Prior research indicates that cognitive impairment is common in patients with MS and that worsening of cognitive function greatly impacts a patient’s daily living and quality of life. Therefore, assessment of cognition as a marker for disease progression and disease activity may lead to more timely and targeted treatment interventions. However, despite the importance of measuring cognition and the existence of a large variety of cognitive measures, monitoring of cognition is not yet part of standard care for patients with MS. Furthermore, digital monitoring tools for cognition are not yet employed in clinical practice due to various challenges, including the lack of standardization.

Understanding these challenges, my colleagues and I hypothesized that smartphone-based assessment of information processing speed in patients with MS in an everyday environment could better reflect real-life cognitive functioning than periodic momentary neuropsychological assessment performed in a clinical setting. While assessment of cognition via the use of wearable devices, such as tablets and smartphones, has been previously researched, no prior studies investigated metrics regarding optimal frequency or time of day of assessment. More importantly, to the best of our knowledge, all previous reports of smartphone-based Symbol Digit Modalities Test (SDMT) applications were performed on a standard and/or preconfigured smartphone provided by the study, whereas our study was performed using participants’ own smartphones.

Morning & Evening Estimates Both Reliable

For a paper published in Multiple Sclerosis Journal, we investigated a smartphone-adapted cognitive test for reliability and validity to assess information-processing speed, the most frequently affected cognitive domain in MS. At baseline, clinical assessments for patients with MS and healthy controls (HCs) included the Expanded Disability Status Scale (EDSS). During a 28-day follow-up, participants used the MS Sherpa® app (Orikami) to perform the smartphone-adapted SDMT (sSDMT) every 3 days on their own smartphone. With the SDMT, the number of correct pairings of symbols and digits by the patient during 90 seconds were scored, reflecting information-processing speed. We then analyzed the reliability of repeated smartphone measurements and the validity of the sSDMT on score differences across different groups and on correlations to information-processing speed as measured with the clinical SDMT.

Our findings indicate high reliability estimates for the sSDMT. Measurements repeated in short time intervals and reliability estimates were similar, whether the smartphone SDMT was assessed during the morning or evening. In addition, reliability was high whether single measurements were used or averages of multiple measurements. For validity, sSDMT scores were highest in the HC group, followed by patients with MS who were cognitively preserved (MS-CP), according to a cutoff on the clinical SDMT score, and lowest in patients with MS who were cognitively impaired (MS-CI). Receiver operating characteristics curve analyses, ie, plotting the sensitivity of 1-specificity for the differentiation between groups using the smartphone SDMT, shows curves statistically significantly different from 0.50, even for differentiation between MS-CP patients and HCs (Figure).  Lastly, smartphone SDMT scores were strongly correlated with the clinical SDMT scores and moderately correlated with the EDSS.

Cognition Reliably Assessed by Patients

Our study illustrates that cognitive function can be reliably self-assessed by patients with MS, remotely and more frequently in their own environment using their personal smartphone. Although the clinical SDMT was assessed orally and the sSDMT is performed by tapping on the device, no significant impact on the relation between the smartphone and clinical SDMT scores was found when adjusting for demographic covariates, arm function, EDSS, or smartphone screen size. This indicates that the relation between clinical and sSDMT is not confounded by these investigated covariates.

For future research, the study team would like to see further examination of clinical properties, for instance, the sensitivity of the instrument to clinically meaningful change and the ecological validity. Furthermore, the relation between smartphone SDMT measurements over time, disease activity (MS relapse and MRI activity) and disease progression (MRI atrophy measures) is also highly relevant for the translation of this digital biomarker into clinical practice.

Author