[Interobserver agreement in using a checklist for the safe administration of enteral nutrition]

Stella Marys Rigatti Silva,Ana Paula Almeida Corrêa,Michelle Dornelles Santarém,Michelli Cristina Silva de Assis,Mariur Gomes Beghetto

DOI: https://doi.org/10.20960/nh.03578

2021-10-13

Abstract:Objective: to describe an evaluation of interobserver agreement in the use of a checklist related to the use of nasoenteral tube (NSS) as a presupposition for quality in obtaining data. Method: a methodological study conducted in 2018 in a Brazilian hospital, preceding the data collection of an open-label clinical trial. Independent observers, blinded to the evaluation of their peers, evaluated patients with NSS through a 25-item checklist. The data collected by eight previously trained research assistants (RA) were compared to those obtained by an experienced nurse (reference standard). Agreement was measured using the kappa coefficient and PABAK. Results: a total of 451 observations were made in pairs. Considering the total items on the checklist there was almost perfect agreement (k > 0.80) in all observation pairs (nurse vs. each RA): RA 1 (k = 0.91; 95 % CI = 0.89-0.93); RA 2 (k = 0.83; 95 % CI = 0.80-0.85); RA 3 (k = 0.92; 95 % CI = 0.90-0.94); RA 4 (k = 0.83; 95 % CI = 0.80-0.86); RA 5 (k = 0.94; 95 % CI = 0.92-0.96); RA 6 (k = 0.94; 95 % CI = 0.92-0.96); RA 7 (k = 0.96; 95 % CI = 0.95-0.98); RA 8 (k = 0.73; 95 % CI = 0.70-0.77). However, for isolated items, and in specific RAs, there were fair agreements, unacceptable to effectively collect data from a clinical trial. Retraining and supervision of RAs were able to improve agreement between observers. Conclusion: an evaluation of interobserver agreement proved to be fundamental to ensure the reliability of data collection and, therefore, to avoid measurement biases.