Evidências de validade da escala de avaliação da segurança do paciente renal crônico em sessão de hemodiálise

Abstract

The scale construction and content and semantics validation were performed in a previous study, in which it was completed with 57 items, divided into six dimensions. It was methodological research conducted with the following steps: content validation and internal structure validation (pilot test, construct validation, temporal stability, equivalence, internal consistency and parameterization). For this purpose, the scale was applied to 240 chronic renal patients undergoing hemodialysis in three clinics in Fortaleza-CE, Brazil, data collection occurred from September 2019 to May 2021. Research Ethics Committee of the Federal University of Ceará approved the study according to CAAE: 03379918.9.0000.5054. At this stage of the research, the scale was evaluated by 11 researchers from safety, hemodialysis, and instrument validation, and then passed through the sieve of three researchers also from safety and hemodialysis. At this stage, 12 items were excluded, three were included, and 36 were rewritten, leaving the scale with 48 items. In the pilot test, 16 items were changed. In construct validation and mean and standard deviation analysis, 24 items were excluded. In the correlation matrix, five items were eliminated. In the verification of internal consistency, four items were deleted, obtaining a final alpha value of 0.78. When performing the matrix of principal components and Scree Plot, the model was obtained with 72.41% of total explained variation and three dimensions. Items 01, 04, 07, 11, 28, 40 formed factor 1, with an alpha of 0.93; items 06, 10, 14, 35, 37 formed factor 2, with an alpha of 0.86; and items 27, 29, 30, 42 formed factor 3, with an alpha of 0.49. Factor loadings ranged from 0.44 to 0.993. The dimensions (factors) were renamed: Factor 1 – Strategic procedures for error prevention; Factor 2 – Health management organizational processes; and Factor 3 – Care practices for infection control. In the test-retest, kappa coefficient ranged from 0.24 to 1.00, with four items scoring below the recommended value. In reproducibility, ten items had negative kappa coefficient, which ranged from -0.089 to 1.00. Regarding parameterization, the cutoff point defined was at score 34. It was concluded that the safety assessment scale for chronic renal patients undergoing hemodialysis treatment is valid and reliable for application to the target audience. In addition to offering scores capable of concluding whether the assessed institution has safety in its service, it was evidenced that the scale has some temporal stability, except in some items that are related to care procedures. As for reproducibility, it needs to undergo further analysis.