Remote mental state examination (MSE): how to document a video assessment differently

A mental state examination (MSE) conducted over video should not be documented the same way as one conducted in person. Several core MSE domains, gait, olfactory cues, lower-body psychomotor activity, are structurally unobservable on camera, and others are distorted by compression, latency, and lighting. Accurate documentation records the modality, describes the technical conditions, and qualifies each observation against what was actually observable, distinguishing "unassessed" from "absent."

Why remote MSE documentation needs a different approach

The Royal College of Psychiatrists describes the MSE's purpose as providing a clear, objective snapshot of someone's mental functioning at a given point in time. That snapshot depends on what the clinician can directly perceive across the standard domains: appearance and behaviour, speech, mood and affect, thought, perception, cognition, and insight.

In a consultation room, the clinician has a continuous, three-dimensional sensory field: gait as the patient enters, body odour that may signal self-neglect, tremor in the hands and lower limbs, full-body posture, micro-expressions without compression artefacts.

Over video, a significant share of these inputs is absent or degraded. Documenting a remote MSE as though it were an in-person one doesn't just understate a methodological limit. It produces an inaccurate clinical record.

UK regulators have been clear on the principle. GMC and RCPsych digital guidance states that the standards expected of doctors apply equally to digital and conventional consultations, and that doctors must give consideration to the potential limitations of the medium used. Practice hasn't fully caught up.

An audit at Nottinghamshire Healthcare NHS Trust, published in BJPsych Open, found that only 81% of remote consultation records documented the modality at all, consent for remote consultation was recorded in just 57%, and MSE domains were adequately recorded in only 50–70% of notes. Those are exactly the gaps through which medico-legal risk enters.

Larger-scale work is under way. Researchers at South London and Maudsley NHS Foundation Trust and King's College London are applying natural language processing to around five million clinical documents from roughly 80,000 patients to analyse how the content of remote assessments differs from in-person ones, with the working hypothesis that video narrows the gap more than telephone or text does.

What the camera frame cannot capture: a domain-by-domain breakdown

• Appearance and behaviour. Visible from the shoulders up, in the patient's chosen lighting and environment. Body habitus, clothing below the frame, footwear, hygiene, and nutritional status are unobservable. Telephone consultations remove appearance entirely, which BJPsych Open notes is a core MSE component.

• Psychomotor activity. Gait is unassessable unless the patient stands and moves within frame. Akathisia and lower-limb restlessness may be invisible; a patient who appears still on screen may be markedly agitated below it. Tremor and subtle asymmetric movement can sit below consumer camera resolution, and poor frame rate can create the impression of movement that isn't there.

• Olfactory cues. Completely absent. Signs of alcohol use, self-neglect, or poor hygiene that carry real clinical weight in person cannot be assessed.

• Speech. Generally assessable, but audio compression, microphone quality, and latency can mask or mimic features. Slowed speech may reflect genuine psychomotor slowing or a patient pausing for lag; pressured-sounding speech may be normal speech through a microphone compressing dynamic range.

• Mood and affect. Subject to compression artefacts, frame-rate drops, and audio latency that flatten or delay perceived emotional response and disrupt the reading of congruence between words and expression. A patient who appears blunted may be blunted, or may be transmitting through a poor connection in a dim room.

• Thought, perception, cognition, insight. Largely preserved, since these rest on verbal content, though early comparative research on remote cognitive testing in older adults found lower scores in the remote condition, particularly where hearing loss interacted with audio quality. The medium can affect patient performance, not just clinician observation.

• Rapport and eye contact. Structurally ambiguous. A patient looking at the clinician's face on screen appears to look slightly down or aside; one looking into the camera appears to hold eye contact but cannot see the clinician. Body orientation, proxemics, and postural mirroring are largely absent. Standard in-person language for eye contact doesn't transfer without qualification.

What the camera shows that a clinic room can't

One dimension runs the other way. A video MSE conducted in the patient's home can reveal clinically relevant information a standardised clinic room never would: visible domestic disorganisation relevant to self-care or cognitive functioning, signs of isolation or of other people present, factors affecting privacy and safety.

These observations belong in the record where they bear on the assessment, with the same care about scope: what was visible in frame, not an inference about the whole home. Third parties visible or audible during the session should be noted, both for clinical context and for consent.

The distinction that matters most: absent versus unassessed

A clinician who saw no gait disturbance during a video consultation has not established that gait is normal. They've established that gait was not observed. These are clinically and medico-legally different statements, and the record must reflect the difference.

The same applies to olfactory cues, lower-body motor activity, and anything else the frame excludes. Applied consistently, this single distinction protects the accuracy of a remote MSE record more than any other change.

When a psychologist documents that a patient appeared 'casually dressed and well-groomed,' that statement carries an implicit claim about the patient's overall presentation. In an in-person assessment, that claim is grounded in full-body observation under consistent lighting. In a video consultation, it is grounded in something considerably narrower.

How to write qualifying language into the record

Qualifying language is not a sign of clinical uncertainty. It's clinical precision, and it doesn't need to be lengthy.

Appearance: "Appearance assessed from the shoulders upward via video consultation. Patient appeared neatly dressed in casual clothing; hair appeared groomed. Hygiene, full-body presentation, and gait were not assessable in this modality."

Psychomotor activity: "Upper body movement appeared within normal limits. Gait, lower-limb activity, and fine motor signs were not observable; psychomotor assessment is therefore partial."

Speech: "Speech normal in rate and volume as perceived; audio quality was variable during the first ten minutes, which may have affected the reliability of prosody observations."

Affect: "Affect appeared euthymic and congruent with reported mood. Session audiovisual quality was good; no technical factors were identified that would be expected to distort affect perception." Or: "Affect appeared constricted throughout. Note: frame-rate instability during the first fifteen minutes may have limited accurate assessment of facial expressivity."

Eye contact: "Patient appeared attentive and engaged. Eye contact could not be assessed using standard in-person criteria due to camera positioning; patient consistently oriented toward the screen and responded appropriately to conversational cues."

Olfactory: "Olfactory assessment was not possible in this remote consultation format."

A brief statement at the start of the MSE section acknowledging the modality and conditions, followed by domain-specific qualifications only where they apply, is sufficient. For clinicians using an AI scribe to draft the consultation notes, it's worth checking that the draft carries these qualifications rather than defaulting to unqualified in-person phrasing, and that the modality is stated explicitly. The clinician's review is where that gets caught.

When a remote MSE isn't enough

Remote assessment isn't appropriate in every context, and the record should say so when it isn't. In-person assessment is generally indicated where:

• Active risk involving suicidality or self-harm requires observation of behavioural and physical cues that video can't reliably capture

• Suspected psychosis or first-episode presentation demands assessment of thought disorder and the congruence between verbal content and non-verbal behaviour at a standard remote delivery can't meet

• Suspected organic or neurological presentation needs physical examination and full-body observation of movement, cognition, or physical illness presenting as psychiatric symptoms

• The technical environment is inadequate, so the assessment's validity is compromised from the outset

• Significant sensory or cognitive impairment interacts with the medium itself, as the older-adult cognitive testing research suggests

Where in-person follow-up is clinically indicated, document the recommendation and the reason for it. That entry is itself part of a defensible record.

Why this carries medico-legal weight

In personal injury proceedings, capacity determinations, disability assessments, and fitness-to-practise hearings, the record is read as a factual account of what was observed and how.

An MSE documented without reference to its remote modality will be read as equivalent to an in-person assessment. If it later emerges the assessment was remote, the absence of qualifying language raises questions about the clinician's awareness of their own method's limits.

The risk runs both ways. An unqualified remote record can overstate the reliability of findings that were genuinely limited, leading a later clinician to place unwarranted confidence in them.

It can also understate risk by silently omitting observations that could only have been made in person, such as physical neglect below the frame or gait abnormalities pointing to an organic cause.

A counterpoint is worth stating plainly: a remote MSE by an experienced clinician with an established therapeutic relationship can yield observations of real clinical value. The argument for qualifying language isn't that remote assessment is inferior. It's that the record must accurately reflect the conditions under which observations were made, so their weight can be judged by whoever reads them next.

Where the guidance stands in the UK and Europe

Guidance is developing unevenly. The RCPsych has published broad remote-consultation guidance. The British Psychological Society goes furthest, stating that psychologists conducting remote assessments must explicitly note the modality in reports and consider whether the conditions were adequate to support the conclusions drawn.

The European Federation of Psychologists' Associations has framework-level ethics but no technical documentation guidance, and national bodies in Germany and the Netherlands haven't reached the procedural specificity of the BPS. In that gap, many clinicians reference the American Psychological Association's 2024 telepsychology guidelines as the most detailed international standard, particularly its requirement that clinicians be prepared to explain differences between remote and in-person results. The defensible position is to apply the most rigorous available standard and document transparently.

A defensible baseline for the remote MSE record

• Record the modality, including platform where relevant to consent and data security.

• Describe the technical conditions: "video call; audiovisual quality stable throughout" is enough when there's nothing to report.

• Qualify each domain against what was and wasn't observable, rather than an implied in-person standard.

• Distinguish absent from unassessed, every time, recording unobservable domains as "not assessable via remote consultation" rather than leaving them blank.

• Note disruptions at clinically significant moments: frame-rate drops, audio dropout, connection loss.

• Document the environment where it bears on safety, privacy, or clinical judgement, including any third parties present.

• State when in-person follow-up is indicated, and why.

• Retain session metadata (date, time, duration, platform) where data protection rules permit.

• Document consent for remote assessment as a distinct step from consent to treatment, the gap the Nottinghamshire audit found most consistently.

• In formal reports, state overtly that equivalence between remote and in-person assessment is not guaranteed, rather than defaulting to in-person reporting conventions.

Applied consistently, these standards make the record something that can be read, relied upon, and scrutinised with confidence, whatever note format the service uses around it.

Frequently asked questions

▶ Why does documenting a remote mental state examination require a different approach to an in-person one?

Because an MSE depends on direct observation of posture, gait, olfactory cues, affect, and motor activity, and over video many of those inputs are absent or degraded. Using the same unqualified language as an in-person assessment produces a record that's imprecise at best, and clinically misleading and medico-legally indefensible at worst.

▶ Which mental state examination domains are most affected by remote delivery?

Psychomotor activity and gait, since patients are usually visible only from the chest up. Olfactory cues are absent entirely. Appearance is limited to the visible frame, affect is distorted by compression and latency, and eye contact can't be judged by in-person criteria because the camera and screen aren't in the same place. Thought, perception, and cognition are largely preserved, as they rest on verbal content.

▶ What is the difference between documenting a sign as 'absent' versus 'unassessed' in a remote mental state examination?

A clinician who saw no gait disturbance on video hasn't established that gait is normal, only that it wasn't observed. These are clinically and medico-legally different statements. Anything structurally unobservable in the remote context should be recorded as unassessed, never absent.

▶ How should affect be documented when technical issues affect the quality of a remote session?

Describe the observed affect in standard clinical language, note the audiovisual quality, and flag where technical factors may have shaped your impression. For example: "Affect appeared constricted throughout. Note: frame-rate instability during the first fifteen minutes may have limited accurate assessment of facial expressivity."

▶ Why can't eye contact be documented the same way in a remote mental state examination as in an in-person one?

A patient looking at your face on their screen appears, from your side, to be looking slightly down or to the side. A patient looking into the camera appears to hold eye contact but can't see you at the same time. That structural asymmetry means eye contact in video consultations always needs qualifying in the record.

▶ What are the medico-legal risks of an unqualified remote mental state examination record?

In personal injury proceedings, capacity determinations, and fitness-to-practise hearings, the record is read as a factual account of what was observed. An MSE with no reference to its remote modality is read as an in-person assessment. If it later emerges that it wasn't, the missing qualification raises questions about the clinician's awareness of their own method's limits.

▶ What professional guidelines exist for documenting remote mental state examinations?

GMC and RCPsych guidance establishes that the same standards apply to digital consultations and that clinicians must consider the limitations of the medium. The British Psychological Society requires the modality to be stated in reports. Beyond that, UK and European guidance lacks technical detail, so many clinicians reference the American Psychological Association's 2024 telepsychology guidelines as the most operationally specific international standard.

▶ What should every remote mental state examination record include as a minimum standard?

The modality and platform; the audiovisual quality and any disruptions; each domain qualified against what was and wasn't observable; unassessed domains recorded as such; and consent for remote assessment documented separately from consent to treatment. A Nottinghamshire NHS audit found consent for the remote modality recorded in only 57% of notes, the most consistent gap.

▶ Does using qualifying language in a remote mental state examination record undermine its clinical value?

No. It's a sign of precision, not uncertainty. A remote MSE by an experienced clinician with an established therapeutic relationship can yield observations of real value. The point of qualifying language isn't that remote assessment is inferior; it's that the record must accurately reflect the conditions the observations were made under, so their weight can be judged by whoever reads them next.