The cognition section of a mental status exam (MSE) covers six core domains: orientation, attention, memory, language, visuospatial ability, and executive function (including abstraction and judgment). Documenting it well means testing each domain with a specific bedside task and then describing what you found in clear, precise language. Here’s how to approach each domain and put it into writing.
The Six Neurocognitive Domains
The DSM-5 framework organizes cognition into six areas: complex attention, learning and memory, executive ability, language, visuoconstructional-perceptual ability, and social cognition. A standard MSE doesn’t require formal neuropsychological testing of every domain, but it should touch on each one enough to flag problems. When cognition appears normal across all areas during a routine interview, many clinicians document this as “cognition grossly intact.” That shorthand is fine for a healthy patient with no cognitive complaints, but any abnormality deserves domain-specific detail.
Orientation
Orientation is typically tested across three parameters: person, place, and time. You ask the patient their name, where they are, and today’s date. Some clinicians add a fourth sphere, situation (“Why are you here today?”), making it “oriented x4.” Orientation largely reflects recent memory function, so deficits here often signal broader memory problems.
In your documentation, specify which spheres are intact and which are not. Writing “oriented x2 to person and place, disoriented to time” is far more useful than “partially oriented.” If the patient gets the month right but not the date, note that level of detail. It helps the next clinician gauge whether things are stable or worsening.
Attention and Concentration
Attention is the gateway to every other cognitive domain. If a patient can’t focus, memory testing and language testing become unreliable. Two classic bedside tasks are serial sevens and digit span.
For serial sevens, you ask the patient to count backward from 100 by sevens (100, 93, 86, and so on). This tests both sustained attention and working memory. Note how many subtractions the patient completes and where errors occur. One important caveat: serial sevens has significant educational bias. A patient with limited math background may struggle for reasons that have nothing to do with attention. Alternatives like spelling “WORLD” backward, reciting days of the week in reverse, or reciting months of the year backward can be more equitable.
For digit span, you read a string of numbers and ask the patient to repeat them forward, then backward. Most healthy adults can repeat five to seven digits forward and four to five backward. In your note, record the longest string repeated accurately in each direction. You might write: “Attention intact; completed serial sevens to 72 with one error; digit span 6 forward, 4 backward.”
Another quick screen for sustained attention is a letter-tapping task: you read a string of random letters aloud and ask the patient to tap the table every time they hear a target letter, like “A.” Errors of omission (missing the target) or commission (tapping on the wrong letter) both suggest attentional difficulty.
Memory: Immediate, Short-Term, and Remote
Memory is typically broken into three time frames. Immediate memory (also called registration) is tested by giving the patient three or four unrelated words and asking them to repeat the words right away. This confirms they encoded the information. Short-term (recent) memory is tested by asking them to recall those same words after a delay of three to five minutes, during which you perform other parts of the exam. Remote (long-term) memory is assessed through questions about verifiable personal history, like where the patient went to school or significant life events.
When documenting, specify the time frame and what happened. For example: “Registered 4/4 words immediately; recalled 2/4 after 5 minutes without cues, 3/4 with category cues; remote memory intact for personal history.” The detail about cueing matters because a patient who retrieves a word with a hint (“it was a fruit”) has a retrieval problem, while one who can’t recall even with cues likely has an encoding or storage problem.
Language
Language assessment during the MSE covers several components: fluency, comprehension, repetition, and naming. You can gather much of this passively during the interview. Is the patient speaking in full sentences with appropriate word choice (fluency)? Do they follow your questions and instructions (comprehension)? Can they repeat a phrase like “no ifs, ands, or buts” (repetition)? Can they name common objects when you point to them, like a watch or a pen (naming)?
Document what you observe specifically. “Speech fluent with intact comprehension and repetition; naming intact for high- and low-frequency objects” paints a clear picture. If there are deficits, describe them: “Patient used circumlocutions for common objects (called a watch ‘the thing that tells time’), suggesting word-finding difficulty.” Avoid vague terms like “language impaired” without specifying which component is affected.
Visuospatial Ability
The clock drawing test is the most widely used bedside screen for visuospatial and executive function. Give the patient a blank sheet of paper and ask them to draw a clock face with all the numbers, then set the hands to a specific time (10 minutes past 11 is the standard prompt). This single task tests spatial planning, number placement, and the ability to translate an abstract instruction into a motor output.
When describing the result, note specific errors rather than just scoring it as pass or fail. Common problems include clustering all the numbers on one side of the clock (suggesting spatial neglect), placing numbers outside the circle, or drawing the hands incorrectly. You might write: “Clock drawing notable for numbers 1 through 12 crowded into the right half of the clock face; hands correctly placed at 11 and 2.” Copying a pre-drawn clock is an easier version of the task and can help distinguish whether the problem is with planning (impaired on spontaneous drawing but intact on copy) or with basic visuospatial processing (impaired on both).
Executive Function, Abstraction, and Judgment
Executive function refers to higher-order abilities like planning, problem-solving, mental flexibility, and impulse control. Abstraction is one measurable piece of this. Two traditional bedside tests are proverb interpretation and similarities.
For proverb interpretation, you ask the patient what a common saying means, such as “Don’t cry over spilled milk.” An abstract response (“Don’t worry about things you can’t change”) suggests intact abstraction. A concrete response (“Clean up the milk”) may suggest impairment, but there’s an important bias to recognize: familiarity with a proverb strongly influences how well someone interprets it. Patients from different cultural, racial, or educational backgrounds may be unfamiliar with English-language proverbs, leading clinicians to incorrectly conclude they are thinking concretely. Similarities testing (“How are an apple and an orange alike?”) is generally more reliable because it depends less on cultural knowledge. An abstract response is “both are fruits”; a concrete response is “both are round.”
Judgment can be assessed with hypothetical scenarios (“What would you do if you found a stamped, addressed envelope on the sidewalk?”) or by evaluating the patient’s real-world decision-making during the interview. Document your findings with the specific test used: “Abstraction intact; provided abstract interpretations for 2/2 similarities. Judgment fair, as evidenced by appropriate help-seeking behavior.”
Thought Process vs. Cognition
One common source of confusion is the overlap between thought process and cognition. Thought process describes how a patient’s ideas connect to one another during conversation and is documented separately from the cognitive exam, though the two can influence each other. Circumstantial thinking means the patient goes off on tangents but eventually circles back to the point. Tangential thinking means they go off-topic and never return. Perseveration means they return to the same subject regardless of the question. Flight of ideas involves rapid shifts between loosely connected topics. Thought blocking involves sudden interruptions in the flow of thought, often seen in psychosis. These descriptors belong in the “Thought Process” line of your MSE, not the cognition section, but recognizing them during the interview helps you interpret cognitive test results more accurately.
Standardized Screening Tools
When a more structured assessment is needed, two validated tools dominate clinical practice: the Mini-Mental State Examination (MMSE) and the Montreal Cognitive Assessment (MoCA). The MMSE is scored out of 30, with 24 or above considered normal. The MoCA is also scored out of 30, with 26 or above considered normal, and it adds one point for patients with fewer than 12 years of formal education. The MoCA is generally preferred for detecting mild cognitive impairment because it includes more demanding tasks for attention, executive function, and visuospatial ability. The MMSE tends to have a ceiling effect, meaning patients with early or subtle deficits can still score in the normal range.
When you use a standardized tool, document the total score, the version used, and any domain-specific subscores that stand out. “MoCA score 22/30 (normal ≥26), with points lost in delayed recall (1/5) and visuospatial/executive tasks (2/5); attention and orientation intact” gives the reader a meaningful cognitive profile in a single sentence.
Putting It All Together in a Note
A well-written cognition section moves systematically through the domains and uses specific, descriptive language. Here’s what a normal documentation might look like:
“Cognition: Alert and oriented to person, place, time, and situation. Attention intact; completed serial sevens without error. Registered 3/3 words immediately, recalled 3/3 at 5 minutes. Language fluent with intact naming, repetition, and comprehension. Clock drawing normal. Abstraction intact per similarities testing. Judgment and insight good.”
And for a patient with deficits:
“Cognition: Alert, oriented to person and place, disoriented to time (stated month as March, actual June). Attention impaired; completed serial sevens to 86 with two errors, then discontinued. Registered 3/3 words, recalled 0/3 at 5 minutes, 1/3 with category cues. Language fluent but with frequent word-finding pauses and circumlocutions. Clock drawing notable for misplaced numbers and incorrect hand placement. Interpreted proverbs concretely. Judgment poor, as evidenced by inability to describe a safe discharge plan. MoCA 18/30.”
The goal is for any clinician reading your note to reconstruct the patient’s cognitive profile without having been in the room. Name the test, give the result, and describe the quality of the performance. That combination of structure and specificity is what separates a useful MSE from a checkbox exercise.