Context: The Disablement in the Physically Active (DPA) scale is a patient-reported outcome instrument recommended for use in clinical practice and research. Analysis of the scale has indicated a need for further psychometric testing.
Objective: To assess the model fit of the original DPA scale using a larger and more diverse sample and explore the potential for a short-form (SF) version.
Design: Observational study.
Setting: Twenty-four clinical settings.
Patients or other participants: Responses were randomly split into 2 samples: sample 1 (n = 690: 353 males, 330 females, and 7 not reported; mean age = 23.1 ± 9.3 years, age range = 11-75 years) and sample 2 (n = 690: 351 males, 337 females, and 2 not reported; mean age = 22.9 ± 9.3 years, age range = 8-74 years). Participants were physically active individuals who were healthy or experiencing acute, subacute, or persistent musculoskeletal injury.
Main outcome measure(s): Confirmatory factor analysis was conducted to assess the factor structure of the original DPA scale. Exploratory factor, internal consistency, covariance modeling, correlational, and confirmatory factor analyses were conducted to assess potential DPA scale SFs.
Results: The subdimensions of the disablement construct were highly correlated (≥0.89). The fit indices for the DPA scale approached recommended levels, but the first-order correlational values and second-order path coefficients provided evidence for multicollinearity, suggesting that clear distinctions between the disablement subdimensions cannot be made. An 8-item, 2-dimensional solution and a 10-item, 3-dimensional solution were extracted to produce SF versions. The DPA SF-8 was highly correlated (r = 0.94, P ≤ .001, R2 = 0.88) with the DPA scale, and the fit indices exceeded all of the strictest recommendations. The DPA SF-10 was highly correlated (r = 0.97, P ≤ .001, R2 = 0.94) with the DPA scale, and its fit indices values also exceeded the strictest recommendations.
Conclusions: The DPA SF-8 and SF-10 are psychometrically sound alternatives to the DPA scale.
Keywords: covariance modeling; measurement.