Development and Implementation of the Maryland Adult Protective Services Assessment Tool

Background:

Adult Protective Services (APS) in Maryland is mandated by state statute to serve adults ages 18 and older who may be vulnerable to abuse, exploitation, neglect, and self-neglect due to a range of functional, developmental, or intellectual disabilities and self-care limitations, including cognitive impairments, behavioral health conditions, physical illnesses or injuries, or frailty due to advanced age. Each year Maryland APS conducts more than 7,000 investigations, a number projected to increase as the population ages. Similar to many states across the country, Maryland APS faces challenges related to limited resources, increasing caseloads, inadequate staffing, and the absence of evidence-based practice. This poster will describe a collaboration between social work academic researchers and the Maryland Office of Adult Services to design and implement an evidence-informed risk assessment tool to guide APS investigations, a 2-year project funded through the Administration for Community Living’s Grants to Enhance State Adult Protective Services.

Methods:

During the first year of the project, we developed a pilot Maryland APS Assessment Tool based on: 1) an extensive review of the empirical and conceptual literature, 2) previous work by the state of Colorado in their own researcher-practitioner partnership to develop an assessment tool, and 3) input from an Advisory Group comprised of APS supervisors and investigators from across the state. During the second year, we piloted the tool in 8 jurisdictions in Maryland, using stratified random sampling to identify local departments in counties both rural and urban and representing different regions of the state. Over the 3-month pilot study, APS staff conducted 461 investigations. Supervisors and staff also participated in focus groups to discuss the feasibility and utility of the new assessment tool, and gave additional input on changes prior to statewide implementation.

Results:

Through this process, we developed the Maryland APS Assessment Tool, which examines 75 items across eight domains of potential risk: physical functioning, cognitive functioning, behavioral health, health conditions, physical environment, social environment, financial circumstances, and primary caregiver.  Response options include: no risk, minor risk, moderate risk, severe risk, and unable to assess. Both the pilot assessment data and focus groups indicate the new tool improves the ability of APS to make a determination of whether abuse, neglect, or exploitation has occurred and put interventions in place to reduce the risk of re-victimization. Modifications to the assessment tool and training were made based on risk factors with a higher proportion of ‘unable to assess’ responses (10-20%) and feedback from the focus groups.

Conclusions and Implications:

We will conclude with next steps following statewide implementation with version 2 of the assessment tool in 2019, including plans to test the internal consistency of domains, examine inter-rater reliability through case scenarios, assess concurrent validity via associations between risk factors and case determinations, and examine predictive validity via associations between risk factors and re-occurrence. We will discuss the challenges and benefits of designing an assessment tool that is evidence-informed and meets the needs of APS investigators in the field, and implications for similar efforts in other states.