Accurately predicting the length of an operation has benefits beyond simply informing patients and their families. Doing so means improved patient satisfaction and has the potential to prevent over- or under-utilization of operating rooms, which has significant implications on both resources and staff. Ultimately, that leads to improved quality of care for patients, improved well-being for surgical teams, and reduced costs for hospitals.
With that in mind, Mayo Clinic researchers have shown that by looking at patients’ specific traits they can more accurately predict the length of an operation.
“From an efficiency stand-point the current systems are often unreliable and contribute to costly overestimation and underestimation of surgery length,” says Cornelius Thiels, D.O., M.B.A., a Mayo Clinic resident and a lead author for the research.
Making the most of available resources means more patients can get the care they need more efficiently and with less overtime for hospitals.
“Patients do not want to wait weeks to have their surgery due to a backlog, but patients also do not want to wait all day for their surgery to start due to poorly planned operating room schedules,” says Dr. Thiels, who did the research as a fellow in the Mayo Clinic Robert D. and Patricia E. Kern Center for the Science of Health Care Delivery.
The research team published a pair of papers to show which factors can help predict surgery length. The team found male gender, obesity, younger age, abnormal liver function, and patients with other significant health problems are all possible predictors of longer surgeries for a common procedure: minimally invasive gallbladder removal. The model they used predicted the operation length more accurately, and it also better identified groups of patients who are more likely to have extremely short or long operations.
“Predicting these outliers is equally important, as outliers are the most costly to the system,” the researchers wrote. “Even small gains can have significant impact.”
The first paper, published recently in The American Journal of Surgery, used 24,099 cases in the American College of Surgeons National Surgical Quality Improvement Program database. The paper looked at patients who underwent minimally invasive gallbladder removal surgery to find which factors can help predict operative duration, and which aren’t as useful, including smoking, cardiovascular disease, chronic obstructive pulmonary disease (COPD), diabetes, and an abnormal white blood cell count. The second paper, published recently in the journal Surgical Endoscopy, confirmed the findings using 1,801 cases at Mayo Clinic from 2007 through June 2013, and 11,842 cases between 2005 and 2013 from the database. That paper also determined the time impact for each useful patient factor.
“Using the American College of Surgeons National Surgical Quality Improvement Program database, we were able to broaden our patient sample and strengthen our prediction model beyond what would be possible when limited to our internal data,” says Bethany Lowndes, Ph.D., a health sciences researcher in the Center for the Science of Health Care Delivery and a lead author for the research.
The research was done in collaboration between the Mayo Clinic Division of Subspecialty General Surgery, and the Surgical Outcomes and Health Care Systems Engineering programs in the Center for the Science of Health Care Delivery.
“This type of cross-divisional research has the greatest potential to be not only scientifically fruitful but also is more likely to result in translational research,” says Dr. Thiels.
An aging U.S. population and an increasing percentage of obese patients also complicate scheduling predictions, and make a model that accounts for those factors more useful.
Current scheduling systems are fairly inaccurate, and typically predict the average operation length, but don’t account for significantly longer or shorter cases, the researchers wrote. Other studies have looked at patient factors related to operative duration, but they haven’t been translated into a model that can be used in practice yet.
This research began with that in mind, and because it was initiated by clinicians looking to solve that problem, it should help with execution.
“When the research is complete, the desire to implement it into practice has already been fostered,” says Dr. Lowndes.
To improve their model’s accuracy, researchers suggest using other potential factors for this operation – which weren’t available in the database – including the presence of gallstones or previous abdominal scarring, surgery or tenderness. Researchers also will need to test and validate the model for other procedures.
“This will allow us to apply it to various surgical procedures and improve predictability,” says Dr. Lowndes.
At that point, the researchers hope it can be implemented into the surgical scheduling process and improve access for patients.
“Given that every hospital could benefit financially from improved operative time prediction, we feel this first step is important,” the researchers wrote.