Use of Telephone Interview for Cognitive Status (TICS) for the Detection of Dementia in Primary Care

Abdulrahman et al. assessed the utility of the Telephone Interview for Cognitive Status (TICS) in screening for dementia remotely in the Netherlands. In our current healthcare era, the importance of studying telemedicine modalities is crucial for shaping medicine after a pandemic. The authors reported TICS to be a valid tool and proposed utilizing a cutoff score of 29 to screen for dementia in the elderly. This yielded a sensitivity of 65% and specificity of 88% when compared to a diagnosis of dementia in the patients’ health records. The promise of this research is to save clinicians time and money while also increasing access to dementia screening for patients.

This study certainly has clinical implications, and it is a relatively low risk proposition to utilize telephone dementia as initial screening before an in-person visit. However, we do not believe these results support a recommendation of exclusively screening for dementia over the phone due to several threats to the study’s validity. Firstly, of the 1473 invited to participate there were only 810 who consented. A 55% enrollment rate is standard but leaves the door open for selection bias, especially when studying a medical condition that affects mental status and mental health. There is certainly potential for patients with dementia to decline or ignore a request to participate in this trial, lowering the representation of patients with dementia in the study. The study showed that the self-reported average TICS score was higher than previously documented levels indicating this bias may have occurred.

The group discussed the statistical analysis of the data. Relying on the calculated sensitivity based on the non-imputed data the authors report a sensitivity of 100% for dementia with a cutoff TICS score of 29. When the cut off is set at a TICS of 28 the sensitivity drops to 77%. Clearly there was at least one person with dementia that had a TICS score of 28 and no one with a TICS score of 29. However, of the 655 with TICS scores over 30 there were 472 who were not verified (72%). We do not know how many of the 472 people may have had dementia and whether this would impact the findings. Including imputed data changes the reported sensitivity of 65% from the study, far from the previously calculated value of 100%. This large effect of imputation leaves the reader to question the accuracy of the imputation, whether the non-imputed data was accurate, whether a finer scale is warranted, and where the true value of the test's sensitivity may lay.

While the researchers did include confounding demographic factors, they did not report significant p value differences for age, median MMSE, education level, and incidence of stroke. Patients with history of stroke and with an older age were noted to perform worse on TICS. Stroke leading to vascular dementia and older people performing worse due to cognitive decline sufficiently explain those phenomena. Additionally, the correlation of the Mini-Mental State Examination (MMSE) with TICS supports the validity of TICS in screening for dementia. Finally, the greater representation of people with less than 12 years of education in the TICS under 30 group demonstrates a previously described trend of lower education levels leading to worse performance on dementia screening.

The findings of this study support the inclusion of TICS in medical practice. While more research should be performed to confirm an optimal cutoff value, and what the sensitivity, specificity, negative, and positive predictive values are, this tool could have significant implications to clinical practice in the United States. Additionally, the diversity of the population in the United States lends itself well to assessing whether there are underlying cultural or socioeconomic factors that influence the results of TICS. Many people in the United States struggle with access to healthcare, for example due to living far from providers, not having reliable transportation, or the ability to use that transportation. Telemedicine can reach some people that in person visits cannot. However, there are also people that telemedicine cannot reach: those without internet access, those with trouble using internet access, and people with visual or auditory disabilities that struggle to use the telephone or video chats. An additional benefit of this technology is its privacy. Patients do not necessarily need to live their home to access telemedicine which could help with the stigmatization against mental health. This may be significantly different in the Netherlands where there is a policy of universal health care. Those administering the exam would need to be trained, and it should be offered in multiple languages. Additionally, as it stands the screening cannot be performed on those with impaired hearing, though maybe a video chat option could be developed with sign language interpreter. We look forward to utilizing TICS and future research on its effectiveness in a new era of medicine.