July 1, 2005

Wait Control

A queuing theorist calls for a 21st-century approach to hospital management.

Trained primarily as a queuing theorist, I was drawn to health care because it is riddled with delays. Almost all of us have waited for days or weeks to get an appointment with a physician or schedule a procedure, and upon arrival we wait some more until being seen. In hospitals, it is not unusual to find patients waiting for beds in hallways, and delays for surgery or diagnostic tests are common.

Causes of Poor Performance

Why are there so many examples of poor performance in an industry with per capita expenditures of about $5,000 a year? Consider the biggest source of care and expenditures: hospitals. Until recently, hospitals were paid on a fee-for-service basis, prices were controlled by the states and there was little or no competition. Since most care is paid for by insurance, customers exert little or no pressure on costs. Quality ratings are based primarily on the status of affiliated physicians, creating a culture that is physician-oriented rather than patient-oriented. Though hospitals are beginning to focus more on efficiency, service and quality, they are still often managed using intuition and rules of thumb rather than service standards, data and decision-support systems.

Emergency Department Overcrowding

A lot of the problems prevalent in hospital care show up in emergency departments (EDs). A survey of EDs in the United States in 2002 found that more than 90 percent of directors cited overcrowding as a problem, with almost 40 percent reporting overcrowding on a daily basis. When EDs experience significant overcrowding, they often go on diversion, redirecting ambulances to other EDs. A nationwide survey of hospitals in 2002 found that nearly 13 percent of urban hospitals were on diversion more than 20 percent of the time. However troubling on the surface, these reports are even more ominous given the current environment of terrorist threats. So what can be done to improve this situation?

Emergency department directors report that patients back up primarily because no inpatient beds are available. Yet, there has been a widely held perception in the health care community that there are too many hospital beds. This belief is primarily due to the discrepancy between what has historically been considered the optimal hospital occupancy figure of 85 percent and the actual average occupancy rate for nonprofit hospitals, which recently has been about 66 percent. Largely because of this perception, the number of hospital beds has decreased almost 25 percent in the last 20 years.

Inappropriate Measures and Dangerous Delays

My research has shown that determining bed capacity based on occupancy levels can result in very long waiting times for beds. This is of particular concern for intensive care units (ICUs), which are used for the most critically ill patients.

Using data from 1997 on all ICU units in New York State, I found that the average occupancy level was 75 percent, which might indicate to some that there was an excess of beds in these units. But given the critical condition of a patient who needs an ICU bed, there should be enough beds to insure that the probability of having one available when it is needed is very high. Based on this criterion, my findings indicated that between 74 percent and 95 percent of hospitals in New York State did not have enough ICU beds.

This apparent shortage was consistent with national reports that the longest ED delays are for patients waiting for a critical care bed. Yet there was a 20 percent decrease in ICU capacity between 1995 and 2001!.

Financial Pressures and Misconceptions

Why have hospitals cut beds, particularly ICU beds, thereby subjecting patients whose needs are urgent to potentially dangerous delays? There are many likely reasons. Since most hospitals don’t collect or analyze data on patients’ delays, they may be unaware that the policy of high bed use results in long delays for beds. Nursing shortages and prospective payment systems that create disincentives to treat certain patients also may play a role. But probably the biggest reason is that hospital managers are under unrelenting financial pressure and believe that this means they must operate clinical units, particularly expensive ones like ICUs, at high occupancy levels.

Ironically, this approach could be costing the hospital money. How? I have found that hospitals with an insufficient number of beds in one type of unit (e.g., ICU and cardiology) may have more beds than needed in another unit. This misallocation often results in patients being placed in the “wrong” bed temporarily and then transferred, frequently resulting in a longer hospital stay. And each additional day costs the hospital money since most “payers,“ including Medicare and many insurers, base compensation on a fixed, diagnosis-based fee schedule regardless of the length of stay. Also, backups in the ED ultimately lead to ambulance diversions, resulting in loss of potential patients and their associated revenues.

Is There a Doctor in the House?

The most critical component of ED delay may be the initial wait to be seen by a physician. Yet in many hospitals, patients spend hours waiting before they see a doctor, and many leave without being seen (LWBS). Studies have shown that LWBS patients are indeed sick, and one study estimated that 46 percent of LWBS patients require immediate medical attention.

Though these delays may be due to understaffing, my research demonstrates that another important factor is the staffing pattern. ED arrivals are very unpredictable, and the volume changes dramatically over the course of the day and the week. Hospital managers, while aware of the variability, usually don’t collect and analyze demand patterns and instead allocate staff based on general perceptions and intuition. As a result, they don’t do a good job of matching staffing levels to arrivals.

The value of basing staffing on data and decision-support tools was demonstrated in a study I conducted in collaboration with the directors of a New York City hospital ED. Using a queuing model I developed, we were able to reduce delays and decrease the number of LWBS patients by almost 20 percent without adding any physicians. This result was even more impressive given that the arrival volume had increased by 7.3 percent over the study period.

Patients at Risk

No one likes to wait, but do delays endanger patients? To explore this issue, my colleagues and I examined the connection between ambulance diversions and mortality from myocardial infarction (heart attack) in New York City.

We collected data on all ambulance diversions and all deaths from myocardial infarction in New York City for 1999 and 2000. During the study period, on average three hospitals a day citywide were on critical adult diversion status, with each diverting ambulance admissions for approximately five hours. The number and length of diversions were significantly greater in the winter, during the week and in the evening hours. Even worse, hospitals tended to be on diversion at the same times, resulting in increased ambulance travel times for patients and increased unavailability to pick up new patients.

We found that high levels of diversion can be deadly. Using multinomial regressions, our analysis showed that on days when more than 20 percent of a borough’s total ED hours were spent on diversion, fatalities from myocardial infarction increased 46 percent boroughwide; and when at least 25 percent of a borough’s hospitals were on diversion simultaneously, these deaths increased 16.5 percent.

My research in emergency responsiveness leads me to believe that most hospitals and communities are not well equipped to handle everyday emergencies. Yet hospitals must also handle demand surges resulting from natural occurrences, such as HIV/AIDS, SARS and flu, and, in our current environment, from terrorist attacks.

Though hospital administrators and policymakers are becoming increasingly aware of the need for significant changes in the ways that hospitals are managed and financed, the threats raised by terrorism require us to rethink fundamentally how capacity decisions are made and resources are managed. Incentives and education are necessary to bring hospital management into the 21st century and to develop coordinated care across hospitals and other health care facilities to more effectively deal with patient demand, whether it be the everyday or the extraordinary.

To read more about Professor Green‘s research, visit Columbia Ideas at Work.

Make a Gift

Bring your generosity full circle with a gift to Columbia Business School.


Post a Class Note

Looking for an easy way to keep in touch with your network? You're a click away.

Post a Note

Pan-Asia Reunion 2014

Celebrate Columbia Business School’s global alumni network and thought leadership at the 2014 Pan-Asian Reunion in Singapore, featuring a two-day symposium on emerging business trends with discussions led by preeminent faculty members and alumni leaders.

Visit Website >


jQuery(document).ready(function() { if(GetParameterValues('item') && GetParameterValues('item') !='') { var show1 = GetParameterValues('item'); jQuery('#fullProfile li').hide(); jQuery('#fullProfile li#' + show1).show(); } function GetParameterValues(param) { var url = window.location.href.slice(window.location.href.indexOf('?') + 1).split('&'); for (var i = 0; i < url.length; i++) { var urlparam = url[i].split('='); if (urlparam[0] == param) { return urlparam[1]; } } } });