Validation of the Dutch version of the quick mild cognitive impairment screen (Qmci-D)

Abstract

Background: Differentiating mild cognitive impairment (MCI) fromdementia is important, as treatment options differ. There are few short(<5 min) but accurate screening tools that discriminate between MCI,normal cognition (NC) and dementia, in the Dutch language. The QuickMild Cognitive Impairment (Qmci) screen is sensitive and specific indifferentiating MCI from NC and mild dementia. Given this, we adaptedthe Qmci for use in Dutch-language countries and validated the Dutchversion, the Qmci-D, against the Dutch translation of the StandardisedMini-Mental State Examination (SMMSE-D).Method: The Qmci was translated into Dutch with a combined qualitativeand quantitative approach. In all, 90 participants were recruited from ahospital geriatric clinic (25 with dementia, 30 with MCI, 35 with NC).The Qmci-D and SMMSE-D were administered sequentially but randomly bythe same trained rater, blind to the diagnosis.Results: The Qmci-D was more sensitive than the SMMSE-D indiscriminating MCI from dementia, with a significant difference in thearea under the curve (AUC), 0.73 compared to 0.60 (p = 0.024),respectively, and in discriminating dementia from NC, with an AUC of0.95 compared to 0.89 (p = 0.006). Both screening instrumentsdiscriminated MCI from NC with an AUC of 0.86 (Qmci-D) and 0.84(SMMSE-D).Conclusion: The Qmci-D shows similar,(good) accuracy as the SMMSE-D inseparating NC from MCI; greater,(albeit fair), accuracy differentiatingMCI from dementia, and significantly greater accuracy in separatingdementia from NC. Given its brevity and ease of administration, theQmci-D seems a useful cognitive screen in a Dutch population. Furtherstudy with a suitably powered sample against more sensitive screens isnow required.