Supervision is where the money hides, and the authorisation does not price it. Aletta is licensed for adults, in outpatient settings, under supervision [4], and nothing in the announcement says how many machines one supervisor may cover. Until that ratio is written down, the staffing case is a sketch with the operating cost left blank.
Start with what the human labour is actually worth. The Bureau of Labor Statistics puts the median phlebotomist wage at about $43,660 [c13b]. Spread across a 2,080-hour year that is roughly $21 an hour [2], which puts the labour content of a five-minute draw at about $1.75 [3]. No capital device earns its keep by recovering $1.75 a stick, and Vitestro's own framing does not ask it to.
The number that does the work is churn. BLS counts 139,700 phlebotomists in 2024 and 6% growth by 2034, with roughly 18,400 openings a year [c13a] [c13c]. Six percent of 139,700 is about 8,382 net new jobs over the decade [5], so 184,000 openings against that gain is around 95% replacement hiring [6]. That is the mechanism behind the delays Michelle Tarver of the FDA's device centre pointed to when she cited a growing shortage of trained phlebotomists [12] [11]. A machine is bought against vacancy and turnover, not against wages.
On throughput the arithmetic is clean. At a 109-second median against roughly 300 seconds by the company's own reckoning [1] [2], that is 33 arms an hour versus 12 [4]. The gain only appears if check-in, tube labelling and the lab can absorb it, and the trial evidence says nothing about that part of the room.
The clinical numbers are more modest than the clock. A 95% first-stick rate against a manual benchmark of 93 to 97% in the A.D.O.P.T. trial is parity, sitting mid-range of human performance [6]. Haemolysis at 0.6% against a 2% best-practice ceiling is the stronger result [7], alongside mild adverse events in 1.3% of participants with none serious or moderate [8].
The special controls are the part with a long tail. Future manufacturers must show performance comparable to or better than trained phlebotomists across patients with varying health statuses, difficult vein access, and varying skin tones [9]. TNW notes that near-infrared light and Doppler ultrasound do not read pigment the way an eye does, and whether that closes the failure gap in routine practice is now an empirical question rather than a marketing one [14]. Because the controls also cover labelling, performance testing and clinical testing, the next entrant files a 510(k) instead of a De Novo [10].
Vitestro has been at this since 2017 [5], which is roughly how long it takes to make an ordinary procedure legible to a regulator. What comes out the other end is a clinical robot whose business case fits on one line, with one blank still in it.
