To evaluate the effects of different testing environments and repeat testing using a computerised, automatically scored, self-administered MS cognitive battery: MSreactor (http://www.msreactor.com).Methods
MSreactor monitors processing speed, attention and working memory. Participants confirm a ball’s appearance (Simple Reaction Test, (SRT)), react to the ball’s colour (Choice Reaction Test (CRT)) and detect if consecutive cards are identical (One Back Test (OBT)). Participants were offered 6 months clinic, or home-based testing every 1 to 3 months. Linear mixed models were constructed where the dependent variable was reaction time outcomes (SRT, CHRT, OBK), and independent variables were location or the home test number/order. Fixed effects included: EDSS, time of day and sex.Results
Three-hundred and seventy four MS patients completed baseline clinic testing; 166 had repeated clinic testing at the time of extraction. 84.5% of the cohort agreed to home and clinic testing with repeat testing compliance of 80%. Reaction times for all tests were 3% faster (p<0.001) at first home compared with the first clinic test, with no difference detected between second clinic/home test. Reaction times for SRT, CRT and OBK was 5.7% to 6.8% slower in the morning (p<0.001) regardless of location. EDSS was associated with slower reaction times in the first clinic/home comparison only. Reaction times during repeat home testing increased by 1.5%–2% with each unit increase in EDSS. Analysis of repeat home testing (first 5 home tests) showed no practice effect on SRT and CHRT scores but OBK reaction times decreased slightly by 0.8% (p<0.001) for each additional home test.Conclusions
Our results demonstrate the reliability of MSreactor scores with minimal testing environment or practice related effects observed on SRT and CHRT reaction times. MSreactor provides a feasible testing platform for cognitive monitoring in MS and our results will assist in defining clinically relevant changes in cognitive monitoring scores over time.