Impact of Implementing a Real Time Frequent Admitter Risk Score (FAM-FACE-SG) on Readmission Rates

Overview

In an earlier study using electronic health records (EHR), the investigators have identified nine factors to be significantly associated with FA risk. These nine predictors include Furosemide intravenous 40 milligrams or more; Admissions in the past one year; Medifund status; Frequent emergency department use; Anti-depressants treatment in past one year; Charlson comorbidity index; End Stage Renal Failure on dialysis; Subsidized ward stay and Geriatric patient. The investigators have combined these nine predictors into the FAM-FACE-SG score for FA risk (defined as 3 or more inpatient admissions in the following 12 months). The FAM-FACE-SG risk score has the advantage of being deployed in our hospital's enterprise data repository known as Electronic Health Intelligence System or eHINTs for short, on a real-time or near real-time basis. On a daily basis, data from multiple data sources are extracted, transformed and loaded onto the eHINTS system. The system can be programmed to run every midnight to provide risk scores the following morning for patients admitted the previous day. In this trial, the intervention is to combine the FAM-FACE-SG risk score in addition to a decision making algorithm to guide referrals to various transitional care services based on needs assessment on nursing and function. The primary objective is to evaluate the impact of our intervention in improving healthcare utilization (hospital readmissions, emergency department (ED) attendances, length of stay up to 90 days post-discharge).

In an earlier study using electronic health records (EHR), The investigators have identified nine factors to be significantly associated with FA risk. These nine predictors include Furosemide intravenous 40 milligrams or more; Admissions in the past one year; Medifund status; Frequent emergency department use; Anti-depressants treatment in past one year; Charlson comorbidity index; End Stage Renal Failure on dialysis; Subsidized ward stay and Geriatric patient. The investigators have combined these nine predictors into the FAM-FACE-SG score for FA risk (defined as 3 or more inpatient admissions in the following 12 months). The FAM-FACE-SG risk score has the advantage of being deployed in our hospital's enterprise data repository known as Electronic Health Intelligence System or eHINTs for short, on a real-time or near real-time basis. On a daily basis, data from multiple data sources are extracted, transformed and loaded onto the eHINTS system. The system can be programmed to run every midnight to provide risk scores the following morning for patients admitted the previous day. In this trial, the intervention is to combine the FAM-FACE-SG risk score in addition to a decision making algorithm to guide referrals to various transitional care services based on needs assessment on nursing and function. The primary objective is to evaluate the impact of our intervention in improving healthcare utilization (hospital readmissions, emergency department (ED) attendances, length of stay up to 90 days post-discharge). The aims of this cluster RCT are to: (1) evaluate the impact of implementing the FAM-FACE-SG risk score in addition to a decision making algorithm to guide Patient Navigator (PN) referrals to various transitional care services based on needs assessment on nursing and function on improving healthcare utilization (hospital readmissions, emergency department (ED) attendances, length of stay up to 90 days post-discharge); (2) measure the implementation of the risk score (Fidelity of the PNs in adhering to the protocol in recruiting patients according the score priority; Referral rate of the PNs to various transitional care services; Qualitative feedback from PNs on the perceived benefits and behavior change after receiving the scores); (3) conduct an economic analysis of the cost-benefit of implementing the risk score.