Is remote bilateral-stimulation therapy as effective as in-person?

Nobody can say yet. Pooled analyses of video-delivered psychotherapy — across mental disorders generally, and for PTSD specifically — have not detected a difference from in-person care, at low-to-moderate certainty and with wide prediction intervals; but no adequately powered head-to-head trial of eye-movement or bilateral-stimulation therapy delivered remotely versus in person has been published. The one large direct comparison is a retrospective chart review, and it favoured in-person on effect size.

The published remote studies are small and largely uncontrolled, and they report few adverse events. That is reassuring but it is not an established safety profile: adverse events were rarely a pre-specified outcome, and most cohorts were pre-screened by the clinician for suitability. Some modality developers and training bodies do not endorse remote delivery. The decision is clinical, client by client, and the preparedness guidance is far more settled than the efficacy evidence.

What should I confirm at the start of every remote session?

Where the client physically is right now (the input to any emergency response, and the jurisdiction you must be licensed for), a phone number they can be reached on if the call drops, that the space is private and who else is in the home, and that you can both see and hear each other. Professional telehealth guidance treats verifying and documenting the client’s location as required whenever feasible; check your own board’s rule.

Clinician guide

Is remote bilateral-stimulation therapy safe? What the research says, and how to prepare.

Clinicians disagree about delivering eye-movement and bilateral-stimulation work over video, and so do the organizations that train them. This is a neutral read of the published evidence on both sides, what professional guidance actually requires, and a checklist for the clinicians who decide to do it — written so you can decide for yourself.

Disclosure. Psy180 makes browser-based bilateral stimulation software. We are not a training organization, we do not certify clinicians, and we have no outcome data on our own product. This article summarizes the published literature and publicly available guidance from professional bodies. Nothing here is clinical, legal, or licensure advice; verify every regulatory point against your own licensing board.

What we can and can’t tell you — the limitations up front

The evidence for delivering bilateral-stimulation (BLS) therapies remotely is real but thin — thinner than the evidence for remote psychotherapy in general. The first systematic review of online eye-movement therapy for PTSD (Lenferink et al., 2020, PMID 32905864) found exactly one trial, uncontrolled, and concluded that effects needed examining before wider dissemination was warranted — adding that remotely delivered cognitive-behavioural therapy seemed the preferable PTSD treatment in the meantime. Since then the literature has grown, but almost entirely through service evaluations, open cohorts, single-arm pilots, therapist surveys, and a few small group-protocol trials.

No adequately powered head-to-head randomized trial of remote versus in-person individual eye-movement or BLS therapy has been published. Two are registered: a youth non-inferiority trial (NCT06617663, protocol PMID 41327248) and a three-arm in-person vs remote vs waitlist trial (NCT06776744, protocol PMID 40676651).
The total published sample for remote individual delivery is in the low hundreds, spread across protocols that are not interchangeable.
Adverse events were a pre-specified, defined outcome in only a small minority of studies. Most others report “no adverse events” without describing how they looked.
Selection bias runs through the cohort literature. In nearly every positive study the clinician chose which clients were suitable for remote work. The clients judged too dissociative, too at-risk, or without a private space are invisible in the results — so “no adverse events” in a pre-screened sample is not a finding about the clients who were screened out.

What “remote bilateral-stimulation therapy” means here

Clinician-delivered, synchronous, video-based sessions in which alternating stimulation — eye movements, tones, or taps — is administered in discrete sets under the clinician’s direction. That is distinct from self-guided apps and from unsupervised self-administration, which this article does not cover. The field also uses many protocols — standard, recent-episode, group, blind-to-therapist, brief variants — and they are not interchangeable in the evidence base.

The broader telehealth-psychotherapy evidence — the part that is reasonably solid

For psychotherapy delivered by video in general, pooled analyses have not detected a difference from face-to-face care — across mental disorders, and for PTSD specifically — with caveats that matter.

A 2026 multilevel meta-analysis of 12 randomized trials (900 participants across six diagnoses including PTSD; PMID 42555822) found no significant difference between video and face-to-face (g = −0.09, 95% CI −0.52 to 0.33) — but a 95% prediction interval of −1.65 to 1.46, meaning a future trial in a new context could still land on either side.
A 2025 meta-analysis of 29 trials (PMID 38254285) found no detectable difference in PTSD outcomes by video (d = 0.06, 95% CI −0.17 to 0.28; moderate quality) — and depression outcomes that favoured in-person (d = 0.28, 95% CI 0.03 to 0.54; low quality).
A 2022 meta-analysis of 13 trials in primary care (PMID 35617629) found no difference in PTSD severity at six months (SMD −0.11, 95% CI −0.28 to 0.06) or in therapeutic alliance at three months (SMD 0.04, 95% CI −0.51 to 0.59); the authors note the evidence is heavily veteran-weighted.
A 2022 rapid review of 22 randomized trials (PMID 34368853) found seven of eight non-inferiority trials null — and two studies in which higher-severity subgroups did worse on telehealth.
In an equipoise-stratified randomized trial of Cognitive Processing Therapy (120 service members and veterans, 57% of whom declined one arm; PMID 35038985), telehealth was the least-refused format and produced large symptom change (d = 2.0, against 2.1 in-home and 1.3 in-office) — not eye-movement therapy, but instructive.

The evidence specific to eye-movement and bilateral-stimulation therapy delivered remotely

Strelchuk et al., 2023 (PMID 38073540) — an NHS service evaluation comparing 33 in-person with 45 online clients over consecutive years. No evidence of a difference in therapy completion, drop-out, or adverse events; weak evidence that online completers finished with slightly lower PTSD scores. Clients named feeling more in control and not travelling as benefits, and lack of privacy and loss of “transition time” between therapy and daily life as concerns. Not randomized; completers-only analysis. One of the few studies here that treated adverse events as an explicit outcome, and the only one that interviewed clients.
Mischler et al., 2021 (PMID 34675854) — 23 therapists reporting 102 video sessions. Distress reduction comparable to published in-person figures. The strongest predictor of reduction was the type of stimulation used: eye movements outperformed tapping. About a third of therapists had clients who declined remote work entirely. Therapist-reported, retrospective, unblinded.
McGowan et al., 2021 (PMID 34763697) — 33 therapists reporting on 93 clients, adults and young people, with meaningful reductions on all four measures. Self-selected therapist sample, no control group, no adverse-event capture.
Morris et al., 2023 (PMID 36548084) — 144 residential-care staff, recent-episode protocols delivered by telehealth in 2020 versus face-to-face in 2018–19: no significant difference, no adverse reactions reported. Retrospective, historical control, occupational sample.
Bongaerts et al., 2021 and 2022 (PMID 34025912, PMID 36186161) — four-day, fully remote intensive programmes: first six patients, then 73 with PTSD (nearly half meeting complex-PTSD criteria). Very large effects (CAPS-5 d = 2.12), maintained at six months, no drop-out and no adverse events. Single-site, uncontrolled, delivered by the programme’s own developers, no blinded adverse-event assessment.
Farrell et al., 2023 (PMID 37063579) — a randomized delayed-treatment trial of a group protocol delivered by videoconference to 85 frontline workers: significant reductions in trauma, anxiety, and depression scores (d = 0.71 on depression). And a null result the authors report plainly: no effect on moral injury through six months.
EMERALD, 2026 (PMID 42518933) — a randomized feasibility trial across three NHS hospitals with therapy delivered face-to-face or online: no treatment-related serious adverse events, and an explicit statement that its estimates are hypothesis-generating and should not be read as evidence of effectiveness.
Single-session and group variants (PMID 33052727, PMID 40250156, PMID 38868491, PMID 40417270) report reductions in small samples, with honest negatives alongside — one found anxiety rose again between the second and third assessment.
Other modalities. One remote study of the Flash Technique (PMID 40405468) was self-administered in a non-clinical sample with no control — not a licence for unsupervised self-use, and not evidence about any tool. We found no peer-reviewed study of Brainspotting delivered remotely.

The evidence and arguments on the other side

The largest direct comparison in this specific therapy — a retrospective chart review of 279 veterans (PMID 40424141) — found in-person produced greater PTSD improvement (d = 0.85) than telehealth (d = 0.66) after adjustment, with no difference in completion or reliable change. Non-randomized and almost certainly confounded by severity and preference, but it does not favour remote.
Higher-severity subgroups did worse on telehealth in two trials within the 2022 rapid review (PMID 34368853); depression favoured in-person in the 2025 meta-analysis (PMID 38254285).
Attrition can be severe in online group trauma work: in one 2026 study, 31 women were eligible and 9 completed both phases and assessment — the report does not say how many started (PMID 42529059).
Bilateral-stimulation fidelity is an uncontrolled variable online (PMID 39564248): clinicians have used sticky notes, knee taps, auditory beeps, free web tools that show advertisements mid-processing, and phone-sized screens that do not cover the client’s visual field — and the therapist usually cannot see what is happening on the client’s screen. Where stimulation type was measured, it mattered.
The client’s physical environment can be the limiting factor, independent of the therapy: a review of remote imagery-focused work found support for telehealth being no less effective overall, yet reported it was not viable for many clients in thin-walled or high-density housing, or with complex presentations (PMID 33725200); for clients experiencing intimate-partner violence, privacy and honest disclosure at home may be impossible (PMID 36195844).
A dissenting expert-consensus review (PMID 42284478) holds that eye-movement therapy has not demonstrated superiority over controls — a minority position that diverges from major guidelines and from network meta-analyses such as Hoppen et al., 2025 (pediatric PTSD, ages 19 and under; PMID 39630422) — but one a careful reader should know exists.

Some modality developers and training bodies do not endorse remote delivery, citing safety and fidelity concerns. The published literature does not currently settle that question in either direction, because the trials that would settle it have not been run.

Where clinicians say it goes wrong in practice

In EMDRIA’s 2020 Virtual Training and Therapy Task Group survey of roughly 1,600 members (task group report), 53% were willing to deliver or consider virtual sessions; safety was the single most common free-text concern, with 263 mentions, ahead of relationship and attunement (212 mentions) and technical problems (88 mentions).
Bursnall et al., 2022 (PMID 35136713) — 562 therapists on five continents and 148 of their clients: reluctance fell from 54% at the start of distancing to 11% a year later; 88% of clients were very or extremely comfortable. Named barriers: digital exclusion linked to deprivation, clinical severity, connectivity disrupting sessions, and the need for specific training. The authors’ conclusion: further research is needed to confirm that online delivery is clinically non-inferior.
Vitiello & Sowa, 2022 (PMID 36063291) — 111 clinicians in an academic psychiatry department: only 23.3% did proactive safety planning in more than half of their virtual visits. Those who faced a real emergency reported being unable to find the local magistrate, the right staff, the paperwork, or a secure channel, and confusion over which police agency and court applied when the client was elsewhere in the state. Clients could also simply move out of camera view.

What the guidance bodies say — and which parts are actually required

The preparedness guidance is far more settled than the efficacy evidence, and it converges across bodies. Read the “force” column carefully: the American Telemedicine Association’s joint document with the American Psychiatric Association uses graded language where “shall” means required whenever feasible and practical; the American Psychological Association’s Guidelines for the Practice of Telepsychology (2013, revised 2024) describe themselves as aspirational; and only state licensure law and HIPAA are binding statute.

Verify and document the client’s location at the start of video treatment
Source · ATA / American Psychiatric Association, Best Practices in Synchronous Videoconferencing-Based Telemental Health (2022) — “shall”
Force · Required whenever feasible and practical, in the document’s graded language. Not a universal licensing-board mandate in those words — check your board and your payer.
Practise only where you are licensed for the client’s physical location
Source · State licensure law; APA Guidelines for the Practice of Telepsychology (2013, revised 2024) — interjurisdictional practice; NASW/ASWB technology standards §2.02; ACA Code of Ethics, distance counseling section
Force · Generally binding, and the hardest requirement in the set — but rules vary by profession and state, and interstate compacts (PSYPACT, the Counseling Compact, the Social Work Licensure Compact), temporary-practice allowances, and federal-employment exceptions can change the answer. Confirm with your own board.
Full HIPAA compliance for remote care — BAA, risk analysis, security controls
Source · HHS Office for Civil Rights; COVID-era telehealth enforcement discretion expired with the 90-day transition period on 9 August 2023
Force · Legally binding for covered entities and business associates.
A backup plan for technology failure (for example, a phone)
Source · ATA 2022 — “shall”; APA Guidelines for the Practice of Telepsychology (2013, revised 2024) — emergencies; EMDRIA virtual-delivery guidance (2020)
Force · Required whenever feasible (ATA); recommended (APA, EMDRIA).
Emergency-management provisions in your written protocol; privacy at both ends; identify everyone present at both sites
Source · ATA 2022 — “shall”
Force · Required whenever feasible.
Coordinate directly with emergency services; do not tell the client to hang up and call 911; stay connected until responders arrive
Source · ATA 2022
Force · Strongly recommended; organizational training on the hand-off plan is “shall”.
Obtain an emergency contact person and local emergency resources; document a refusal and reconsider whether remote service is safe
Source · APA Guidelines for the Practice of Telepsychology (2013, revised 2024 — explicitly aspirational); EMDRIA 2020; NASW/ASWB §2.13
Force · Recommended.
Screen for dissociation, suicidality, substance use, violence history, support system, and distance to the nearest emergency facility before remote trauma work
Source · ATA 2022 (mixed); APA; EMDRIA; ISSTD 2011 for the clinical basis — note ISSTD’s guideline predates telehealth and does not address it
Force · Mixed; distance-to-facility and support-system assessment are “shall”.
A pre-agreed disconnection plan and an abreaction/dissociation plan, in the consent
Source · EMDRIA 2020 (platform → phone on mute → local support person); APA Guidelines for the Practice of Telepsychology (2013, revised 2024) — informed consent
Force · Recommended.
A grounding and closure protocol for remote sessions
Source · Universally recommended; nowhere standardized
Force · Write your own.

Two details worth getting right. First, 911 reaches the dispatcher for the caller’s location — a therapist in one state calling 911 does not reach the client’s town. Second, 988 (the Suicide & Crisis Lifeline) has routed by approximate geography since the FCC’s 2024 georouting order, but it does not receive the caller’s location and cannot dispatch to an address. It is a crisis-support line, not a dispatch mechanism; in-person response needs the address and the client’s local emergency number.

Deciding case by case

Guidance frames this as questions the clinician answers, not rules a vendor sets. Remote trauma work is probably not right, today, where there is no reliably private space; where a household member’s presence makes disclosure unsafe; where the connection is unstable; where the nearest emergency facility is far; where there is no support person and the client will not identify one; where the client declines to share their location; where severity or dissociation is beyond what you judge you can manage at a distance; where active substance use makes in-person screening matter; or where crises recur. Telehealth guidance also asks that clients be told in advance that services may be discontinued if they can no longer be safely managed at a distance.

A pre-session preparedness checklist

Most items trace to a cited guidance document above; the rest are practice notes drawn from the studies cited earlier. Adapt it to your protocol and your board’s rules.

Before the first session

Licensure verified for the state or jurisdiction where the client will physically be.
Written consent covering emergency procedures, technology-failure procedures, limits of confidentiality, location-of-care guidance, and the possibility of discontinuing remote work if it becomes unsafe.
The client’s address and phone number recorded; an emergency contact named — and the client has actually spoken to that person about being available during session times.
The client’s local emergency number, nearest emergency department, local mobile crisis team, and 988 recorded. (911 reaches the dispatcher for the caller’s location, not the client’s; 988 routes by approximate geography but cannot dispatch to an address.)
Distance to the nearest emergency facility and the adequacy of the support system assessed.
Screening for dissociation, suicidality, substance use, and violence or self-injury history.
Any refusal to supply the above documented, with a reconsideration of whether remote service can be offered safely.
A disconnection plan agreed: the platform first, a phone on mute beside the computer second, a pre-briefed local support person third.
An abreaction or dissociation plan agreed with the client in advance.
A written emergency hand-off plan — rehearsed before it is needed.

At the start of every session

Confirm where the client is right now.
Confirm the phone number they can be reached on right now.
Confirm the room is private, and who else is in the home.
Confirm both of you can see and hear each other; re-confirm the plan if they have disconnected before.

During

Stimulation in discrete, clinician-directed sets, with graduated targets and distress monitoring.
Watch for the client moving out of frame.
Reserve time for closure regardless of where processing has reached.

If the connection drops mid-processing

Call the phone immediately; if no answer, contact the pre-agreed support person.
If there is a safety concern, contact the client’s local emergency services directly using the address on file, hand off structured information, give them your contact details, and stay connected to the client until responders arrive. Do not tell the client to hang up and call 911.

After

Document the location, the emergency plan as shared, any technical disruption, and any emergency actions taken.

What would change the picture

Adequately powered non-inferiority trials of remote versus in-person individual delivery; adverse events as a pre-specified, defined outcome; head-to-head comparison of stimulation delivery methods online; clinician-administered versus self-administered stimulation; and a per-session record of what stimulation actually reached the client’s screen. Until then, the honest position is the one above: modest evidence, strong and concrete guidance, and a decision that belongs to the clinician and the client.

How to read this article

Psy180 renders bilateral stimulation on both screens rather than streaming it through the video feed. Nothing in this article is evidence that Psy180 treats anything, and no software feature makes a remote session safe — tools support a clinician’s own protocol; they do not replace it. Where this article paraphrases a guidance document, read the document; where it cites a study, read the study.

Why video lag breaks eye movements Therapies that use bilateral stimulation All guides