Measurement Protocol v1.0

On this page 16 sections

1. Current status
2. Why a clinic is publishing a protocol at all
3. What can be measured, and what is a judgement
4. Defining transection
5. Donor density
6. Punch
7. Classifying hair pattern
8. Skin thickness and firmness
9. Complications
10. Outcome at 6, 9 and 12 months
11. Photograph standardisation
12. Missing data
13. Inclusion and exclusion
14. Retrospective and prospective data are never mixed
15. Changes to this protocol
16. Comments

This is a measurement protocol, published before any data has been collected. No study has started. Nothing on this page is a result.

We publish it now, in advance, for one reason: a protocol written after the numbers are in is not a protocol. It is a description of whatever happened to be favourable. Putting the definitions on the record first is the only way anyone — including us — can tell the difference later.

Why a clinic is publishing a protocol at all

While researching this site we went looking for a published transectionTransection Cutting through a follicle during extraction, damaging or destroying the graft. rate specific to tightly coiled hair. We did not find one. The closest study measured it in six coiled-hair patients out of 152, and concluded that scalp thickness and firmness affected the rate more than curl did.

That gap is recorded on our verification page as something we could not source. But we are a clinic that performs these procedures. We are in a position to measure the thing we could not find measured — and continuing to write “no data exists” while sitting on an unpublished dataset would be a strange kind of honesty.

So this is the protocol we intend to work to. It is deliberately conservative, and several of its rules make our numbers look worse rather than better. That is the point.

What can be measured, and what is a judgement

Confusing these two is the most common failure in clinic-published data. A number that came out of someone’s opinion is not a measurement, however many decimal places it carries.

Rule: judged variables may be recorded and reported, but never presented as measurements, and every one of them carries the name of the person who made the judgement.

Defining transection

This is the definition most easily bent, so it goes first.

Transection is not one event. A study we read separates graft damage into five kinds — transection, paring, fracture, dermal papilla injury and partial bulb injury.^[1] A single “transection rate” hides four of them.

More importantly, the same study examined the same grafts twice: 7.40% under a loupe, 6.34% under a microscope. The instrument changes the number. A transection rate published without its magnification cannot be compared to anyone else’s, including ours.

Transected follicle: a hair follicle whose integrity has been disrupted by cutting during extractionGraft extraction A distinct step in the procedure, attributable to a named person., visible at ×10 magnification or greater. Paring, fracture and papilla injury are recorded separately and are not counted as transection.

Mandatory fields with every rate: magnification · instrument (loupe or stereomicroscope) · who counted.

The denominator

A follicular unit contains one to four hairs. If one hair of a three-hair unit is cut, is that a lost graft? The literature answers this inconsistently, which is why rates from different clinics are rarely comparable.

We will report both:

• Primary: transected follicles ÷ total follicles in the grafts examined
• Secondary: grafts containing ≥1 transected follicle ÷ grafts examined

Sampling: 100 randomly selected grafts per case. Counting all of a three-thousand-graft session is not realistic, and pretending otherwise would be the first lie in the dataset.

Blinding: the person who extracted a graft does not count its transection. A different technician counts, without knowing who performed the extraction. This is the minimum condition under which the number means anything.

Donor density

Recorded with instrument, magnification, measurement sites and unit. Two units, separately, because they are routinely confused: follicular units per cm² and hairs per cm².

Sites: occipital midline plus two lateral points, against fixed anatomical references.

Measured on trimmed or shaved scalp. Visual counting through unshaved coiled hair is not reliable, and a density figure taken that way is a guess with a number attached.

Punch

• Outer diameter in millimetres. Not gauge — gauge is ambiguous between systems.
• Type: sharp / dull / hybrid / flared / curved
• Motion: rotary / oscillating / manual
• Speed, where applicable
• If the punch changes mid-case, both are recorded. A single “punch diameter” field would hide the most interesting thing that happened.

Classifying hair pattern

We use the 3A–4C scale across this site because readers use it. It is worth being clear about what it is: a scale devised for a hair-care product line, not a clinical instrument, and its agreement between observers is poor.

That makes it acceptable for communication and unacceptable as a study variable. So three things are recorded:

1. Walker type — judged, with the name of whoever classified it
2. An objective proxy: complete turns per centimetre on a stretched lock, or curl diameter in mm
3. Hair shaft cross-section, where available

Skin thickness and firmness

Two published series report that these affected transection more than curl did. A protocol that does not record them is broken before it starts.

We will record an ordinal grade together with the name of the person grading, and state plainly that it is an operator judgement rather than an instrument reading. Where a published grading scale is used, we will name it and note that its author has a commercial interest in the devices it was developed alongside.

Complications

A complication rate without a time window is meaningless. Each is defined with the point at which it is assessed:

Poor growth is not a complication at three months. It is the expected course. Assessing it early inflates the rate and tells the patient something untrue about their own result.

Outcome at 6, 9 and 12 months

Twelve months is the minimum for a final assessment. Six and nine months are interim and will never be presented as final.

Two independent axes, because they diverge more often than clinics admit:

• Objective: hair counts in a defined recipient zone from standardised macro photographs with a reference marker in frame
• Subjective: blinded independent assessors, plus patient-reported outcome

Assessors will not be told which photograph is before and which is after, and images will be presented in randomised order where practical.

Photograph standardisation

Without a standard, a twelve-month photograph looks better simply because the hair is longer. Fixed:

• Distance, focal length and lighting — a fixed rig, not a phone held at arm’s length
• Five to seven set angles, in the same order every time
• The same hair length. This is the most frequently broken rule and the most misleading.
• A scale marker in frame
• Same background; no flash variation; wet-or-dry state held constant

Missing data

Blank fields are not allowed. Every absence is coded with the reason for it:

NOT_MEASURED · NOT_RECORDED · PATIENT_DECLINED · LOST_TO_FOLLOWUP · NOT_APPLICABLE

No imputation. Missing data carries information — often about the cases that went least well — and filling it in destroys exactly that.

Inclusion and exclusion

Written before collection begins. Criteria added afterwards are not criteria; they are selection.

Included: Afro-textured hairAfro-textured hair Tightly curled hair type, the umbrella descriptor used across this site. as defined by the objective proxy above · primary FUEFollicular unit excision Extraction of individual follicular units from the donor area with a punch. · consent recorded

Recorded as separate cohorts, never merged into the main analysis:

• Active scarring alopeciaScarring alopecia Alopecia in which the follicle is destroyed and replaced by scar tissue. — with biopsy inflammation status recorded
• Previous transplantation in the same area (repair cases behave differently)
• Active traction alopeciaTraction alopecia Hair loss caused by sustained tension on the follicle, concentrated at the hairline and temples. where the causative tension is ongoing

Retrospective and prospective data are never mixed

This is the strictest rule here, because it is the easiest one to break by accident.

Permitted use of retrospective records: feasibility only — estimating how many eligible cases exist, and designing the collection form. Always labelled as feasibility.

Not permitted: publishing any transection rate or outcome figure derived from retrospective records. There is no way to make such a number trustworthy after the fact, and presenting one alongside prospective figures would contaminate both.

Changes to this protocol

Each version is published with a number and a date. If a definition changes after collection has begun, the change is shown here with its reason, and any affected figures are reported under both definitions.

We would rather publish an awkward protocol history than a clean one nobody can check.

Comments

If you work in this field and think a definition here is wrong, tell us before we start collecting. Corrections are far cheaper now than they will be in twelve months, and a protocol improved by someone outside the clinic is worth more than one written entirely inside it.

Clinical references

1. Park JH, You SH. Various Types of Minor Trauma to Hair Follicles During Follicular Unit Extraction for Hair Transplantation. Plastic and reconstructive surgery. Global open. 2017;5(3):e1260. view source →

References are checked on the dates recorded in our evidence library. If a source has been superseded, tell us and we will update it.