Development and Validation of a Screening Protocol to Identify Elder Abuse in United States Emergency Departments

Statement of the Problem: Elder abuse is a common and morbid problem that is often unrecognized. The prevalence of elder abuse is particularly high among older adults receiving care in emergency departments (EDs). Unfortunately, the potential of EDs to identify elder abuse is not currently realized in part because no protocols to screen for elder abuse have been designed specifically for this setting. We propose to develop and validate an ED-specific protocol to screen for patients at high risk for elder abuse. The protocol will be easy to use by the emergency nurses who will administer it and will significantly improve upon available detection methods. A positive screen will prompt further assessment by a social worker or emergency provider to determine the need for referral to adult protective services.

Subjects and Partnerships: Study subjects will be individuals aged 65 years and older presenting to EDs located in Chapel Hill, North Carolina; Birmingham, Alabama; and Ypsilanti, Michigan.

Research Design and Methods: In Phase 1 (12 months), we will develop a protocol to screen for individuals at high risk for elder abuse. The initial draft of the protocol will be based on a review of existing protocols and input from national-experts in geriatric emergency medicine (Platts-Mills), geriatric social work (Zimmerman), dementia (Sloane), sexual assault (McLean), emergency nursing triage (Travers), ED-based screening (Haukoos), and psychometric assessment (Reeve). Emergency nurses will then administer this protocol to patients (N¡Ý50), and revisions will be made based on debriefing interviews with patients and nurses. In Phase 2 (6 months), we will conduct a pilot study (N=100) to finalize the methodology to be used in Phase 3. In Phase 3 (18 months), we will conduct a multi-center validation study (N=1800). All positive screens and a random sample of negative screens will receive a structured evaluation by a social worker. This social worker evaluation, the patient¡¯s medical records, and available reports from local Adult Protective Services will be reviewed by a panel of experts whose decision on the presence or absence of abuse will serve as a reference standard.

Analysis: The sample size for the validation study (Phase 3) has been calculated in order to obtain a lower bound confidence interval for the estimate of sensitivity of less than 5%. Standard methods will be used to calculate sensitivity and specificity.

Products, Reports, and Data Archiving: Expected deliverables will include peer-reviewed publications describing: 1) the derivation of the protocol and 2) estimates of the sensitivity and specificity of the protocol in identifying elder abuse. Data from all three phases will be made publicly available. Subsequent studies will assess effectiveness (including cost-effectiveness)and establish optimal methods for dissemination and implementation. Collectively, this work will substantially improve the identification of abuse among older adults in the United States.

Note: This project contains a research and/or development component, as defined in applicable law.

ca/ncf