intensive patient care processes, helping clinicians prevent or mitigate harmful and potentially costly adverse events.

Medtronic’s solution gathers patient physiological data from a variety of wireless and bedside devices made by Medtronic or other vendors. Vital Sync is a connectivity and remote patient monitoring software solution from Medtronic that links electronic medical record systems to a variety of medical devices, including ventilators, capnography monitors, pulse oximeters, depth-of-consciousness monitors and other devices. Clinicians can remotely view results on any web-enabled device, including smartphones.

Patient involvement

Disruptive technology also gives patients an opportunity to take a more active role in checking on their health and participating in their care.

For example, Zipnosis is a virtual care platform that works with a client health system—using its physicians and labeled with the system’s name—and acts as “an entrée and an entry point” to the health system, says Jon Pearce, the its CEO.

Whereas traditional telemedicine has to put both clinicians and patients in front of a camera at the same time, the Zipnosis platform works like this: A patient with a simple condition, such as a bladder infection, can go to the hospital’s website and answer a series of software-guided diagnostic questions, enabling the service to capture a full history. It then applies rules to suggest the appropriate level of care for that patient—ranging from online advice with no need to talk to anyone, to a video consultation to a clinic visit, which the service can schedule.

And smartphones laced with other advanced technologies are enabling consumers to take more of their diagnostic and care needs into their own hands.

For example, Duke University researchers have developed a handheld device for cervical cancer screening that produces high-quality images on a smartphone or laptop, part of an initiative to make screenings more accessible, easier to conduct and less costly than studies using expensive traditional equipment.

The wandlike device, which is portable and simple to use, captures high-quality images of the cervix. In fact, the pocket colposcope rivals the image quality of the best colposcopes on the market but at a fraction of the weight, size and cost, contends Nimmi Ramanujam, the Robert W. Carr, Jr., Professor of Biomedical Engineering at Duke.

“The mortality rate of cervical cancer should absolutely be zero percent because we have all the tools to see and treat it—but, it isn’t,” says Ramanujam. “Women do not receive screening or do not follow up on a positive screening to have colposcopy performed. We need to bring colposcopy to women so we can reduce this.”

Not just for tech’s sake

Many experts believe that innovation needs to continue to focus on making changes that improve healthcare, and not just make clinicians’ lives more complex.

Healthcare’s plodding processes need reinvention, but innovation without a plan of adoption won’t make a difference in care transformation. Digital disruption only works if it solves longstanding problems and makes clinicians’ lives easier, Slavitt says.

For example, the healthfinch service uses a cloud-based rules engine integrated with an EHR to detect a task—refill request, appointment or some other trigger—and review a patent’s chart for a variety of activities, such as the last appointment, medicines taken, recent problems, recent lab tests and vital signs. An algorithm determines what should be done, and routes the task to the right person.

With technology designed to take the rules-based protocol from the nurses and automate all the steps, the resulting work done by the system is handed off to the nurse, who then has the information to be assured that it met the protocol and is ready for action, Berkowitz says. For example, he explains, conditions were met for a refill, conditions weren’t met for reasons that required extra scrutiny or physician intervention, or it was in a gray area in which not all data could be found but a nurse could fill in the rest.

All told, the automation was able to reduce the workload of nurses by 80 percent, he says, which made it possible to redeploy the excess nursing staff to other centralized activities where they were needed.

Whatever the promise of advanced technology, however, digital disruption has its limits in the current culture, experts warn. Physicians have warmed to the idea of following evidence-based or recommended rules for diagnosis and treatment, but trusting conclusions generated by artificial intelligence or machine learning is another matter.

For reprint and licensing requests for this article, click here.