Initial Validity and Reliability of the Banff Patella Instability Instrument

Abstract

Background:

Patellofemoral instability is a well-recognized problem, but there are currently no published patient-reported quality of life outcome measures that are disease specific for the treatment of this population.

Purpose:

To establish the content validity, initial construct validity, and initial reliability of the Banff Patella Instability Instrument (BPII).

Study Design:

Cohort study (diagnosis); Level of evidence, 2.

Methods:

The content of the BPII was validated using a modified 3-stage Ebel procedure and analysis of floor and ceiling effects. As a measure of internal consistency, the Cronbach α was utilized to assess how reliably the 32 items of the questionnaire measured a similar construct. Test-retest reliability of the BPII was calculated using an intraclass correlation coefficient (ICC). Construct validity was evaluated on 150 questionnaires completed by patients with a confirmed diagnosis of patellofemoral instability. A one-way between-group analysis of variance was employed to determine if the BPII was able to differentiate between patients presenting at the initial orthopaedic consultation relative to patients presenting at 6 months and 12 months postoperatively.

Results:

Content validity was clearly established as each item in the BPII achieved a minimum of 83.3% agreement (range, 83.3%-100%) for relevance among the expert panelists. The average agreement was 96.9%; 24 items achieved 100% agreement. There was no evidence of floor or ceiling effects. Reliability (internal consistency) of the BPII was established at the initial orthopaedic consultation (α = .91), 6 months postoperatively (α = .97), and 12 months postoperatively (α = .97). Test-retest analysis resulted in an ICC of 0.98 between tests. Construct validity was established as there was a statistically significant difference in BPII scores at the initial orthopaedic consultation and 6-month and 12-month postoperative appointments (F_2,146 = 75.62; P < .001).

Conclusion:

The BPII demonstrates content validity, strong initial reliability, and a statistically significant level of construct validity in patients with patellofemoral instability. This population includes patients with recurrent patellofemoral instability as well as surgically stabilized patients.