82.
RASCH ANALYSIS TO CONFIRM THE PSYCHOMETRIC
PROPERTIES OF A NEW PATIENT-RELATED OUTCOME SCORE
FOR ANTERIOR CRUCIATE LIGAMENT DEFICIENCY
Jonathan Comins, Michael Krogsgaard, Svend Kreiner, John Brodersen
Section for Sportstraumatology M51, Bispebjerg University Hospital, and The
Research Unit and Sectio University of Copenhagen ; Section for
Sportstraumatology M51, Bispebjerg University Hospital , University of
Copenhagen ; Department of Biostatistics, Institute of Public Health,
University of Copenhagen ; Research Unit and Section of General Practice,
Department of Public Health, University of Copenhagen
Background:
Patient reported outcome (PRO) questionnaires are increasingly
used to measure treatment effects in orthopedic surgery. Confirmation of
content coverage and relevancy using patient interviews is essential to ensure
content validity. Construct validity can be assessed using Modern Test Theory
and Rasch models. No instruments used to assess outcome in patients with
ACL deficiency have been developed using patient interviews and Rasch
analysis.
Purpose / Aim of Study:
The objective of this study was to investigate the
psychometric properties of the KNEES-ACL, a newly developed 55-item PRO
for patients with deficiency of the ACL, using Rasch analysis.
Materials and Methods:
The psychometric properties of the KNEES-ACL
were assessed using the Partial Credit Rasch Model and confirmed using
Graphical Log-linear Rasch Modeling. Unidimensionality of the reported
domains, local repsonse dependency, and differential item functioning (DIF)
was scrutinized across three groups of patients: group I were pre ACL
reconstruction; group II were 4 to 16 months after isolated ACL-reconstructive
surgery; and group III were at least 28 months after isolated ACL-
reconstruction. Other covariates for DIF analyses were gender and age group.
Findings / Results:
The pilot questionnaire was completed by 242 patients:
group I (n=62); group II (n=87); and group III (n=93). Forty-one of the 55
items distributed in seven scales fit the Rasch model and exhibited
unidimensionality. The Rasch scales were ADL, Psychosocial Consequences,
Looseness, Slackness, Sports Behavior, and Sports Physical Performance.
Thus, the pilot questionnaire was modified based on these results.
Conclusions:
The KNEES-ACL is the first PRO instrument specifically
constructed and validated for patients with ACL-deficiency using Rasch
analysis.
