"Playing God": The Unwieldy and Unfortunate Positions of Professional Homelessness Workforce Using VI-Spdat

Abstract

Background and Purpose

The Vulnerability Index Service Prioritization Decision Assistance Tool (VI-SPDAT) is widely used within U.S. homeless services systems to assess mortality risk and prioritize individuals for limited housing resources based on their resultant score. Communities nationwide have voiced concerns about the tool's efficacy. An emergent body of research on the VI-SPDAT has identified issues of algorithmic bias that impact housing outcomes for minoritized populations. However, there is limited research on VI-SPDAT implementation and how service providers experience using this tool and negotiate the limitations of an algorithmic prioritization system to aid the most vulnerable. This study is a sub-study of the Coordinated Entry System Triage Tool Research and Refinement (CESTTRR) study conducted from 2020-2023 in Los Angeles. This paper examines accounts of the workforce administering the VI-SPDAT and explores opportunities for addressing the challenges presented by the current prioritization system.

Methods

One-on-one, semi-structured interviews were conducted via Zoom in June and July 2021 with twenty service provider staff responsible for implementing the VI-SPDAT in Los Angeles County, CA. Participants were recruited through a list of contacts generated by the local Continuum of Care, a countywide entity responsible for coordinating a region's response to homelessness, and a Community Advisory Board affiliated with the CESTTRR study. Participants represented a range of community-based organizations serving various subpopulations and included frontline staff and management. Interviews solicited participants' experiences administering the VI-SPDAT and recommendations for best practices. Interviews were automatically transcribed by Zoom and thematically coded using ATLAS.ti. Researchers adopted an inductive approach to thematic analysis and engaged in an open coding process, calibrating codes through memo-writing and co-coding.

Findings

Findings indicate that service providers exercise considerable discretion in implementing the VI-SPDAT. Providers described grappling with decisions concerning where, when, and who to assess, how to communicate the tool's purpose, whether to disclose scores to clients and how to ensure scores reflected the reality of clients' vulnerabilities. Participants also relayed the complex traumatic effects of tool administration—both on themselves and clients, and often situated these effects within broader socio-environmental and cultural contexts. Staff expressed concern when working with subpopulations particularly those who are disproportionally represented in homelessness due to histories of structural racism. Suggestions to make prioritization tools like the VI-SPDAT more equitable and ways to mitigate their traumatic impacts were also shared.

Implications and Conclusions

Findings suggest that service providers are often negatively impacted by the experience of using assessment tools in the context of their work with people experiencing homelessness. Interviews uncovered many challenges posed by implementing algorithmic prioritization systems to address homelessness and point toward potential solutions for mitigating the unintended consequences of tools like the VI-SPDAT. These findings suggest that as new tools are developed and implemented, engaging in more community-driven processes and integrating trauma-informed practices to support both clients and tool administrators may be helpful. This could promote more equity and consequential community inclusion within current homeless services systems.