Development and psychometric testing of the Bakas Caregiving Outcomes Scale

Background: Family caregivers of stroke survivors experience more depressio n, emotional problems, social inactivity, and general ill-health than nonca regiving individuals. While a number of instruments measure these variables in family caregivers, they are often too global, indicating the need for a situation-specific instrument measuring life changes resulting from provid ing care. Objectives: To develop and psychometrically test the Bakas Caregiving Outco mes Scale (BCOS) measuring changes in family caregiving outcomes in the str oke population. Method: 48 items were developed, with 27 being judged as I content valid by a panel of five experts. Psychometric testing with two convenience samples of family caregivers of stroke survivors (ns = 92,:104) was conducted usin g item analysis. Cronbach's alpha, factor analysis, and hierarchical multip le regression. Results: Item analysis in sample no. 1 resulted in a 12-item scale with alp ha = .90. The scale was further shortened to a 10-item scale in sample no. 2 with alpha =.77. Unidimensionality was supported by factor analysis in bo th samples. Using hierarchical multiple regression, 63% of the variance of the 12-item BCOS in sample no. I, and 45% of the variance of the 10-item BC OS in sample no. 2 was accounted for by the constructs in the conceptual mo del. Criterion-related validity was supported in both samples by significan t correlations with the LIFE-8 and a criterion variable with the same respo nse format as the BCOS. Significant BCOS correlations with he SF-36 Health Survey subscales in sample no. 2 provided further evidence of criterion-rel ated validity. Conclusion: The 10-item BCOS is a brief, easy to administer instrument that has evidence of reliability and validity in family caregivers of stroke su rvivors. The 10-item BCOS could, serve as a valuable measure in research, a s well as an assessment tool to identify family caregivers in need of inter vention.