All that follows is from the above resource.

• Abstract
• 1. INTRODUCTION
• 2. METHODS
• 3. DISCUSSION
• 4. CONCLUSIONS
• AUTHOR CONTRIBUTIONS
• CONFLICT OF INTEREST STATEMENT
• Supporting information
• ACKNOWLEDGMENTS
• Contributor Information
• REFERENCES
• Associated Data

Abstract

US clinical practice guidelines for the diagnostic evaluation of cognitive impairment due to Alzheimer’s Disease (AD) or AD and related dementias (ADRD) are decades old and aimed at specialists. This evidence‐based guideline was developed to empower all—including primary care—clinicians to implement a structured approach for evaluating a patient with symptoms that may represent clinical AD/ADRD. As part of the modified Delphi approach and guideline development process (7374 publications were reviewed; 133 met inclusion criteria) an expert workgroup developed recommendations as steps in a patient‐centered evaluation process. The workgroup provided a summary of validated instruments to measure symptoms in daily life (including cognition, mood and behavior, and daily function) and to test for signs of cognitive impairment in the office. This article distills this information to provide a resource to support clinicians in the implementation of this approach in clinical practice. The companion articles provide context for primary care and specialty clinicians with regard to how to fit these instruments into the workflow and actions to take when integration of performance on these instruments with clinical profile and clinician judgment support potential cognitive impairment.

Keywords: Alzheimer’s disease, cerebrospinal fluid, dementia, diagnosis, frontotemporal dementia, Lewy body dementia, magnetic resonance imaging, mild cognitive impairment, molecular biomarkers, positron emission tomography, vascular cognitive impairment

1. INTRODUCTION

The Alzheimer’s Association convened a Diagnostic Evaluation, Testing, Counseling and Disclosure Clinical Practice Guideline Workgroup (the DETeCD‐ADRD CPG Workgroup). Our emphasis is on good clinical practice for the process of evaluating a patient presenting with an illness (i.e., symptoms, obtained through history, and signs, obtained through examination) that may represent the clinical manifestations of common brain diseases, especially Alzheimer’s disease (AD) and AD related dementias (ADRD)—in some cases with exacerbating medical conditions or factors. While this guideline applies to a patient with any severity of cognitive or behavioral impairment, it does not consider individuals who do not have symptoms; therefore, it does not address the topic of screening in asymptomatic people. ¹ , ² , ³ , ⁴ , ⁵ , ⁶ This DETeCD‐ADRD CPG seeks to empower all clinicians, including those in primary, specialty, or subspecialty care, to implement a structured yet individualized patient‐centered approach to diagnostic evaluation that includes clear communication with the patient and an informant or care partner(s).

For any given individual, differentiation of what is a cognitive–behaviorally impaired versus an unimpaired state requires clinical judgment. ² , ⁷ , ⁸ , ⁹ , ¹⁰ , ¹¹ , ¹² The determination that a person has mild cognitive impairment (MCI) or dementia (or Mild versus Major Neurocognitive Disorder in Diagnostic and Statistical Manual of Mental Disorders Fifth Edition terminology) is the first step of a diagnosis that requires the clinician to integrate reliable history regarding the types and trajectory of changes in cognitive, activities of daily living (ADL), and mood and behavioral functions (from the individual and an informant) with the patient’s performance on tests of cognitive function in multiple domains (attention, memory, language, executive function, visual function, socio‐emotional behavior). ⁷ , ⁸ , ⁹ The reported symptoms and performance on tests are both influenced by a variety of individual factors that have to be considered, including education, occupation, culture, living situation, family or other relationship dynamics, developmental history, and medical and psychiatric comorbidities.

For example, the diagnosis of dementia may be straightforward in a formerly high‐functioning patient whose family reports insidious onset of impaired memory and executive function with impairment in instrumental activities of daily living (IADLs) and who scores a 20/30 on the Montreal Cognitive Assessment (MoCA). ¹³ Yet this first step in the diagnostic formulation may be very challenging in a symptomatic and highly educated person who reports memory loss with an impact on occupational function but who performs in the unimpaired range on a brief cognitive assessment test like the MoCA; such an individual may require a neuropsychological evaluation to document impaired performance or to establish a current baseline that can serve to track changes. Unlike the diagnosis of anemia, hyperkalemia, or proteinuria, there is no test value that determines the diagnosis of MCI or dementia—it requires integration of multiple layers and types of information and, importantly, clinical judgment.

In some cases, a patient with cognitive concerns may be documented to perform normally on detailed neuropsychological evaluation; such a person might be classified as having “subjective cognitive decline,” a clinical construct being studied extensively by an international research community. ¹⁴ In other cases, a patient has developed an acquired change in personality or behavior but is found to be cognitively intact; such a patient might be classified as having “mild behavioral impairment,” another clinical construct being evaluated by the research community. ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸

The second step—determination of cognitive–behavioral syndrome—facilitates communication about the specific types of impairments the patient has, regardless of the severity of those impairments (i.e., MCI or dementia). While some patients present classically with one of the recognizable cognitive–behavioral syndromes (see Table 2 of companion manuscript), ¹⁹ others may not fit so clearly into these syndromic diagnostic criteria. In these cases, additional information from informants may be helpful, or additional office‐based assessments of cognitive, behavioral, and sensorimotor function may be necessary, or consultation with a specialist(s).

1.1. History of present illness

Regardless of whether the patient or a family member initiated the medical contact, the history of present illness (HPI) is the cornerstone of the approach to medical diagnosis. In the era of the electronic medical record in which more activities are relegated to templates and checkboxes, the artful elicitation of the HPI is an interactive exercise in spontaneity, unexpected responses, and nimble redirection. In practice, the initiation of an evaluation for cognitive or behavioral concerns, particularly when symptoms are relatively more prominent, is more likely to have come from a family member or close friend rather than the patient, because of the impairment or loss of awareness and insight that often accompany acquired cognitive and behavioral syndromes. The family member or friend who often prompts the evaluation becomes a confidante or “informant” who provides key observations to the clinician, who should compassionately elicit, compile, and ultimately integrate and interpret what the patient and informant(s) describe. A substantial body of evidence indicates that—in the setting of what ultimately is diagnosed as a likely neurodegenerative form of MCI or dementia—informant reports provide added value to the history as taken from the patient ²⁰ , ²¹ , ²² , ²³ , ²⁴ , ²⁵ and to cognitive test performance. ²⁵ , ²⁶ , ²⁷ , ²⁸

The goal is to obtain a comprehensive description of the patient’s principal cognitive and behavioral symptoms and their impact on daily function, interpersonal relationships, and comportment; the time course of those symptoms; the existence and evolution of other relevant symptoms; and the pertinent medical history and risk factors. The interaction between historian (i.e., the clinician) and the patient and informant(s) almost always begin with the query that has the general form: “What is the main reason you are here to see me and what would you like to accomplish from the visit today?”

1.2. History of cognitive symptoms

In the context of a suspected cognitive–behavioral syndrome in an older adult, the potential for complexity, ambiguity, or misdirection of the response to this question is predictably unpredictable. Because diminished insight is common in individuals with a syndrome of cognitive–behavioral impairment, the patient and their care partner (informant) may have divergent opinions about the nature of the symptoms and their consequences. The likelihood of a potential disagreement in perspectives can be communicated up front and be acknowledged as a useful clue for the clinician (e.g., “This is a safe place and time when you should feel free to disagree with each other: it can help me understand and advise you better”). It may be valuable to interview the patient and informant(s) separately because of discomfort with honest reporting or overt friction. Therefore, the clinician needs to be flexible and to encourage and pursue all lines of the story and, informed by these and other information gathered during or after the visit (post‐visit phone calls are often helpful), integrate perspectives and information into an initial narrative that represents the most likely approximation of the actual events.

The meaning of words like “memory loss” or “confusion” used by the patient and informant may differ substantially from the clinician’s sense of those words. The clinician must therefore encourage the patient or informant to elaborate by giving specific examples. The loss of episodic memory that occurs in typical MCI or dementia due to AD involves difficulties with learning and recalling newly acquired information and recent life events. Sometimes patients or informants may use the term “memory loss” when referring to word‐finding difficulty, inattention, loss of geographic orientation, or loss of the ability to perform step‐by‐step tasks. It can be very challenging to distinguish the early stages of cognitive decline due to neurodegenerative disease from normal aging. It is important to skillfully communicate that changes which may be common in individuals with advancing age are not always normal and could benefit from further diagnostic evaluation. Several validated instruments offer structured questionnaires to assist in the organization and reporting of symptoms of cognitive impairment (some of these also include mood/behavior and/or daily function (Table 1)).

TABLE 1.

Validated instruments to assist in the structured reporting of symptoms of cognitive impairment.

Instrument	Purpose	Features	Comments
IQCODE 29 , 30 , 31	The first informant‐based questionnaire to rate change in cognitive function from the person’s previous ability.	Original had 26 items; short version has 16 questions that measure cognitive decline from premorbid level. Each item is rated on a 5‐point scale from 1 (“much better”) to 5 (“much worse”) and ratings are averaged, with 3 representing no change. Validated in people with dementia against other measures of cognitive decline. Not influenced by education, pre‐morbid ability, or language proficiency, but is affected by informant characteristics and the quality of the relationship between the informant and the subject. Less sensitive to MCI.	Available at https://nceph.anu.edu.au/research/tools‐resources/informant‐questionnaire‐cognitive‐decline‐elderly Information from the IQCODE and the MMSE can be combined in the DemeGraph (https://biostats.com.au/Demegraph/) to aid in assessing for dementia.
AD8 32	Brief screening (2–3 minutes) interview that can differentiate between individuals with and without cognitive impairment.	A patient or informant rates yes/no questions about memory, orientation, judgment, and everyday function. The AD8 is a valid and reliable dementia screening measure compared to the expert clinical judgment and neuropsychological assessments. The AD8 is an appropriate screening tool for dementia but may not be sensitive to other more acute or subtle forms of cognitive dysfunction.	Available at https://www.alz.org/media/Documents/ad8‐dementia‐screening.pdf
QDRS 33	10 item questionnaire completed by informant, rating change from premorbid baseline on an ordinal scale from 0 to 3, which when summed aim to capture the types and severity of cognitive and behavioral symptoms and impact on daily function.	The QDRS is a free screening and staging tool (not a diagnostic tool). It can be used as a structured screen for cognitive, behavioral, and functional changes and symptoms as well as for staging severity. QDRS scores correlate with the longer CDR (see Table 3). Takes 7‐10 min of informant time. Score range interpretations: normal 0‐1; MCI 2‐5; mild dementia 6‐12; moderate dementia 12‐20; severe dementia 21‐30.	Available at https://umiamibrainhealth.org/downloads/the-quick-dementia-rating-system-qdrs-patient-and-informant-versions/
AQ 34 , 35	Developed as primary care tool to detect cognitive impairment due to AD.	The AQ is an informant‐based assessment consisting of 21 yes/no questions that can be administered in ≈ 3 minutes. The individual items are divided into the domains of memory, orientation, functional ability, visuospatial, and language. Items that receive a “yes” response are given 1 point; six items particularly associated with clinical AD more weighted and are given 2 points. The total AQ score ranges from 0 to 27; normal ≤4; MCI 5–14; AD dementia ≥15.	Available at: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3207359/pdf/nihms325035.pdf 35
ECog 36 , 37	Assesses functional abilities that are linked to specific cognitive abilities.	39‐item questionnaire can be given to informants and separately to patients. Scoring produces one global factor and six domain‐specific factors. Subsequent studies support validity of short‐form ECog‐12 in discriminating against people with dementia from cognitively normal individuals, but less sensitivity for MCI. Validated against other measures of functional and neuropsychological impairment.	The original ECog and ECog‐12 are detailed in Farias et al. 36 and Tomaszewski Farias et al. 37
CFI 38	The CFI was developed to facilitate evaluation of cognitive symptoms in dementia prevention studies.	14‐item questionnaire focused on change in cognitive and functional abilities over the previous year, which is completed by patient or separately by informant. Total scores range from 0 to 14 (yes = 1, no = 0, and maybe = 0.5), with higher scores indicating greater subjective cognitive complaints. In people with MCI, informant rating is useful for prognosis. 39	Full questionnaires are available in Li et al. 39
Cognitive Change Index 40	Originally developed from measures to detect subjective cognitive decline	A 20‐item questionnaire asking patients and informants to separately rate change in cognitive function compared to the previous 5 years on a scale of 1 (no change) to 5 (severe change). Questions cover memory, executive function, and language. Validation study showed that scores from informant reports are abnormally elevated in people with MCI or dementia.	Available from the authors upon request.
Cambridge Behavioural Inventory 41 , 42 , 43	Developed to assist in the differential diagnosis of different forms of dementia.	A 45‐item informant‐completed questionnaire that obtains information about a range of cognitive, mood and behavioral, and daily functional symptoms. May be best for specialty settings or some general practice settings.	Available at https://www.sydney.edu.au/brain-mind/resources-for-clinicians/dementia-test.html
SIST‐M 44 , 45	Developed as an interview (≈ 25 minutes) for history of symptoms of cognitive and functional impairment	SIST‐M has high reliability against the CDR score in people with MCI and mild dementia. An informant‐rated questionnaire has also been developed. Best for specialty settings.	Available in Okereke et al. 44 , 45

Abbreviations: AD, Alzheimer’s disease; AD8, Eight‐item Information nterview to Differentiate Aging and Dementia; AQ, Alzheimer’s Questionnaire; CDR, Clinical Dementia Rating; CFI, Cognitive Function Instrument; ECog, Everyday Cognition; IQCODE, Informant Questionnaire on Cognitive Decline in the Elderly; MCI, mild cognitive impairment; MMSE, Mini‐Mental State Examination; QRDS, Quick Dementia Rating Scale; SIST‐M, Structured Interview and Scoring Tool–Massachusetts Alzheimer’s Disease Research Center.

A critical element of the compilation of a history of cognitive or behavioral symptoms is its profile of characteristics, intensity, temporal course, and impact. Patients and informants may struggle to explain how symptoms first appeared; how they may have evolved over time in frequency, duration, and intensity; or whether the symptoms were episodic or ever‐present but may have become more noticeable or troubling. It is very common for patients and companions to frame their history‐telling around an event such as a surgery or a major psychosocial trauma that they concluded was solely causal. It can be very challenging for a clinician to dissociate the description of symptoms or behaviors of the patient from the patient’s or care partner’s view of a possibly (but in many cases not likely solely) causal mechanism. Patients, informants, and many clinicians with limited proficiency with dementia assessment often attribute changes in cognition, daily activities, behavior, or sensorimotor function to “normal aging” or to anxiety, mood, or sleep disorders. While acquisition of some of the critical historical information could be captured using algorithmic approaches, the clinician’s understanding of a history of the insidious development of mid‐ to late‐life cognitive or behavioral symptoms is an iterative process, truly an art anchored in clinical experience, diligence, and judgment. Proficiency in this art can be facilitated by a comprehensive and structured approach, but for it to be patient centered and most beneficial, this triadic dialogue, which also provides ample opportunities for psychoeducation, cannot be reduced to an algorithmic inquisition.

1.3. History of mood and/or behavioral symptoms

In many patients, behavioral or mood‐related (neuropsychiatric) symptoms are an early feature of neurodegenerative disease and may or may not be recognized by patients or informants as part of the illness under evaluation. ⁴⁶ , ⁴⁷ In many cases, the patient or care partner may not recognize them as being related to cognitive decline, a condition or brain disease, and the clinician must probe for these or other neuropsychiatric symptoms. In addition to providing diagnostic information, symptoms such as depression, anxiety, delusions, hallucinations, agitation, or obsessive‐compulsive behavior may offer targets for symptomatic treatment. Yet the clinician must also make sure that the words being used by the patient or informant to report on symptoms are consistent with the observed changes in behavior. Concerns about “personality change” need to be discussed to determine whether they arise from apathy, depression, anxiety, hallucinations, delusions, disinhibition, impulsivity, compulsive behavior, or loss of empathic concern. Similarly, an informant’s impression that a patient who is no longer interested in previous activities is depressed might arise when the patient’s change in affect and behavior is rather due to apathy and executive dysfunction. A clinician experienced with these problems who carefully probes the history will often be able to differentiate apathy and executive dysfunction from symptoms consistent with major depressive disorder. Correspondingly, reports of the patient appearing anxious, hiding items and then not knowing where they are, being avoidant, and “forgetting” or resisting to take medications or to eat may, upon a deeper dive by a probing clinician, be discovered to be due to paranoid delusions or hallucinations (e.g., see recently updated International Psychogeriatric Association [IPA] criteria for psychosis in major and mild neurocognitive disorders ⁴⁸ ) instead of anxiety and memory dysfunction. The approaches to psychoeducation, counseling, management, and care differ greatly based on these different impressions and conditions.

Neuropsychiatric symptoms often precede cognitive decline, and they commonly increase in frequency and intensity as neurodegenerative dementing conditions progress. ⁴⁹ The 13‐item Neuropsychiatric Inventory (NPI) ⁵⁰ , ⁵¹ is the most widely used instrument to survey many of these symptoms. Other validated instruments offer structured questionnaires that can assist in the organization and reporting of symptoms of mood or behavioral change

TABLE 2.

Validated instruments to assist in the assessment of neuropsychiatric symptoms in AD/ADRD.

Name	Format	Considerations
General instruments for neuropsychiatric symptoms
NPI‐Q 51	12‐item questionnaire completed by informant; each item is first rated as “present/absent” (Yes/No); if present, severity is rated on 3‐point scale from mild (1) to severe (3). Severity total range: 0–36 (none‐max) Symptom Distress (how much it affects informant/care partner): rated on 6‐point scale (0–5). Distress range: 0–60 (none‐max)	Covers broad range of symptoms/behaviors. It is a modified (abbreviated) version of the NPI. The 12 neuropsychiatric domains assessed are: delusions; hallucinations; agitation/aggression; depression/dysphoria; anxiety; elation/euphoria; apathy/indifference; disinhibition; irritability/lability; motor disturbance (e.g., pacing, picking, repetitive motor behaviors); night‐time behaviors; and appetite/eating. Suitable for both detection and tracking progression/monitoring. Provides information regarding severity of symptoms (how noticeable it is in the patient) and the amount of distress it is causing the care partner/informant. Some modified versions (e.g., NACC UDS) only gauge symptom severity. It can take 5 to 10 minutes depending on proficiency of administrator and familiarity of informant and whether both severity and distress are elicited. A total score can also be derived by multiplying severity score and distress score for each item and summing across all items. Available at: https://www.alz.org/media/Documents/npiq‐questionnaire.pdf
NPI 52	12 items administered in a structured interview to informant with ratings similar to those above (NPI‐Q)	Requires training and proficiency to administer. Used widely in clinical research and more suitable administration to specialist setting; NPI‐Q provides good proxy in clinical setting. Available at: https://www.alz.org/media/Documents/npiq‐questionnaire.pdf
BEHAVE‐AD 53 , 54	25 item scale administered to informant; presence of symptoms and impact on patient for each item rated from 0 to 4 (not present to severe) Global impact severity on care partner/informant is rated for each item from 0 to 4 (not troubling to severely troubling)	Covers broad range of symptoms/behaviors. Mostly used in clinical research setting and more suited to specialist setting; ≈ 20 to 25 minutes to administer; solicits presence of behavioral symptoms and their impact. Suitable for detection, staging, and tracking progression/monitoring.
MBI‐C 17	34‐item questionnaire structured to be consistent with the five domains in the MBI criteria: decreased motivation, emotional dysregulation, impulse dyscontrol, social inappropriateness, and abnormal perception or thought content. If a question is endorsed as present, it is rated mild, moderate, or severe.	Questionnaire to be used primarily by family members or other close informants to systematically measure behavioral changes exhibited by older adults that might precede the onset of dementia. It was specifically designed to: (1) operationalize the MBI concept; (2) measure a selected list of neuropsychiatric symptoms which may help identify prodromal illness; and (3) help predict risk of dementia due to AD or ADRD. The primary goal is case detection of a behavioral pre‐dementia state not better captured by other diagnostic criteria.
Depression instruments
GDS 55	15 items Patient self‐administered (but can be administered to patient) Yes/No responses	Quick (3–5 minutes) screening tool for depressive symptoms and depression in older adults, public domain. Scores of 5–8 suggest potential for mild depression; 9–11 moderate depression; 12–15 severe depression. Suitable for detection (and abbreviated monitoring) in MCI and mild dementia. Less suitable for more advanced and severe dementia and individuals with poor comprehension; and for monitoring severe depression.
PHQ‐9 56	9 items Patient self‐administered (clinician verified) Each item rated 0–3: 0 = not at all; 1 = several days; 2 = more than half the days; 3 = nearly every day Range 0–27 (no depression–severe depression)	Quick (3–5 minutes) screening, diagnostic, and monitoring tool for depressive symptoms and depression in older adults widely use in primary care. Has been validated in individuals with MCI/dementia Scores of 5–9 suggest mild depression; 10–14 moderate depression; > 14 moderately severe/severe depression. Also available in shorter 4‐ and 2‐item versions. Suitable for detection and monitoring in MCI and mild dementia. Less suitable for more advanced and severe dementia and individuals with poor comprehension. Available download from: https://www.mdcalc.com/calc/1725/phq9-patient-health-questionnaire9
CSDD	19 items Administered to patients and care partner (patient does not need to be able to answer for scale to be completed. Each item rated: 0 = absent; 1 = mild to intermittent; 2 = severe Scores range from 0 to 38 (none–max depressive symptoms)	Well suited for detecting, tracking progression, monitoring depression, and depressive symptoms across severity spectrum of MCI‐dementia. Scores of > 11 suggestion probable depression.
Anxiety Instruments
PSWQ‐A 57	8 items Patient self‐administered (can also be administered to care partner/informant) Ratings for each item are on a 1–5 Likert scale (1 = not at all typical of me; 5 = very typical of me) Range 8–40 (no anxiety–severe anxiety)	A widely used abbreviated version of the 16‐item PSWQ that was developed as a screening tool to assess worry symptoms and anxiety in older adults. It is in the public domain. Cut‐off of 17 has been suggested for detection of significant anxiety in individuals with mild/moderate dementia. 58
GAI 59	20 items Patient self‐administered (can also be administered to care partner/informant) All items answered dichotomously Yes (1) or No (0) Range: 0–20 (none–max severity)	Was developed to screen for anxiety symptoms in older/geriatric population, has been subsequently studied in mild to moderate AD dementia (a cut off score of 8 has been suggested to detect significant in mild/moderate dementia 58 ). Suitable for detection of symptoms. It is copyrighted and fee may be required for clinical use. A short form with five items (GAI‐SF) is also available for very brief screening.
Agitation instruments
CMAI	Presence and frequency of 29 behaviors Does not rate severity Administered to care partner informant.	Broadly covers presence of agitation and related disruptive behaviors such as verbal aggression, repetitiveness, screaming, hitting, grabbing, and sexual advances. Requires training to administer, more suited to specialist setting and clinical research.

Abbreviations: AD, Alzheimer’s disease; ADRD, Alzheimer’s disease related dementias; BEHAVE‐AD, Behavioral Pathology in Alzheimer’s Disease Depression Scale; CMAI, Cohen Mansfield Agitation Index; CSDD, Cornell Scale for Depression in Dementia; GAI, Geriatric Anxiety Inventory; GDS, Geriatric Depression Scale; MBI‐C, Mild Behavioral Impairment Checklist; MCI, mild cognitive impairment; NPI, Neuropsychiatric Inventory; NPI‐Q, Neuropsychiatric Inventory‐Questionnaire; PHQ‐9, Patient Health Questionnaire; PSWQ‐A, Penn State Worry Questionnaire Abbreviated.