CHICAGO -- Using a special care team and dedicated space can help , experts said here.
, clinical performance process director for hospital medicine and critical care at Banner Health, a 28-hospital chain spanning seven states, explained to an audience at the annual meeting of the Healthcare Information and Management Systems Society how Banner improved its process for putting patients in observation.
Banner's odyssey began in December 2012, when the hospital leadership approached Figueroa's department about addressing the large number of patients who stayed under observation for long periods of time -- sometimes for more than 24 hours -- before either being admitted or discharged.
Lots of Silos
"In the majority of cases, a decision on whether to discharge a patient from the hospital can be made in 24 hours; in rare cases it spans more than 48 hours," Figueroa said. However, getting the number of hours down to the least possible was not that simple. "There were many silos working in observation across the system -- lots of fragmentation and different approaches in care for observation patients."
Of the 28 hospitals in the system, one in particular -- Banner Baywood Medical Center in Mesa, Ariz. -- was taking an approach to observation patients that seemed to be working, explained Figueroa. Banner officials decided to implement Baywood's approach in all 28 hospitals -- and, as with most company-wide changes that Banner makes, they planned to do it all on the same day at exactly the same time in all facilities.
The Baywood approach that Banner officials chose involved a care team that was dedicated to observation patients, dedicated space for an observation unit, standardized criteria for placing patients in observation, a review of results every 4 hours, and standards for transition to an appropriate care setting. A "tracking board" was also implemented to monitor the progress of each patient as they progressed through the system.
The observation team included physicians available at all hours to see observation patients, as well as a case manager, a social worker, and an observation RN, Figueroa said.
Chest Pain, Syncope Most Common Diagnoses
When the Banner team started researching the issue, it found that the most common diagnoses for observation patients were chest pain and syncope, followed by transient ischemic attack; abdominal pain; nausea, vomiting, and diarrhea; headache; and atrial fibrillation. The team set as its goal to reduce the average length of stay under observation from 27 hours to 18 hours.
To design the implementation of the new plan, Banner held a 3-day event involving representatives from the information technology, business intelligence, case management, finance, nursing, processing engineering, providers, and ancillary services (laboratory, pharmacy, imaging) departments.
Banner also formed teams at each facility that included the facility's chief financial officer, chief medical officer, physician, lead, and ancillary services representatives, among others. Each facility team formed a workgroup to decide where the observation unit would be installed, how big it would be, and how to schedule nurses and other care team members for the unit, she noted.
Changing to the new observation model required a culture shift at many Banner facilities, said Figueroa. "Part of the culture at Banner Baywood was that the practice managed observation patients similar to a long-stay emergency department [visit], contributing to a shorter length of stay," she said. In addition, hospitalists at other facilities had to understand that "their day will be very different managing observation patients in the new model -- given that they'll be required to manage them whether in dedicated units or cohort areas -- multiple times during their shift."
To help standardize observation care for chest pain and syncope, the team created guidelines and a tool kit for the care team, explained , process engineer program director at Banner Health. The information included criteria for placing patients under observation care as well as an observation order set.
Materials for Patients
The team also developed procedures for working with patients and their families, she continued. Once the decision is made to place the patient under observation, the emergency department nurse provides a letter to the patient and their family giving them an overview of what observation status means and what they can expect; the letter says that the patient will probably go home within 24 hours and will need a follow-up visit with their cardiologist or primary care physician. The team also developed a brochure and a video containing the same information.
In the case of an observation patient with chest pain, a workup order is placed within 60 minutes, Bucco said. The nurse does the workup -- including an initial EKG -- and assesses what home care needs, if any, the patient has. The care team then discusses a plan of care for the patient, and during the next 4 to 8 hours, a second EKG is performed; the provider then determines if any other work needs to be done.
During hours 8-12, additional diagnostics are completed, and the provider can reach out to a cardiologist for a referral and to discuss a care plan. "By hours 12-16, [the team] is waiting for reassessment, results, and the patient's disposition," Bucco said. "If all goes well, the patient is discharged [during] hours 16-20."
The team follows each patient on a tracking board, with special icons to flag patients whose length of stay has gone over 18 hours or who are returning within 30 days of a hospital stay.
The new approach has had results, Bucco said; from January 2013 to December 2014, the proportion of observation patients with a stay of less than 24 hours increased by 32%. In addition, the percentage of observation patients with lengths of stay less than 18 hours rose from 26.4% to 29.9%.
There has also been a 10% decrease in provider consults and a 52% increase in observation care set utilization -- "truly a remarkable improvement" considering that before that there had not been any standardized care sets across the Banner system, she said.
In terms of costs, there was a 20.5% decrease in ancillary cost per case in the areas of laboratory, imaging, and pharmacy. Imaging costs in particular dropped 57.4%, "primarily due to the reduction of doing the stress tests in the observation patients," Bucco said, noting that stress tests are still included in the order sets so providers can perform them if they decide they are necessary.