Validation of memorial delirium assessment scale

• Published in: Journal of critical care Vol. 24; no. 4; pp. 530 - 534
• Main Authors: Shyamsundar, G., MBBS, Raghuthaman, G., MD, DPM, Rajkumar, Anto P., MD, DPM, DNB, Jacob, K.S., MD, MRCPsych, PhD
• Format: Journal Article
• Language: English
• Published: United States Elsevier Inc 01.12.2009; Elsevier Limited
• Subjects: Adult; Aged; Critical Care; Delirium; Delirium - diagnosis; Female; Hospitals; Humans; Male; Mental Disorders - diagnosis; Middle Aged; Mortality; Observer Variation; Psychiatric Status Rating Scales; Reliability and factor analysis; Reproducibility of Results; Sensitivity; Specificity; Studies; Validity; India; Delirium; Validity; Sensitivity; Specificity; Reliability and factor analysis
• ISSN: 0883-9441; 1557-8615
• DOI: 10.1016/j.jcrc.2008.12.016

Abstract Purpose Delirium is common but underdiagnosed in critical care units, which results in increased morbidity and mortality. This study aimed to evaluate the reliability and validity of diagnosing delirium with the Memorial Delirium Assessment Scale (MDAS) in India. Materials and Methods One hundred twenty consecutive patients admitted to medical and cardiac intensive care units of a tertiary care hospital, who were neither mute nor intubated, participated in this study. A trained junior resident screened for delirium using MDAS. A consultant psychiatrist established the reference diagnosis of delirium in accordance with International Classification of Diseases, 10th Revision diagnostic criteria for research. Results Interrater agreement and test-retest reliability of MDAS were 0.92 (95% confidence interval, 0.81-0.96) and 0.93 (95% confidence interval, 0.83-0.97), respectively. The MDAS had good internal consistency, with Cronbach α of .89 and Guttman split-half coefficient of 0.71. Factor analysis revealed a 2-factor structure, namely, cognitive disturbances and behavioral abnormalities. A receiver operating characteristic curve obtained the optimal threshold for screening as MDAS total score of at least 10. The sensitivity, specificity, positive predictive value, and negative predictive value were 100%, 95.45%, 75%, and 100%, respectively. Conclusions Our findings suggest that training junior residents to use a brief screening instrument to diagnose delirium is a reliable and valid option in resource-poor critical care settings.