Reduction in patient burdens with graphical computerized adaptive testing on the ADL scale: tool development and...

Area of Science:

• Rehabilitation Medicine
• Health Informatics
• Gerontology

Background:

• Stroke patients often undergo lengthy assessments for activities of daily living (ADL).
• Traditional non-adaptive testing (NAT) can be time-consuming and burdensome for patients and healthcare professionals.
• Computer adaptive testing (CAT) offers a potential solution to improve assessment efficiency.

Purpose of the Study:

• To verify the effectiveness and efficiency of CAT in reducing time and burden for patients, nurses, and occupational therapists.
• To compare the test length and ability estimation accuracy of CAT versus NAT for ADL assessments in stroke patients.

Main Methods:

• Developed a VBA-Excel CAT module using item banks from the Barthel Index (BI) and Frenchay Activities Index (FAI).
• Simulated CAT to compare average test length against NAT.
• Illustrated CAT's multimedia interface on a tablet PC for data collection and error detection.
• Demonstrated quality control using fit statistics to identify and reconfirm response errors.

Main Results:

• CAT significantly reduced the number of endorsed items compared to NAT (13.42 vs. 23 items), achieving 41.64% efficiency in test length.
• No significant differences were found in averaged ability estimations between CAT and NAT.

Conclusions:

• A programmed VBA-Excel CAT module, particularly via mobile nursing services at the bedside, can reduce patient burden and save time.
• CAT presents a more efficient alternative to traditional paper-and-pencil NAT for ADL assessments in stroke patients.