An analysis of the diagnostic ability of graduating baccalaureate nursing students using computer clinical simulations

This exploratory, descriptive study informs curriculum and instructional decisions in nursing higher education. Benner's novice to expert paradigm and information processing theory were used to explore 23 graduating generic baccalaureate students' ability to develop nursing diagnoses (NDX). Subjects completed 2 clinical simulations (PMP-Ellis and PMP-Brown) via NEMAS$\sp\copyright$ software on IBM$\sp\copyright$ microcomputers. Selected subjects were interviewed following the simulations. Quantitative descriptive and qualitative theme analyses were conducted to answer the questions: (1) how do subjects vary in pertinent item marking and in categories of etiologic statements developed; (2) what percentage identifies the best, reasonable and not-reasonable nursing diagnostic conditions and defining characteristics (DC) which match the conditions; (3) to what extent do diagnostic statements follow the PES (Problem, Etiology, Signs & symptoms) format; (4) what differences exist in the NANDA Taxonomy I patterns of NDX identified; (5) what subject background characteristics were related to diagnostic performance; and (6) how do subjects describe and explain the bases for their decisions? Although NEMAS$\sp\copyright$ can restrict independent data gathering, its use allows any number of subjects to be evaluated on the same nursing diagnostic cognitive tasks. Ten DC were theoretically defined and expertly confirmed for each case. Subjects identified significantly ($t$ = 5.21) more items as pertinent in Ellis (COPD) than Brown (newly diagnosed hypertension.) For the 3 best Ellis NDX, subjects correctly matched: 41% (ineffective airway clearance), 34% (alteration in comfort), and 43% (ineffective breathing pattern) of the defining characteristics. Subjects correctly matched 60% of the DC in each best Brown NDX (anxiety: mild-moderate, activity intolerance, and ineffective individual coping). Difficulty indices calculated to further examine subject accuracy with defining characteristics revealed that: in the 3 Ellis NDX, 6/10, 7/10 and 6/10 DC were identified by fewer than half the subjects; in the Brown case the respective proportions were 5/10, 7/10 and 6/10. Among the study's recommendations are the need for research on: faculty consensus on NDX; faculty development in NDX language acquisition, hueristics and teaching; the use of difficulty indices as a specific means of evaluating student NDX ability; and the role of NEMAS in evaluating diagnostic ability.