Objectives: To determine the psychometric properties of the Sydney Psychosocial Reintegration Scale (SPRS), an instrument developed to quantify disability and handicap in persons with traumatic brain injury (TBI).
Design: Descriptive correlational study.
Setting: Brain Injury Rehabilitation Unit and Brain Injury Outpatient Clinic.
Participants: Two samples, a "subacute" group (n = 20) and a "long-term" group (n = 40), were studied to examine responsiveness (subacute group), reliability, and validity (long-term group) of the SPRS.
Main outcome measure: The SPRS is a 12-item questionnaire measuring three domains of everyday living commonly disrupted after severe TBI: occupational activities, interpersonal relationships, and independent living skills.
Procedure: Patients in the subacute group were rated with the SPRS by a clinician at admission to the rehabilitation unit and again three months later or at discharge from the unit (whichever occurred first). For individuals from the long-term group attending the outpatient clinic, a close relative was interviewed with the SPRS and other validating measures. The SPRS was readministered one month later.
Results: Internal consistency of the SPRS was high (alpha coefficient = .90), as was agreement between raters and stability over a one-month period (r(i) = .95 and .90, respectively). Reliability and stability coefficients for the three domains of the scale were also high, ranging from.86 to.94 for reliability and.77 to.93 for stability. Preliminary evidence for construct validity was established with a number of standard instruments, with evidence of both convergent and discriminant construct validity from the Sickness Impact Profile (SIP). The SPRS was sensitive to group differences on the Glasgow Outcome Scale (GOS) and to changes occurring during the period of active recovery.
Conclusions: The results suggest that the SPRS has sound psychometric properties, being a reliable, stable, sensitive, and valid instrument. It is potentially useful in both clinical and research settings.