Chris J. Edgar

Cognitive dysfunction is a prevalent and impactful feature of Parkinson's disease (PD), with substantial heterogeneity in onset, severity, and progression, and a high lifetime risk of dementia. Despite its clinical importance, effective treatments and fit-for-purpose outcome measures remain limited. This article reports outcomes from a multi-stakeholder roundtable convened by The Michael J. Fox Foundation for Parkinson's Research, outlining priorities to advance patient-focused endpoints and improve clinical trial design to accelerate therapeutic development for cognitive dysfunction. In the context of the current understanding of cognitive impairment across the PD/dementia with Lewy bodies (DLB) spectrum and existing neuropsychological assessment approaches, recommendations are made regarding necessary literature reviews, including qualitative data and related conceptual models, disease models including pathophysiology and neuropsychological profile, and clinical outcome assessment (COA) reviews, including content validity, psychometrics, and fitness-for-purpose. Additionally, novel conceptual models of cognitive impairment across the normal, mild cognitive impairment (MCI) and dementia spectrum are needed and may require further qualitative evidence.

Background: Remote and decentralized approaches to clinical outcome assessments in clinical trials can reduce patient and trial burden, improve recruitment and retention, and lower barriers to trial participation. Remote assessment may be facilitated through smartphone‐based cognitive assessment, especially in Bring‐Your‐Own‐Device (BYOD) trials, and can also enable novel designs, for example, the use of high frequency ‘burst’ assessments. Application of such digital assessments in remote or unsupervised settings requires understanding of potential error from test administration and delivery platform (e.g., smartphones vs. computer).Method: A series of studies explored the modification of the well‐validated Cogstate Brief Battery (CBB) to include practice trials with dynamic feedback to support unsupervised cognitive assessment, and adaptations for smartphone delivery. Data were drawn from healthy populations in (1) the Healthy Brain Project, where middle‐aged adults (n = 1,594) completed unsupervised cognitive assessments on a computer, (2) a pilot study of young adults (n = 60) who completed both smartphone and computer administration, and (3) a large (n = 35,000) study of adults who completed unsupervised smartphone‐based cognitive assessments in a BYOD context.Result: Data from middle‐aged adults enrolled in the Healthy Brain Project indicated that adaptation of the CBB for unsupervised computer‐based assessment had high levels of acceptability (98% complete data) and usability (95%), with data meeting criteria for low error rates and ease of understanding). In young adults, rates of test completion were high and comparable (>98%), and performance accuracy (d’s 0.4) on smartphone than computer. Smartphone usability was high (e.g., >85% found text and button size to be ‘just right’), but there was not a strong preference for smartphone versus computer (56% preferred smartphone) and different issues of fatigue, distraction, and use of keyboard input were raised for the platforms. Similar findings of outcome consistency across platform were observed in the large BYOD sample (N>35,000).Conclusion: These results indicate that smartphone assessment has high acceptability, good reliability, and that accuracy of performance is equivalent between smartphone and computer versions. The data also raise important issues for the successful adaptation and modification of cognitive tasks for smartphone administration.

Background: The Cogstate Brief Battery (CBB) is a computerized cognitive test battery found to detect and confirm cognitive deficits related to Alzheimer’s Disease (AD). As the clinical AD and normative samples used to understand the sensitivity of the CBB to AD have been relatively small and varied in terms of selection criteria, it is important to examine the ability of all CBB measures, and potential composites, to discriminate adults with mild cognitive impairment (MCI) and dementia due to AD from carefully selected cognitively unimpaired (CU) adults.Method: All speed and accuracy measures from the CBB were examined and both theoretically and statistically derived composites were created, from a sample of 4969 CU adults and 185 adults who met clinical criteria for MCI (clinical dementia rating, CDR = 0.5) or dementia (CDR > 0.5) due to AD.Result: Individual CBB measures of learning and working memory showed high discriminability for AD‐related cognitive impairment for CDR 0.5 (AUCs ∼ 82‐.85), and CDR > 0.5 (AUCs ∼ .90‐.95). There was also high discrimination ability for theoretically derived CBB composite measures, particularly for the Learning and Working Memory (LWM) composite measure (CDR 0.5 AUC = .83, CDR > 0.5 AUC = .97). Various statistically derived linear composite measures showed discrimination abilities similar to the LWM composite.Conclusion: In older adults, the CBB is an effective instrument for objectively discriminating cognitive impairment due to MCI or AD‐dementia from unimpaired cognition, with the LWM composite being a near‐optimal linear CBB composite for that purpose.

Background: The International Shopping List test (ISLT) is a verbal word list learning test set within a realistic shopping list context, supporting ease of cultural adaptation, and addressing content validity and patient relevance. Improvements to and the wide adoption of natural language processing (NLP) software has enabled the possibility for self‐administered tests of verbal learning. Lila^TM a self‐administered, smartphone‐application (app) was developed, using the virtual assistant and NLP. We have previously reported acceptability and usability data from qualitative analysis of concept elicitation interviews. We now seek to provide preliminary validation data around its use.Method: A convenience sample of healthy young adults (n = 40), M(SD)age = 35 (16), were invited to install and complete one assessment of the Lila^TM app. A usability survey was also administered to explore participants’ experiences. The sample was predominantly female (70%) and of Asian ethnicity (62.5%). Data were collected in Australia, using an English word pool developed for the USA.Result: Acceptability was generally high with N = 38 completing immediate recall rounds and N = 32 completing the delayed recall round. Eight‐two percent found it extremely or somewhat easy to hear and understand what items were on the shopping list and 95% indicated that it was extremely or somewhat easy to understand what they had to do to complete the assessment. Scoring accuracy was 88% at the word item level, with reprocessing to address accent, context setting, and other NLP approaches to improve accuracy underway e.g., homophones biased to the grocery item.Conclusion: Whilst acceptability was high, the importance of localization of test stimuli was confirmed as well as the need for additional data processing options to account for accent. Continued optimization of data processing to score participant responses and word item pools to reduce ambiguity is expected to further improve accuracy e.g., accounting for regional accents and consideration of homophones and compound words. However, specific characteristics of speech including in those for whom English may be a second/foreign language will be an area of focus for future development.

Background: The Alzheimer’s Disease Assessment Scale‐Cognitive subscale (ADAS‐Cog) and Mini‐Mental State Examination (MMSE) are used as clinical outcome assessments (COAs) to investigate Mild Cognitive Impairment (MCI) and Alzheimer’s disease (AD). The Cogstate Brief Battery (CBB) is a digital assessment of psychomotor function, attention, visual learning, and working memory. Digital assessments may overcome limitations of traditional assessments but score conversions haven’t been undertaken.Method: Data were aggregated for baseline visits from the A4, ADNI, and AXON studies. Participants included healthy/cognitively normal (HC), MCI, pre‐clinical AD, and mild AD. An equipercentile linking routine for single‐group designs, using bootstrap‐resampled standard error (SE; variability of estimated CBB scores for each COA score) and bias (mean differences of resampled results from full model) metrics, was employed to fit score distributions. Reaction time (RT; log10 ms) and accuracy (proportion correct) for each of the CBB tests and two composite accuracy (LWM) and reaction time (ATT) scores were fit to ADAS‐Cog and MMSE totals.Result: Data for 8,878 participants were analyzed (HC = 4,021; PC = 773; MCI = 420; AD = 414; Unclassified/Mixed = 3,250). Scores ranged from 0‐39 (ADAS‐Cog13), 0‐27 (ADAS‐Cog11), and 9‐30 (MMSE). Estimates of CBB scores by COA scores showed expected associations between assessments. CBB composites for Accuracy and RT at MMSE > = 26 were estimated at approximately 0.7‐0.9 and 2.8‐2.6 respectively. SE in ADAS‐Cog and MMSE totals from CBB scores were larger for scores which reflect worse performance and were from smaller samples (accuracy: > 0.10; RT > 1.0). MMSE RT SE estimates were significantly smaller (all < 0.2) than ADAS‐Cog estimates. Results of both ADAS‐Cog versions were virtually identical.Conclusion: Tables for converting ADAS‐Cog and MMSE totals to CBB scores will be provided. Results can be reliably converted when performance is clinically better. Lack of samples in lower performance ranges presents an issue for reliable conversion. Demonstrating how traditional COAs compare to digital assessments can aid clinical interpretation and the understanding of limitations in each assessment. Using score equating tables may help establish a robust method for exploring associations and differences between assessments and scores.

The One Card Learning Test (OCL80) from the Cogstate Brief Battery—a digital cognitive test used both in-person and remotely in clinical trials and in healthcare contexts to inform health decisions—has shown high sensitivity to changes in memory in early Alzheimer's disease (AD). However, recent studies suggest that OCL sensitivity to memory impairment in symptomatic AD is not as strong as that for other standardized assessments of memory. This study aimed to improve the sensitivity of the OCL80 to AD-related memory impairment by reducing the test difficultly (i.e., OCL48). Experiment 1 showed performance in healthy adults improved on the OCL48 while the pattern separation operations that constrain performance on the OCL80 were retained. Experiment 2 showed repeated administration of the OCL48 at short retest intervals did not induce ceiling or practice effects. Experiment 3 showed that the sensitivity of the OCL48 to AD-related memory impairment (Glass's Δ = 3.11) was much greater than the sensitivity of the OCL80 (Glass's Δ = 1.94). Experiment 4 used data from a large group of cognitively normal older adults to calibrate performance scores between the OCL80 and OCL48 using equipercentile equating. Together these results showed the OCL48 to be a valid and reliable test of learning with greater sensitivity to memory impairment in AD than the OCL80.

Background: There is a need for feasible, scalable assessments to detect cognitive impairment and decline. The Cogstate Brief Battery (CBB) is validated for Alzheimer’s disease (AD) and in unsupervised and bring your own device contexts. The CBB has shown usability for self-completion in the home but has not been employed in this way in a multisite clinical trial in AD. Objective: The objective of the pilot was to evaluate feasibility of at-home, self-completion of the CBB in the Alzheimer’s Disease Neuroimaging Initiative (ADNI) over 24 months. Methods: The CBB was included as a pilot for cognitively normal (CN) and mild cognitive impairment (MCI) participants in ADNI-2, invited to take the assessment in-clinic, then at at-home over a period of 24 months follow-up. Data were analyzed to explore acceptability/usability, concordance of in-clinic and at-home assessment, and validity. Results: Data were collected for 104 participants (46 CN, 51 MCI, and 7 AD) who consented to provide CBB data. Subsequent analyses were performed for the CN and MCI groups only. Test completion rates were 100%for both the first in-clinic supervised and first at-home unsupervised assessments, with few repeat performances required. However, available follow-up data declined sharply over time. Good concordance was seen between in-clinic and at-home assessments, with non-significant and small effect size differences (Cohen’s d between -0.04 and 0.28) and generally moderate correlations (r = 0.42 to 0.73). Known groups validity was also supported (11/16 comparisons with Cohen’s d≥0.3). Conclusion: These data demonstrate the feasibility of use for the CBB for unsupervised at-home, testing, including MCI groups. Optimal approaches to the application of assessments to support compliance over time remain to be determined.