IntroductionBackground

According to the United Nations, a country is considered old when ≥7% of its population is aged >60 years [,]. However, some countries exceed this percentage [-]. The increase in life expectancy and the growing population of older adults represent some of the most significant demographic changes that society is experiencing today. In all countries, individuals aged ≥60 years are beginning to constitute a large segment of the population. It is estimated that the proportion of this population group will increase 2-fold, rising from 11% to 22% by the year 2050 []. Although there are significant variations between countries and continents, the segment of the population aged 60 years is the fastest growing []. In addition, 10.4% of adults aged >60 years have cognitive impairments associated with aging. In this group, 20.9% of people aged >80 years have this condition []. This affects the social and economic aspects of a country [,].

However, it is not the process of aging itself that is causing alarm among current governments and societies, but rather the burden of dementia that is associated with this aging population. The probability of developing dementia increases with age, doubling approximately every 5 years. In general, it is estimated that there are approximately 40 million people with dementia worldwide, with approximately 9 million people in Europe alone. Furthermore, this global figure is estimated to double every 2 decades, reaching 131.5 million by 2050 [,].

Due to the aforementioned reasons, there has been an increasing interest in cognitive training and other interventions that can mitigate or reverse these degenerative changes in older adults. There are several research papers and recent literature on the outcomes of cognitive interventions for brain training in older adults [-], but their scientific positions vary, and all authors emphasize the need for more empirical evidence. Furthermore, cognitive decline and changes in cognitive status [] can easily go unnoticed in clinical settings. Cognitive assessments are often time-consuming and require a trained health care specialist, such as a neurologist or gerontologist [,], to provide detailed information on the patient’s health []. However, conducting these evaluations poses significant access barriers, as many older adults are unable to attend promptly due to physical or cognitive limitations, fear of going out (due to the pandemic), long waiting times, or long travel distances. These barriers constitute important obstacles in determining the initial stages of cognitive decline.

Thus, this study aimed to provide an updated literature review of the main specialized digital cognitive tests for older adults. This review classified the tests based on their types and characteristics. In addition, it included comparative tables that highlight the technological aspects, cognitive domains evaluated, tasks, activities, and psychometric parameters used in each test.

Cognitive Problems in Older Adults

As a normal component of aging, many people experience a decline in their cognitive functions. When the decline becomes more significant, pathological processes may occur. Different levels of cognitive impairment were observed. As cognitive degeneration progresses, cognitive and functional decline reach a threshold, and the person is clinically diagnosed with probable dementia []. At present, it is unlikely that neuronal damage in the brain can be reversed, and most of the recent treatments available only provide symptom relief rather than a cure for the disease. However, the disease progression can be effectively controlled if dementia is detected at an early stage. Therefore, the most effective strategy is to detect dementia in its early stages and initiate an intervention. Theoretical changes in cognitive function in a person as a function of age (toward possible dementia) are shown in [].

‎Figure 1. Changes in cognitive functions in a person.

In recent years, the concept of mild cognitive impairment (MCI) has been applied to describe the transitional stage between normal aging and the early stages of dementia []. It refers to a “mild” condition in which there is objective memory impairment without functional impairment [,]. In general, individuals with MCI have a high probability of gradually progressing to dementia. This means that they are at a higher risk of developing dementia than healthy individuals. Approximately 12% to 15% of individuals with MCI develop clinical dementia with functional disability within 1 year; therefore, the early detection of dementia depends on an accurate diagnosis [,].

According to specialists [], there are various types of cognitive impairment. Currently, amnesic and nonamnesic MCI are distinguished. This distinction is based on the presence or absence of deterioration in the mnemic function. In addition, it is possible to differentiate MCI according to the number of affected cognitive domains. Some individuals have unidomain MCI, whereas others have multidomain MCI, which involves impairment in >1 cognitive domain. Although memory impairment is the most representative symptom of MCI, several cognitive domains other than memory are compromised in most individuals with MCI [,,].

Theoretical Background

A summary is presented on the theoretical background of cognition and cognitive functions, and then it delves into the traditional evaluation tests as well as the digital cognitive tests that are available. The scheme of the work is shown in .

‎Figure 2. Structure of the theoretical background. Cognitive Domains and Functions

For specialists, cognition is defined as “the set of psychological skills that account for all mental life,” and it is composed of “cognitive domains” (sets of cognitive processes or functions). Cognitive functions are mental processes that allow us to perform any task. They enable an individual to actively participate in the processes of receiving, selecting, transforming, processing, storing, and retrieving information, thus enabling them to function effectively in their surroundings. Cognitive skills are continuously used to learn and remember information, integrate personal history and identity, manage information related to the individual’s location and destination, maintain and distribute attention, recognize different sounds, process different stimuli, perform calculations, and mentally represent an object [,].

According to the literature, the most important cognitive functions include attention [,,], orientation [,], memory [,], perception (or gnosis) [,,], executive functions [,], praxis [,,], language [,], social cognition [,], and visuospatial skills [,]. Cognitive disorders affect the cognitive functions of individuals who experience them.

A special area of interest (for older adults) turns out to be memory, which deserves further explanation. It is the most frequently mentioned cognitive function, the lack of which is attributed to making most daily errors. Apparently, everything is attributed to the “lack” of memory, which is difficult to define. However, in general terms, memory refers to the ability to acquire, store, and retrieve various types of information [,]. At each stage of the memory process, the entire brain is involved [], encompassing several phases: registration, encoding, storage, recall, and recognition of information. There are different forms and types of memory [,], including sensory or iconic memory; short-term memory (STM); working memory (also called operational memory); and long-term memory, which can be divided into 2 groups [-]: explicit or declarative memory and implicit or nondeclarative memory.

Traditional Neuropsychological Tests

Neuropsychological evaluation is used to determine an individual’s cognitive status. It is conducted when there are symptoms of cognitive syndromes such as aphasia or dementia. The evaluation consists of performing cognitive tests to establish the presence of these syndromes. For example, if it is necessary to determine a patient’s language proficiency, a cognitive test assessing language skills should be conducted. Similarly, if it is necessary to determine a patient’s memory status, a cognitive test evaluating their memory should be performed. Therefore, neuropsychological evaluation can determine the presence of cognitive or behavioral syndromes and suggest the etiology of the pathological condition as well as the possible evolution. This knowledge is crucial for determining appropriate rehabilitation measures for patients [].

Several cognitive assessment techniques have been developed to assess adults in clinical settings. They can be distinguished between tests for cognitive screening, which diagnose possible levels of cognitive impairment or degrees of dementia, and specialized tests that evaluate one or more specific cognitive domains [,]. A previous literature review established that the most common screening types are the Montreal Cognitive Assessment (MoCA) [,], the Mini-Mental State Examination (MMSE) [,,], the Addenbrooke’s Cognitive Examination-Revised [], the Mini-Cog [] and the Abbreviated Mental Test (AMT) [,]. Among the tests used to evaluate cognitive abilities in specific domains and in patients with special characteristics are the following: the Stroop test, which is an attentional test that detects neurological and brain problems and assesses the ability to classify information from the environment and react selectively to it []; the Corsi Cubes test, which evaluates visuospatial STM and allows the study of the effect of emotions on STM []; the Trail Making Test (TMT), a neuropsychological test that measures visual attention and task switching (consisting of 2 parts) []; and the Rey-Osterrieth Complex Figure test, which provides information about a person’s neuropsychological functioning [] in terms of attention, concentration, coordination, and visuospatial abilities, among others.

Most cognitive tests are evaluated by specialists by using psychometric parameters. Psychometry is a branch of experimental psychology that is responsible for measuring and quantifying a person’s psychological processes and cognitive abilities []. The most commonly used measures to assess the quality of psychometric instruments, as well as the measurement parameters that are normally used in cognitive tests, include reliability [], validity [], sensitivity [,], specificity [,], and receiver operating characteristic (ROC) curves []. These measures are briefly explained in .

Digital Cognitive Tests

New computing technologies and platforms, including tablets and smartphones, offer many opportunities to create interactive tasks and experiences that can be used to infer the cognitive status. Cognitive assessment software packages are available [,], and they offer computerized versions of traditional tests that can be self-administered. In addition, various websites perform tests and training for certain cognitive abilities. Examples include Cognifit [], NeuronUP [], and Stimulus [].

An analysis of previous literature reviews yielded 4 papers by the following authors: Zygouris and Tsolaki [] in 2015, Aslam et al [] in 2017, Marques-Costa et al [] in 2018, and Tsoy et al [] in 2021. These reviews included 11 to 17 papers that were published in 2012, 2015, 2017, and 2019, respectively. The most recent study by Tsoy et al [] focused on only 3 cognitive domains: attention, memory, and language. From the above, it can be deduced that there is a need for an up-to-date state-of-the-art technology. Previous reviews have mainly focused on computer tests and digital test batteries, neglecting the inclusion of emerging technologies such as virtual reality (VR), video games, gamification, and artificial intelligence (AI). Tsoy et al [] used this as an exclusion criterion for his systematic review.

Previous reviews provide different levels of detail regarding the characteristics of different digital tests. These include the hardware used [,,], input mode or data capture [,], test time [-], and administration modality [,]. However, none of them offer details on the instructions and how they are delivered to the patient, the environment or place where the test is performed, or the sequence of tasks that the participants must perform. This lack of information makes it difficult to compare the results within the same cognitive domain.

One aspect analyzed by all these reviews was the quality of the cognitive tests and the psychometric properties applied. They agree on highlighting important drawbacks in terms of the replicability of the studies [], existence of well-structured psychometric data [], and evaluation of various psychometric properties []. However, a limitation of these reviews is that they do not include the sample size in the tests, details of how many participants had MCI or dementia, or psychometric parameters used for each test in their comparative tables. All these aspects are addressed in this review.

The main specialized digital cognitive tests for older adults are presented, classifying and comparing the domains and cognitive tasks evaluated. This complements the missing characteristics in the analyses of the previous reviews.

MethodsObjectives and Research Questions

To investigate the current state of applications or digital tests for cognitive evaluation in older adults, 4 research questions were formulated:

RQ1: What are the different technological alternatives that are currently used in digital devices to assess the cognitive abilities of older adults?RQ2: What are the different types and characteristics of computerized (digital) cognitive tests?RQ3: What are the main characteristics of the subtests and tasks used in different digitized cognitive tests?RQ4: What are the main effects, personal traits, and psychometric parameters considered for validating digitized cognitive tests for older adults?Eligibility Criteria

A literature review was conducted on digital systems for detecting cognitive problems. The aim was to gather updated information on technological solutions that could help overcome the possible barriers and difficulties of traditional psychometric tests. To do this, the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) methodology was used. Through a standardized review process, information is delivered in a flowchart that considers 4 stages: identification, screening, eligibility, and inclusion.

To search primary studies (articles), we searched 3 databases based on the search strategies: Web of Science, PubMed, and Scopus. These databases were chosen because they have peer review processes in which experts approve the publications. We combined the keywords with logical operators to obtain the following search expression:

(“cognitive assessment” OR neuropsych*) AND (computer* OR web OR “digital test” OR evaluation) AND (“older adult” OR adult) AND (“cognitive impairment”) AND valid*.

Due to the high number of publications (although a large percentage of articles appear on several sites), the following criteria were considered for the first selection and review of the most relevant articles:

Inclusion criteria were articles and research papers that were published 2015 onward and were written in both English and Spanish; those articles in which titles and summaries (abstracts) included terms that addressed any of the research questions. The articles should be published by publishers with a website and should fall under the categories of scientific articles, conferences (proceedings), or book chapters.Exclusion criteria were articles in which titles and abstracts were not related to the objective of the study or the research questions, those that were repeated in another language, and those that were not related to older adults.Data Collection

On the basis of the proposed methodology, we searched for articles from the 3 databases. The process is shown in the flowchart in . The explanation for each phase of the PRISMA methodology is as follows:

‎Figure 3. PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) scheme of the systematic review (own elaboration). Identification

A systematic search of the literature was performed using 3 detailed databases, resulting in the retrieval of 203 articles from Web of Science, 216 articles from PubMed, and 151 articles from Scopus.

Screening: duplicates were eliminated, leaving 397 articles. The searches were then filtered by title, abstract, and systematic review, resulting in a total of 147 reports.Eligibility: after conducting the first superficial reading of the articles (n=101), those that did not meet the inclusion criteria were excluded, resulting in 46 studies.Inclusion: these 46 documents were reviewed again to identify points that did not align with the study objectives. Finally, 20 articles were selected. All articles corresponded to scientific articles; there were no conference proceedings, book chapters, or university theses.

In a previous search, various validated methodologies were found concerning the so-called “gold standard,” or reference test. These tools allow for different methodologies to be used in conducting structured tests to quantify the affected cognitive domain. was created using 20 articles discovered. Each article included the first author’s name, year of publication, country, instrument name, and the technological alternative that met certain quality criteria.

Table 1. Articles analyzed.Study, yearCountryArticle titleTypes of technologyKöstering et al [], 2016GermanyAnalyses of rule breaks and errors during planning in computerized tower tasks: insights from neurological patientsComputerized cognitive testsScharre et al [], 2017United StatesDigitally translated Self-Administered Gerocognitive Examination (eSAGE): Relationship with its validated paper version, neuropsychological evaluations, and clinical assessmentsDigital versionWong et al [], 2017ChinaComputerized Cognitive Screen: a self-administered computerized test for screening for cognitive impairment in community social centersComputerized cognitive batteryValladares-Rodriguez et al [], 2017SpainDesign process and preliminary psychometric study of a video game to detect cognitive impairment in senior adultsComputerized cognitive gameRapp et al [], 2018United StatesComputer simulations for assessing cognitively intensive instrumental activities of daily living in older adultsSimulator of daily activityPossin et al [], 2018United StatesThe Brain Health Assessment for detecting and diagnosing neurocognitive disordersCognitive web platformGroppell et al [], 2019United StatesA rapid, Mobile Neurocognitive Screening Test to aid in identifying cognitive impairment and dementia (BrainCheck): cohort studyCognitive web platformKhaligh-Razavi et al [], 2019United StatesIntegrated Cognitive Assessment: speed and accuracy of visual processing as a reliable proxy to cognitive performanceComputerized cognitive testsEraslan Boz et al [], 2019TurkeyA new tool to assess amnestic mild cognitive impairment in Turkish older adults: Virtual SupermarketVirtual reality environmentTakahashi et al [], 2019JapanDevelopment and validity of the Computer-Based Cognitive Assessment Tool for intervention in community-dwelling older individualsComputerized cognitive batteryIchii et al [], 2019JapanCogEvo, a cognitive function balancer, is a sensitive and easy psychiatric test battery for age-related cognitive declineComputerized cognitive batterySchulz-Heik et al [], 2020United StatesEvaluation of adding the CANTABa computerized neuropsychological assessment battery to a traditional battery in a tertiary care center for veteransComputerized cognitive batteryCahn-Hidalgo et al [], 2020United StatesValidity, reliability, and psychometric properties of a computerized, cognitive assessment test (Cognivue)Computerized cognitive batteryTsoy et al [], 2020United StatesBHA-CSb: A novel cognitive composite for Alzheimer’s disease and related disordersCognitive web platformChin et al [], 2020KoreaA validation study of the Inbrain CSTc: a tablet computer-based Cognitive Screening Test for Elderly People with cognitive impairmentComputerized cognitive batteryLunardini et al [], 2020ItalySupervised digital neuropsychological tests for cognitive decline in older adults: usability and clinical validity studyComputerized cognitive testsNoguchi-Shinohara et al [], 2020JapanA new computerized assessment battery for cognition (C-ABC) to detect mild cognitive impairment and dementia around 5 minCognitive web platformChan et al [], 2020ChinaElectronic cognitive screen technology for screening older adults with dementia and mild cognitive impairment in a community setting: development and validation studyCognitive web platformRodriguez-Salgado et al [], 2021CubaA brief digital cognitive assessment for detection of cognitive impairment in Cuban older adultsCognitive web platformBottiroli et al [], 2021ItalyThe smart aging platform for assessing early phases of cognitive impairment in patients with neurodegenerative diseasesComputerized cognitive game

aCANTAB: Cambridge Neuropsychological Test Automated Battery.

bBHA-CS: Brain Health Assessment-Cognitive Score.

cCST: Cognitive Screening Test.

ResultsOverview

Some preliminary statistics could be obtained regarding the publication dates of the articles, such as the countries where the studies were conducted (). Subsequently, the technological alternatives were classified and explained. For example, differentiating between digital tests, that is, if they are cognitive batteries and computerized cognitive tests, or computerized platforms (web), or simulators of daily activity, using a VR environment, or computerized cognitive games. Possible factors that influenced each psychometric instrument were also characterized. For example, the technical aspects of the tests such as the administration time, data capture mode, modality (technology used), operation (instructions) of the software, and how the tests are administered.

In addition, comparative tables were created to provide a synthesized description of the digitized cognitive tests, differentiating some characteristics such as the cognitive domains covered, the number of tests or stages available, the activities involved in each test, and the evaluation scoring system. Finally, the articles were analyzed to validate their psychometric properties in comparison with traditional tests.

Statistical data of the articles studied are shown in (based on the year of publication) and (based on the country where the study was conducted).

Table 2. Papers by the year of publication.Year of publicationPapers (n=20), n (%)20161 (5)20173 (15)20182 (10)20195 (25)20207 (35)20212 (10)Table 3. Papers by the country of study.Countries for studyPapers (n=20), n (%)United States8 (40)Europe5 (25)Japan3 (15)Chinese2 (10)Korea1 (5)Cuba1 (5)RQ1: What Are the Different Technological Alternatives That Are Used in Digital Devices to Assess the Cognitive Abilities of Older Adults?Overview

To answer the first research question (RQ1), we conducted a preliminary analysis of the selected articles, identified differences in categories of digital tests, and briefly described the identified classes. shows the number of articles on each application (according to the type of technology).

Table 4. Papers by the type of digital test (technology).Types of digital test (technology)Papers (n=20), n (%)Test battery6 (30)Web platform6 (30)Computerized test3 (15)Games2 (10)Digital version1 (5)Simulator IADLa1 (5)Virtual reality1 (5)

aIADL: Independent Activities of Daily Living.

Cognitive Batteries and Computerized Cognitive Tests

These are a set of tests and tasks that allow the evaluation of multiple cognitive domains, such as language, executive function, attention, and memory [,]. The tests are based on and validated using psychometric methods. They present a greater advantage than traditional tests (pencil and paper), as they lead to a detailed cognitive profile, reducing possible errors caused by administration bias []. In addition, they provide information about the testing process, such as reaction times and the sequence of answers (good and not good). In addition, automatic scoring helps professionals improve clinical diagnoses. The small difference between test batteries [,-,] and computerized tests [,,] is that the former has several tests and tasks designed to assess >4 different cognitive domains, whereas the latter includes fewer tests and focuses on a single cognitive domain. In general, computerized cognitive batteries, which are used on PCs, notebooks, or tablets, serve as supportive tools in clinical and community settings.

Simulators of Daily Activity

At the onset of MCI, the ability to perform daily activities remains unaffected. However, as the deterioration progresses, the performance of these activities decreases. Activities such as shopping, taking medications, and using telephone lines become difficult. To assess the condition, it is necessary to invest in devices and time and to consider the burden it places on the patients. As an alternative, the use of simulators for daily activities offers enormous possibilities. For example, in Simulation-Based Assessment of Cognition [], the patient can interact with software to complete tasks such as withdrawing money at a virtual automated teller machine or making a call to a virtual pharmacy to indicate the needed medicine. These activities are associated with executive function. Therefore, it is an instrument with many advantages that require further development to include the full range of activities commonly performed by older adults, vary the performance depending on the particular neuropathology, and adapt to different cultural and socioeconomic settings.

Computerized Platforms (Web)

These are cognitive batteries, tasks, or adaptations of traditional tests that allow comprehensive evaluations of cognitive impairment using mobile devices or devices connected to the internet. The cognitive evaluation begins after patients enter their sociodemographic data. Among these options, the platforms can be configured with an algorithm that enables the questions to be displayed and prompts the participant to select the correct answer. Meanwhile, the information is stored on a centralized computer server, which allows access to cognitive detection to be faster, more efficient, and automatic; and it is possible to perform the test in the comfort of one’s home (or any appropriate place) [,,,-].

Digital Version

These are digital cognitive assessments that are equivalent to paper versions and allow for greater flexibility. In addition, the tool (a single test, not a battery) can increase the screening of individuals who are being evaluated through self-administration using technological devices. The technological solution is built based on the questions from the original version but with the added advantage of automatically measuring the time it takes for participants to answer the questions. In addition, it can determine the frequency with which participants return to previous pages based on the subject. Finally, the evaluation is accessible through web, and the results obtained are delivered in a digital format [].

VR Environment

Currently, there is an increase in the use of VR technology for evaluating cognitive dysfunction. An individual can enhance their interaction in a simulated environment by following the instructions of a traditional cognitive test. For example, in a Virtual Supermarket (VSM) [], before the exercise, age, gender, occupation, years of education, and any possible memory complaints are registered. VSM generates a randomized list of products for a daily shopping activity. The individual is expected to locate the items on the list, place them in the shopping cart, take them to the register and pay the correct amount. Furthermore, the participant must navigate the VSM by touching green footprints on the screen while pushing the shopping cart. It is an exercise designed to examine multiple cognitive domains, such as visual and verbal memory, executive functions, attention, and spatial navigation. People with cognitive impairment require more time and make a greater number of errors than healthy individuals []. For example, patients with cognitive impairment will not be able to remember a list of instructions. The authors mentioned possible limitations if all the adults had similar experiences and functioning in daily life, especially in tasks related to purchases. They also note that the payment in euro currency could have added additional complexity and cognitive load.

The use of VR technologies allows for reduced costs and decreased administration time due to automatic scoring. In addition, participants may be able to self-test in the comfort of their homes without the supervision by a specialist, who will be consulted only if the test detects signs of possible deterioration. However, further studies that include participants with different degrees of familiarity with new technologies, especially tablets, are needed.

Computerized Cognitive Games

These applications can be categorized into 2D games and 3D scenario generation. Games (or game batteries) are clinically useful resources that allow for the detection of deficiencies in multiple domains. For example, Episode Gamification [] is a game that involves taking a virtual walk through a medium-sized city where everyday objects are displayed. The challenge is to remember the maximum number of items to be displayed while avoiding any interfering objects. Another example is Smart Ageing [], a 3D game that features a loft with a kitchen, bedroom, and living room area. Participants use a touch screen monitor to navigate and interact with the environment, performing five tasks related to daily life: (1) find a list of objects in the kitchen after exploring it; (2) water the flowers while listening to the radio, pressing the space bar each time the word “sun” is heard; (3) make a phone call using the phone book and the phone that are placed on the nightstand, remembering to turn on the television after dialing the number; (4) identify the 12 objects presented in task 1 from a 2D screen with 24 images of objects; and (5) find each of the objects searched for in task 1 while being in the kitchen. Therefore, cognitive computer games have emerged as a novel approach for assessing the cognitive state of people, allowing them to simulate recurring tasks and sensory stimuli while collecting information on the patient’s reaction time in certain tests [,].

One problem with several traditional cognitive tests is that they exhibit a “learning bias” [], meaning that an individual’s cognitive performance improves with repeated exposure to the test, solely because of learning the task, without any actual change in their cognitive ability. Consequently, this bias reduces the reliability of a test when it is used repeatedly (for example, when monitoring performance over time). Computerized cognitive tests can overcome this difficulty by randomly tailoring different task contents to participating adults. In addition, to adopt a psychometric evaluation, it is necessary to normalize the target population according to their specific context and ensure semantic agreement in the tasks, considering both language and culture. For example, in this review, we found only 2 platforms that could be adapted to Latin American population [,].

Digital cognitive tests have the potential to be objective, standardized, and most importantly, repeatable. Computerized testing applications provide ideal formats for generating alternative tests, thereby improving test-retest reliability during repeated administration in long-term monitoring. As screening and monitoring tools for serious diseases, computerized cognitive tests are being developed, with emphasis on ensuring their comprehensiveness, validity, and reliability.

Regarding new technologies, although the selected articles were published between 2016 and 2021, only a few used games, virtual or augmented reality, or simulators of daily life for cognitive purposes. Only 4 applications moved in this direction [,,,], although we expect more to come in the near future. This is especially relevant for cases in which one wants to measure slight cognitive differences over time (whether improvements or deterioration) instead of simply discriminating between a healthy adult, someone with MCI, and a certain degree of cognitive impairment. Our review also found a study that used machine learning techniques []. The review by Marques-Costa [] also observed the need to include item-response theory techniques associated with automatic assessment. The item-response theory can help adapt the difficulty level of cognitive tests to older adults’ personal characteristics and context. This makes us expect a greater inclusion of data analysis and AI techniques in future research, especially those aimed at improving the accuracy of diagnosis and instrument reliability.

With regard to VR technology use, we highlight its absence in cognitive tests. From previous reviews [-], the only case mentioned was the Computer Assessment of Mild Cognitive Impairment application, which included a VR driving task []. In our systematic review, we found only one study based on supermarkets []. It can be argued that this approach allows designers to contrast their batteries with classic paper-and-pencil tests and compare their concurrent validity. At the same time, the higher cost and potential risk associated with developing a VR- or gaming-based test from scratch might also influence the decision to adopt a more “conservative” approach and rely on proven testing instruments.

Finally, an important limitation of most digitized cognitive applications is that they require reliable internet connection. It is necessary for these applications to be able to work offline, saving progress and not depending only on synchronous communication. Future work should explore solutions that can function without a stable internet connection to enhance the accessibility of such tools and encourage their use in rural areas. In addition, in the face of any network contingency, applications that can work offline [] would perform all its functionalities on a local computer in the face of any network contingency.

RQ2: What Are the Different Types and Characteristics of Computerized (Digital) Cognitive Tests?Overview

On the basis of the collected information, a characterization of the different digital cognitive tests was conducted to answer the second research question (RQ2). Next, the technical aspects of the tests, such as the administration time, data capture mode, modality, mode of operation (or software instructions), administration method, and location, are detailed. Some criteria or parameters allowed for comparing the different tests included in the selected articles, based on which was prepared.

Table 5. Main characteristics of the cognitive digital tests.Cognitive digital testTime (min)Input modeModalitySoftware instructionsExam administrationLocationTOLa []8Touch Screen, peripheral, or PC mouseComputer with touch monitorVisual instructions (text and time limit)Administered by examiner psychologistFace to face; laboratory roomeSAGEb []17Touch screenTablet or web-basedVisual instructionsSelf-administrationFace to face; community-clinical settingsCoCoScc []15Touch screen or headphonesComputer with touch monitorVisually or verbally (audio)Self-administered or browserFace to face; housing and community centersEpisodix Gamification (CVLTd) []30-40Touch screen or PC peripherals (joystick, mouse, and Kinect)Android or computer (Windows, Linux, or iOS)Instructions are provided in audio and text format.Personal clinical support assistanceFace to face; community and university centerSIMBACe (IADLf) []10Touch screen or PC peripheralTablet or computerModules with visual (text) and verbal (voice recorder) instructionsSelf-managed or trained technicianFace to face; medical care centerUCSFg Brain Health []10Touch screenSoftware platform or iPad 9.7 inchInstructions in the software; examiner evaluatingAutomated scoring; cannot be self-administeredFace to face; diagnosed in university centersBrainCheack Inc []21 (mean)Mobile touch screeniPad, iPhone, or desktop browserInstruction by examinerResearch staffFace to face; community centerCGN_ICAh []5Touch screeniPad, Raspberry, or web—iSelf-administered or examinerFace to face or distance; clinic or homeVSMj []25Touch screen and computer peripheralsTablet (10-inch) or PCInstructions by examinerSelf-administered or personal assistance in repeating instructionsFace to face; institutes and medical centersCompBased-CATk []10-15Touch screen, PC peripherals, or headphonesTablet (Asus) or computer (Windows 10)Visual on-screen instructions and voice with external noise-cancelling hearing aidSelf-administered or minimal assistance in instructingFace to face; institutes and geriatric hospitalsCogEvo []10Touch screenComputer OSlAudiovisual with home iconAdministered by examinerFace to faceCANTAm []45-60Touch screenTouchscreen computer (Windows)Verbal instructions from the instructorAdministrated with trained supervisorFace to face; adult centersCognivue []10PC peripheralsComputer OSAutomated instructions and test subbatterySelf-administered or assisted by nonclinical support staffFace to face or clinical establishment. No specific placeBHA-CSn []10Touch screenSoftware and tablet, TabCAT Pad de 9,7—Managed by examinerFace to face; adult and Alzheimer centersInbrain CSTo []30Touch screenTablet or OS Microsoft Windows 10Verbal (written)Minimum attendanceFace to face or distance; private room in clinicTrail Making Test y Bells Test []5Touch screenTab A6 con S pen or Webserver nubeVerbal assistant, virtual supervision of the testUnsupervised environment; virtual onlyDistance; Geriatric Foundation or HomeC-ABCp []5Touch screenComputer (OS) with touch screen (80×60)On-screen text and verbal description with headphones on PCAssistance by a psychologist if neededFace to face; Memory ClinicEC-Screenq []5Touch screenWeb or tabletReading questions answerAutoadministrated or assistedFace to face or distance; geriatric community settingsBHA []10-5Touch screenTablet or webDigital surveyNeurologist examiner or neuro psychologistFace to face or institute community centersSGr (IADL) []10-30Touch screenTouch screen computerVisual instructions (and examiner)Administered in the presence of a neuropsychologistFace to face or neuropsychology unit, communities

aTOL: Tower of London.

beSAGE: Self-Administered Gerocognitive Examination.

cCoCos: Computerized Cognitive Screen.

dCVLT: California Verbal Learning Test.

eSIMBAC: Simulation-Based Assessment of Cognition.

fIADL: Independent Activities of Daily Living.

gUCSF: University of California, San Francisco.

hCGN-ICA: Cognitivity Neurosciences-Integrated Cognitive Assessment.

iNot available.

jVSM: Virtual Supermaket.

kCompBased-CAT: Computer-Based Cognitive Assessment Tool.

lOS: Multiple Operating System

mCANTAB: Cambridge Neuropsychological Test Automated Battery.

nBHA-CS: Brain Health Assessment-Cognitive Score.

oCST: Cognitive Screening Test.

pC-ABC: computerized assessment battery for cognition.

qEC-Screen: Electronic Cognitive Screen.

rSG: serious game.

Time

The administration time of the tools found in the literature varied from 5 to 45 minutes. Compared with the traditional MMSE assessment that relies on professional training, digital cognitive tests take approximately 10 minutes, excluding the time needed to score the participant. In contrast, if the examination is performed by a technician or someone without specialized training, it will take much longer []. Computerized batteries allow for more accurate measurements; for example, the computerized assessment battery for cognition [] can be administered in a short amount of time (approximately 5 min). It is a sensitive battery that detects not only cognitive impairment but also dementia. Finally, positive results should be considered supportive methods and not definitive diagnoses. Therefore, patients should be referred for a more comprehensive evaluation by health professionals [].

Input Mode (or Capture)

The data-acquisition mode was identified using a touch interface and computer peripherals. In a study of digital games, it was found that older adults preferred a touch interface to computer peripherals such as a keyboard and mouse. In addition, in the digital version of the Self-Administered Gerocognitive Examination (eSAGE) [], the participants did not use a stylus but instead used their fingers to draw or write the requested answers. Finally, the use of hearing aids allows for the cancellation of external noise and helps patients avoid distractions [].

Modality (of the Device)

Psychometric instruments were implemented on desktop computers, laptops, tablets, iPads or iPhones. The use of larger screens helps individuals with visual impairments access and complete the tests. These digital methods allow us to capture the response time with better precision. For example, the Electronic Cognitive Screen [] integrates the clock task, which reflects the speed of processing, and executive function. This test, when displayed on a tablet, allows for the detection of a person’s lesser fine motor control; it is easier than the paper version for older adults.

Software Instructions

The mode of operation (or software instructions) can be entered by examiners or more easily integrated into the software. For the “Simulation-Based Assessment of Cognition” simulation software, the instructions consist of voice recordings and text files []. Each module incorporates specific instructions. The platforms read the questions and then prompt the participants to select the correct answer. The “TMT and Bells” tests [] incorporate AI to detect the participants’ voices and dictate the task guidelines. In the virtual game “VSM,” the guidelines are shown visually and auditorily on the screen for each activity in the game’s virtual environment []. The administration of the digital version is minimally assisted because each item comes with simple written instructions. Finally, the role of the examiner must be clear because providing clues during the test can introduce bias into the evaluation.

Test Administration

The administration of a psychometric instrument can influence the results. The advantage of administering digital cognitive tests in a laboratory is the potential to reduce the frequency of errors, software failures, and interruptions, but the psychometric instrument can be conditioned to exclude patients []. The Computerized Cognitive Screen [] allows for self-administration with automated scoring while still requiring minimal assistance for older people. The automated calculation of scores in the “Cognivue” battery [] is more efficient and consistent than traditional tests. In Episodix [], standardized administration makes data collection and response time capture more efficient. In the digital version [], web-based administration is useful for people living in rural regions with limited resources and a lack of access to health care providers. The test supervisor only observed and served as a guide or security during the evaluation. The greatest advantage found in VR environments is allowing older adults to check their cognitive functioning at home and only visit a specialist if necessary []. The Integrated Cognitive Assessment visual categorization test [] was found to be self-administered due to its simple design with no language and culture barriers. The platform [] has an automatic scoring algorithm that helps alleviate the burden on the professional staff. Thanks to automatic scoring, not requiring individuals with specialized knowledge makes it easier to eliminate human error and reduce the duration of examinations.

Location (or Place)

Participants were recruited from clinics, memory study centers, hospitals, and neurological institutions. Diagnostic tests in community centers are beneficial because they allow digital cognitive tests to be performed on a larger scale and are easily accessible, as they require minimal professional demand. In addition, for some older adults, interacting with a computer can be more pleasant than visiting a health care facility. It can also be administered by individuals with a lower level of education, although guidance from professionals is necessary for those in community centers and primary health care clinics. Home-administered cognitive tests can provide a relaxing, nonintrusive, and familiar environment for patients [,]. The flexibility of the applications in their electronic versions also facilitates their administration in the comfort of one’s home using a mobile device or tablet. In the case of the TMT-Bell digital test [], the authors emphasized the need to strengthen the collaboration between technicians and clinicians. They also suggested the development of novel indicators that could further enhance the utility of digitized tests. In addition, digitized tests were performed in a clinical setting under the supervision of a trained professional. Therefore, they proposed conducting more usability and validity studies on their entire platform in a domestic environment. Despite the digitization of cognitive tests, only 4 applications have declared that they can be used remotely (web-based) without the presence of a supervisor [,,,].

From the reviewed articles, we can conclude that there are clear advantages to conducting (digital) tests using technological devices. These tests allow psychometric evaluations to be executed in the comfort of one’s home or community centers [,,,,-], with minimal intervention from a specialist professional. In addition, most psychometric tests were implemented on equipment with touchscreen devices, such as notebooks, tablets, and smartphones. However, there was one exception: the Cognivue application [] was carried out with a desktop computer. In cases where an older person is not familiar with technology, assistance can be provided by a family member or the clinical supervisor.

In addition, computerized cognitive tests provide new opportunities to remotely monitor cognitive changes and detect early dementia. Without the need to visit a trained practitioner or counselor, computerized cognitive tests can be distributed over the internet and self-administered [,,,,,] or eventually supported by a supervisor [,,,]. Some tests need to be administered by neurologists, psychologists, or specialized clinical personnel [,,,,].

Technology benefits psychometric tests in numerous ways. For example, it significantly reduces the time that elapses between the application of the evaluation instrument and the reporting of results, sometimes providing immediate feedback. In addition, technology helps to eliminate qualification errors that are common among human beings. It is also possible to implement technological security measures, such as using passwords, and finally, it enables the adaptation of the content of the test according