Completed NA INTERVENTIONAL NCT07383870

Is as good as enough to replace? — an AI-and-ultrasound venipuncture device tested arm against arm — a clinical trial (ClinicalTrials.gov)

Seoul National University Hospital Updated 2026-05-28

A single-center randomized crossover trial comparing a device that assists venipuncture using AI and ultrasound against manual venipuncture. Each participant is punctured in both arms, one by the device and one by hand, with the same operator performing both. The primary measure is first-attempt success rate, and the trial is registered as a non-inferiority confirmatory design. Target enrollment is 190.

Trial overview (primary data)

  • StatusCompleted
  • ConditionsVenipuncture
  • InterventionsDEVICE: AIIV L2, PROCEDURE: Manual venipuncture
  • SponsorSeoul National University Hospital
  • Target enrollment190 participants
  • Period2026-01-26 〜 2026-02-12

Key points

  • A single-center randomized, crossover, open-label non-inferiority confirmatory trial comparing an AI-and-ultrasound venipuncture device against manual puncture, with a target enrollment of 190.
  • Each participant is punctured in both arms, one by device and one by hand, and the same operator performs all procedures, removing both individual vein variation and operator skill from the comparison.
  • The primary measure is first-attempt success rate; secondary measures are procedure time, hemolysis, pain on a numeric rating scale, within-subject pain comparison, and acceptability.
  • A non-inferiority design requires a tolerable margin fixed in advance, but the specific margin is not stated in the record.
  • The registered duration is 17 days, against a median of 534 days across the 803 records this site holds as of 2026-09-04, because every assessment completes immediately after the procedure.

1A trial that does not need to win

When something new is tried, the usual aim is to show it is better than what came before. What this trial sets out to show is that it is not worse. A device that assists venipuncture with AI and ultrasound is compared against puncture by hand. Drawing blood and placing a line are procedures nearly everyone needs at some point, and they are already well established.

The value of a device therefore lies not in beating the human hand but in holding the same level while bringing something else.

2Superiority and non-inferiority ask different questions

The point of comparisonA superiority trialA non-inferiority trial
The questionIs it better than the standard?Is it not worse than the standard?
What it requiresEnough size to detect a difference if one existsDeciding in advance how much difference is tolerable
How the conclusion is usedA reason to replaceGrounds to allow replacement when other advantages exist

A non-inferiority trial does not hold together unless the tolerable margin is fixed beforehand. Set the margin wide and anything passes; set it narrow and the exercise becomes a proof of superiority in all but name. This registration states that the design is a non-inferiority confirmatory trial, but the specific margin is not part of the record.

3Comparing one arm against the other

Target enrollment for this study190Single center, randomized crossover
Records mentioning non-inferiority among the 803 this site holds as of 2026-09-04121.5% of the total
Records mentioning a crossover design, of those same 803121.5% of the total

Each participant is punctured in both arms, one by the device and one by hand. How easily a vein can be found varies enormously between people, so comparing separate individuals never puts the two methods on equal ground. Compared within one person, that variation disappears, and because the same operator performs every puncture, differences in individual skill stay out as well.

The primary measure is whether the first attempt succeeded, with procedure time, hemolysis, pain and acceptability as secondary.

4What the brevity says

The registered duration is 17 days. Against a median of 534 days across the records held here that is extraordinarily short, and the reason is that every assessment finishes moments after the procedure. Take a procedure that needs no follow-up, and even a trial asking whether replacement is acceptable can close this quickly.

Read the other way, long-term safety and behavior in the hands of operators who are not yet used to the device stay outside what this design can show.

Why it matters

Where an established procedure is the thing being replaced, the question asked is non-inferiority rather than superiority. Because non-inferiority means nothing until the tolerable margin is set, the choice of margin decides how strong the conclusion can be. Comparing within one person and holding the operator constant is a useful pattern wherever the comparator is human skill.

FAQ

What does non-inferiority mean?
It means showing not that the new method is better, but that it is not worse by more than a margin fixed in advance. Such results are used as grounds for replacement when the new method offers other advantages.
Why compare one arm against the other?
How easily a vein can be found differs greatly between people, so comparing separate individuals would not put the methods on equal footing. Comparing within one person, with a single operator, narrows the sources of difference.
Has the device been shown to beat the human hand?
No. The design is a non-inferiority trial, and this record does not report results.

Sources (primary)

Source: ClinicalTrials.gov (U.S. NIH/NLM, public domain). This site does not provide medical advice. Verify the latest and exact details with the official source. This site is not endorsed or certified by the NIH/NLM.

#Clinical trials#AI#Healthcare#Medical devices#Non-inferiority trial
Disclaimer: This site independently summarizes and classifies information based on official data sources. Always verify the latest and accurate information with the official sources. Content on finance, health, legal, and security is information, not advice. This site is not an official website of the U.S. government.