Remote CPT Testing: Validity and Environment Controls

Can you trust a continuous performance test (CPT) that a patient takes at home? The research says yes, with conditions. Studies comparing home and lab administration have found significant correlations between settings, and they found that ADHD and control groups still differ when tested at home. Browser-based timing is less precise than a lab setup, but it is precise enough for reaction-time measures. What decides whether a remote CPT is valid is the setup: a standardized environment, a consistent device, a check on effort, and a note of anything that went wrong.

Photo by Brandy Kennedy on Unsplash

This guide is written for PMHNPs and other prescribers who evaluate and follow ADHD patients over telehealth, where an in-office test isn’t an option. Psychologists adding a remote measure to a battery will find the same controls apply. If you’re new to the test itself, start with what a continuous performance test is. If you’re weighing formats, see remote vs. in-office ADHD testing. This page covers the next question: how to make a remote result defensible. We’ll also show where objective ADHD testing with AxonCPT fits. To compare the remote tools themselves, see best remote ADHD tests for telehealth practices.

What does the research say about remote continuous performance test validity?

The evidence is modest in size, but it points in a consistent direction:

• Young adults, home vs. lab. Participants took an online CPT once at home and once in the lab, a week apart. Home and lab scores showed acceptable reliability, and the online scales correlated modestly to highly with a gold-standard CPT. In a second study, participants with ADHD made significantly more omission and commission errors and had greater response-time variability than controls [1].
• Children, lab then home. In 73 children aged 6–13, split-half reliability was high in both the lab and at home. Scores correlated significantly across settings, and the ADHD and control groups differed in both settings [2].
• Adolescents and adults, home vs. clinic. In a validation study of the QbCheck, an online test taken at home correlated highly with the clinic-based QbTest. It discriminated ADHD from controls with a sensitivity of 82.6% and a specificity of 79.5% [3]. Two of the authors were employed by the test’s manufacturer, which should be weighed when reading the results [3].

Two cautions apply. First, these studies used specific tests, so their results don’t transfer automatically to every remote CPT. Second, CPTs have only moderate diagnostic accuracy in any setting. A JAACAP meta-analysis in children and adolescents found pooled sensitivity of 0.75 and specificity of 0.71 for CPT total or ADHD scores, and concluded that CPTs should be used only within a comprehensive diagnostic process [4]. Remote administration doesn’t fix that limitation. The goal is not to make it worse.

Is browser timing accurate enough?

Reaction time and its variability are central CPT metrics, so timing precision matters. The largest independent test of experiment software, the “timing mega-study” in PeerJ, measured stimulus and response timing across lab and online platforms with dedicated hardware [5]. Its findings:

• Online studies were less precise than lab systems, with somewhat more variability in every measurement.
• Even so, for response times, most packages achieved precision under 10 ms in all browsers tested, and the best achieved under 3.5 ms.
• Precision varied considerably between operating system and browser combinations.

For clinical use, this means device and browser differences are a real source of noise, but a small one compared with the differences CPTs are designed to detect. The practical fix is consistency: have each patient use the same device and browser for every test.

What threatens validity in an unsupervised test?

Threat	What it looks like	Control
Distracting environment	Interruptions, TV, other people in the room	Written pre-test instructions, and an interruption question afterward
Device and input	Switching between phone and laptop, or a touchscreen one time and a keyboard the next	Same device and input method at every visit
Timing noise	Browser and OS variability	Consistent browser; close other apps and tabs
Patient state	Poor sleep, illness, caffeine, dose timing	Record sleep, time of day, and time since last dose
Effort and credibility	Very slow, erratic, or implausibly poor performance	Interpret alongside other validity data (see below)
Who took the test	Someone else completes it	Send the link to the patient directly, and confirm at the next visit
Practice effects	Scores improve with repetition	Space the tests out, and compare against the patient’s own baseline

The Inter-Organizational Practice Committee’s teleneuropsychology guidance makes the same general point. It says remote assessment may be reliable and valid, but clinicians should report modifications to standard procedures and state limitations on their conclusions [6].

Effort and performance validity in remote testing

Unsupervised testing makes some clinicians worry that effort will drop. CPTs can help here. A 2025 meta-analysis of 18 studies found that CPTs distinguish credible from noncredible performance with a moderate effect size (g = 0.73). The effect was smaller in criterion-group studies (g = 0.55) than in simulation studies, and the authors cautioned against relying on a single CPT [7]. A study of CPT-3 embedded validity indicators found that response-time variability and related scores flagged invalid performance, but with low sensitivity. They should be used to flag problems, not to rule them out [8].

In practice, a remote CPT result that clashes with the interview, the collateral information, and daily functioning is a reason to ask questions, and possibly to retest under better conditions. It is not a reason to discount the test completely, and it is not a reason to trust it blindly.

Try it with your next patient. AxonCPT gives you 2 free tests every month — no card required. Send a link, your patient finishes in ~16 minutes from home, and you get a norm-referenced report instantly. Create your free account →

A pre-test checklist to send patients

Copy this into your patient instructions:

1. Use a laptop or desktop if you have one, and use the same device every time.
2. Choose a quiet room with the door closed and your phone on silent and out of reach.
3. Close other apps and browser tabs.
4. Take the test at about the same time of day as last time. If you take ADHD medication, take the test at the time your clinician specified relative to your dose.
5. Don’t take the test when you’re sick, very short on sleep, or in a hurry.
6. Complete the test yourself, in one sitting.
7. If you’re interrupted, finish the test anyway and tell your clinician what happened.

How to interpret and document a remote result

• Record the conditions. Note the device, time of day, medication timing, sleep, and any reported interruptions.
• Read the time course. A vigilance breakdown across test quarters shows whether performance changed during the test. That is useful context when you’re deciding whether a lapse reflects sustained-attention problems or a one-off interruption.
• Compare like with like. For titration, compare each result with the same patient’s earlier tests under matched conditions, not only with the norm.
• State limitations. One sentence is enough, for example: “Administered remotely, unsupervised; patient reported no interruptions.”
• Never use the test alone. Put the result alongside the interview, rating scales, and collateral information.

How this looks in practice

1. A PMHNP sees a new adult patient by video and decides objective data would help.
2. The PMHNP sends an AxonCPT link along with the checklist above. The patient takes the test in their browser on a phone, tablet, or computer, with no app or login, in about 16 minutes.
3. The report arrives when the patient finishes. It shows five metrics compared with age- and sex-matched norms from a 20,000+ reference dataset, plus a Q1–Q4 vigilance breakdown. See what the report looks like.
4. At the next visit, the PMHNP asks about the testing conditions, reviews the results with the patient, and documents how they fit with the other evidence.
5. At follow-up, the patient retests on the same device at a matched time relative to their dose, and the PMHNP compares the results over time.

For where this sits in the full workup, see how to run a telehealth ADHD evaluation.

Frequently asked questions

Is an at-home CPT as accurate as one done in the clinic? Studies of home vs. lab administration show significant cross-setting correlations and preserved group differences [1][2][3]. Online timing is somewhat less precise than a lab’s [5]. The best evidence supports using home results with standardized conditions and documented limitations, rather than treating them as identical to a proctored test.

Should patients use a phone or a computer? Either can work, but consistency matters more than which device. Because timing and input vary across devices [5], use the same device and input method for every test in a series.

Can patients fake a remote CPT? Anyone can underperform on purpose, in the clinic or at home. CPT metrics can flag noncredible performance with moderate accuracy, but they should never be the only validity check [7][8].

Does remote testing affect billing? Coverage and code requirements for remotely administered tests vary by payer. See our ADHD testing CPT code guide and confirm with each payer.

Start testing your patients with AxonCPT

Remote CPT testing is only as good as the conditions you set and the context you interpret it in. AxonCPT makes the testing part fast, so PMHNPs can concentrate on the clinical reasoning and the documentation that supports stimulant decisions. Create your free AxonCPT account with 2 free tests every month and no card required, or learn more at axoncpt.com.

References

1. Raz S, Bar-Haim Y, Sadeh A, Dan O. Reliability and validity of the online continuous performance test among young adults. Assessment. 2014;21(1):108–118.
2. Bart O, Raz S, Dan O. Reliability and validity of the Online Continuous Performance Test among children. Assessment. 2014;21(5):637–643.
3. Ulberstad F, et al. Objective measurement of attention deficit hyperactivity disorder symptoms outside the clinic using the QbCheck: Reliability and validity. Int J Methods Psychiatr Res. 2020;29(2):e1822.
4. Arrondo G, et al. Systematic Review and Meta-Analysis: Clinical Utility of Continuous Performance Tests for the Identification of ADHD. J Am Acad Child Adolesc Psychiatry. 2024;63(2):154–171.
5. Bridges D, Pitiot A, MacAskill MR, Peirce JW. The timing mega-study: comparing a range of experiment generators, both lab-based and online. PeerJ. 2020;8:e9414.
6. Bilder RM, et al. InterOrganizational practice committee recommendations/guidance for teleneuropsychology (TeleNP) in response to the COVID-19 pandemic. Clin Neuropsychol. 2020;34(7–8):1314–1334.
7. Toptas P, et al. Evaluating Continuous Performance Tests as Embedded Measures of Performance Validity in ADHD Assessments: A Systematic Review and Meta-Analysis. Assessment. 2025.
8. Ord AS, et al. Examining embedded validity indicators in Conners continuous performance test-3 (CPT-3). Clin Neuropsychol. 2021;35(8):1426–1441.