Autor(es): Oliveira-Kumakura, Ana Railka de Souza ; Caldeira, Sílvia ; Simão, Talita Prado ; Camargo-Figuera, Fábio Alberto ; Cruz, Diná de Almeida Lopes Monteiro da ; Carvalho, Emília Campos de
Data: 2018
Identificador Persistente: http://hdl.handle.net/10400.14/20933
Origem: Veritati - Repositório Institucional da Universidade Católica Portuguesa
Assunto(s): Nursing diagnosis; Psychometry; Validation studies
PURPOSE: To analyze the knowledge related to the use of the Rasch model in validation of nursing diagnoses. METHODS: Integrative literature review with search in LILACS, PUBMED, CINAHL, and SCOPUS. FINDINGS: Five studies comprised the sample, which analyzed unidimensionality, local independence, item calibration, item reliability, separation of items and people, and differential item functioning for analyzing nursing diagnoses. CONCLUSIONS: The Rasch model seems to be a useful method to validate nursing diagnoses and probably also for the validation of nursing outcomes in the Nursing Outcomes Classification. The use of this model is promising, considering the advan- tages that it can be used in studies with several methodological designs. IMPLICATIONS FOR NURSING: Methods that are able to provide more robust evidence of nursing diagnosis validity are needed to support highly accurate diag- nostic findings in clinical practice.
