Every Number We Have Comes From the Man Who Made the Tool

The standard explanation for why hair transplantation is harder in tightly curled hair is that the follicle curves under the skin. You cannot see the curve, the punch travels straight, and the graft is cut through — transected — before it comes out.

It is a good explanation. It is repeated everywhere. And the largest study of the procedure in patients of African descent found that it is not the main thing going on.

What the two series report

In 2016 a retrospective case series of 18 patients with tightly curled Afro-textured hair compared three punch types in sequence — conventional sharp and dull rotary punches, then a two-pronged curved non-rotary punch — with several needle gauges, until transection rates were acceptable.

The curved non-rotary punch achieved a transection rate below 5 percent in all instances. The rotary punches, sharp and dull, completely failed or produced excessive transection in eight of the 18.

The paper opens by stating that patients with tightly curled hair using conventional rotary punches frequently have unacceptably high transection rates, and that these patients are unsuitable candidates for the procedure. Which is the case for the new tool, made in the first paragraph.

In 2023 came the larger one: 64 patients of African descent — 45 men and 19 women — across seven clinics. Twenty-eight of them fell into the authors’ Class V, defined as the highest degree of hair curliness, skin thickness and firmness. Only 19 or 18 gauge punches were used.

Mean transection rate 3 to 6 percent, highest in the Class V patients.

And then the sentence that reframes the subject: skin thickness and firmness had a greater effect on the maximum transection rate than hair curliness.

Both papers have the same author. Umar is the sole author of the 2016 paper and the first author of the 2023 one, and he developed the two-pronged curved non-rotary punch that both papers are evaluating.

I want to be careful about what that does and does not mean. It does not mean the numbers are wrong. Device developers publish their own early series in every surgical field, because at the start nobody else has the device. It does mean there is no independent replication — no series in tightly curled hair by anyone who did not make the instrument — and a 3 to 6 percent transection rate is exactly the kind of figure that ought to have one before it is quoted as a standard.

The comparison numbers, and the population nobody named

To know whether 3 to 6 percent is good, you need something to hold it against.

The cleanest comparator is a single-blind study in which each of ten patients had three rows of test areas harvested, one punch type per row, all at 0.9 millimeters, with every follicular unit checked under a microscope. Sharp punch: 23.9 percent transection. Serrated: 18.8. Blunt: 14.5. Hair yield ran the other way, highest with the blunt punch.

A companion study from the same group found the mid-occipital zone had the lowest transection rate, and that the right side of the patient ran slightly higher than the left — 17.7 against 16.3 percent — which is a nice reminder of how much of this is the surgeon’s hand rather than the patient’s biology.

The ethnicity and phototype of those twenty patients is not stated in either paper.

So the comparison is between a device series in African-descent patients written by the device’s inventor, and a set of general transection numbers from an unnamed population. That is the state of the evidence, and it is thinner than the confidence with which the procedure is marketed.

Complications: a good review with one thing missing

A 2025 scoping review pulled together 43 publications on complications in modern hair transplantation, from 1985 onward.

Two large series put the overall complication rate at 1.2 and 4.7 percent. Bleeding requiring intervention ran up to 8 percent, persistent numbness up to 11, infection up to 11 — including two reports of Kaposi varicelliform eruption and one of mucormycosis. Shedding at the donor site reached 4.1 percent and at the recipient site 6.5.

The most common donor-site complication was hypertrophic scarring or keloid formation after strip harvesting, at up to 15.1 percent.

At the recipient site the numbers get wide and soft — crusting up to 54.8 percent, frontal swelling up to 50, sterile folliculitis up to 53.3 — and the authors say so directly, describing these as poorly defined with a broad range of incidences.

None of it is stratified by ethnicity, ancestry or phototype.

That 15.1 percent scarring figure is the one to be careful with. It is a strip-harvest number in an unstratified population, and it will be tempting for anyone writing about keloid-prone skin to reach for it. It is not a figure for Black patients. Nobody has published one.

What exists for keloid after the newer extraction technique is two case reports, one patient each. Neither states the patient’s ethnicity in its abstract. One of them opens by noting that keloid scarring is a rare complication of transplantation and usually associated with strip harvesting rather than extraction — which is the honest framing, and also an admission that the denominator is unknown.

And graft survival, the number a patient actually cares about — how many of the transplanted follicles are still growing a year later — has never been measured and published in Afro-textured hair at all. The transection rate tells you how many grafts survived the harvest. It says nothing about how many survive the transplant.

Transplanting into scarred scalp

The evidence gets genuinely small at exactly the point many Black women are asking about surgery.

Central centrifugal cicatricial alopecia destroys the follicle and leaves scar. Traction alopecia at the margin, past a certain point, does the same. Both are conditions where the question is not thinning but replacement.

The flagship paper on transplantation for CCCA reports two patients. Two African American women, each with five to six years of progressive crown loss, who had a scalp biopsy and a test session before the procedure. Hair growth was visible at the recipient sites in both, beginning between four and five months after the test session, with no postoperative scarring in either.

The authors state their own evidence base plainly: there is a paucity of published prospective studies reporting successful surgical therapy for CCCA. And they attach a condition that gets dropped in the retelling — the approach is described as safe in women with end-stage CCCA who histologically display a lack of inflammation on scalp biopsy. Not CCCA in general. Burnt-out CCCA, confirmed under a microscope first.

Widening the lens to every primary scarring alopecia, a systematic review searching back to 1960 found fifteen reports and 34 patients in total. Twenty-six had moderate to positive results and eight had negative results or recurrence of disease. Positive outcomes were reported in CCCA, morphoea, discoid lupus, pseudopelade and folliculitis decalvans; lichen planopilaris and frontal fibrosing alopecia produced both. The authors add that the data must be interpreted with caution due to concern for positive-result publication bias.

Thirty-four patients, nearly sixty years, and the authors warning you that the good results are the ones that got written up.

For traction alopecia there is no cohort study at all — a 1986 paper on surgical correction of traumatic marginal alopecia in Black women, a single case of extensive loss from a ponytail, and a single case in a beard.

The gap between the need and the evidence

Traction alopecia was found in 17.1 percent of African schoolgirls and 31.7 percent of women in the population studies behind a 2008 analysis of 1,178 people. In that analysis, the risk was highest when traction was applied to relaxed hair — an odds ratio of 3.47 against natural hair — and only 18.9 percent of those with traction alopecia had never had symptoms related to hairdressing. I went through the tension mechanism and what is reversible in the piece on the school years.

Set that prevalence beside a surgical literature of one 1986 paper and two case reports and you have the shape of the problem. A common, partly preventable, permanently scarring condition, and essentially no published evidence on what surgery achieves for it.

Punch, test session, biopsy

I would ask what punch, at what gauge, and what transection rate the surgeon measures in their own practice on hair like mine — because the published figures come from two papers by one interested party, and any surgeon doing this work regularly should have their own numbers.

I would ask whether a test session comes first. In the CCCA literature it is not optional; it is the design.

And if the loss is scarring rather than thinning, I would want the biopsy result in hand before anything else, because the only published condition on which transplantation into CCCA was called safe is the absence of active inflammation under the microscope. A surgeon who does not ask for one is not following the two-patient study everyone is citing.

The honest summary is that the technique appears to work, in the hands of the people who developed it, on the patients they selected. That is a real starting point. It is not the same as knowing what it does.

This article is educational and is not a substitute for care from someone who can look at your skin. Anything new, changing, painful or spreading should be assessed rather than researched.