The number of people with dementia will rise in the upcoming years, which will have large impact on patients, their caregivers, health care systems, and society at large. This emphasizes the need for efficient and accurate early diagnostics of dementia. An increasing number of people with cognitive signs and symptoms are nowadays being referred to specialized diagnostic memory services (Gruters et al., Reference Gruters2019), in which further assessment of cognitive functions can be done. In a memory clinic setting, the neuropsychological assessment (NPA) is one of the major pillars of the routine clinical workup for a diagnosis of dementia (Schmand et al., Reference Schmand, Eikelenboom and van Gool2011). An NPA gives important insight into the diagnosis of the subtypes, the individual profile, including the cognitive strengths and weaknesses, and has high ecological validity, thus outperforming biomarkers, such as MRI and markers in the cerebrospinal fluid.

On the other hand, a full NPA is time-consuming and therefore relatively expensive. The resources available for a NPA depend on the setting and health care organization and may vary among countries. In many services, they are limited and the NPA may then not be part of the routine clinical workup. Therefore, triage with cognitive screening tests can be helpful to identify for which patients a broader assessment is warranted.

In this issue of International Psychogeriatrics, Dautzenberg et al. (Reference Dautzenberg, Lijmer and Beekman2022) describe a study in which the Montreal Cognitive Assessment (MoCA) was especially used for this preselection. The MoCA is a screening test that was originally designed to identify people with Mild Cognitive Impairment (MCI) (Nasreddine, et al., Reference Nasreddine2005). The MoCA is increasingly being used in clinical practice, as it covers several cognitive domains, is available in many languages, and is one of the best-studied screening instruments (de Roeck et al., Reference de Roeck, De Deyn, Dierckx and Engelborghs2019). It has well-validated cutoff scores and a high sensitivity for diagnosing MCI (de Roeck et al., Reference de Roeck, De Deyn, Dierckx and Engelborghs2019; Tsoi et al., Reference Tsoi, Chan, Hirai, Wong and Kwok2015, Reference Tsoi2017). In contrast, specificity is often a topic of debate and dependent on different cutoff scores and demographic characteristics (Carson et al., Reference Carson, Leach and Murphy2018; Davis et al., Reference Davis, Creavin, Yip, Noel-Storr, Brayne and Cullum2015). The generally used cutoff score of the MoCA (25/26) focuses on the differentiation between patients with MCI and those without cognitive impairment. Previous studies, however, recommended a lower cutoff, as this originally defined cutoff often resulted in increased false-positive rates (Carson et al., Reference Carson, Leach and Murphy2018).

The topic of an efficient use of diagnostic tools has been the theme of a recent and relevant special issue of International Psychogeriatrics, in which the use of enhanced assessment instruments for older adults with major neurocognitive disorders. Screening instruments are always striving to find an optimal balance between sensitivity and specificity. The use of an artificial binary cutoff score in gradually progressive neurocognitive disorders in which no categorical differences exist implies that we necessarily have to accept false positives and false negatives. Dautzenberg et al. (Reference Dautzenberg, Lijmer and Beekman2022) introduced a pragmatic approach to deal with this issue, by introducing the use of a double threshold in the MoCA. The study compared different strategies to select patients for further diagnostics, in this case an NPA. They suggested a second-stage selection process using the MoCA with a double threshold as an add-on after initial assessment of the optimal strategy. The study evaluated and provide comparisons of different selection strategies to select patients in need of an NPA in a relevant department of old-age psychiatry, resulting in three pathways, that is only NPA for those with a high likelihood of having dementia, monitoring for those suspected of MCI, and no further testing for people with no cognitive impairment. A double threshold resulted in three groups: those with a score <21 receive an invitation for NPA; those with a score between 21 and 25 qualify for active monitoring; and for patients with a score ≥26, no follow-up is deemed necessary. The approach of a double cutoff combines strategies to reduce rates of both false positives and false negatives. As the authors point out, this approach can be used for differentiating between healthy states from medical conditions whatsoever and between subthreshold conditions versus frank dementia. The resulting three groups (normal cognition, versus MCI, versus dementia) are, as they correctly stated, intuitive.

We agree with the authors that the identification of intermediate states between cognitively healthy and frank dementia gives the opportunity to monitor patients at increased risk. The authors do not define “active monitoring,” but it probably means repeating the MoCA over time. The authors suggest a strategy of administering a full NPA only in older adults with a high likelihood of having dementia. This approach may, however, be a too restricting, as the added value of a NPA has shown to be highest in people with a diagnosis of MCI rather than dementia (Jansen et al., Reference Jansen2017). In addition, in memory clinic settings, the proportion of patients with milder cognitive disorders, such as MCI or Subjective Cognitive Decline has increased over the last 20 years, and may nowadays be larger than the proportion of people with dementia (Gruters et al., Reference Gruters2019). This means that people are seeking for help in earlier stages of cognitive impairment. Performing timely and accurate diagnostics of people in prodromal phases of dementia is important to not delay an accurate diagnosis of a neurocognitive disorder. Moreover, differentiated, domain-specific cognitive profiles cannot be taken from screening tests, such as the MoCA (Coen et al., Reference Coen, Robertson, Kenny and King-Kallimanis2016), but are important to give more insight into the patient’s individual profile of cognitive strengths and weaknesses, which can be used as a first step in a personalized rehabilitation care. In addition, the identification of prodromal phases of neurodegeneration can help to provide patients and their caregivers with timely care and counseling and prepare them and their family members with their changing roles (de Vugt and Verhey, Reference de Vugt and Verhey2013). Therefore, diagnosing prodromal stages of neurocognitive disorders may offer opportunities related to avoiding delay, counseling, and improving quality of life.

Dautzenberg and colleagues wanted to avoid a NPA in patients without cognitive disorders, who in their cohort mostly likely suffered from a primary psychiatric disorder. This raises the question of the benefits of NPA in patients with low a priori likelihood of primary cognitive disorders, for example, in people who are referred because of a major depressive episode. Many of them do have cognitive complaints, and for them too, an NPA can be an important tool to know whether these can be objectified or not and to reassure them when there are no such impairments. On the other hand, the lower cutoff of 21 as a score for identifying people suitable for an NPA could be problematic, as the added value of an extensive NPA in people with severe dementia is questionable, and only will burden the patient in that phase of the disease.

As was clearly pointed out in the recent special issue of International Psychogeriatrics, the selection of an optimal screening instrument will remain challenging and dependent on purpose, resources, setting, and disease stage. The implementation of standardized screening instruments along the patient journey, starting in primary care, is important to timely identify and refer patients for relevant diagnostics (Nguyen and Lee, Reference Nguyen and Lee2020). The use of a double cutoff as a more differentiating screening tool such as the MoCAwill clearly give opportunities for a more adequate referral for further diagnostics, such as an NPA. In the end, however, clinical judgment remains the most important diagnostic. In this way, we fully agree with the authors that the MoCA (and screening instruments in general) should not be used as a stand-alone assessment, and diagnosing (prodromal) dementia should principally be based on a (multidisciplinary) evaluation of the full clinical context. Furthermore, functioning in daily life is an essential component to take into account when diagnosing dementia, an aspect that is generally not integrated in cognitive screening instruments at all. Also in the selection of functional scales, more sensitive instruments will be helpful in identifying people at risk in earlier phases of the disease, for example by integrating more advanced instrumental activities of daily living (Kalligerou et al., Reference Kalligerou2020; Luxenberg, Reference Luxenberg2020).

In the early diagnostics of cognitive disorders, we endorse the use of a short, standardized neuropsychological test battery, which is shorter than an extensive and therefore expensive NPA, but which is also more conclusive than the general impression we can get from a cognitive screening instrument, and which is more robust against the weaknesses of a screening instrument (highly sensitive to age, education, ethnicity, other prominent cognitive deficits except memory, etc.). In this respect, attention should be paid on further harmonization in the neuropsychological field. First steps on national (Ramakers et al., Reference Ramakers2020) and international (Costa et al., Reference Costa2017; Boccardi et al., Reference Boccardi2022) harmonization strategies were taken in the last years and can be used as a first stage to make key cognitive batteries more widely available, and also when recourses are limited. In addition, digitalization and remote assessment forms can also help in improving this part of early diagnostics and monitoring of dementia, a topic that was realized even more since the COVID-19 pandemic.