RESUMO
OBJECTIVES: The purpose of this study was to create, standardize, and validate a new instrument, named 4-DSD, and determine its diagnostic accuracy in the diagnosis of delirium in subjects with moderate to severe dementia. DESIGN: Multicenter cross-sectional observational study. SETTING AND PARTICIPANTS: Older patients consecutively admitted to acute and rehabilitation hospital wards. MEASURES: The DSM-5 was used as the reference standard delirium assessment. The presence and severity of dementia was defined using the AD8 and the Global Deterioration Scale (GDS). The 4-DSD is a 4-item tool that ranges from 0 to 12. Item 1 measures alertness, item 2 altered function, item 3 attention, and item 4 acute change or fluctuation in mental status. RESULTS: A total of 134 patients were included in the study. Most of the patients were enrolled in acute hospital wards (60%), with 40% in rehabilitation settings. A minority of the patients were categorized with moderate dementia, with a GDS score of 5 (4%). Most of the patients were in the moderate-severe stage with a GDS score ≤6 (77%); 19% were classed as severe, with a GDS score of 7. A 4-DSD cutoff score ≥5 had a sensitivity of 80% and specificity of 80% with a positive predictive value (PPV) of 67% and a negative predictive value (NPV) of 89%. In the subgroup with moderate-severe dementia (n = 108), the sensitivity and the specificity were 79% and 82%, respectively, with a PPV and NPV of 62% and 92%. In the subgroup with severe dementia (n = 26) the sensitivity was 82% and the specificity 56% with a PPV of 78% and a NPV of 63%. CONCLUSIONS AND IMPLICATIONS: The availability of a specific tool to detect delirium in patients with moderate-severe dementia has important clinical and research implications, allowing all health care providers to improve their ability to identify it.
