In 2008, the California Department of Transportation (Caltrans) had a problem. Interstate 405, connecting north and south Los Angeles, had been declared the busiest highway in the nation, and traffic congestion was widely understood to be hellish. In the face of rising public frustration, Caltrans took the obvious next step. It launched the ambitious $1.6 billion I-405 Sepulveda Pass Improvement Project, which added an additional lane of traffic. A year after completion, the results were in: traffic along the widened stretch had actually worsened by 1 minute. The I-405 had become a victim of induced demand.
Induced demand is economist-speak for the concept that building more of something in high demand can actually increase demand for that thing. And it’s well-documented in the transportation industry. The idea is that the additional lanes incentivize drivers to use those lanes, who then take more trips and trips at peak hours. Congestion might briefly improve but quickly returns to previous levels. In fact, in 2009, two economists characterized the “fundamental law of road congestion” and demonstrated an almost perfect correlation between lane miles built and vehicle miles driven.
Induced Demand in the ED
In the same vein, it’s a mistake to add beds to a dysfunctional healthcare system.
Today, U.S. emergency departments (EDs) face a similar conundrum — seemingly limitless demand for emergency services. Beginning in 2012, the annual number of ED visits nationally jumped from 131 to 150 million in just 7 years, easily outstripping population growth. A recent study in California found that from 2011 through 2019, ED visits increased 23.4% while the state’s population grew by only 5%.
Concurrently, measures of ED overcrowding and dysfunction have also skyrocketed. From 2017 through 2021, the median percentage of ED patients nationally who left without being seen (LWBS) doubled from 1% to 2%. Another study found that by the end of 2021, boarding in a sampling of academic EDs had surged 40% beyond pre-pandemic levels. Many EDs now find themselves in a state of gridlock worthy of the I-405.
The natural instinct is to try to build our way out of this overcrowding, but mounting evidence suggests that EDs — like the transportation industry — are increasingly subject to the paradox of induced demand. In 2010, the University of California Davis expanded its ED from 33 to 53 beds. A year post-expansion, mean daily volumes had risen by 13%, the percentage of LWBS visits remained unchanged, and boarder hours had significantly increased. A study in Australia found that opening a new ED resulted in an 18% jump in volume (with 3% growth in the local population) with no change to ambulance offload times, time to see a doctor, or ED length of stay.
During my 20 years as a frontline emergency medicine physician, my healthcare organization has added both a hospital and an urgent care to my service area, created additional shifts at each of our facilities, and ramped up the speed at which physicians treat patients. After a decade of well-intended capacity expansion, our ED volumes remain at an all-time high. In the context of a national healthcare worker shortage, staffing often becomes the rate-limiting step that ultimately kills further expansion attempts. As the authors of the UC Davis study gamely note, “ED expansion alone does not appear to be an adequate solution to ED crowding.”
Out-of-the-Box Solutions
Induced demand is, understandably, a hard sell. But if we’re able to suspend disbelief, a host of out-of-the-box solutions become possible, including lessons borrowed from the transportation industry. These solutions hinge on the concept that, in an era of limited supply and endless demand, it is crucial to appropriately match resources with need.
For instance, one effective approach to decreasing traffic relies on the concept that one of the most effective deterrents to congestion is congestion itself. Posting highway travel times sets realistic expectations and helps drivers make wiser decisions about whether it’s worth it to continue their trip. For some, a different form of transportation or timing becomes more attractive; others choose to continue. Some evidence suggests that encouraging drivers to use alternate forms of travel, such as public transportation, during peak hours can positively impact traffic, particularly along congested routes.
This may be what the Academic College of Emergency Physicians (ACEP) was thinking when it proposed posting wait times and triaging low acuity patients to a non-ED setting. ACEP noted, “Diversion of lower acuity patients from the ED may result in increased ED acuity and decreased volume…anecdotal evidence supports the public posting of wait times as a means to distribute flow around a system or geographical area.” The goal was to allow patients with non-emergent conditions to make more informed decisions about when and how to seek care. A proliferation of alternate care options such as urgent care and telemedicine appointments now make this concept of redistributed flow all the more realistic.
Another strategy to address traffic congestion focuses on dynamic, rather than static, management of road conditions to match supply with demand. The idea is to recognize where and when the bottlenecks occur and, again, encourage redistribution of traffic throughout the system. Examples of dynamic management include the use of metering lights, carpool lanes, and dynamic pricing, and it’s proven remarkably effective at decreasing congestion. In fact, it’s been effective in many industries, including energy companies (think smart meters), restaurants (happy hour), and airlines (midweek specials).
It’s well known that certain times, days of the week, months, and holidays are prone to higher congestion in EDs. Why not move from the usual static approach to staffing, responsibilities, and roles toward a more adjustable model that acknowledges this reality? A recent review found that accurate assessment of peak ED hours in combination with flexible staffing levels and roles (achieved through cross-training staff) was associated with increased efficiency and patient throughput. In an era when ED bandwidth is stretched precariously thin, a dynamic model of operations can enable more appropriate resource allocation, improve patient-centric care, and maximize safety of patients, visitors, and staff.
Finally, there is some evidence that induced demand works in reverse. Although removal of a freeway inevitably sparks public outcry, horrific traffic jams don’t usually ensue. In 2003, San Francisco demolished its Central Freeway and replaced it with a surface-level boulevard. The project has been hailed as a huge success, and traffic generally flows well.
Humor me here — I’m not engaging in burn-it-down nihilism. Clearly, reverse induced demand can only be accommodated to a point. We can’t simply close down EDs haphazardly any more than we can convert the I-5 into a one-lane road and not bring traffic to a standstill. But we can begin engaging in creative solutions and we can leverage strategies that we know work — such as robust primary care systems and an emphasis on preventive care — rather than defaulting to the way we have always done it. More is not always more.
So, the next time you’re languishing in a packed ED, or stuck in seemingly endless traffic, it’s worth pausing to consider the principle of induced demand.
Mary C. Meyer, MD, MPH, is a frontline emergency medicine physician and the former director of emergency management for the Permanente Medical Group, Kaiser Permanente Northern California. This perspective is the author’s alone and does not necessarily reflect that of any institutions or companies with which she is affiliated.
Please enable JavaScript to view the