S. Bala  |  02/27/2009

Lean Triage for Hospital ERs

Saving patients, time, and money can happen all at once.

The United States spends 16 percent of its gross domestic product (GDP) on health care, more than any other nation. Although that investment has produced medical experts and breakthroughs envied the world over, a great majority of U.S. citizens are unhappy with the end results. When the nonpartisan Commonwealth Fund conducted a poll of U.S. health care consumers last year, 69 percent expressed strong dissatisfaction with the current health care system. In a 2007 survey, the same group found U.S. respondents twice as likely to support a complete overhaul of their system than those from Canada, Germany, the Netherlands, New Zealand, Great Britain, and Australia--all nations that spend half as much GDP as the United States on health care.

Lesson: If the health care delivery process is broken, no amount spent on superior health care professionals and resources will change the quality of outcome--or the dissatisfaction that patients have with their treatment experiences. My firsthand experience consulting hospital medical staff and administrators in the United States supports this conclusion. Problems in the delivery of care are often caused by certain systemic pressure points becoming overtaxed. The most glaring example is the way the high demand for immediate care is tightly funneled through the typical hospital emergency room (ER). This funneling routinely compromises the speed, satisfaction, and outcome for patients entering the health care system through ER doors.

For the estimated 40 million U.S. citizens without health insurance, local ERs that can’t by law refuse them treatment have morphed into the medical option of first resort for both primary and emergency care. Even among the 80 percent of U.S. citizens who are insured, the ER frequently ends up being their first stop. Sometimes the insured begin there because their medical condition first occurs after normal doctor-office hours, and the ER is their only practical option for determining if it’s serious or not. Such behaviors explain the Center for Disease Control’s 2006 finding that 40 percent of U.S. ERs are overcrowded, and how, according to one Harvard study, average ER wait times rose 36 percent from 1997 to 2004. Half of ER patients waited in excess of 30 minutes to be examined. Even more disturbing, half of all heart attack patients waited more than 20 minutes.

Given the volume of ER patients and variety of their needs, improving satisfaction and the quality of treatment hinges greatly on how successfully the incoming are triaged. Quality care isn’t a question of treating all patients faster. It’s a matter of determining--as fast as possible--which patients need to be treated more quickly and which can endure a modest wait, and, in either case, how all unnecessary wait time can be stripped from the process flow. The goal is to simultaneously improve the rate of speed and accuracy at which debilitating conditions are diagnosed and prioritized. Faster, correct identification of the condition at hand (and the appropriate course of treatment it demands) allows for a more customized, right-timed response that is neither recklessly rushed nor dangerously delayed.

No magic pill but…

We can’t instantly optimize the myriad pathways of care that patients--from indigent to insured, infant to elderly--should ideally walk. A wealth of steps can improve the speed and quality of the intake-and-treatment process. For an ER in an award-winning U.S. hospital at which my colleagues and I worked, the suite of solutions ranged from state-of-the-art (a $100 million database that allowed return-patient histories to be factored into assessments with the swipe of a bar-coded key-chain tag) to the lowest of low tech (new exterior signage that ensured the shortest possible trip from parking lot to the walk-in registration desk). Being able to discreetly access a patient’s prior ER history with an “e-card,” for instance, allows the triage nurse to know about a patient’s past asthma attacks even if the patient is having trouble catching his breath to speak. This solution not only reduced the intake process for return e-card patients to three questions, but also the total paperwork associated with the initial processing of e-card patients is now 11 pages, down from 22. In addition to these solutions, process stakeholders within the ER and the hospital at large were trained in lean to encourage continuous improvements.

Thanks to the many process improvement recommendations made, the average expected throughput time for an ER patient, from entry to release or admittance, has been trimmed to 75 minutes. Situation-specific improvements are more impressive. Obvious “emergency” patients (e.g., those brought in by ambulance) have a wait time of 20 minutes. For less serious “urgent” patients, the expected wait is down to 90 minutes. Note that a visit to a doctor’s office typically runs more than an hour, from entry to exit. That means an “urgent” patient whose doctor’s office is closed can get the same or better care, in nearly the same time frame, at this retooled ER.

The goal in reengineering care in the emergency arena is to quickly yet correctly segregate patient populations in a way that facilitates a more focused, higher-quality care option. At this hospital’s ER, triage nurses placed incoming patients into five well-defined categories of escalating severity. After fast-tracking the 7 percent of patients who stood out as the obvious first and second in severity, the remaining third, fourth, and fifth underwent a more deliberative evaluation. During this step, one in four patients was rerouted to a rapid care center. This center attends to “urgent, nontraumatic” conditions needing quick care to reverse or stabilize a situation that could conceivably progress to serious if unaddressed too long. Another one in five of the third, fourth, and fifth were designated as “nonurgent trauma,” defined as conditions that won’t grow worse for any wait, but merit a same-day treatment response. (Think broken bones in need of casting.) Teasing out and redirecting “urgent, nontraumatic” and “nonurgent, traumatic” patients thinned ER traffic 42 percent and trimmed expected wait time for those who remained by 20 percent.

Addressing side effects

Improved utilization of commonplace equipment, it turned out, had a profoundly positive side effect on wait time, satisfaction, and quality of outcome. Previously, patients with possible fractures saw an ER doctor before and after getting an X-ray. A small change in protocol--empowering nurses to order the X-rays before the doctor ever saw a possible-fracture patient--dramatically cut the time spent waiting for a definitive diagnosis. This policy change freed doctors to help more ER patients in less time and spend more of that time on treatment. Because X-rays would have been required regardless--determining if there was a break and what kind--the change didn’t cause diagnostic costs to increase.

Scheduling radiologists to work more of their full-time workweek during hours when trauma injuries where historically highest, and less when they were lowest, further truncated the wait time for X-ray patients. The rescheduling also eliminated costly overtime. (A similar weighting of triage-nurse schedules had a comparable benefit.) When X-rays revealed a fracture during peak hours, it was recommended that patients have the option of immediately setting the break with a temporary splint, and then returning at a later off-peak time to have permanent casting done. Patients would wait a little longer to conclude treatment, but they could carry out most of the wait at home--or for a child, perhaps at a favorite ice cream parlor.

Something as simple as speeding the turnover of specialty ward beds--which must be cleared before patients slated for overnight stay can leave the ER--could also have a subtle yet significant effect on overall ER cycle time. Most hospitals, like hotels, know roughly when a guest is expected to check out, but they don’t know the exact time until after checkout happens. My colleagues and I determined that a little bit of technology-induced “visibility” and a lot of interdepartment and cross-functional coordination could loosen the bottleneck. To speed bed turnover after checkout confirmation, the hospital network was set up so all parties involved could be notified in real time. This enabled bed prep to be done in a lightening-quick “pit crew” fashion.

A lower-drama trauma

At the top of the hospital’s triage pyramid is a smaller, far more serious subset--“severe trauma” ER patients. It’s a group that often arrives by helicopter, and for whom life and death is usually measured in seconds, not minutes. Every item required to stabilize these patients must be at arm’s length and all procedure- critical materials organized for easy location. Unfortunately, different doctors and nurses are perpetually restocking and rearranging trauma-area supplies in conflicting and ultimately chaotic ways as they move frenetically from one patient to the next. This can quickly unwind even the most organized layout.

Our answer to this problem was to train trauma staffers in the 5S methodology of continuous improvement: sort, straighten, sweep, standardize, and self-discipline. Special attention was paid to self-discipline, which, it was stressed, would be fundamental to staff collectively keeping the area in order. Where all surgical gloves were once stuffed indiscriminately into one small area, they were now clearly separated by size and granted a larger portion of real estate in the trauma area “cockpit.” This prevented precious seconds from being wasted looking for a proper fitting pair. Tellingly, making more room for such an inexpensive but essential item meant removing from the trauma-area inventory many more expensive but far less frequently needed items that had been taking up valuable space.

Dedicating more space and preeminent position to high-use, trauma-area essentials--at the expense of pricier nonessentials--unlocked cash that had been trapped in overstock, even as it quickened the pace of life-saving treatment. Organizing materials into procedure-specific prefab kits allowed team members to better address those trauma situations they confronted most routinely. The catch? Realizing any of these cost- and life-saving dividends required that team members surrender a lot of individual preferences for how specific materials and kits are configured. It meant asking a strong-willed and accomplished doctor or nurse to forego irrational but understandable attachments to having a certain item always at his or her fingertips just because it proved critical once--but wouldn’t prove useful 999 times out of 1,000.

Long-term cure

From my experience, the right elements and motivations are there for the United States to consistently offer the highest-quality health care to all. Hospitals want to attain the best health care infrastructure so they can become the first choice to the most patients and best doctors. Health care professionals want to deliver the best individual patient care and outcomes. Insurers want to provide affordable quality coverage so they can attract and keep a large, profitable base of policyholders. The problem is that each of these three constituencies naturally see their goal as the most crucial, when, for the system to function optimally, no one goal can ever be more important than the other two.

Long term, doctors must see themselves less as welcome guest workers at the hospital hosting them, and more as cohosts with a vested interest in the facility’s broader success. An ER with faster throughput cycle time, after all, benefits not only patients, but also the financial health of doctors and hospitals alike. Insurers, for their part, must see the folly of punishing good process-reengineering deeds by paying less for a procedure if it’s done in a rapid response center than when it’s done in an ER. Finally, hospitals must see the need for an on-site “primary care” option for indigent patients. If care must be provided regardless of ability to pay, common sense dictates it be delivered in the most cost-efficient fashion, and providing primary care in an emergency care setting is as far from cost-efficient as one can get.

There’s nothing wrong with wanting the U.S. health care system to undergo a radical overhaul for the better, as one-third of U.S. citizens desire. But until that day dawns, hospitals must do their best to address unfunded patient care costs, which won’t disappear anytime soon, and that, increasingly, will be unfairly shifted onto the insured in the form of higher premiums, deductibles, and out-of-pocket expenses. Without wholesale reform, we might not be able to make quality health care affordable to all. However, with the application of lean and continuous improvement methodologies, we can make the treatment experience and outcome a lot better and more affordable for many more people than anyone could have ever imagined possible. That’s a firm step in the right direction.


About The Author

S. Bala’s picture

S. Bala