Sign Up
Login
Try for Free
Login
The T.O.V.A. (Test of Variables of Attention) is an FDA-cleared computerized continuous performance test measuring sustained attention and response inhibition across a 21.6-minute go/no-go task. ADHD evaluators use it as an aid to assessment, never as a standalone diagnostic. Documentation must name the version, conditions, and validity before scores. This page covers how to document T.O.V.A. results, with a fictional sample.
Psychologists, neuropsychologists, psychiatrists, and pediatric clinicians; trained staff may administer, qualified professionals interpret
ADHD evaluation teams, prescribing clinicians monitoring treatment, schools and disability offices, payers and reviewers, families
200 to 500 words for the results section · the standard administration runs 21.6 minutes (10.8 for the preschool visual form), plus setup and interpretation
Computerized go/no-go continuous performance test (visual and auditory versions; standard scores, mean 100, SD 15, by quarter, half, and total)
ADHD evaluations as adjunctive performance data, baseline and on-medication treatment monitoring (visual version), targeted differential questions about attention
FDA 510(k)-cleared as an aid to assessment (the visual version also treatment evaluation); The TOVA Company; no stimuli, norms, or thresholds reproduced here
The T.O.V.A. (Test of Variables of Attention; The TOVA Company; originated by Lawrence M. Greenberg, MD, with current manuals crediting Leark, Dupuy, Greenberg, Kindschi, and Hughes) is a computerized go/no-go continuous performance test delivered through licensed software and a dedicated microswitch the manufacturer specifies at roughly millisecond timing precision. The task runs in two contrasting halves: targets are infrequent in the first, taxing sustained attention, and frequent in the second, taxing response inhibition. The current product is T.O.V.A. 9 on the 9.1 software line, and the visual and auditory versions are distinct tests, not presentation options: separate norms, different age ranges (visual roughly 4 through 80, auditory 6 through 29), different FDA-cleared indications, and an Attention Comparison Score generated only for the visual test. Reports return standard scores (mean 100, SD 15, age- and sex-referenced) for omission errors, commission errors, response time, response-time variability, and d-prime response sensitivity, organized by quarter, half, and total, alongside anticipatory-response and validity information. The standard administration takes 21.6 minutes; a shorter preschool visual form takes 10.8.
The load-bearing fact is regulatory. The system is FDA-cleared under 510(k) K173915 (March 2018, product code LQD, predicate QbTest) as an aid: the visual version aids assessment and treatment evaluation, the auditory version aids assessment only, and the cleared labeling restricts interpretation to qualified professionals, with the manual adding that the device does not diagnose any disorder. Clearance is a substantial-equivalence finding, not a validation of diagnostic accuracy, and professional guidance in the US, Canada, and Australia uniformly treats continuous performance tests as adjuncts: a recent meta-analysis of 19 commercial-CPT studies put pooled discrimination at an area under the curve of roughly 0.70 to 0.80, with sensitivity near .75 and specificity near .71. So the documentation job is to record a constrained laboratory performance sample honestly, aid-not-diagnosis framing intact, and feed it to the synthesis covered on the ADHD evaluation report page, alongside the rating-scale evidence covered on the Conners 4 and Vanderbilt pages.
Psychologists and neuropsychologists fold the T.O.V.A. into ADHD evaluations as performance-based evidence beside history, collateral report, and rating scales; psychiatrists and pediatric prescribers use the visual version's treatment-evaluation indication for baseline and on-medication comparisons during titration; and clinics with technician-administered testing use it because administration and interpretation can be split, provided the interpretation stays with a qualified professional. The note's readers are unusually varied. A payer reviewer may be applying a policy that classes computerized attention testing as experimental (several major US policies do), so the note must show the test as one component of a medically necessary evaluation rather than a standalone billable diagnostic. A school or testing agency reading an accommodations request needs multimethod evidence, because no single CPT score establishes the need for extended time. A prescriber reading a titration retest needs the medication name, dose, and dose-to-test interval before the score means anything. And a family often arrives having read that this is an "objective ADHD test," which the documentation must gently correct. Adult self-report screening lives on the ASRS page, adult observer ratings on the CAARS-2 page, and everyday executive-function ratings on the BRIEF-2 page; this page owns the CPT session itself.
No authority prescribes a T.O.V.A. note format, but the instrument's design dictates what a defensible entry must contain: exact identification, the conditions that move CPT scores, medication timing, validity before numbers, the pattern in prose, the comparison score framed as an aid, and an integration statement. Each section below carries the pitfall that most often undermines it.
Instrument identification, exactly. Record T.O.V.A. 9, the modality (visual or auditory), the software or build identifier printed on the patient's report, the administration date and time, and the purpose: initial adjunctive assessment or treatment-evaluation retest. The modality line is not pedantry: the two versions carry different norms, age ranges, and FDA indications, and only the visual test generates the Attention Comparison Score. Pitfall: "TOVA administered" or "TOVA positive." Unversioned, unmodalitied results cannot be interpreted, compared at retest, or defended to a reviewer.
Testing conditions. Chart time of day, the prior night's sleep, caffeine, nicotine, or other substance exposure, acute illness, pain, or mood state, vision or hearing correction, room interruptions and observed engagement, motor or hand limitations affecting the microswitch, and prior exposure to the T.O.V.A. or another CPT. These are the variables that move CPT scores without any attention disorder present. Pitfall: A score table with no conditions. Sleep deprivation or a 4 p.m. slot can manufacture a poor performance; a practice effect can manufacture an improvement.
Medication status, precisely. State the medication name and dose, whether testing was on or off medication, the date and time of the last dose, and the elapsed interval, plus whether the status reflected usual treatment, a prescriber-directed hold, or a planned comparison. Whether to hold a stimulant is the prescriber's clinical and safety decision for the question at hand; no generic washout duration applies across formulations. Pitfall: "Tested unmedicated per protocol" with no prescriber plan, or a patient told to stop a stimulant because "the TOVA must be done off meds."
Validity before scores. Resolve interpretability first: anticipatory responses, sufficient usable responses, interruptions or hardware warnings, instructions followed, and the unusual-pattern or performance-validity output where it applies (visual testing, ages 17 and older, and only under certain data conditions). A flag means caution and clinical adjudication. Pitfall: Scores interpreted past an unresolved validity problem, or a validity flag written up as malingering. The analysis identifies unusual patterns; it does not diagnose deception.
The standard-score pattern, in prose. Describe the variables by the periods they came from: omissions concentrated in the low-target first half, variability worsening across later quarters, commissions rising under the inhibition demand, response speed slow but stable, d-prime discrimination reduced. Quarter, half, and total results are separate aggregations of the underlying responses, not averages of each other, and every variable has nonspecific causes. Pitfall: A bare number grid, or one outlying quarter promoted into a disorder. "Inattention" and "impulsivity" are shorthand for omission and commission patterns, not proven constructs.
The Attention Comparison Score, framed as an aid. For visual administrations, report the ACS with its meaning attached: a composite of first-half response speed, second-half response sensitivity, and overall response-time variability, compared against the manufacturer's independently diagnosed ADHD sample and its normative sample. Below zero means more similar to the ADHD comparison group; at or above zero means more similar to the normative group and does not exclude ADHD. It was formerly called the ADHD Score, a name worth retiring. Pitfall: Zero treated as a diagnostic threshold, the ACS read as an ADHD probability, or an auditory report searched for an ACS it does not produce.
Integration and retest documentation. Close with a synthesis: the T.O.V.A. finding was integrated with developmental history, cross-setting symptoms and impairment, informant ratings, interview, and observation, and it neither establishes nor excludes ADHD. For treatment-evaluation retests, document matched conditions (time of day, sleep, hardware), the interval, prior exposure and expected practice effects, and the medication timing, and conclude in performance terms: more consistent under the medicated condition. Pitfall: A CPT overruling the clinical picture in either direction, or a retest gain attributed wholly to medication with practice effects unmentioned.
T.O.V.A. RESULTS DOCUMENTATION SKELETON Client: [initials] Age: [ ] Date + time: [ ] Version: [T.O.V.A. 9, visual / auditory] Build (from report): [ ] Purpose: [initial adjunctive assessment / treatment-evaluation retest] Administered by: [ ] Interpreted by: [ ] Conditions: [sleep hours; time of day; caffeine/nicotine/substances; illness, pain, mood; vision/hearing correction; interruptions; motor factors; prior T.O.V.A. or CPT exposure] Medication: [name + dose; on/off; last dose date + time; interval; usual regimen / prescriber-directed hold / planned comparison] Validity review (before any interpretation): [anticipatory responses; usable responses; unusual-pattern / performance- validity output if applicable (visual, 17+); interruptions; hardware; instructions followed] Interpretable: [yes / with caution / no] Score pattern (standard scores, prose by period): Omissions: [ ] Commissions: [ ] Response time: [ ] RT variability: [ ] d-prime: [ ] Pattern notes: [which half / quarters carried the findings] Attention Comparison Score (visual only): [value relative to zero + aid-to-assessment framing; not a probability or threshold] Integration: [convergent / discordant with history, ratings, observation, impairment; neither establishes nor excludes ADHD] Retest notes (if applicable): [interval; matched conditions; practice effects; medication timing; performance-level wording] Plan: [how the finding feeds the evaluation or titration] Clinician signature / credentials: Date:
Free to use and share, no signup. The PDF includes a one-page cheat sheet with section-by-section pitfalls and a pre-sign checklist; the DOCX is the blank documentation skeleton, ready to adapt. Neither reproduces stimuli, timing parameters, norms, thresholds, or report content.
Scenario: an adult ADHD evaluation in which a valid off-medication visual administration converges with the clinical picture, written so the score aids the assessment without becoming the diagnosis. All details are fictional.
Patient: J.H., 34 years · Setting: Outpatient ADHD evaluation, adjunctive CPT session · Clinician: K. Osei, PsyD, licensed psychologist · Note date: 08/21/2026
Identification and purpose: Visual T.O.V.A. 9 administered 08/21/2026 at 9:20 a.m. as an adjunct to a comprehensive adult ADHD evaluation; the software build identifier from the patient report is filed with the raw data. This was an initial assessment administration, not a treatment-evaluation retest. Administration was monitored by trained staff; interpretation is by the undersigned psychologist.
Conditions and medication: J.H. reported approximately seven hours of sleep, one morning coffee two hours before testing, no nicotine, no acute illness or significant distress, and corrected vision in use. He has no prior exposure to the T.O.V.A. or any other continuous performance test. He had not taken stimulant medication that day: the hold was directed by the prescribing clinician as part of a planned baseline assessment, and the last dose, the prior morning, is documented in the medication record with the elapsed interval.
Validity: The administration completed without interruption or hardware warning, instructions were followed, usable responses were sufficient, and anticipatory responding was not clinically significant. The performance-validity analysis applicable to visual administrations at this age raised no concern. The administration was judged interpretable.
Results: Response speed was broadly adequate but became progressively less consistent: response-time variability worsened across the later quarters, omission errors increased late in the low-target first half, and d-prime target discrimination weakened under the second half's inhibition demand, while commissions remained only mildly elevated. Standard scores by quarter, half, and total are retained with the report. The visual Attention Comparison Score fell below zero, meaning this laboratory performance pattern was more similar to the manufacturer's independently diagnosed ADHD comparison sample than to its normative sample. The ACS is an aid to assessment; it is not a diagnostic probability, a severity rating, or a standalone determination.
Integration and plan: The T.O.V.A. pattern converges with the evaluation's other evidence: documented childhood-onset disorganization and sustained-attention difficulty, current occupational impairment, collateral history, and elevated self- and observer ratings. The finding was integrated with that history, the interview, and observations, and was not used as the sole basis for any conclusion; a normal result would likewise not have excluded the diagnosis. The diagnostic formulation and recommendations appear in the full evaluation report. If stimulant treatment is initiated, a treatment-evaluation retest at a matched morning time, with dose-to-test interval documented and practice effects acknowledged, may inform titration alongside symptom ratings, function, and tolerability.
This sample is fictional and for educational purposes. It does not describe a real patient or record; the details are invented to show documentation structure and are not clinical guidance. No stimuli, timing parameters, norms, thresholds, or proprietary report content are reproduced.
Writing these after every session? BastionGPT drafts complete notes from bullets, dictation, or a transcript.
Generate a note from bulletsUnited States: the regulatory frame is precise and routinely inflated, so the note should get it exactly right. The system is FDA-cleared, not approved, under 510(k) K173915 (March 2018, product code LQD, regulatory class listed as unclassified, predicate device QbTest): the visual version is cleared to aid assessment and treatment evaluation, the auditory version to aid assessment only, and the cleared labeling restricts interpretation to qualified professionals; clearance is a substantial-equivalence finding, not a validation of diagnostic accuracy (REGULATORY STATUS). Professional guidance points the same way: the AAP's ADHD guideline rests diagnosis on clinical criteria, its 2024 evidence review concluded sole reliance on continuous performance tests cannot be recommended, and the AACAP practice parameter states testing is not mandatory for routine diagnosis; no federal FDA or DEA rule requires any CPT before stimulant prescribing (prescribing law concerns controlled-substance registration, monitoring programs, and telemedicine rules, with the DEA's telemedicine flexibilities currently extended through December 2026), so a clinic's "objective test required" policy is a clinic, payer, or institutional rule, not law (CONVENTION and LAW). Payers are openly skeptical: Aetna's ADHD policy classes computerized attention and vigilance testing as experimental, Centene finds insufficient evidence for routine diagnostic use, Cigna's neuropsychological-testing policy (effective May 2026) does not establish uncomplicated-ADHD CPT testing as a covered indication, and a 2026 UnitedHealthcare policy names comparable technologies unproven, so coverage should be verified, never assumed (PAYER POLICY). Billing distinguishes the automated-instrument code 96146 from professional administration and scoring (96136 to 96137), technician administration (96138 to 96139), and evaluation services (96130 to 96133); the right code follows who did what, and no code creates medical necessity. For ADA accommodations, review is individualized and multimethod, and no CPT score is dispositive (LAW and CONVENTION).
Canada and Australia register the device modestly and position it identically in practice. In Canada the system is supplied under a Health Canada Medical Device Establishment Licence (number 9655), an establishment-level licence for low-risk devices that should never be written up as a product approval, and revised federal establishment-licensing guidance takes effect in December 2026; clinically, the CADDRA Canadian ADHD practice guidelines (version 4.1) make the clinical interview and validated rating scales the mainstays, caution that standardized testing can be normal despite real impairment, and decline to let computerized results determine diagnosis, severity, disability, or accommodation eligibility, with coverage varying by province, plan, and whether testing sits inside an insured service (REGULATORY STATUS and CONVENTION). In Australia the device appears on the Australian Register of Therapeutic Goods as cognitive-assessment software (identifier 471105, Class IIa, entered November 2024), a registration rather than an efficacy endorsement; the national evidence-based ADHD guideline finds neuropsychological and computerized assessment not required for diagnosis and not shown to improve diagnostic accuracy enough for routine use, and the March 2026 Medicare Benefits Schedule contains no continuous-performance-test item, so any reimbursement rides on the practitioner's broader assessment service (REGULATORY STATUS, CONVENTION, and PAYER context). In all three countries the same sentence serves the reader: the T.O.V.A. contributes a standardized performance sample to a clinical evaluation that no jurisdiction requires and no instrument replaces.
Instrument facts, psychometric honesty, and rights complete the page. Version discipline first: the current product is T.O.V.A. 9 on the 9.1 software line, official 2026 manuals reference different releases, and licensed installations vary, so the build identifier printed on the patient's report is the citable version. The visual and auditory versions keep separate norms (roughly 1,714 and 2,680 participants in the FDA filing) and the detailed manual describes historically regional, demographically homogeneous components in the visual norms, including child, adolescent, and adult samples from Minnesota described as approximately 99 percent White, so results deserve integration with the examinee's linguistic, cultural, educational, and sensory context. Reliability is stronger for response time and response-time variability than for individual error counts, commission performance shows practice effects on repeat administration, and diagnostic accuracy is genuinely moderate: an early clinical study reported sensitivity near 80 percent with specificity near 72 percent, a pediatric study found the test flagged about 30 percent of control children, response-time variability is the most reproducible ADHD-associated CPT finding while remaining nonspecific, and the pooled commercial-CPT meta-analysis landed at an area under the curve of 0.70 to 0.80; no current study supplies a T.O.V.A.-specific false-negative rate for high-ability or compensated presentations, a gap worth stating rather than filling with a guess. On rights: a licensed practice may print and save its own patients' reports into the chart, export is by file rather than any public API, and the licence bars redistribution of manuals, norms, report templates, and scoring logic, so public pages and tools may describe the paradigm but not clone the test, its stimuli, or its thresholds. T.O.V.A. is a registered trademark of The TOVA Company. BastionGPT is not affiliated with, or endorsed by, The TOVA Company. This page reproduces no stimuli, timing parameters, norms, thresholds, or report content.
The numbers behind these errors are specific. The 510(k) clearance record limits the device to an aid role with asymmetric visual and auditory indications; a meta-analysis of 19 commercial-CPT studies pooled sensitivity near .75 and specificity near .71; and Aetna's ADHD policy classes computerized attention testing as experimental for the indication. The BastionGPT Clinical Advisory Board sees the same errors most often in T.O.V.A. documentation reviews:
BastionGPT is specifically trained, tuned, and clinically tested on ADHD evaluation reports.
See how clinicians use it day to day on the AI therapy notes page.
Many BastionGPT users report saving more than 90 minutes per day on documentation.
HIPAA-compliant with a signed BAA on every plan. Your data is never used to train models. BastionGPT drafts, you review and sign.
Performance on a 21.6-minute computerized go/no-go task (10.8 minutes for the preschool visual form), split into a low-target half that taxes sustained attention and a high-target half that taxes response inhibition. The report returns age- and sex-referenced standard scores (mean 100, SD 15) by quarter, half, and total for omission errors (missed targets), commission errors (responses to non-targets), response time, response-time variability (the most reproducible ADHD-associated CPT variable), and d-prime response sensitivity, plus anticipatory-response and validity information; halves and totals are separate aggregations of the underlying responses, not averages of quarter scores. Each variable has nonspecific causes: omissions can reflect fatigue or sensory access as well as inattention, commissions can reflect a speed-accuracy tradeoff as well as impulsivity, and variability moves with sleep, mood, pain, and engagement. It measures a constrained laboratory sample of attention and response control, never symptoms, impairment, onset, or the DSM criteria.
They are distinct tests, not two skins on one instrument. The visual version is normed for roughly ages 4 through 80, carries the FDA indication to aid both assessment and treatment evaluation, generates the Attention Comparison Score, and hosts the performance-validity analysis for ages 17 and older. The auditory version is normed for ages 6 through 29, is cleared to aid assessment only, and produces no ACS. Their norms are separate, so their standard scores are not interchangeable, and a person can perform differently across modalities. Use the auditory version when a defined clinical question concerns auditory attention or when visual interpretation is limited, not routinely to generate more data: no strong evidence shows that administering both materially improves diagnostic accuracy over a good multimethod evaluation. Whichever is used, the note says which, because every downstream fact (norms, ages, indication, ACS) hangs on that word.
A visual-only composite, formerly called the ADHD Score and earlier the Attention Performance Index, built from three components: response speed from the first half, response sensitivity (d-prime) from the second half, and overall response-time variability. It compares the examinee's pattern with two manufacturer reference groups: a score below zero means the performance profile is more similar to the independently diagnosed ADHD comparison sample, and at or above zero more similar to the normative sample. What it is not: a DSM threshold, an ADHD probability, a severity rating, a medication rule, or a measure of real-world impairment, and a result above zero does not exclude ADHD, because many diagnosed people perform normatively on structured tasks. A value near zero is a borderline similarity finding with limited clinical meaning. Auditory reports mark it not applicable, and the proprietary equation stays in the manual. Defensible phrasing: "more similar to the manufacturer's ADHD comparison sample; an aid to assessment, not a diagnosis."
No, on both words. It is FDA-cleared, not approved: 510(k) K173915 (March 2018) found the system substantially equivalent to a predicate device, the QbTest, under product code LQD, with the regulatory class listed as unclassified; a 510(k) clearance addresses intended use and equivalence, not diagnostic accuracy. And the cleared indication is an aid: the visual version aids assessment and treatment evaluation, the auditory version aids assessment only, and results are to be interpreted by qualified professionals, with the manufacturer's manual stating the device does not diagnose any disorder. Abroad the registrations are similarly modest: a Health Canada establishment licence and an Australian ARTG software registration, neither an efficacy endorsement. Independent evidence matches the labeling: pooled commercial-CPT discrimination sits at an area under the curve of roughly 0.70 to 0.80, and guideline bodies conclude CPTs cannot carry a diagnosis alone. "FDA-cleared aid to the assessment of attention deficits" is the whole defensible claim.
It depends on the question, and the prescriber decides. For an untreated baseline, an off-medication administration can be appropriate when the prescriber has planned the hold and it is safe; for treatment monitoring, test on the usual regimen or at a deliberately chosen post-dose interval, commonly near expected peak effect. Either way the chart records the medication name and dose, on-or-off status, the last dose's date and time, the elapsed interval, and whether the status reflected usual treatment, a prescriber-directed hold, or a planned comparison. Two practices to avoid: instructing a patient to stop a stimulant because "the TOVA must be done unmedicated" (medication decisions belong to the prescriber and the clinical question), and applying a generic washout duration across formulations, which pharmacology does not support. A dose-to-test interval that goes unrecorded quietly invalidates the retest comparison the whole exercise was for.
No. A normal result shows that performance stayed broadly age-referenced during one short, quiet, novel, externally structured session; it says nothing about regulation across prolonged, distracting, self-directed real-world demands, which is where ADHD lives. False negatives are an accepted feature of the instrument class: guideline bodies note that many people who meet criteria perform normally on standardized tasks, high-ability and compensated adults do so with particular frequency (no T.O.V.A.-specific false-negative rate exists, a gap to state rather than fill), and diagnosis rests on developmental history, cross-setting symptoms, impairment, and informant evidence. Defensible documentation: the result was discordant with, but does not invalidate, the cross-setting clinical evidence, and the T.O.V.A. was considered non-exclusionary rather than used to overrule the history, ratings, records, and observed impairment. The reverse discipline also holds: an abnormal result in someone without the clinical picture flags a performance difference to explain, not a diagnosis to award.
Anything that breaks standardization or the response record: excessive anticipatory responses (responses too fast to be stimulus-driven, suggesting guessing, rhythmic responding, or misunderstanding), too few usable correct responses, room interruptions, hardware or administration errors, observed failure to follow instructions, or behavior that undermines the norms' assumptions. The current software adds an unusual-pattern or performance-validity analysis with deliberately restricted applicability: visual administrations, ages 17 and older, and only when the data conditions for it are met, so its absence from a report is often correct rather than an omission. Two rules govern the write-up. Validity is resolved before any score is interpreted, because a score table cannot outrank an unresolved standardization problem. And a flag is a caution requiring clinical adjudication, not a finding of malingering or deception; if effort or reporting style genuinely concerns you, that conclusion needs converging evidence from the broader evaluation, never a single CPT output.
The visual version carries exactly that FDA-cleared indication, and the literature supports a group-level story: response-time variability, the variable most associated with ADHD, improves under stimulant treatment. The individual-level discipline is what the note must carry. Compare against a true baseline with matched conditions (time of day, sleep, hardware, setting), document the interval, prior exposure, and expected practice effects (commission performance in particular improves on repetition), and record the medication, dose, and dose-to-test timing, since a peak-effect test and a trough test are different experiments. Reliability is stronger for timing measures than for individual error counts, so small error-count changes prove little. Conclude in performance terms: "more consistent under the medicated condition," integrated with symptom ratings, classroom or work function, side effects, sleep, appetite, and vitals, and never as the sole basis for dose selection. There is no validated retest schedule; retest when the result could change a management decision and the comparison can be interpreted.
Yes. Give it the facts (modality and build, purpose, testing conditions, medication and last-dose timing, validity findings, the standard-score pattern by quarter and half, the ACS where applicable, and the surrounding evaluation) and it drafts the results section: identification exact, validity resolved first, the pattern described in prose, the Attention Comparison Score framed as an aid, and the integration statement running both directions, ready for your review. It can also cross-check a finished note for approved-versus-cleared inflation, diagnosis or exclusion language, an ACS treated as a threshold, a missing modality or medication interval, or a retest gain with practice effects unmentioned, and it can draft the family explanation and titration-retest comparisons. BastionGPT is HIPAA-compliant with a signed BAA on every plan, and your data is never used to train models.
The instrument facts and compliance claims on this page trace to these sources, last verified August 2026:
Educational content, not legal or billing advice. Sample notes are fictional. Follow your organization's policies and your board, payer, and jurisdiction requirements.