Doctor of Philosophy (PhD) in Physical Therapy
All rights reserved. This publication is intended for use solely by faculty, students, and staff of Nova Southeastern University. No part of this publication may be reproduced, distributed, or transmitted in any form or by any means, now known or later developed, including but not limited to photocopying, recording, or other electronic or mechanical methods, without the prior written permission of the author or the publisher.
College of Health Care Sciences - Physical Therapy Department
M. Samuel Cheng
Publication Date / Copyright Date
Nova Southeastern University
Michelle Lynn Peterson. 2020. Comprehensive Physical Function Measure for the Intensive Care Unit: The use of Rasch Analysis and Item Response Theory. Doctoral dissertation. Nova Southeastern University. Retrieved from NSUWorks, College of Health Care Sciences - Physical Therapy Department. (194)
Introduction: This dissertation study provides an in-depth examination of current intensive care unit (ICU) physical function assessment measures and synthesizes these into one comprehensive measurement tool that addresses multiple areas of function.
Background: A recent systematic review identified 26 measures being used in ICU clinical research, although a subsequent systematic review revealed only 14 of those physical function assessment measures have psychometric properties evaluated specifically for the ICU setting. A robust physical function assessment measure for use in the ICU, allows for determination of efficacy of treatment, comparison of studies across settings, and broader interpretation of results.
Specific Aims: Aim 1: Identify physical-function measures currently utilized in the ICU that have been psychometrically tested.
Aim 2: Analyze all measure constructs to determine redundancies and appropriateness for use in the ICU setting according to Rasch analysisand item response theory.
Aim 3: Create a comprehensive, robust functional measurement tool for use with patients in the intensive care unit.
Methodology: Rasch analysis was used for individual activity task evaluation, ranking of task difficulty, and removal of duplicate tasks. IRT Rasch analysisincluded: item fit, hierarchy, reliability, dimensionality, DIF, and probability. Receiver operating characteristics curve was conducted for predictive validity.
Results: Fifteen items out of a total of 53 met the requirements for an optimal rating scale. The items were ranked according to difficulty and there was no misfit. The reliability indexes were 5.13, α = .96 and 21.52, α = 1.00 for person and item scores respectively confirming scale hierarchy. The Cronbach Alpha (KR-20) person raw score "test" reliability was 0.96 with SEM 2.72. DIF was deemed non-significant and the probability curves were well delineated and ordered. The comprehensive physical function measure was found to have predictive validity for discharge to home with an optimal ICU admission cut-off score of 42 raw & 51 equal-interval, (sensitivity 71.7%, specificity 67.7%), and an optimal ICU discharge cut-off score of 54 raw & 61 equal-interval, (sensitivity 81.6%, specificity 82.0%).
Summary: A robust, reliable, and valid 15-item comprehensive physical function measure for use in the ICU was developed through Rasch analysisand item response theory.
Intensive Care Unit, Item response theory, Outcome measures, Rasch analysis, Physical function, ICU patients