Abstract
With the recent approval of disease‐modifying treatments for mild cognitive impairment (MCI) and mild Alzheimer's disease (AD) by the United States Food and Drug Administration (FDA), Medicines and Healthcare products Regulatory Agency (MHRA), European Medicine Agency's Committee for Medicinal Products for Human Use (EMA/CHMP) entities, there is a growing sense of urgency and renewed efforts to reassess and understand what constitutes a clinically meaningful benefit in the context of new treatments for AD care, despite the discordance between regulatory entities in regulatory decision‐making. While the concept of minimal clinically important difference (MCID) was introduced many years ago, there remains an ongoing debate about how best to evaluate and define clinical benefit in the context of emerging and new therapies for dementia. In this perspective piece, we assess how MCID can be applied to common endpoints and identify areas where MCID application or generation could be useful to enable a better valuation of therapeutic innovation. We offer recommendations for greater consistency in measures used to define MCID, and encourage the prioritized use of patient‐reported measures in early AD to build fieldwide consensus for MCID estimation methods and application in AD.
Highlights
There is no gold standard or field‐wide consensus on what constitutes a clinically meaningful change in Alzheimer's disease (AD) progression trajectories.
Anchor‐based minimal clinically important difference (MCID) may be used as a tool that can be leveraged for greater contextualization of the clinical relevance of a treatment effect.
Patient‐reported outcomes (PROs) should be used to define MCID, particularly within mild cognitive impairment (MCI), prodrome/mild AD groups.
Greater consistency is needed in the outcome measures used to detect cognitive and functional change to define MCID. This will enable MCID comparisons and support replications of MCID estimates across AD populations.
Observational data can augment the clinical characterization and impact of treatment effect and help establish a “ground truth” MCID.
MCID estimates for AD outcomes may be used in regulatory submissions to help contextualize the importance of a statistically significant treatment effect.
Keywords: clinically meaningful benefit, cognitive impairment, dementia, disease‐modifying treatments, minimally clinically important difference
1. BACKGROUND
The prevalence of Alzheimer's dementia (AD) is increasing at an alarming rate globally, with an estimated 50 million people living with some form of dementia. 1 , 2 Characterized by cognitive and functional deterioration and changes to behavioral abilities that worsen over time 3 , 4 , 5 dementia progression places a considerable burden on individuals, their care partners, families, healthcare systems, and society at‐large. With the increasing global increase in dementia and rising costs associated with AD treatments and care, there is a palpable sense of urgency to accelerate the development of treatments that demonstrate a beneficial impact on progression of the disease.
Determining what represents a meaningful treatment effect when statistical significance is achieved but clinical relevance is uncertain remains a hotly debated topic within the AD landscape. Randomized clinical trials (RCTs) rely on rigorous, well‐powered study designs to assess the clinical efficacy of a treatment by demonstrating superiority, non‐inferiority, or equivalence compared to placebo. The challenge is that a statistically significant treatment effect does not offer insight into the magnitude or clinical importance of an observed treatment effect. There are two disease‐modifying therapies (DMTs) now available in the United States for early AD, but the clinical relevance of these DMTs has been called into question 6 , 7 due to conflicting study results, drug safety risk, and the unclear relationship between Aβ reduction and cognitive improvements. Given this controversy and the uncertainty associated with clinical trial outcomes for other treatments in the pipeline, there is renewed interest in defining what constitutes a clinically relevant benefit, mainly focusing on the patient perspective now central to evaluating the therapeutic value of development pursuits. 8
Minimal clinically important difference (MCID) is defined as the smallest change on a measure that is reliably associated with a meaningful change in how a patient feels, functions, or thrives. 8 While the concept was coined decades ago, 9 AD researchers are increasingly using various MCID methods, notably anchor‐based, distribution‐based, Delphi panel, MCID methods, a combination, or triangulation 10 , 11 of methods to establish if the efficacy of a determined treatment intervention or newly approved therapy has clinical relevance. Anchor‐based methods relate patient perceptions to the change on a target measure where the anchor is preferentially patient‐reported measures although anchors based on clinician‐reported outcome measures, that is, an assessment completed by a clinician, are often used. Distribution‐based methods evaluate the scale properties and spread of scores for a given measure, but the approach does not provide information about the magnitude of an observed effect. Delphi methods rely on expert and clinician input to define meaningful change thresholds. Establishing whether a treatment achieves statistical significance and demonstrates a beneficial treatment response that is clinically relevant or effective relative to a yardstick could provide invaluable information about what is important to patients, health providers, and other stakeholders. 12 , 13 However, reported MCID estimates for AD are limited and seldom applied to, or later validated in other samples beyond the ones used to establish the MCID, limiting their utility for clinical decision‐making and generalizability to other AD populations.
Currently, there is no fieldwide consensus or gold standard for determining the smallest, detectable amount of change in an AD trial outcome that is meaningful across the AD continuum, signifying a need to establish MCID estimation standards for AD therapeutic intervention initiatives. In this piece, we propose using MCID estimation to assess the value and perceived importance of emerging and new therapies for AD. We propose a framework to standardize the methods used for estimating MCID and to build consensus in AD research. We recommend the consistent use of measures to define MCID along the AD continuum. Furthermore, we recommend the prioritized use of patient self‐report measures as the anchor for anchor‐based MCID methods applied to early and mild AD progression groups to enhance the patient voice and help determine the true value of a treatment. Implementing such recommendations will advance multi‐stakeholder decision‐making and enable the later inclusion of MCID estimates as an instrumental component of the body of evidence required for regulatory consideration, approval, later reimbursement, and treatment management.
2. CONTEXT OF MCID USE: FRAMEWORK FOR APPLICATION OF MCID IN INTERVENTIONAL TRIALS
Clinical trials are central to determining treatment efficacy, high‐frequency safety signals, and are essential for obtaining drug approvals and making treatments widely accessible. In AD randomized controlled trials, reliable clinical outcome assessment measures 14 are routinely used as AD trial endpoints to detect and describe a potential treatment effect, primarily focused on comparisons between groups and relying on a p‐value threshold (i.e., p < 0.05). However, this threshold does not provide insight into the clinical importance of an observed effect. The value of defining MCID for AD populations is to facilitate interpretation of the clinical relevance from study results. If the observed change is greater than the MCID in the context of improvement, it is considered meaningful change; if MCID is smaller than the observed effect, it is not clinically meaningful. Furthermore, MCID estimates provide a threshold range, or smallest numerical value, “cutoff” point that is clinically meaningful rather than a relative value evidenced by a proportional change in a measure compared to a reference point. Typically, MCID values for improvement are smaller than MCID for deterioration when defining MCID for a specific context 15 . A recent review of outcome measures used in RCTs (n = 91) of non‐pharmacological interventions for patients with symptomatic AD found that only 22% of the outcome measures were shared by more than one of the trials included in the review, 16 and is indicative of the need for greater consistency of measures used in AD RCTs. Standardizing the use of reliable measures to detect change is an important component in establishing MCID standards for AD.
In this manuscript, we identify areas that require further investigation, and propose a framework for establishing MCID standards for AD clinical research.
2.1. Consideration of self‐report measures to define MCID
A notable limitation in the AD MCID literature is the lack of patient‐reported outcome (PRO) measures used to define MCID. Clinical judgment is important in determining the course of treatment, but the patient's perspective must be factored more prominently into the AD MCID discourse. A PRO measure, as defined by the United States Food and Drug Administration (FDA), 8 is an assessment or report of a person's health coming directly from that person, without interpretation of the response by a clinician or anyone else. As people with preclinical AD and MCID/AD prodromes can accurately self‐report on their own condition, it is not appropriate to solely rely on clinician‐reported outcome measures or observer‐reported outcome measures (i.e., an assessment completed by an informed observer or caregiver to determine the course and change in treatment response) in this population. In instances where individuals are functioning and maintaining their activities of daily living (ADLs) and responsibilities, PRO measures serve as valuable tools for defining MCID in earlier stages of AD progression.
Anti‐amyloid DMTs have previously included quality of life, 17 , 18 and behavioral measures such as the clinician‐reported, quality of life in AD (QoL‐AD) 19 (range 13–52; higher scores indicate better QOL) and the Neuropsychiatric Inventory – 10 items (NPI‐10) (sum score range 0–24, with higher scores indicate more behavioral disturbance‐has also been used, 20 and self‐reported measures such as the EuroQoL –5 Dimensions (EuroQoL‐5D) 21 versions (health index with 1 representing full health and 0 representing a state as bad as death), World Health Organization – Quality of Life (WHOQoL) (range 0–100, with higher scores indicating better health status). 22 Despite the use of these measures in non‐pharmacological and in DMTs and RCTs, they have not been used to define MCID. There is a need to establish MCID estimates using these measures to contribute to the body of evidence. Following the FDA's patient‐focused drug development guidance to incorporate the patient perspective in clinical trial designs, 8 it is no longer sufficient to merely include PRO measures as an exploratory measure in the schedule of assessments. In cases where there is sufficient sample size to support the detection of a treatment effect, MCID estimates can be defined using clinician‐derived MCID and/or patient‐reported ‐derived MCID to evaluate whether the magnitude of an impact is perceived similarly. Clinician‐derived MCID estimates enable clinician‐patient dialogue for making better informed treatment decisions. Patient‐derived MCID could be instrumental in informing medicinal product labeling and reimbursement considerations to help determine what is communicated about the product's effectiveness.
Proposed areas ripe for anchor‐based MCID application and replication are summarized in Table 1.
TABLE 1.
Proposed areas for anchor‐based MCID replication and application.
| Measure a (range, interpretation) | MCID methods | Reported MCID thresholds for improvement | Reported MCID thresholds for deterioration | Areas for further investigation |
|---|---|---|---|---|
| Clinician‐reported outcome measures | ||||
| CDR‐SB 34 (0–18; higher = worse) | Anchor‐based | X |
|
|
|
ADAS‐Cog 35 (0–70; higher = worse) |
Anchor‐based | X |
|
|
|
iADRS 36 (0–144; lower = worse) |
Anchor‐based | X |
|
|
|
MMSE 37 (0‐30; lower = worse) |
Anchor‐based | X | 1‐ to 3‐point decrease for deterioration 22 , 23 , 24 , 31 | Not recommended for AD MCID estimation |
|
QoL‐AD 19 (13–52, higher = better) |
X | X | X |
|
|
NPI‐10 20 (0–120, higher = worse) |
X | X | X |
|
| Patient‐reported outcome measures | ||||
|
EQ‐5D‐5L 21 (0–100, higher = better) |
X | X | X |
|
|
WHOQoL 22 (0–100; higher = better) |
X | X | X |
|
Note: X, not reported/not established.
Abbreviations: AD, Alzheimer's disease; ADAS‐Cog, Alzheimer's Disease Assessment Scale‐Cognitive Subscale; CDR‐SB, Clinical Dementia Rating Scale‐Sum of Boxes; EQ‐5D‐5L, EQ‐5 Dimensions‐5 Levels; iADRS, Integrated Alzheimer's Disease Scale; MCI, mild cognitive impairment. MCID, minimal clinically important difference; MMSE, Mini‐Mental State Examination; NPI‐10, Neuropsychiatric Inventory‐ 10 items; QoL‐AD, Quality of life‐Alzheimer's Disease; WHO‐QoL, World Health Organization‐Quality of life.
Number corresponding to relevant reference.
2.2. Using MCID to determine responders versus non‐responders
Determining the level of treatment response sufficient to produce a beneficial effect is an essential aspect of clinical trial conduct. The treatment response can inform Phase III trial planning and patient selection criteria. In comparative studies, treatment response is generally defined as a mean difference, change, or improvement in status from baseline. 23 In these studies, the treatment effect is typically assessed using a binary outcome indicative of treatment response.
Level of responsiveness can be influenced by various factors including sex, concurrent illnesses, and long‐term use of other medications, 24 , 25 and may impact treatment outcomes, making MCID estimation much more complex. Given these complexities coupled with approximately one‐third of AD patients not responding to pharmacotherapies, 26 it is important for the AD research community to prioritize and explore patient‐derived and clinician‐derived MCID for populations with varying clinical characteristics.
In cases where a defined MCID for improvement is not met, estimates for MCID for disease‐modifying and symptomatic therapies are important to assess an individual's status. When symptoms remain unchanged or are too small to achieve MCID, this can be interpreted as an indication of preservation of function in the context of a treatment response. Stabilization is an important goal, particularly from the patient and caregiver perspective, and could be perceived as an important and beneficial outcome, given there is no demonstrable decline in an individual's status. Knowing the proportion of individuals who respond to a treatment compared to those who do not respond positively to the treatment needs to be examined more fully. This would enable individual‐level MCID threshold interpretation of a treatment response. A goal for the AD research community is to ascertain the probability of benefit 27 where an absolute difference in outcomes is computed by subtracting the percentage of patients who improved with placebo from the percentage who improved with treatment and would complement MCID definitions and applications. Other statistical methods for determining responsiveness can be used to complement interpretation of MCID estimates but will not be highlighted here given the focus of this manuscript on MCID methodologies for application in AD populations.
2.3. May MCID be established by proxy? Incorporating the caregiver perspective
The perspective of caregivers is crucial, yet is often overlooked when defining MCID for AD. Caregivers can offer valuable insights as knowledgeable informants, particularly regarding ADLs and changes in function and cognition over time. However, the limited use of caregiver input partly stems from concerns that they could be biased in their perceptions of the patient's condition, which may lead to an overestimation or underestimation of the actual meaningful change experienced by the patient. Although the FDA discourages the use of proxy measured for symptoms only known by the patient, 8 there are instances where caregiver information is useful. Observer‐reported outcome measures often capture different information and concepts compared to a clinician or a PRO measure and can therefore be recommended as a proxy‐derived MCID (P‐MCID). A P‐MCID may only be estimated when (i) clinician and patient‐derived MCID differ, or fall closely outside of the MCID thresholds estimated; (ii) an observer‐reported outcome version of the scale used in the study contains the same construct and concepts as the clinician‐reported outcomes and PROs; (iii) ratings are largely focused on observable traits of a patient within the context of an interventional trial. However, not all individuals with AD have a knowledgeable informant to report on their behalf. In these circumstances, clinician‐derived MCID should be prioritized. Some studies suggest using various sources of information to triangulate information to derive guidelines for interpreting change scores on health outcome measures, 10 , 11 and defining a proxy‐derived MCID could complement study results, clinical judgement, and the patient perspective. Caregivers provide a unique perspective to the AD lived experience, but caregiver‐focused, observer‐reported outcome measures may be used exploratorily in conjunction with other clinician‐reported measures to complement MCID‐derived estimates for therapeutic development programs and clinical trial designs.
2.4. Using observational data to define MCID along the AD continuum
Reported MCID for AD varies by stage of AD severity. Evidence indicates that MCID values differ depending on the severity of illness. (See Hamilton et al, for an example in Huntington's disease 28 ). Therefore, individuals in the early stages of AD disease progression may have smaller MCID estimates compared to those in later stages of disease. For example, a two‐point MCID on a given measure in the direction of deterioration in early AD may herald a change to one's work status, daily activities, or responsibilities whereas in later stages of disease progression, a two‐point change may not be impactful to warrant a change to one's existing care.
Within the AD landscape, there is a need to replicate previously reported MCID based on AD severity to assess and establish group equivalence. This is where observational data could prove useful. Leveraging observational data in the absence of a given treatment may provide a framework, or “ground truth” for defining clinical relevance in the absence of a specific treatment. Analyses could use age‐ and AD severity‐matched groups to define and compare MCID. This would be invaluable in informing sample size estimates for clinical trial planning and defining the time point for when a therapeutic intervention is introduced.
2.5. Clinical meaningfulness from the payer perspective
Economic evaluation frameworks play an important role in evaluating the economic implications of a given treatment. They help determine the utility, effectiveness, and benefits of a treatment, which are essential factors in making informed decisions regarding access, reimbursement, and healthcare coverage worthiness. Such evaluations consider various factors, including the treatment's impact on patient outcomes, aggregated healthcare costs, and societal benefits. While health economists and healthcare payers define clinical meaningfulness in terms of the health benefit and cost effectiveness of a treatment, 29 the most used health economic measure of effectiveness is to express the effect of a treatment intervention in terms of a single index, preference‐weighted generic health outcome, or quality‐adjusted life‐year (QALYs). The use of QALYs raises ethical questions about its appropriateness for determining coverage, reimbursement, and incentives. The United States government has shown reluctance to use QALYs, opposing their use in health‐care decision‐making and banning their use in federal programs. This is due to concerns about the discriminatory nature of placing a monetary value on one's life. 30 Using the same PRO measure in MCID and QALYs estimates in AD populations could complement and contextualize the value and meaningfulness of a treatment, but further research is needed.
2.6. Using MCID as a component of the regulatory submission package
In preparation for regulatory submission and approval, it is important to consider how MCID can be used as part of the regulatory engagement discussion. Based on 2018 FDA draft guidance, a treatment for mild to moderate dementia is considered effective if there is an improvement on measures of cognition, function, and a clinician‐based global impression. 8 For MCI groups, a statistically significant change on a specified composite measure of cognition and function would be needed to demonstrate substantial effectiveness. CHMP/ EMA's requirements, although seemingly aligned with the FDA 31 have proven in real life much more stringent. At the time of writing, none of the anti‐amyloid disease modifying treatments have been approved.
The FDA has also indicated support of anchor‐based methods to determine the meaning of individual‐level change, or responsiveness. The 1998 guidance was issued in response to the FDA's Modernization 23 Act of 1997 (FDAMA) (Pub. L. 105‐115), which states the substantial evidence requirement for effectiveness, which had been largely interpreted as 2 well‐controlled trials, could also be met by a single trial plus confirmatory evidence. 14 Therapeutic development programs require a thorough risk‐benefit assessment to demonstrate confirmatory evidence. However, MCID estimation information could be leveraged and included in the submission package to better contextualize a treatment effect, and serve as an important component of the substantial evidence of treatment effectiveness 32 needed to support regulatory consideration and approval.
We will briefly discuss and summarize clinical outcome assessments often used to define MCID and identify areas that need to be explored more fully
3. CLINICAL OUTCOME ASSESSMENTS AS ENDPOINTS FOR DETERMINING CLINICAL MEANINGFULNESS FOR AD
Based on a recent review of studies that derived MCID for AD trials by Muir et al., 33 the Clinical Dementia Rating Scale Sum of Boxes (CDR‐SB), 34 Alzheimer's Disease Assessment Scale‐Cognitive Subscale (ADAS‐Cog), 35 Integrated Alzheimer's Disease Rating Scale (iADRS), 36 and Mini‐Mental State Examination (MMSE) 37 are consistently used as outcomes, endpoints in AD trials, and used for anchor‐based MCID. These clinician‐reported measures assess language, comprehension, memory, orientation, and visual‐spatial abilities with varying levels of specificity and sensitivity, with the CDR‐SB being most preferred followed by the ADAS‐Cog 38 , 39 , 40 , 41 and iADRS measures. ADAS‐Cog measures have also been used to define detectable change, 42 whereas MCID detects the smallest, important change and is an important distinction. For this paper, we focused on MCID methods and use in AD research only.
The empirical evidence for meaningful change estimates using these measures along the AD continuum is summarized in Table 1 and highlights areas where further investigation is needed. For example, of the studies using these measures to define MCID for MCI and AD, the majority relied on clinician‐reported outcome measures to define the anchor, linking the statistical significance of a given treatment effect to the clinical judgement of a treatment response, not the patient's perception, of that response. In preclinical AD and MCI/AD prodromes, individuals earlier in the disease course notice changes in their abilities, however subtle, and should have a say in the course and direction of their care and empower individuals to have a voice in their well‐being until they are no longer able to knowledgeably report on how they feel, function and thrive. Undoubtedly, in moderate to more severe AD progression, it is appropriate to rely on clinician‐derived MCID thresholds and to a lesser extent, observer‐reported measures, that is, an assessment completed by an informed observer or caregiver to determine the course and change in treatment response.
We briefly highlight some important considerations regarding the use of these measures for MCID applications.
The CDR‐SB (range 0–18, with higher scores suggesting the presence of dementia), 10 , 34 is favored by regulatory entities and researchers for MCID estimation given it has been used in pivotal trials and shown to be sensitive to change. 43
The ADAS‐Cog (score range 0–70, with higher scores indicating greater cognitive impairment) is commonly used in AD trials and more complex to administer but is more sensitive and less influenced by level of education and language skills than the MMSE. 44 Thus, the MMSE is not as readily used as a primary endpoint, partly due to susceptibility to varying education levels 45 and unstable inter‐rater reliability 46 but is still used as a secondary, or supplementary outcome measure in AD trial designs, However, in case where other ADAS‐Cog versions are used, researchers must reference the applicable ADAS‐Cog version associated with established MCID estimates, as much and where able, to better contextualize a predicted or observed treatment response. In cases where the expanded, more responsive version is used, clinical meaningfulness could be derived from the measure where group equivalence could be determined. This would require an appraisal of the study's enrollment criteria and demographic and clinical characteristics to identify similarities in MCID thresholds across AD groups.
The iADRS (range 0–144, with lower scores indicative of greater severity) is a combined assessment that includes scores from the ADAS‐Cog and the AD Cooperative Study—instrumental Activities of Daily Living (ADCS‐iADL). iADRS has acceptable psychometric properties and shown to be effective in capturing disease progression and is a reasonable cognitive and functional endpoint that can be used in clinical trials. 47
Replicating MCID estimates remains a challenge given the variability in methodologies, AD samples, and measures used to define MCID. However, we can use established evidence to further advance the use and integration of MCID methods in AD trial designs and clinical care.
Table 2 summarizes our recommendations for MCID estimation to promote consistency and standardization in estimation methodologies used and applied to AD severity groups. This framework is designed to provide a common reference and standards to enable broader application to inform decision‐making.
TABLE 2.
Recommended standards for AD MCID estimation.
| Measure a (range, interpretation) | |
|---|---|
| Sample size | Adequate sample size (N > 100) |
| Scale selection | Identified scale is shown to be reliable, valid and sensitive to change 15 and suitable for use in the intended, target population |
| Anchor‐based considerations |
|
| Distribution‐based considerations |
Should be used when:
|
| Triangulation 11 | Consult combination of anchor‐based and distribution‐based methods and available sources of relevant information (e.g., PoB, 26 ) to assess the convergence and/or divergence of the observed effect and across MCID estimates to support statistical inference |
| Prioritized Domains |
Preclinical/early MCI/mild AD
Moderate to Severe AD
|
| Interpretation 28 |
|
| Reporting standards |
|
Abbreviation: AD, Alzheimer's disease; ADLs, activities of daily living; iADLs, instrumental Activities of Daily Living; MCI, mild cognitive impairment; MCID, minimal clinical important difference; POB, probability of benefit; PRO, patient‐reported outcome; QALYs, quality‐adjusted life years; SD, standard deviation.
Number corresponding to relevant reference.
4. CONCLUSION
MCID is a tool for establishing the value of a treatment intervention and is crucial for evaluating the effectiveness of treatments and interventions on AD progression and quality of life. Using a quantifiable measure of meaningful change can be vital for the development of new and emerging treatments and the advancement of AD research.
Greater consistency in the measures used to define MCID in AD research is needed. Most measures used to define MCID are clinician‐reported outcome measures in more advanced stages of progression. In the early stages of AD, the field is ripe for the prioritization and use of sensitive PRO measures to amplify the patient perspective and to determine the value of a treatment. This can help define MCID to improve our understanding of the earlier stages of AD progression. Additionally, it can complement the cost‐effectiveness and cost‐utility analysis of interventions by using consistent measures to assess the value and potential benefit of treatments. Reported MCID values in the extant AD literature are seldom replicated making it difficult to determine what MCID estimates are suitable for decision‐making. The use of variable MCID thresholds makes it challenging to draw meaningful comparisons with other study samples that do not share the same clinical characteristics or level of severity. Therefore, it is important to carefully consider the population's characteristics when estimating MCID for specific progression groups.
To address these limitations, we proposed a clear set of standards that can be used as a reference for AD clinical research and identified areas where further investigation is needed. As the methodology of calculating MCID improves and available datasets become larger and more diverse, MCID thresholds will better reflect the patient perspective. Researchers must continue to refine, establish, and replicate MCID outcomes to improve care and clinical trial outcomes for individuals living with AD.
CONFLICT OF INTEREST STATEMENT
The authors declare no conflicts of interest. Author disclosures are available in the Supporting Information.
Supporting information
Supporting Information
ACKNOWLEDGMENTS
J.L.H., R.L.M.F., C.S., and N.S. are employees of CHDI Management, the company that manages the scientific activities of CHDI Foundation. CHDI Management was not involved in the writing of the article.
Hamilton JL, Fuller RLM, Modi N, Sampaio C. Utilizing MCID for evaluating clinical relevance of AD therapeutic interventions. Alzheimer's Dement. 2025;11:e70138. 10.1002/trc2.70138
REFERENCES
- 1. Gustavsson A, Norton N, Fast T, et al. Global estimates on the number of persons across the Alzheimer's disease continuum. Alzheimers Dement. 2023;19(2):658‐670. [DOI] [PubMed] [Google Scholar]
- 2. 2021 Alzheimer's disease facts and figures. Alzheimers Dement. 2021;17(3):327‐406. [DOI] [PubMed] [Google Scholar]
- 3. Cohen S, Cummings J, Knox S, Potashman M, Harrison J. Clinical trial endpoints and their clinical meaningfulness in early stages of Alzheimer's disease. J Prev Alzheimers Dis. 2022;9(3):507‐522. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Schneider LS, Goldberg TE. Composite cognitive and functional measures for early stage Alzheimer's disease trials. Alzheimers Dement. 2020;12(1):e12017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Cummings J. Meaningful benefit and minimal clinically important difference (MCID) in Alzheimer's disease: open peer commentary. Alzheimers Dement. 2023;9(3):e12411. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Mullard A. FDA approval for Biogen's aducanumab sparks Alzheimer disease firestorm. Nat Rev Drug Discov. 2021;20(7):496. [DOI] [PubMed] [Google Scholar]
- 7. Alexander GC, Emerson S, Kesselheim AS. Evaluation of aducanumab for Alzheimer disease: scientific evidence and regulatory review involving efficacy, safety, and futility. JAMA. 2021;325(17):1717‐1718. [DOI] [PubMed] [Google Scholar]
- 8. US Department of Health and Human Services, Food and Drug Administration, Center for Drug Evaluation and Research, Center for Biologics Evaluation and Research, and Center for Devices and Radiological Health . Guidance for industry: patient‐reported outcomes measures: use in medical product development to support labeling claims: draft guidance. Health Qual Life Outcomes. 2006;4:79. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Jaeschke R, Singer J, Guyatt GH. Measurement of health status. Ascertaining the minimal clinically important difference. Control Clin Trials. 1989;10(4):407‐415. [DOI] [PubMed] [Google Scholar]
- 10. Borland E, Edgar C, Stomrud E, Cullen N, Hansson O, Palmqvist S. Clinically relevant changes for cognitive outcomes in preclinical and prodromal cognitive stages: implications for clinical Alzheimer trials. Neurology. 2022;99(11):e1142‐e1153. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Leidy NK, Wyrwich KW. Bridging the gap: using triangulation methodology to estimate minimal clinically important differences (MCIDs). COPD. 2005;2(1):157‐165. [DOI] [PubMed] [Google Scholar]
- 12. Molnar FJ, Man‐Son‐Hing M, Fergusson D. Systematic review of measures of clinical significance employed in randomized controlled trials of drugs for dementia. J Am Geriatr Soc. 2009;57(3):536‐546. [DOI] [PubMed] [Google Scholar]
- 13. Webster L, Groskreutz D, Grinbergs‐Saull A, et al. Core outcome measures for interventions to prevent or slow the progress of dementia for people living with mild to moderate dementia: systematic review and consensus recommendations. PLoS One. 2017;12(6):e0179521. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Food and Drug Administration . Patient‐focused drug development: selecting, developing, or modifying fit‐for‐purpose clinical outcome assessments: guidance for industry, food and drug administration staff, and other stakeholders. Food and Drug Administration; 2022. [Google Scholar]
- 15. Hays RD, Woolley JM. The concept of clinically meaningful difference in health‐related quality‐of‐life research. How meaningful is it? Pharmacoeconomics. 2000;18(5):419‐423. [DOI] [PubMed] [Google Scholar]
- 16. Couch E, Lawrence V, Co M, Prina M. Outcomes tested in non‐pharmacological interventions in mild cognitive impairment and mild dementia: a scoping review. BMJ Open. 2020;10(4):e035980. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Angioni D, Cummings J, Lansdall CJ, et al. Clinical meaningfulness in Alzheimer's disease clinical trials. A report from the EU‐US CTAD task force. J Prev Alzheimers Dis. 2024;11(5):1219‐1227. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Landeiro F, Mughal S, Walsh K, et al. Health‐related quality of life in people with predementia Alzheimer's disease, mild cognitive impairment or dementia measured with preference‐based instruments: a systematic literature review. Alzheimers Res Ther. 2020;12(1):154. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Logsdon RG, Gibbons LE, McCurry SM, Teri L. Quality of Life‐Alzheimer's Disease Scale (QOL‐AD). APA Psych Test; 1999. [Google Scholar]
- 20. Kaufer DI, Cummings JL, Ketchel P, et al. Validation of the NPI‐Q, a brief clinical form of the Neuropsychiatric Inventory. J Neuropsychiatry Clin Neurosci. 2000;12(2):233‐239. [DOI] [PubMed] [Google Scholar]
- 21. Stolk E, Ludwig K, Rand K, van Hout B, Ramos‐Goñi JM. Overview, update, and lessons learned from the international EQ‐5D‐5L valuation work: version 2 of the EQ‐5D‐5L valuation protocol. Value Health. 2019;22(1):23‐30. [DOI] [PubMed] [Google Scholar]
- 22. Development of the World Health Organization WHOQOL‐BREF quality of life assessment. The WHOQOL Group. Psychol Med. 1998;28(3):551‐558. [DOI] [PubMed] [Google Scholar]
- 23. Rombach I, Knight R, Peckham N, Stokes JR, Cook JA. Current practice in analysing and reporting binary outcome data—a review of randomised controlled trial reports. BMC Medicine. 2020;18(1):147. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Briggs R, Kennelly SP, O'Neill D. Drug treatments in Alzheimer's disease. Clin Med. 2016;16:247‐253. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Claxton A, Baker LD, Wilkinson CW, et al. Sex and ApoE genotype differences in treatment response to two doses of intranasal insulin in adults with mild cognitive impairment or Alzheimer's disease. J Alzheimers Dis. 2013;35(4):789‐797. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Wang D, Ling Y, Harris K, Schulz PE, Jiang X, Kim Y. Characterizing treatment non‐responders vs. responders in completed Alzheimer's disease clinical trials. medRxiv. 2023. [PMC free article] [PubMed] [Google Scholar]
- 27. McGraw K.O, Wang SP. A common language effect statistic. Psychol Bull. 1992;111:361‐365. [Google Scholar]
- 28. Hamilton JL, Mills JA, Stebbins GT, et al. Defining clinical meaningfulness in Huntington's disease. Mov Disord. 2023;38(6):1036‐1043. [DOI] [PubMed] [Google Scholar]
- 29. Sánchez Fernández I. Understanding cost effectiveness in neurology. J Neurol Sci. 2023;455:122787. [DOI] [PubMed] [Google Scholar]
- 30. United States. Congress. House . Committee on Veterans' Affairs. Committee on Veterans' Affairs Member Day : Members Day hearing before the Committee on Veterans' Affairs, U.S. House of Representatives, One Hundred Sixteenth Congress, first session, Washington, USA, 10 September, 2019. 106. U.S. Government Publishing Office; 2023. [Google Scholar]
- 31. Morant AV, Vestergaard HT, Lassen AB, Navikas V. US, EU, and Japanese regulatory guidelines for development of drugs for treatment of Alzheimer's disease: implications for global drug development. Clin Transl Sci. 2020;13(4):652‐664. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Center for Drug Evaluation and Research,Center for Biologics Evaluation and Research,Oncology Center of Excellence,Office of the Commissioner . Demonstrating substantial evidence of effectiveness with one adequate and well‐controlled clinical investigation and confirmatory evidence guidance for industry draft guidance. Food and Drug Administration; 2023. [Google Scholar]
- 33. Muir RT, Hill MD, Black SE, Smith EE. Minimal clinically important difference in Alzheimer's disease: rapid review. AlzheimersDement. 2024;20(5):3352‐3363. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Berg L. Clinical Dementia Rating (CDR). Psychol Bull. 1998;24:637‐639. [PubMed] [Google Scholar]
- 35. Kueper JK, Speechley M, Montero‐Odasso M. The Alzheimer's Disease Assessment Scale‐Cognitive Subscale (ADAS‐Cog): modifications and responsiveness in pre‐dementia populations. A narrative review. J Alzheimers Dis. 2018;63(2):423‐444 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Wessels AM, Rentz DM, Case M, Lauzon S, Sims JR. Integrated Alzheimer's disease rating scale: clinically meaningful change estimates. Alzheimers Dement (N Y). 2022;8(1):e12312. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Folstein MF, Folstein SE, McHugh PR. “Mini‐Mental State”. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res. 1975;12(3):189‐198. [DOI] [PubMed] [Google Scholar]
- 38. Howard R, Phillips P, Johnson T, et al. Determining the minimum clinically important differences for outcomes in the DOMINO trial. Int J Geriatr Psychiatry. 2011;26(8):812‐817. [DOI] [PubMed] [Google Scholar]
- 39. Knopman DS, Jones DT, Greicius MD. Failure to demonstrate efficacy of aducanumab: an analysis of the EMERGE and ENGAGE trials as reported by Biogen, December 2019. Alzheimers Dement. 2021;17(4):696‐701. [DOI] [PubMed] [Google Scholar]
- 40. Sims JR, Zimmer JA, Evans CD, et al. Donanemab in early symptomatic Alzheimer disease: the TRAILBLAZER‐ALZ 2 randomized clinical trial. JAMA. 2023;330(6):512‐527. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Lansdall CJ, McDougall F, Butler LM, et al. Establishing clinically meaningful change on outcome assessments frequently used in trials of mild cognitive impairment due to Alzheimer's disease. J Prev Alzheimers Dis. 2023;10(1):9‐18. [DOI] [PubMed] [Google Scholar]
- 42. Schrag A, Schott JM. What is the clinically relevant change on the ADAS‐Cog? J Neurol Neurosurg Psychiatry. 2012;83(2):171‐173. [DOI] [PubMed] [Google Scholar]
- 43. O'Bryant SE, Waring SC, Cullum CM, et al. Staging dementia using Clinical Dementia Rating Scale Sum of Boxes scores: a Texas Alzheimer's research consortium study. Arch Neurol. 2008;65(8):1091‐1095. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44. Kaufmann DM, Geyer HL, Milstein MJ, Rosengard J. Kaufman's Clinical Neurology for Psychiatrists. Elsevier; 2023:111‐158. [Google Scholar]
- 45. Lezak MD, Howieson DB, Loring DW. Neuropsychological Assessment. 4th ed. Oxford University Press; 2004:xiv, 1016 p. p. [Google Scholar]
- 46. Liu KY, Schneider LS, Howard R. The need to show minimum clinically important differences in Alzheimer's disease trials. Lancet Psychiatry. 2021;8(11):1013‐1016. [DOI] [PubMed] [Google Scholar]
- 47. Andrews JS, Desai U, Kirson NY, Zichlin ML, Ball DE, Matthews BR. Disease severity and minimal clinically important differences in clinical outcome assessments for Alzheimer's disease clinical trials. Alzheimers Dement (N Y). 2019;5:354‐363. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supporting Information