Validation of the MISSCARE-BRASIL survey - A tool to assess missed nursing care 1

ABSTRACT Objective: to analyze the metric validity and reliability properties of the MISSCARE-BRASIL survey. Method: methodological research conducted by assessing construct validity and reliability via confirmatory factor analysis, known-groups validation, convergent construct validation, analysis of internal consistency and test-retest reliability. The sample consisted of 330 nursing professionals, of whom 86 participated in the retest phase. Results: of the 330 participants, 39.7% were aides, 33% technicians, 20.9% nurses, and 6.4% nurses with administrative roles. Confirmatory factorial analysis demonstrated that the Brazilian Portuguese version of the instrument is adequately adjusted to the dimensional structure the scale authors originally proposed. The correlation between “satisfaction with position/role” and “satisfaction with teamwork” and the survey’s missed care variables was moderate (Spearman’s coefficient =0.35; p<0.001). The results of the Student’s t-test indicated known-group validity. Professionals from closed units reported lower levels of missed care in comparison with the other units. The reliability showed a strong correlation, with the exception of “institutional management/leadership style” (intraclass correlation coefficient (ICC)=0.15; p=0.04). The internal consistency was adequate (Cronbach’s alpha was greater than 0.70). Conclusion: the MISSCARE-BRASIL was valid and reliable in the group studied. The application of the MISSCARE-BRASIL can contribute to identifying solutions for missed nursing care.

capable of investigating the extent and nature of this phenomenon is essential, making it possible to study staff perceptions of care omission and reasons for such omission.
To this end, the MISSCARE survey has been developed to measure missed nursing care and to analyze its causes. The tool was developed and validated in the United States with a sample of 1,098 nursing professionals and consists of items that comprise 24 basic nursing care elements in part A and 17 reasons for missed care in part B (4) .
In Brazil, the MISSCARE survey has already been culturally adapted and tested for internal consistency (5) .
A different study expanded the original instrument to include other items associated with missed nursing care and reasons for missed care specific to the Brazilian context (unpublished data). This instrument is called

MISSCARE-BRASIL.
The aim of this study was to validate the MISSCARE-BRASIL survey to enable investigations about the phenomenon in the Brazilian context.

Method
The study was methodological and crosssectional. The MISSCARE-BRASIL survey was validated by assessing psychometric properties related to its construct validity and reliability via confirmatory factor analysis, known-groups validation, convergent construct validation, analysis of internal consistency (Cronbach's alpha) and test-retest reliability.

MISSCARE-BRASIL is a self-completion survey.
After complying with the required ethical procedures, the questionnaire was given directly to participants in a brown envelope. The surveys were returned at a later time, and anonymity was ensured. However, a code was assigned to each participant, so that those selected to participate in the retest phase could be located. Retesting To test construct validity through known groups, mean scores for missed nursing care (per participant) were calculated and analyzed. This total score was the mean missed care score identified for each of the nursing care elements presented in part A. The hypothesis was that nursing professionals who worked in closed sectors presented higher levels of job satisfaction, were not planning to leave their positions, and would present fewer missed nursing care elements. Other items related to missed care were treated as dichotomous variables. Care elements were considered missed if the answers were "occasionally," "rarely" or "never." The Student's t-test for independent samples was used to compare the existence or lack of differences among group means or the study criteria. Furthermore, Cohen's d was used to classify the distance between means as small (d<0.20), medium (≥0.20 to <0.50) or large (≥0.50).
analyze the instrument's stability, the ICC was used; values below 4 indicated low reliability, between 0.4 and 0.74, moderate to good, and equal to or greater than 0.75 indicated excellent reliability. All inferential analyses were conducted based on a 5% significance level (α=0.05).

Results
Three hundred thirty nursing professionals  Table 1 shows the results of CFA, i.e., factor analysis for the MISSCARE-BRASIL survey.    Table 2.
Test-retest reliability of the MISSCARE-BRASIL survey was measured via factor stability and presented Reliability was also tested by assessing the internal consistency of part A and the five factors in part B.
Cronbach's alpha was greater than 0.70, an acceptable level of internal consistency.

Discussion
The aim of this methodological study was to perform