The Self-Administered Gerocognitive Examination (SAGE) is a free, scientifically validated cognitive screening tool developed at the Ohio State University Wexner Medical Center. As a brief cognitive screening instrument, it helps clinicians detect signs of cognitive impairment early, including those associated with mild cognitive impairment (MCI), Alzheimer's disease, and other forms of early stage dementia. It is unrelated to the SAGES-3, a screener for gifted students, and to school achievement tests that share the SAGE name.
Tools like the SAGE have become more central to primary care and specialty workflows as the burden of cognitive decline grows. The lifetime risk of dementia among Americans over 55 is now estimated at 42%, with cases projected to exceed one million per year by 2060. The SAGE offers a practical way to begin cognitive screening, and like any screening tool, it serves its purpose best when its strengths and limitations are well understood.
The sections that follow examine what the SAGE evaluates, how it is administered, the benefits and limitations of its design, and how it compares to other cognitive screening tools commonly used in clinical practice.
Article Highlights
The SAGE is a screening tool; it cannot diagnose dementia independently and requires a full clinical assessment for confirmation.
The SAGE is a brief, self-administered cognitive screening test originally developed by Dr. Douglas Scharre and his team at Ohio State. It uses 12 paper-and-pencil test questions to evaluate cognitive domains across memory, language, reasoning, executive function, visuospatial ability, and orientation. The full test takes most patients 10 to 15 minutes, has no time limit, and is freely available online.
The SAGE is designed to flag potential cognitive issues for follow-up rather than to confirm a diagnosis. It is not clinician-administered, not a standalone diagnostic tool, and not a substitute for a comprehensive neuropsychological evaluation. Patients who score in the impaired range need a full clinical workup before any condition can be confirmed.
The features that have made the SAGE widely used include:
The SAGE evaluates six cognitive domains through a mix of written and visual tasks:
The SAGE is fully self-administered. A clinician suggests the test, and the patient may download and print the answer sheet from the Ohio State University SAGE page, then take the test independently. There is no specific time limit, but most patients finish within 10 to 15 minutes. The completed test is returned to the patient's doctor, who scores it and interprets the test results.
Scoring is straightforward and based on a 22-point scale:
| SAGE Score | What It Suggests |
|---|---|
| 17 to 22 | Likely normal cognitive function |
| 15 to 16 | Possible mild memory or thinking impairment |
| 14 or below | Possible more severe memory or thinking condition |
A score of 16 or below is generally considered a signal for further evaluation rather than a confirmed diagnosis.
Patients can complete the SAGE at home, or anywhere they feel comfortable, without a proctor or a visit to the doctor's office. In busy practices, this allows the test to be initiated by the patient and reviewed asynchronously, freeing clinician time for interpretation rather than administration.
Unlike the Montreal Cognitive Assessment (MoCA), which requires paid licensing and training, the SAGE is free to download and use. It is available in over a dozen languages, which broadens access for patients in multilingual practices.
Research published in Alzheimer Disease & Associated Disorders found the SAGE to be sensitive in differentiating between healthy individuals and those with MCI. Researchers publishing in Alzheimer's Research & Therapy reported that the SAGE can flag MCI-to-dementia progression at least six months earlier than the Mini-Mental State Examination (MMSE).
The SAGE has four different versions designed to be equivalent, which allows clinicians to retest patients over time without major practice effects. This is useful for establishing a cognitive baseline and tracking change.
The SAGE serves well as a first screen, but it carries limitations that warrant consideration in clinical practice.
A SAGE score is a signal rather than a definitive answer. A low score may indicate the need for more tests. Screening tools are designed to flag potential issues for follow-up with comprehensive neuropsychological workups, where standardized batteries, clinical history, and functional assessment can determine what is actually driving the cognitive change.
Non-cognitive factors can also contribute to lower SAGE scores in patients who are not cognitively impaired. Anxiety, poor sleep, acute illness, depression, and side effects from common medications can all affect test performance. Some of these conditions are treatable disorders that may resolve with intervention, which is one reason a low SAGE score is best viewed as a prompt for evaluation.
A patient's background can shape their SAGE score, and results are best interpreted with this context in mind. Three demographic factors can be particularly consequential:
Each of these factors can raise the false positive rate, flagging cognitively healthy patients as impaired, or produce a false negative, where high baseline ability masks early decline.
While the SAGE is effective for screening, its accuracy varies meaningfully depending on what is being detected. Research published in Alzheimer Disease & Associated Disorders reports a sensitivity of 79% and a specificity of 95% for detecting cognitive impairment in memory-clinic cohorts. A more recent validation reported how those figures shift across different detection targets:
| Comparison | Cutoff | Sensitivity | Specificity |
|---|---|---|---|
| Controls vs. dementia | <18 | 100% | 80% |
| Controls vs. MCI | <20 | 91% | 65% |
| Controls vs. cognitive impairment (combined) | <19 | 89% | 77% |
A normal SAGE result is reassuring but not conclusive, and patients with subtle or atypical cognitive changes can score in the normal range. A Cochrane evidence summary of self-completed cognitive assessment tools found insufficient evidence to rely on any one self-report tool alone for dementia screening, reinforcing the SAGE's role as one input among several.
Self-administration is one of the SAGE's most cited benefits, and it is also a source of variability. Without a proctor present, patients can misinterpret instructions, skip items, or rush through sections they find frustrating. The result is a score that can reflect attentional issues, comprehension difficulties, or motivation rather than memory issues.
This effect is particularly relevant for patients already showing signs of cognitive impairment, who may be least equipped to navigate written instructions independently. In practice, borderline scores benefit from contextual consideration, including whether the patient completed the test in a quiet setting, whether a family member observed any difficulties, and whether anything in the patient's recent history could have affected performance.
Even when self-administered at home, the SAGE does not replace a clinical visit. Patients still need an appointment to interpret results, arrange follow-up testing, and discuss treatment options. While the SAGE reduces the time needed to complete the initial screen, the downstream process can be significant. In many practices, neuropsychological testing can take months to schedule, during which a patient's condition may continue to progress.
Clinicians choosing a screening tool typically consider how long it takes, who administers it, what it costs, and the population it is normed for. The table below shows how the SAGE compares to three of the most widely used alternatives: the MMSE, the MoCA, and the Saint Louis University Mental Status (SLUMS) exam.
| Tool | Time | Self-admin. | Cost |
|---|---|---|---|
| SAGE | 10–15 min | Yes | Free |
| MMSE | 5–10 min | No | Paid license |
| MoCA | ~10 min | No | Paid license |
| SLUMS | 7–10 min | No | Free |
The MMSE is one of the most widely used cognitive screeners worldwide, available in many languages and completed in 5 to 10 minutes. Unlike the SAGE, the MMSE must be administered by a clinician and cannot be self-completed. In a study directly comparing the two, the SAGE was more sensitive to early signs of cognitive decline and was able to predict MCI-to-dementia progression about six months earlier.
The MoCA is another widely used cognitive screener, with adapted versions for patients with hearing or visual impairment. Unlike the MoCA, the SAGE is self-administered, takes less time per appointment, and is completely free. The MoCA does require paid licensing and training.
The SLUMS test is a cognitive screener for older adults aged 60 and up. It tests a similar range of cognitive areas to the SAGE, with additional domains like extrapolation and abstraction. Unlike the SAGE, the SLUMS is not self-administered and is normed specifically for the 60-and-over population, which limits its broader usability.
The challenge with the SAGE often arises after the first screen. When a result raises concern, the traditional next step is a referral for neuropsychological testing, which can take months to schedule.
Digital cognitive assessments developed for clinical use help shorten that gap. Creyos cognitive assessments combine the convenience of self-administration with the depth needed to support clinical decision-making, generating reports comparable to the information gathered in a neuropsychological exam without the wait associated with traditional referrals.
Creyos's brief, self-administered cognitive screener uses two digital tasks that measure visuospatial working memory and attention. The screener takes approximately six minutes to complete and applies a machine-learning algorithm to flag individuals who may need further testing. In a preliminary validation study published in the Journal of Alzheimer's Disease, the two-task screener identified 100% of patients with clinically diagnosed Alzheimer's dementia in the test cohort, with 86% specificity in matched healthy controls.
The 20-minute Creyos assessment provides detailed cognitive profiling across multiple domains. The assessment uses an additional four tasks and four questionnaires to help establish diagnostic criteria for mild or major neurocognitive disorders, based on DSM-5 criteria. The report is comparable in depth to a neuropsychological exam, available immediately, and accessible from any device.
The Creyos platform also includes a cognitive care planning tool based on guidelines from the Alzheimer's Association. Clinicians can build personalized care plans that support clinical decisions across diagnosis, monitoring, and family communication.
Reviewed by Sydni Paleczny, Staff Scientist at Creyos
Sydni earned her MSc in Neurosciences at Western University under Dr. Adrian Owen. Her research explores neuropsychological outcomes after cardiac surgery, with interests in cognitive neuroscience, critical care, and brain health. At Creyos, she supports scientific validity, health technology, and ongoing research.