.png){kind=logo}
Use patient intake data to predict the length of stay at hospitalization.
Annually, U.S. hospitals account for over 35.7 million stays, costing more than $415 billion (1). With an average stay of 4.6 days, there is a significant opportunity to reduce healthcare costs and improve patient outcomes by safely shortening the length of stay. Prolonged hospitalization increases the risk of hospital-acquired conditions, strains healthcare resources, and affects the hospital’s capacity to admit new patients. Efficiently managing and reducing unnecessary extended stays is crucial for enhancing operational efficiency and patient care in healthcare settings. Delays in discharge often lead to prolonged LOS and create clinical and operational burdens on providers. As long as patients continue to occupy beds while awaiting discharge, clinical personnel must attend to them, reducing the amount of time they can spend with other patients that may require more intensive care. This leads to greater scarcity of beds and delays operational processes, such as sanitizing rooms and medical equipment before subsequent use.
Further, extended LOS can increase risk for HACs in more vulnerable patients, and may also result in “access block”—a situation in which patients requiring admission are forced to wait for more than eight hours in the emergency department due to lack of available inpatient beds (2). Access block occurs for approximately 8% of patients and perpetuates extended LOS; it is associated with nearly a day of increased LOS on average. The impact of prolonged LOS on health outcomes is especially pronounced in the ICU setting and is associated with greater incidence of adverse events for vulnerable patients, such as older adults. Elderly ICU patients generally require more resource-intensive treatment, and roughly 55% that experience a prolonged LOS die within six months of discharge (3). These patients also incur approximately seven times the cost of patients that do not experience a prolonged LOS (3).
A predictive model using synthetic data, "StayOptim AI", has been developed to evaluate patients' LOS based on a multitude of clinical and demographic data. This predicts LOS more accurately, improving healthcare facilities' ability to optimize care coordination and improve resource management.
The synthetic data set simulating patient admissions was constructed using insights from medical studies and healthcare analytics. Resources such as the statistical report by Freeman et al. (1) provide an overview of variations in hospital stay, while Bashkin et al. (2) examine organizational factors that affect LOS, and Abd-Elrazek et al. (3) contribute predictions based on general admission characteristics. These enabled the selection of model variables and ensured realistic prediction capabilities.
‍
Annually, U.S. hospitals account for over 35.7 million stays, costing more than $415 billion (1). With an average stay of 4.6 days, there is a significant opportunity to reduce healthcare costs and improve patient outcomes by safely shortening the length of stay. Prolonged hospitalization increases the risk of hospital-acquired conditions, strains healthcare resources, and affects the hospital’s capacity to admit new patients. Efficiently managing and reducing unnecessary extended stays is crucial for enhancing operational efficiency and patient care in healthcare settings. Delays in discharge often lead to prolonged LOS and create clinical and operational burdens on providers. As long as patients continue to occupy beds while awaiting discharge, clinical personnel must attend to them, reducing the amount of time they can spend with other patients that may require more intensive care. This leads to greater scarcity of beds and delays operational processes, such as sanitizing rooms and medical equipment before subsequent use.
Further, extended LOS can increase risk for HACs in more vulnerable patients, and may also result in “access block”—a situation in which patients requiring admission are forced to wait for more than eight hours in the emergency department due to lack of available inpatient beds (2). Access block occurs for approximately 8% of patients and perpetuates extended LOS; it is associated with nearly a day of increased LOS on average. The impact of prolonged LOS on health outcomes is especially pronounced in the ICU setting and is associated with greater incidence of adverse events for vulnerable patients, such as older adults. Elderly ICU patients generally require more resource-intensive treatment, and roughly 55% that experience a prolonged LOS die within six months of discharge (3). These patients also incur approximately seven times the cost of patients that do not experience a prolonged LOS (3).
A predictive model using synthetic data, "StayOptim AI", has been developed to evaluate patients' LOS based on a multitude of clinical and demographic data. This predicts LOS more accurately, improving healthcare facilities' ability to optimize care coordination and improve resource management.
The synthetic data set simulating patient admissions was constructed using insights from medical studies and healthcare analytics. Resources such as the statistical report by Freeman et al. (1) provide an overview of variations in hospital stay, while Bashkin et al. (2) examine organizational factors that affect LOS, and Abd-Elrazek et al. (3) contribute predictions based on general admission characteristics. These enabled the selection of model variables and ensured realistic prediction capabilities.
‍
