Key Question: Is it possible to measure scientifically the accessibility of medical instrumentation?

Motivation. Engineering design goals are often represented by identifying certain indices that are to be maximized or minimized as part of the design process.  Our vision is that a tool, which we call MED-AUDIT (Medical Equipment Device Accessibility and Universal Design Information Tool), can provide effective accessibility metrics that allow the answer to the above question to be an affirmative "YES."  This research program studies possible multi-factorial indices, starting with targeted areas of medical instrumentation.  It will also perform appropriate reliability and validity studies on measures proposed here. If successful, it will become a key part of the design guidelines developed as part of D4.1.

Desired Outputs:

1. To create a MED-AUDIT tool that provides a protocol to examine and rate a medical instrument design to determine the degree of its accessibility.
2. To document the MED-AUDIT development steps content and construct validation. Internal validation steps include piloting its data alongside other RERC-AMI data acquisition approaches from MU-Lab.
3. To perform appropriate reliability and validity investigations to empirically study the capabilities of the MED-AUDIT measurement approach and publish these studies.
4. To seek supplemental sponsored research funding to comprehensively test the MED-AUDIT system.
5. Depending on findings from above, the resulting MED-AUDIT tool will define generic or constrained applications and be considered for integration into D4.1 (Design Guidelines for Medical Instrumentation).

Use of Results: Results of Research Program R2 will be used in Research Program R2 and in Development Programs D2, D3 and D4.

The Team:  This is a joint effort between the R2D2 Center of the University of Wisconsin-Milwaukee and the RERC's Medical Device Usability and Accessibillity Lab within the rehabilitation labs at Marquette University.

Process: Determining the accessibility of a device has multiple dimensions. Three key domains include: a) identifying procedures that provide the activity context for which the instrument is used (task analysis); b) identifying features of a medical instrument that might be accessible or not; and c) rating the degree of accessibility for each device feature and task to specified impairments. The team has been meeting regularly to tackle the challenge of creating taxonomy, structure and content for each of these domains.

Progress:

• Exploration of various programming options including fuzzy logic platforms and other development packages, with a decision made to begin by using an existing OTFACT software platform (for occupational therapy performance assessment) to pilot and explore MED-AUDIT frameworks.
• Development of techniques to replace the embedded OTFACT question and response sets with the MED-AUDIT specific items and elicitation prompts. The strategy leveraged the TTSS (Trichotomous Tailored Sub-branching Scoring) technique of OT FACT and applied its data collection protocol to pilot MED-AUDIT concepts.
• Development of a working version of MED-AUDIT in the OT FACT shell for RESNA 2004 allowed visitors to the RERC-AMI booth to electronically score an early draft of the MED-AUDIT. As depicted in the software screenshot shown below, the range of impairments addressed included blind, low vision, deaf, hard of hearing, impaired sensation, impaired upper, lower, trunk mobility, reading, cognition and behavior.
• A contract with Califex Software, Inc., whose head programmer worked on the original OTFACT programming team, has enabled updates to OT FACT to more fully support the MED-AUDIT question sets for dynamic electronic data collection.
• Created several taxonomy domains and defined taxonomic relationships.
• Developed drafts of Sections I (Task Analysis), II (Device Features), III (Impairments) for testing, and created dual taxonomic strategies with one based on having a user with expertise in disabillity and UD concepts (easily, flexibly, safely) integrate this expertise with task analysis, and a second requiring less user knowledge and using a more integrated and automated predictive concept.