Hospitals must achieve a difficult balance when scheduling nursing shifts. Nurses often make the difference in life-and-death situations and play a central role in quality of patient care. Having too many nurses, however, causes budgeting problems, as nurses’ compensation makes up the largest component of hospital budgets — typically more than 50 percent of all costs.
When assigning nurses to clinical units, a hospital must consider numerous factors, including current patient level, anticipated admissions and discharges, patient severity levels, availability of support personnel and pressure to keep costs down. To complicate matters, hospitals are also confronted with a nationwide nurse shortage.
In the face of these challenges, many hospital administrators implement one of two widely used methods to determine staffing levels.
One method assigns nursing staff by estimating the hours of care each patient must receive from a nurse. The American Nursing Association has criticized this system for its simple quantification, which only takes into account the average time required to care for patients and doesn’t consider the varying needs of patients with different diagnoses, ages and multiple ailments.
The more commonly used method is ratio staffing, which dictates a minimum nurse-to-patient ratio. California mandates ratio staffing, requiring its hospitals to have one nurse staffed per six patients in medical-surgical units, but that rule doesn’t consider other crucial factors — for example, weekends, which are typically slower than weekdays. With fewer patients coming into the hospital on weekends and fewer procedures scheduled, each nurse could take on a larger load. Ratio staffing is also inflexible and doesn’t account for varying patient needs, varying lengths of stay and the differences among hospital units. Still, several other states are considering implementing similar mandates.
Professor Linda Green maintains that ratio staffing is too simplistic for such a complex situation. “The one-size-fits-all approach doesn’t make sense,” she says. “Under some circumstances, it results in understaffing, which isn’t good for patients. In other circumstances, it results in overstaffing, which costs the hospital money.”
One reason for California’s implementation of ratio staffing is that the healthcare field is confronting an ongoing nurse shortage. “Essentially, the law says if you can’t hire enough nurses, you have to admit fewer patients,” Green says. As a result, to maintain the mandated ratios, hospitals may have to cancel elective procedures, divert ambulances to other hospitals and wait longer to admit patients.
Green and doctoral candidate Natalie Yankovic developed a queuing model that will help hospitals to more accurately estimate effective nurse-staffing levels.
“I started to think about what would make sense as a methodology,” Green says, “and was excited about the prospect of measuring the demand and workload for nurses to see how nurse staffing was related to performance, in terms of responsiveness to patients and patient satisfaction.”
When there aren’t enough nurses in each unit, the quality of care suffers in several ways. Immediate attention to patients in need can be jeopardized. The emergency room can fill with patients waiting for beds, even though beds may be available, because there may not be a nurse available to admit patients. Conversely, patients who are ready to leave may have to wait for an available nurse to discharge them.
“A shortage of nurses has been shown by research to result in increased medical errors and deaths,” Green says. “Nurses are one of the primary mechanisms by which errors made by other people, like doctors and pharmacists, are caught and corrected.”
Green and Yankovic’s queuing model bases nurse staffing for a clinical unit on patients’ needs such as admissions, discharges, administration of drugs, preparation for procedures and responding to call buttons; the time it takes to respond to those needs; the number of beds; length of stay and patient movement in and out of the unit.
Obtaining electronic data and sources of information for all the variables was more difficult than Green expected. This may be one reason that queuing theory, which has been used for a century, hasn’t been previously applied to nurse staffing. (Green has used a queuing model to allocate doctor staffing in emergency rooms.)
“A major part of this two-year study, working in conjunction with the Hospital for Special Surgery in New York,” she says, “is trying to collect data and identifying electronic sources of information so that any hospital can replicate this model by plugging data into the equations.”
Green hopes that this will be a model by which all hospitals can easily and effectively staff nurses.
“This is going to be the first methodology,” she says, “that is actually based on nurse performance.”
Linda Green is the Armand G. Erpf Professor of the Modern Corporation in the Decision, Risk and Operations Division at Columbia Business School and a founder and codirector of the Columbia Alliance for Healthcare Management.
Professor Green earned her doctorate in Operations Research from Yale University. Her research, which has focused on the development and application of mathematical models of service systems, has resulted in dozens of publications in the premier technical journals such as Operations Research and Management Science as well as prominent healthcare journals such as Health Services Research, Inquiry and Academic...
Read the Research
Natalia Yankovic, Linda Green
"Identifying Good Nursing Levels: A Queuing Approach"