December 2015

Advice to EMR Vendors: Get Parallel
By Michael Koriwchak, MD
For The Record
Vol. 27 No. 12 P. 28

Thanks in large part to the meaningful use (MU) incentive program, EMR vendors have sold millions of software licenses to medical practices. The most recent number pegs the EMR use rate at approximately 80% of "eligible providers." The Office of the National Coordinator for Health Information Technology has proclaimed victory, declaring that physicians have "seen the light" and embraced the near-miraculous ability of EMRs to rescue medicine from a range of maladies, from unnecessary tests to handwritten prescriptions.

But if you were to walk a mile in a physician's shoes, you would see a totally different world. Forced by law to adopt technology that many in the field regard as promising but not ready for clinical use, physicians are frustrated, bitter, and ready to spit nails.

Physicians have purchased EMRs, jumped through MU hoops, collected incentive payments, and avoided reimbursement penalties. In the process, they have accumulated years of experience using EMRs—and still find the technology to be awkward, inefficient, and difficult to use. EMRs impede workflow, resulting in physicians seeing fewer patients than they did when they employed paper charts. EMR chart notes—even for routine visits—are long and difficult to read, forcing physicians to spend time after-hours completing documentation. Identifying clinically useful information in EMR notes is almost impossible.

It can be argued that for all this expense and frustration, physicians and patients have received absolutely nothing in return. But it doesn't have to be that way.

A Shift in Focus
If EMR vendors would acknowledge the needs of physicians—their true customers—the captive relationship could be rectified to one that is more productive and rewarding for each party. Physicians don't care about MU compliance. They don't want to be conscripted into becoming unpaid Big Data collectors. They don't want patient care to be dictated by clinical decision support tools. They don't want the value of their work to be determined by the number of quality measure check boxes they click. They want value to be measured in terms of their ability to earn a return on investment for their large EMR purchases.

When that EMR comes along, physicians will line up around the block for hours waiting to buy it, regardless of whether there are government incentives or penalties at play.

Where Failures Lie
What are physicians so upset about? All of the problems boil down to how EMRs are viewed differently by the HIT community and physicians. The HIT folks—and the federal government—envision a top-down system in which providers collect data from each care event and feed it to a central database. Through the "miracle" of data analytics, medical breakthroughs magically appear, seemingly falling from the sky. Based on this new knowledge, centralized care is pushed out to individual providers and patients via centrally connected EMRs.

However, physicians don't share this vision. Unlike IT professionals, physicians understand that true quality care starts from the bottom up, not the top down. Quality care begins with a physician, a patient, an exam room, and a conversation. Patients are much more than a statistical profile—becoming familiar with their hopes and fears enhances care. The physician's vision of HIT features the software working for physicians and patients, not the vendors and Uncle Sam.

To fulfill this ideal, an Atlanta practice created an EMR with a one-of-a-kind workflow engine that gathers everything a clinician needs to document and deliver care—and code for that care—into a group of closely linked screens. The engine also includes an MU dashboard (a necessary evil) that allows a clinician to conveniently review the MU criteria for each patient encounter in a timely manner.

Talking Parallel
Central to the design of an effective workflow engine is an understanding of parallel processing, a concept that allows properly designed EMRs to be more efficient than paper charts.

To understand parallel processing, consider two examples. In need of new tires, you take your car to the reliable neighborhood mechanic. Working alone, he can replace only one tire at a time, incapable of moving on to the next wheel until the first is finished. Any interruptions, such as the phone ringing or a parts delivery, will delay the job and may hurt the quality of the work. And the job will take at least all day, maybe longer. This is sequential processing; the steps to completing a task are performed one at a time in sequence. The workflow is slow, inefficient, and prone to error.

Consider the same situation during a car race. The driver, realizing it's time to change tires, makes a pit stop. Like clockwork, a crew of six or seven technicians pounces on the car. In less than 10 seconds, all four tires are changed, plus the gas tank's filled, the driver gets a sip of water, the windshield's cleaned, and small repairs are taken care of. This is parallel processing in which multiple steps occur simultaneously. The workflow is rapid, precise, and effective.

Now let's compare sequential processing with parallel processing in the physician office. Consider the steps involved in getting a patient prepped for surgery. The chart note must be completed with indications for surgery documented. Multiple forms, including surgical consent and presurgical orders, must be filled out and signed. The procedure must be approved by insurance and scheduled. And any preoperative consults, such as cardiology, must be scheduled and completed.

Paper charts cater to sequential processing because the file can only be in one place at a time. But who gets the chart first? The person who schedules the surgery to get the case on the itinerary? Or the medical assistant to get all forms and medical evaluations completed? What if the patient calls the office the next day with a question and the paper chart gets pulled by another staff member? The risk of workflow delay and failure is high.

More than one person can access electronic charts at the same time. This allows parallel processing, which produces a significant increase in workflow speed and greater efficiency. For example, the surgery scheduler and the medical assistants don't have to compete for possession of a single paper chart. Everyone can access the chart and tackle multiple tasks at the same time. Parallel processing also allows great flexibility in workflow design—different people in different locations can perform different jobs.

Neither physicians nor the IT community understands parallel processing well. The idea has received little attention, having been drowned out by the noise surrounding clinical decision support, interoperability, Big Data, MU, and other misguided and misprioritized ideas backed by government bureaucrats and the HIT community.

Should an enlightened EMR vendor come to realize the impact of parallel processing, it would be a game-changer. In fact, the first vendor that has the courage to cast aside government mandates in favor of providing value to physicians will have customers lined up to buy its product.

— Michael Koriwchak, MD, an ear, nose, and throat physician in Atlanta, is vice president of the Docs4PatientCare Foundation.