The Rivermead Behavioural Memory Test (RBMT) outcome measures are standardized assessments of memory designed to quantify problems associated with episodic and prospective memory and orientation difficulties, as well to monitor change following therapeutic interventions (Quemada et al., 2003). The RBMT was originally composed of 12 subtests that represent typical everyday activities, such as remembering an appointment, a name, or a short route as well as picture or face recognition. During assessment the client receives a stimuli and must recall that stimuli or perform certain activities based on recall when asked by the examiner. The original RBMT uses a dichotomous scoring system in which the person either receives a pass or fail, thus more items passed indicates better memory performance. Although the RBMT has shown to be ecologically valid at detecting impairment, especially in low-functioning individuals, the simplified pass/fail scoring system is considered less sensitive to a spectrum of possible scores, which lead to the development of the RBMT-E (extended version) as well as the RBMT second and third editions, whose scoring is more perceptive of the subtler decrements of memory performance (Wills, Clare, Shiel, & Wilson, 2000). The latest version, RBMT-3 is a 10 item restructured edition that includes updated stimuli from the original, improved data and norms, as well as a new novel task subtest, which is a puzzle that has to be solved in a fixed order. It comprises six domains relative to verbal and nonverbal episodic memory, spatial memory, aspects of prospective memory, and procedural memory (Wester, Leenders, Egger, & Kessels, 2013; Wester, van Herten, Egger, & Kessels, et al. 2013). Administration time is less than 30 minutes and in addition to the raw scores on subtests, a Global Memory Index (GMI) can be computed for overall performance (Wester, van Herten, et al., 2013).
Original research by Quemada et al. (2003) found test-retest reliability of the original RBMT to be high at r = 0.96. Reliability was also good for the four parallel forms (the assessment when purchased comes with four separate but similar types) with a range of r = 0.65 to 0.85. With respect to discriminant ability, when comparing mild cognitive impairment and mild neurocognitive disorders of the Alzheimer’s type, it was determined that the profile score correctly classified 90.9% of subjects, while the screening score of the RBMT alone was able to correctly classify 81.8% of subjects (Adachi et al., 2013). Using the revised RBMT-3 in a comparison of healthy controls of those with alcoholism and Korsakoff’s syndrome (KS), which is a type of alcohol related cognitive impairment, showed through ROC analysis that the RBMT-3 GMI (total score GMI) was able to distinguish KS from those with alcoholism at AUC = 0.85 (at the 95% confidence interval [CI]), where AUC is considered the area under the curve (Wester, van Herten, et al., 2013). While at a cutoff score of < 67.5 the RBMT-3 had a sensitivity of 0.80 and a specificity of 0.69 in differentiating the 2 types of alcohol syndromes (Wester, van Herten, et al. 2013). The same study determined that the GMI score could also distinguish those with alcoholism from healthy controls as well at AUC = 0.83 (at the 95% CI) and a GMI cutoff score of 87.5 for this group had a sensitivity of 0.80 and a specificity of 0.62 (Wester, van Herten, et al. 2013). A similar study of 25 subjects with alcohol-related memory impairment and 25 healthy controls found that RBMT-3 scores resulted in less ceiling and floor effects in both groups when compared to the original RBMT (Wester, Leenders, et al. 2013).
Adapted from Wester, A. J., Leenders, P., Egger, J. M., & Kessels, R. C. (2013). Ceiling and floor effects on the Rivermead Behavioural Memory Test in patients with alcohol-related memory disorders and healthy participants. International Journal of Psychiatry in Clinical Practice, 17(4), 286-291.