Accuracy of Diagnostic Tests for Telemedicine Assessment of Dementia and Mild Cognitive Impairment
McCleery J, Laverty J, Quinn TJ Accuracy of diagnostic tests for telemedicine assessment of dementia and mild cognitive impairment. Cochrane Database of Systematic Reviews 2021, No. 7. Art. No.: CD013786. doi: 10.1002 /14651858.CD013786.pub2. Accessed July 27, 2021.
Commentary
This study was conducted to evaluate the diagnostic accuracy of telemedicine assessment of dementia and mild cognitive impairment. It was also intended to identify the quality and quantity of relevant research evidence and to identify and synthesize data on sources of heterogeneity in test accuracy data, patient or clinician satisfaction with telemedicine assessment models, resource use, cost, or feasibility.
As a result,
136 participants and three studies were included in the review.
Two studies (20 and 100 participants) were conducted in community settings in Australia, and one study (16 participants) was conducted in a veteran's home in the United States.
The participants were people with undiagnosed cognitive symptoms and self-screening tests that identified them as being at high risk for dementia.
For all-cause dementia diagnoses, the sensitivity of the telemedicine assessment ranged from 0.80 to 1.00 and the specificity ranged from 0.80 to 1.00.
This can be considered as evidence of very low certainty due to imprecision, discrepancies between studies, and risk of bias. for the diagnosis of MCI, only data from one study giving a sensitivity of 0.71 and specificity of 0.73 were available, while for the diagnosis of cognitive syndrome (dementia or MCI) For the diagnosis of cognitive syndrome (dementia or MCI), data from the same study yielded a sensitivity of 0.97 and a specificity of 0.22.
Since most of the diagnostic discrepancies are due to differences between MCI and dementia, they occur almost equally in both directions.
In addition,
there was insufficient data to assess the accuracy of dementia subtype diagnosis.
A small amount of data was provided in this study indicating a high level of clinician and patient satisfaction with the telemedicine model, but the evidence may be equivocal due to the lack of data on resource use, cost, or feasibility.
The authors concluded that only a few eligible studies with few participants were found, and the key difference between studies providing data for analysis was whether the target condition was dementia only (two studies) or dementia and MCI (one study).
The data suggest that telemedicine assessments may be very sensitive and specific for the diagnosis of dementia of all causes when assessed against the reference standard of traditional face-to-face assessments, but due to small sample sizes and heterogeneity across studies, the estimates are likely to be imprecise.
This may apply primarily to telemedicine models that incorporate a significant amount of face-to-face contact with non-physician health professionals responsible for diagnosis. For the diagnosis of MCI by telemedicine assessment, the best estimates of both sensitivity and specificity were somewhat low, but were based on a single study.
Errors occurred at the boundaries of cognitive impairment / MCI and MCI / dementia. However, they found no evidence that diagnostic discrepancies were more frequent than expected because of the known variability among clinicians' opinions in assigning a diagnosis of dementia.
