What Did The FDA Actually Measure Before Authorizing A Blood Draw Robot?

On August 19, 2026, the U.S. Food and Drug Administration announced on its website that it had authorized Aletta, a robotic blood draw device made by Vitestro, and described it as first of its kind.¹ The agency said the authorization is based on clinical data showing that Aletta achieves successful blood draw rates comparable to or better than trained human phlebotomists, across patients of varying health status, self-reported difficult vein access and skin tone.¹

The FDA's own sentence ends with a condition: "when it proceeds with a stick."¹ Its main cohort had no human comparator, and the paper states that no ancestry information was collected. It reports no skin tone analysis, and it does not say whether skin tone was recorded.⁵ If a skin tone result exists, it is not in any public document examined here.

What did the FDA actually authorize?

The FDA database has the administrative record. The De Novo number is DEN250046, the classification name is robotic peripheral venous blood collection system, the request was received on September 26, 2025, and the decision to grant came on August 19, 2026.³ That is a little under eleven months. The order letter is addressed to Vitestro and names the device as the Aletta ARPD Gen1 US.² The new product code is SIW.⁴

The FDA's definition of the device type describes a machine that scans the patient's limb "using ultrasound and/or other imaging modalities to identify a suitable venipuncture site" and, "if an acceptable site is located," uses a robotic arm to insert the needle.⁴ That conditional is built into the classification itself. The same page lists the submission type as 510(k), which applies to future devices of this type. Aletta itself came through De Novo, according to the FDA release and the De Novo database record.⁴

One detail in the paperwork does not reconcile. The order letter is dated August 19, 2026, but its header reads "Dated: June 5, 2026" and "Received: June 5, 2026", while its body says the FDA received the De Novo request on September 26, 2025.² The database also gives September 26, 2025.³ The June date is probably a later submission. The order does not say.²

What does the 94.5 percent count?

The FDA release gives no numbers and names no trial.¹ The 94.5 percent figure appears in the ADOPT trial, published online in Clinical Chemistry on April 10, 2026. The FDA release does not cite it, and no document examined confirms it was part of the FDA submission. It was sponsored by Vitestro and run at sites in the Netherlands, with data collected from May to September 2025.⁵

The paper's sentence is precise: "The overall first-stick success rate, in patients where the ARPD could identify a suitable vein, was 94.5% (95% CI, 93.3–95.5)."⁵ The predefined subgroups were difficult venous access, at 92.7 percent, obesity, at 97.4 percent, and age 65 or older, at 93.4 percent.⁵ Health status is not named as a subgroup in the abstract or Methods, and the FDA does not say how it defines the term.⁵

Patients the device screened out because it found no suitable vein are not in the 94.5 percent.⁵ The screen-out rate is not in the portion of the full text available for this piece. Without it, the share of all arriving patients who left with a successful first stick cannot be calculated here. That is the number a laboratory manager would want, and it is a different number from the one being quoted.

Comparable to whom?

"Comparable to or better than trained human phlebotomists" reads like a head-to-head result. The FDA attaches it to no stated comparator or margin.¹ In the published trial, cohort 2 was a single-arm study with no human phlebotomist comparator.⁵ The paper's introduction says published first-stick success rates generally range from 80 to 97 percent, and may fall below 80 percent in some settings.⁵ A figure of 94.5 percent falls inside that range. It is not a like-for-like comparison, because the 94.5 percent excludes screened-out patients and the definitions of success may differ.

A single-arm design is not automatically disqualifying. Federal rules accept trials without matched controls as valid scientific evidence, but they say effectiveness evidence should consist principally of well-controlled investigations unless the Commissioner accepts other evidence. Federal rules count "studies and objective trials without matched controls" as valid scientific evidence for device classification.⁷ The rule sets the general standard. It does not say what Vitestro submitted.⁷

The trial registry describes a phase built for the comparison: "Phase B1 is a confirmatory, Pivotal Clinical Study, required for regulatory approval, in which non-inferiority should be demonstrated in comparison to manual blood drawing." The paper does not say which registry phases its two cohorts correspond to.⁶ The record, last updated on August 14, 2026, shows no posted results, lists the study as still recruiting with estimated completion in 2029, and marks it as not an FDA-regulated device study.⁶ It does not say which phase supports the FDA decision.⁶

Was it tested across skin tones?

The FDA release says the results held across skin tones.¹ The peer-reviewed paper cannot confirm it. Its Methods state: "No ancestry information was collected."⁵

The order letter mentions skin tone once, as an example of a subgroup analysis that clinical performance testing must include under the special controls for this device type: "Sub-group analyses based on patient characteristics affecting device performance or safety (e.g., skin tone, body mass index, anticoagulation, or known difficult intravenous access)." The special controls apply to Aletta as well.² The order reports no trial results and no subgroup findings.² If the FDA reviewed skin tone data, it would most likely appear in the agency's decision summary or in Vitestro's labeling. Neither was among the documents examined for this piece.²

Who produced the evidence?

Vitestro sponsored and led the published trial, with independent hospital laboratories as sites. The company is the trial sponsor, the lead authors work in its Clinical Affairs department, and Europe PMC lists it as the grant agency.⁵ The registry entry is submitted and maintained by Vitestro B.V., and the National Library of Medicine does not verify its science.⁶ Sponsor funding is ordinary for device trials. It means the published cohorts are manufacturer-sponsored evidence from Dutch sites, and should be read that way.