Development of a novel computerised version of the Month Backwards Test: A comparison of performance in hospitalised elderly patients and final year medical students

• We examined performance on verbal and computerised versions of the months backward test in medical students and elderly medical inpatients.
• Performance on the two versions of the months backward test was similar but with slower completion times and more errors on the computerised version.
• The computerised version of the months backward test had higher correlation with the MoCA, especially the visuospatial domain.
• The computerised version of months backward test can provide rapid and reliable testing of cognitive function in clinical populations.

Background and aimsThe Months Backwards Test (MBT) is a commonly used bedside test of cognitive function, but there is uncertainty as to optimal testing procedures. We examined performance among hospitalised elderly patients and cognitively intact young persons with verbal and computerised versions of the test.Participants and methodsFifty acute elderly medical inpatients and fifty final year medical students completed verbal (MBTv) and computerised (MBTc) versions of the MBT and the Montreal Cognitive Assessment (MoCA). Completion time and errors were compared.ResultsThirty four participants scored <26 on the MoCA indicating significant cognitive impairment. The mean MoCA scores in the elderly medical group (23.6±3.4; range 13–28) were significantly lower than for the medical students (29.2±0.6; range 28–30: p<0.01). For the verbal months backwards test (MBTv), there were significantly more errors and longer completion times in the elderly medical patients (25.1±20.9 vs. 10.5±4.5; p<0.05). Completion times were 2–3 times longer for the MBTc compared to the MBTv (patients: 63.5±43.9 vs. students 20.3±4.4; p<0.05). There was high correlation between the two versions of the MBT (r=0.84) and also between the MBTc and the MoCA (r=0.85). The MBTc had higher correlation with visuospatial function (MBTc r=0.70, MBTv r=0.57). An MBTc cut-off time of 30 s for distinguishing performance (pass/fail) had excellent sensitivity (100%) with modest specificity (44%) for cognitive impairment in elderly medical patients.ConclusionThe computerised MBT allows accurate and efficient testing of attention and general cognition in clinical populations.