Disablement in the Physically Active Scale Short Form-8: psychometric evaluation

Abstract

Background: Patient-centered care and evidence-based practice (EBP) are core competencies for health care professionals. The importance of EBP has led to an increase in research involving clinical outcomes; current recommendations emphasize collecting patient focused measures, thus increasing the need for psychometrically sound patient reported outcome measures (PROMs) of health. Disablement has been identified as a valuable multi-dimensional construct for patient care. The Disablement in the Physically Active Scale Short Form-8 (DPA SF-8) has been proposed as a tool to be used in the physically active population that assesses a physical summary component of health and a quality of life component however, further analysis is necessary to ensure the instrument is psychometrically sound.

Methods: Confirmatory factor analyses (CFAs) were conducted on the DPA SF-8 at each time point to ensure factor structure. Reliability of the scale and internal consistency of the subscales were assessed, and a minimal detectable change (MDC) calculated. Additionally, a minimal clinically important difference (MCID) was also established, and invariance testing across three time points and groups was conducted.

Results: The CFAs at all three visits exceeded recommended model fit indices. The interclass correlation coefficient value (.924) calculated indicated excellent scale reliability and Cronbach's alpha for subscales PHY and QOL were within recommend values. The MDC value calculated was 5.83 and the MCID for persistent injuries were 2 points and for acute injuries, 3 points. The DPA SF-8 was invariant across time and across subgroups.

Conclusions: The DPA SF-8 met CFA recommendations and criteria for multi-group and longitudinal invariance testing, which indicates the scale may be used to assess for differences between the groups or across time. Our overall analysis indicates the DPA SF-8 is a valid, reliable, and responsive instrument to assess patient improvement in the physically active population.

Keywords: Confirmatory factor analysis; Disablement; Patient reported outcome measures; Quality of life.

Fig. 1

Scale structure of the Disablement in the Physically Active Scale

Fig. 2

Scale structure of the Disablement in the Physically Active Scale Short Form-8

Fig. 3

Confirmatory factor analysis of the Disablement in the Physically Active Scale Short Form-8 visit 1. Chi Sq = Chi Square (χ2), CMIN/DF = the χ2 / degrees of freedom ratio; CFI = Comparative Fit Index; TLI = Tucker-Lewis Index; IFI = Bollen’s Incremental Fit Index; RMSEA = Root Mean Square Error of Approximation, df = degrees of freedom, p = alpha level

Fig. 4

Confirmatory factor analysis of the Disablement in the Physically Active Scale Short Form-8 visit 2. Chi Sq = Chi Square (χ2), CMIN/DF = the χ2/degrees of freedom ratio; CFI = Comparative Fit Index; TLI = Tucker-Lewis Index; IFI = Bollen’s Incremental Fit Index; RMSEA = Root Mean Square Error of Approximation, df = degrees of freedom, p = alpha level

Fig. 5

Confirmatory factor analysis of the Disablement in the Physically Active Scale Short Form-8 visit 3. Chi Sq = Chi Square (χ2), CMIN/DF = the χ2 / degrees of freedom ratio; CFI = Comparative Fit Index; TLI = Tucker-Lewis Index; IFI = Bollen’s Incremental Fit Index; RMSEA = Root Mean Square Error of Approximation, df = degrees of freedom, p = alpha level

Fig. 6

Receiver Operating Curve for Individuals with Persistent Injuries, Visit Two. AUC = area under the curve, MCID = minimal clinically important difference

Fig. 7

Receiver Operating Curve for Individuals with Persistent Injuries, Visit Three. AUC = area under the curve, MCID = minimal clinically important difference

Fig. 8

Receiver Operating Curve for Individuals with Acute and Subacute Injuries, Visit Two. AUC = area under the curve, MCID = minimal clinically important difference

Fig. 9

Receiver Operating Curve for Individuals with Acute and Subacute Injuries, Visit Three. AUC = area under the curve, MCID = minimal clinically important difference