Critical Care Plus: Artificial Variability Research Picks Up Momentum

Boston Area Studies Results are Encouraging

Artificial variability research from eugene Litvak, PhD, professor of health care and operations management, and colleagues at the Boston University School of Management has made considerable progress since we first reported in June 2001.

Litvak and fellow researcher Michael E. Long, MD, adjunct associate professor of health care and operations management at Boston University, recently completed a study funded by the Massachusetts Department of Public Health to find root causes and test hypotheses for emergency room diversion at two large hospitals in the northeast.

The pair tested three primary hypotheses for ED diversion: Patient overuse and an increase in patient ER flow; internal problems in the emergency room such as slow processing, too few cubicles, etc.; and back-up because patients can’t get into ICU or floor units.

Litvak and Long found that only the last of these correlated highly with times most surgical patients left operating rooms and post-operative care units floor beds, which effectively turned ED patients who would otherwise have been moved into ICUs or on to floors into "boarders" who blocked the patient flow. Litvak says that findings support his thesis that emergency department performance itself almost doesn’t play any role in ED patient diversion.

Long notes that he and Litvak have also been looking at other effects of this kind of variability in surgical caseloads throughout the hospital. "One study showed patients receive better care in ICUs, so there’s a quality of care issue here, too, Long says. "Because we can’t staff to the peaks anymore, we know that on some days everyone is going to be overworked, which is when many medical errors occur."

He adds that artificial variability adversely affects nursing morale and retention, thus exacerbating an already problematical nationwide shortage of nurses as it forces them into overtime.

Smoothing Surgical Admissions

Once the patient flow problem is identified as lack of internal beds, Litvak says, the question becomes why there are enough beds at some times but not others. Two forces, the emergency department and the operating rooms, compete for beds, Litvak says. While hospitals can’t control the number of ED patients, altering scheduled surgeries can smooth operating room flow and relieve pressure on the ED and ICU. By smoothing the scheduled surgical admissions hospitals can see what additional resources are needed, Litvak says, and only the uncontrollable unscheduled demand remains, making justifying funding for additional beds a much easier task.

Two facilities Litvak and Long hope to study next are Boston Medical Center and Holy Family Hospital in Methuen, MA. John B. Chessare, MD, pediatrician and chief medical officer at Boston Medical, says his facility has signed on to study its artificial variability and committed to making the changes the study suggests.

Chessare, a devotee of system dynamics, says that very few hospitals are managed scientifically. "ED diversion is an example of a supply big enough to handle the demand but inefficiently used," Chessare says. "Dr. Litvak’s tool is routinely used in every other industry and I think the take-home message here is that the dichotomy of physician and non-physician leaders is very silly."

Michael McManus, MD, anesthesiologist and intensivist and assistant director of Pediatric ICU at Boston Children’s Hospital, has done an article about Litvak’s work for the Journal Anesthesiology. McManus says he expected flows of scheduled surgical patients to be even and non-random and was surprised to find that his ICU arrival rates varied more than the emergency admissions did.

"We had never looked carefully at downstream impact of patient flows from the operating room," McManus says. "We thought more about maximizing OR utilization, but now we’re looking at how we can manage flow to improve capacity if the peaks of demand come in ways that can be slowed."

McManus adds that applying Litvak’s methodology can become more complicated as one gets into it because patient length of stay is also variable. "But even if we can just smooth out the way arrivals arrive, we think the system will work better," he says.

Greg Bird, RN, MS, vice president for nursing at Holy Family, became acquainted with Litvak’s work through a seminar at which Litvak was a breakout speaker. Bird got Litvak to do a presentation for Holy Family’s chiefs of surgery and anesthesia and selected nursing staff. He is currently seeking funding for Litvak to come to Holy Family to study hospital capacity, ED and OR volumes, the number of patients admitted from the ED, OR and physician’s offices, the number of beds, and the available average length of stay. Holy Family will then apply Litvak’s methodology to smooth its surgical schedule.

Clinic Joins Research Study

Jeff Doran, Senior VP of the Lahey Clinic in Burlington, MA, says Lahey’s ED diversion problem has become severely exacerbated over the last couple of years, creating crisis for patients and difficult conditions for staff. Lahey, a regional tertiary referral source for many of the community hospitals north and east of Boston, has an active surgical specialty volume, performing about 18,000 surgical procedures a year. When Doran met Litvak while looking for ways to solve the bed shortage he volunteered Lahey, a multi-specialty group practice with 450 physicians and an attached 269-bed acute care hospital, for Litvak’s study.

Doran assembled an internal support team and Litvak and two of his associates gathered reams of information. Though he says that the resulting report quantified and validated a lot of his intuitive beliefs, Doran credits most of patient flow improvements Lahey has since experienced to opening 19 additional beds, adding nursing staff, and converting med/surg beds into telemetry beds.

"You can’t oversimplify the solution to just leveling patient throughput," Doran notes. "You have to decide who has the biggest priority-the guy who’s waiting to have the tumor taken out or the ED waiting for a patient to arrive, which is where it gets difficult."

He suggests that substantially smoothing out the surgical schedule could find surgeons initiating certain types of surgical cases at hours when they should be sleeping instead of operating.

"People think my research is against surgeons, but it isn’t," Litvak says. "You can actually perform more surgeries when patient throughput increases, and surgeons benefit from that."

But Doran cautions that convincing surgeons to alter their scheduled operations can be tricky. Getting the entire surgical service together to clearly articulate the challenges and issues, what approaches might be and why it would be a good thing for the organization as well as for patients, is a good start because it ensures that everybody is being educated from the same page.

Then look at how surgical cases are scheduled and supported and see if you can make some tangible modifications. "We’re still in the education process on the surgical side," Doran says, which is why some of this data Dr. Litvak has been working on is going to be very useful."

Screen Shots for System Available

Boston University senior systems analyst Abbot Cooper is developing software based on the configuration and number of units for individual hospitals to use to determine their levels of artificial variability. Cooper cautions that the software model is highly sophisticated and requires data most hospitals don’t have readily available. It’s not as simple as plugging in a few numbers and letting it go, Cooper says, because the primary principal data concerns patient’s potential movements as opposed to their actual movements.

"Most hospitals will record patient arrival and departure times but not the length of time in which the patient is ready to leave," Cooper says. "That’s the data you need in order to accurately model the system to distinguish between patient demand and satisfaction of that demand. Hospitals would need to collect more data in order to leverage the software. Every hospital is different, Cooper notes. There’s no cookbook answer that will apply to everybody.

(For more information contact Eugene Litvak at (617) 358-1633; John Chessare at (617) 638-6723; Jeff Doran at (781) 744-3485; Abbot Cooper at (617) 353-8900; Michael McManus at (617) 355-6426; Greg Bird at (978) 687-0156 ext.2176; or Abbot Cooper (617) 353-8900.