The patient’s physical status, the duration of surgery and the surgical specialty are all factors affecting the likelihood of unplanned hospital admission after ambulatory surgery.
“Ambulatory surgery patients with higher ASA [American Society of Anesthesiologists] physical status scores coming in for longer procedures are at greatest risk for needing an unplanned admission to the hospital,” said Caryn Hertz, MD, associate professor of anesthesiology at the University of North Carolina at Chapel Hill (UNCCH) School of Medicine. Dr. Hertz presented the research (poster 5), which was conducted with her colleagues Andrey Bortsov, MD, PhD, and Nathan Woody, at the 2016 joint meeting of the Society for Ambulatory Anesthesia and the ASA.
Ambulatory Care Increasing
UNCCH is a major academic tertiary care center, and its Department of Anesthesiology provides clinical services for more than 60,000 procedures each year. The majority of cases are conducted as ambulatory surgery, and that number is growing every year. Some of these cases require unplanned admissions to the hospital, which can be costly, inconvenient for the patient and disruptive to facility function.
“We wanted to see if we could use discrete population data available in Epic to predict clinical outcomes,” Dr. Hertz said. The research team extracted data from the Epic electronic medical record system for all patients undergoing surgery at UNCCH in October 2015. More than 4,400 procedures were performed at nine different locations. Of these, 54%—or 2,383 cases—were scheduled as ambulatory procedures.
The ambulatory surgical services covered a broad range of specialties: pain, cardiothoracic, dermatology, otolaryngology, family planning, gastroenterology, gynecology, neurology, ophthalmology, oral maxillofacial, orthopedics, pediatric dentistry and surgery, plastic surgery, pulmonary, reproductive endocrinology, surgical oncology, transplantation, trauma, urology and vascular surgery.
Patients ranged from 7 weeks to 95 years of age, with a mean of 47 years and a standard deviation of 22 years. For ASA physical status, 10% of patients were class I, 45% were class II, 40% were class III and 4% were class IV. Body mass index ranged from 10 to 68 kg/m2.
Chi-square and t-test statistics showed unplanned admissions were associated with higher ASA class, duration of surgery (mean of 40 vs. 100 minutes) and specific surgical services. The logistic regression model showed surgical services and ASA class to be good predictors of unplanned admissions.
“For example, a patient with a high ASA score who undergoes a two-hour procedure in urology has a pretty significant chance of being admitted,” Dr. Hertz said. “In this study, we didn’t focus on the ‘why’ but ‘who.’”
In future studies, the research team plans to better define the population of ambulatory surgery patients at risk for unplanned admissions, potentially looking at specific procedures or other discrete variables obtainable through electronic medical record systems.
“A large range of patients fall in the ASA III category, and it can be difficult to determine how to classify a patient,” she said. “If we could break this down more or find another way to classify patients, we may have a better analysis of data.”
In addition, Dr. Hertz would like to develop an algorithm to use preoperative data to predict the likelihood of postoperative admission. “This predictor could be used for better planning,” she said. “If we can figure out how many patients may be admitted related to certain variables, we can tell patients to be prepared for a potential hospital stay.”