Leveraging Artificial Intelligence to Improve Clinical Appropriateness of Inpatient Designation in a Utilization...

Area of Science:

• Healthcare Management
• Clinical Informatics
• Nursing Practice

Background:

• Inappropriate use of observation services impacts hospital reimbursement.
• Existing criteria for patient placement are restrictive and overlook pre-existing conditions.
• Improved utilization management (UM) is needed for accurate observation vs. inpatient decisions.

Purpose of the Study:

• To evaluate an artificial intelligence (AI) tool's impact on observation service rates.
• To assess if AI enhances identification of comorbidities and medical necessity for inpatient appropriateness.
• To determine if AI improves decision-making in a large academic health system's UM registered nurse (RN) department.

Main Methods:

• A pre- and postimplementation study design was used.
• Compared observation vs. inpatient discharge volumes and conversion rates.
• Utilized an AI tool to assist UM RNs in patient care level assessment.

Main Results:

• Observation service discharge rates decreased from 16.69% to 12.75% monthly average postimplementation.
• UM RNs utilized the AI Care Level Score to guide provider discussions.
• AI implementation facilitated better patient placement decisions.

Conclusions:

• AI tools can effectively reduce observation service discharge rates in UM.
• Improved identification of comorbidities and medical necessity enhances decision-making.
• AI supports nurses in advocating for appropriate inpatient admission.