The measurement of delirium: review of scales

This review describes the characteristics and evaluates the psychometric qualities (process of testing and the results) of thirteen delirium instruments. Delirium instruments differ in goal (diagnosis, screening symptoms severity), type of data on which the rating is based (observation, interview or test of patients), the rater qualities required, the number of items and the rating time needed. Most instruments are based on the Diagnostic Statistical Manual criteria and measure signs and symptoms as described by these criteria. Reliability of delirium instruments shows good to excellent results. Validity of the delirium instruments is overall fair to good. Differences exist, however, in the degree to which reliability and validity were tested and the quality of the testing procedures. Most instruments are not further developed and tested after the initial study. Conclusion of this review is that most delirium instruments show promising results but need further testing. Testing is needed in different samples and on a broader range of aspects with regard to reliability and validity. Much emphasis should be given to the procedures used in future studies. Ease of use is an aspect of testing that is so far not taken into account, however, is important for use of instruments in clinical practice. A minority of instruments can be seen as "ready to use" instruments meaning well tested in more than one sample with good results. For screening high-risk, elderly hospitalized patients, the NEECHAM Confusion Scale and the Delirium Observation Screening Scale are recommended. The Confusion Assessment Method is the best diagnostic tool and the Delirium Rating Scale shows best results in screening symptom severity. For ICU patients the CAM-ICU is recommended. The MDAS is well tested in cancer patients. Nurses, however, have not yet tested the DRS and MDAS in practice.