Delirium Screening Instruments Administered by Nurses for Hospitalized Patients - Literature Review

Central European Journal of Nursing & Midwifery

Ševčíková, B., Kubešová, H. M., et al. (2019).

Central European Journal of Nursing & Midwifery, 10(4), 1167-1178.

<div>This review explores the psychometric properties of various delirium screening tools when used for hospitalized adult patients. While this review only includes studies with tools administered by nurses, speech-language pathologists may also administer these tools as part of a multidisciplinary team.&nbsp;</div>

University of Ostrava (Czech Republic)

2008 to 2018

<div>Published studies reporting psychometric properties</div>

24

<div>Diagnostic accuracy varied widely for delirium screening tools when administered by nurses to hospitalized adults. The best predictive value was reported for the Nursing Delirium Screening Scale (Nu-DESC) followed by the Delirium Observation Screening Scale (DoS). The 4 A's Test (4AT) had the highest reliability.</div> <div>&nbsp;</div> <div>Specific psychometric values for the included screening tools were as follows (sorted by the highest reported sensitivity value):</div> <div> <ul> <li><strong>the Nu-DESC:&nbsp;</strong>sensitivity = 25%-100%, specificity = 83%-100%, Positive Predictive Value (PPV) = 5.4%-60%, Negative Predictive Value (NPV) = 78.4%-82.4%, area under the receiver operating characteristic (ROC) = 0.76-0.99, inter-rater reliability = 0.47-0.94 (based on nine studies);</li> <li><strong>the DOS: </strong>sensitivity = 62.2%-97%, specificity = 89%-98.4%, PPV = 53%-95.8%, NPV = 81.8%-99%, ROC = 0.91, inter-rater reliability = 0.73 (based on three studies);</li> <li><strong>the Neelon-Champagne Confusion Scale (NEECHAM): </strong>sensitivity = 100%, specificity = 91%, PPV = 26.3%, NPV = 73.7%, inter-rater reliability = 0.87 (based on two studies);</li> <li><strong>the 4AT:</strong>&nbsp;sensitivity = 83.3%-90%, specificity = 80%-86.3%, PPV = 66.7%-87%, NPV = 80%-94%, ROC = 0.91-0.92, inter-rater reliability = 0.80-0.99 (based on three studies);</li> <li><strong>The Sour Seven Questionnaire:&nbsp;</strong>sensitivity = 89.5%, specificity = 90.0%, PPV = 89.5%, NPV = 90.0%, ROC = 0.92 (based on one study);</li> <li><strong>the modified Richmond Agitation and Sedation Scale (mRAAS):&nbsp;</strong>sensitivity = 85%, specificity = 92%, inter-rater reliability = 0.48 (based on one study);</li> <li><strong>the Delirium Elderly At Risk Instrument (DEAR):&nbsp;</strong>sensitivity = 93.2%, specificity = 41.8%, PPV = 53.4%, NPV = 89.4% (based on one study);</li> <li><strong>the Delirium Detection Scale (DDS):&nbsp;</strong>sensitivity = 71.2%, specificity = 87.1%, ROC = 0.88, inter-rater reliability = 0.77 (based on one study);</li> <li><strong>the Simple Query for Easy Evaluation of Consciousness (SQeeC):&nbsp;</strong>sensitivity = 83%, specificity = 81%, PPV = 39%, NPV = 97% (based on one study); and</li> <li><strong>the Single Question in Delirium (SQiD):</strong>&nbsp;sensitivity = 77%, specificity = 51%, PPV = 42%, NPV = 83% (based on one study).</li> </ul> <div>Limitations to this review include a small number of studies investigating each tool, heterogeneity between studies (i.e., the use of different cut-off scores, variable clinical settings), and a lack of investigation into the quality of each included study. Additional research is indicated.</div> </div>