Increasing the sensitivity of the Story Retell Procedure for the discrimination of normal elderly subjects from persons with aphasia

Background: Clinicians have long recognised the need for assessing language production at multiple levels of complexity and at impairment, participation, and activity levels. Methods for the elicitation of connected spoken language have taken many forms, typically selected with a balance between validly sampling linguistic performance, and reliability and economy of the sampling and scoring procedures. A Story Retell Procedure (SRP) has been proposed as a preferred method for achieving valid, reliable, and economic assessment of connected language (Doyle et al., 2000), and an information unit (IU) metric has been developed for validly and economically capturing important linguistic aspects of the retelling (McNeil, Doyle, Fossett, Park & Goda, 2001). Aims: In keeping with the goal of making assessment procedures as efficient and economic as possible, a study was undertaken to investigate the refinement of the IU metric for increasing the sensitivity of the SRP as an instrument for the detection of connected paragraph-level language production deficits in persons with aphasia. This metric involved the calculation of the percentage of lUs (%IU) produced relative to the time taken to produce them (%IU/Min). Methods & Procedures: A total of 15 persons with aphasia, and 31 normal control individuals without a communication disorder served as participants for this study. Subjects heard, and immediately retold each of 12 stories originally taken from the Discourse Comprehension Test (Brookshire & Nicholas, 1997). The retellings were scored using the procedures outlined by McNeil et al. (2001) with the addition of the %IU calculated over the time of the retelling. Comparisons between subject groups and groups stratified by age, among SRP forms, between scoring methods (%IU vs %IU/Min.), and group misclassification by scoring method were made. Outcome & Results: Application of the %IU/Min with the SRP yielded equivalence among alternate forms as evidenced by non-significant differences and high correlation coefficients among the SRP forms for persons with aphasia. The %IU/Min also decreased the percentage of misclassified aphasic and normal individuals compared to the %IU measure. Older normal subjects were misclassified as aphasic with greater frequency compared to the younger normal subjects. Conclusions: The %IU/Min is a more sensitive metric than the %IU in differentiating individuals with aphasia from older normal controls.