Selection of memory clinic patients for CSF biomarker assessment can be restricted to a quarter of cases by using computerized decision support, without compromising diagnostic accuracy

PLoS ONE

H. Rhodius-Meester, I. van Maurik, J. Koikkalainen, A. Tolonen, K. Frederiksen, S. Hasselbalch, H. Soininen, S-K. Herukka, A. Remes, C. Teunissen, F. Barkhof, Y. Pijnenburg, P. Scheltens, J. Lötjönen, W. van der Flier

Alzheimer’s disease diagnostic criteria allow the use of CSF biomarkers to support diagnosis, but tools are needed to help select patients for which CSF biomarkers have clinical utility. In this study, a computerized decision support system to select patients for CSF biomarkers was evaluated. Simulated CSF biomarkers, in combination with demographic, neuropsychology, and MRI data, in control participants, Alzheimer’s disease dementia, frontotemporal dementia, and vascular dementia were used to estimate the impact on the confidence in diagnosis. The computer-supported prediction restricted CSF testing to only 26% of cases, without compromising diagnostic accuracy.