Measurement tools used to assess unfinished nursing care: A systematic review of psychometric properties

Abstract

Background Different tools have been developed and validated to measure unfinished nursing care. However, no systematic review of the psychometric properties has been performed describing the quality of the methods used to estimate their validity. Aims (a) To identify tools measuring unfinished nursing care that have undergone validation processes; (b) to evaluate critically the quality of the methods used in ascertaining their psychometric properties; and (c) to compare the estimated psychometric properties of these tools. Design A systematic review of the psychometric properties also evaluating their methodological quality was performed by following the COnsensus-based Standards for the selection of health Measurement Instruments (COSMIN) guideline for systematic reviews. Data sources The databases Medline, the Cumulative Index to Nursing and Allied Health Literature, PubMed, Google and Google Scholar were searched up to 30 June 2018. Review methods Data extraction was performed following the predetermined eligibility criteria. Eight properties and their methodological quality were assessed using the COnsensus-based Standards for the selection of health Measurement INstruments criteria. Results A total of 20 studies reported validation data regarding three family of tools: the Missed Care Survey (MISSCARE), the Basal Extent of Rationing of Nursing Care (BERNCA) and the Task Undone scale (TU-7). The most estimated psychometric properties across studies were internal consistency, followed by structural validity, content validity, and cross-cultural validity. The less evaluated psychometric properties were reliability, hypothesis testing and convergent and criterion validity. Conclusion The psychometric properties of the investigated tools showed a more than acceptable quality, as did the methodologies used to estimate these properties. Impact The MISSCARE survey is the most widely used tool validated across countries to date. Evaluating the concurrent reliability of the tools available is strongly recommended to assess their effectiveness in measuring the same phenomenon.