- A new study in JAMA Network Open provides a potential roadmap to hospitals that may be leery to shut down elective surgical procedures while trying to deal with a spike in coronavirus patients. The key is the use of predictive modeling in developing a clinical decision support tool. Although such tools are abundant in healthcare, few are used to determine how likely a patient is to use certain hospital resources.
- According to the study, such a tool is able to separate out elective surgical cases based on length of hospital stay, intensive care length of stay, whether or not a ventilator is required and discharge to a skilled nursing facility. A patient’s age is the biggest factor in most of those determinations, followed by the number of previous inpatient and outpatient visits made by those patients.
- “This work shows the importance of a learning healthcare environment in surgical care, using quantitative modeling to guide decision-making,” the study concluded. Along the same lines, Banner Health in Arizona has been using artificial intelligence to continue performing elective procedures during the pandemic.
As the COVID-19 pandemic hit the United States in March, many hospitals attempted to preserve their bed and intensive care unit capacity by curbing or suspending elective surgical procedures completely. This may have kept the hospitals from being overrun by COVID-19 patients, but it hamstrung its business by grinding virtually all other forms of non-emergency care to a halt. The financial fallout from this practice is still being felt by many hospitals.
The country is currently experiencing a third spike in COVID-19 cases, and hospitalizations will likely follow. Hospitals in some Midwestern states have already reported concerns about shortages and ICU capacity.
In an attempt to create a more financially stable operating environment in the midst of COVID-19, researchers at Duke University and a variety of its affiliates created an algorithm that could be used to decide just how much of a hospital’s resources could be used for elective procedures. They pored over the medical records of more than 42,000 patients who underwent non-emergency surgeries at Duke University Medical Center, Duke Regional Hospital and Duke Raleigh Hospital between January 1, 2017 and March 1. Duke University Medical Center is a tertiary hospital, while the other two are community facilities.
The researchers found that this cohort of Duke patients spent an average 2.3 days in the hospital. Of those patients, 15.2% wound up in the intensive care unit while 3.8% required ventilators. A total of 6.7% were discharged to a skilled nursing facility. The median age of the patients studied was 62, although they ranged in age from 49 to 71. The three most common surgeries performed were knee and hip replacements and shoulder reconstructions — all highly lucrative procedures for most hospitals.
Age, followed by prior hospitalizations, was the biggest predictor in a patient’s length of stay. The type of surgery they received was the most common predictor for going to the ICU. Age tended to best predict a length of stay in the ICU and whether a ventilator was needed. If the patient underwent a coronary artery bypass graft, they were much likely to be discharged to a skilled nursing facility.
Such data could suggest hospitals focusing on younger patients with less complicated medical histories during the COVID-19 pandemic could be a way to manage patient volumes without losing too much revenue.