Split vs Full Thickness, and What a Graft Will (and Won't) Take On
- A skin graft is a FREE transfer of skin with NO intrinsic blood supply, so it depends entirely on the RECIPIENT BED for survival - it takes in three phases: PLASMATIC IMBIBITION (0-48h, nutrients diffuse from the bed), INOSCULATION (48-72h, graft and bed capillaries link up) and REVASCULARISATION (~day 4-7, new vessels grow in).
- Because of this, a graft will only take on a WELL-VASCULARISED bed; it will NOT reliably take on BARE CORTICAL BONE (without periosteum), BARE TENDON (without paratenon), BARE CARTILAGE (without perichondrium), heavily contaminated/infected wounds, or irradiated tissue - such defects need a FLAP. With intact periosteum/paratenon a graft can still take.
- SPLIT-THICKNESS grafts (epidermis + part of dermis) take more reliably and cover large areas, the donor site heals by re-epithelialisation, but they undergo MORE secondary CONTRACTION and give poorer durability, sensation and colour - so in the hand they are used for large or dorsal defects, granulating wounds and temporary cover.
- FULL-THICKNESS grafts (epidermis + the full dermis) undergo LESS secondary contraction and give better durability, sensation and colour match, so they are preferred on the PALM, the FINGERTIPS and ACROSS FLEXION CREASES (where contracture would be disastrous); the trade-off is that they need a cleaner, better-vascularised bed and the donor site must be closed primarily, limiting size.
- DONOR-SITE choice matters: common hand FTSG donors are the wrist/hypothenar crease, antecubital fossa, medial arm and groin; GLABROUS skin (hypothenar eminence, instep) is chosen for PALMAR/fingertip defects to match the special palmar skin; STSG is usually harvested from the thigh.
- THE RECONSTRUCTIVE LADDER has eight rungs worth naming rather than gesturing at: secondary intention, primary closure, delayed primary closure, split-thickness graft, full-thickness graft, local flap, regional/pedicled flap, and free tissue transfer. It is a hierarchy of complexity, NOT an obligation to climb one rung at a time - the RECONSTRUCTIVE ELEVATOR is the better framing, and in the hand the functionally correct answer is often to go straight to a flap rather than accept a graft that will contract across a crease.
- EXPOSED BONE OR TENDON DOES NOT AUTOMATICALLY MEAN A FLAP. A DERMAL SUBSTITUTE laid over the bare structure generates a vascularised neodermis that is grafted at a second sitting, converting an ungraftable bed into a graftable one. In a systematic review of 190 studies, wound DEPTH was the primary indication, with a one-stage approach for deep dermal to full-thickness wounds over 10 cm and a TWO-STAGE approach where donor sites are limited or BONE OR TENDON IS EXPOSED. INFECTION contraindicates the matrix exactly as it contraindicates the graft - so an infected bed rules out both, and the question order is: infected, then bare, then vascularised.
- GRAFT FAILURE is most often due to HAEMATOMA or SEROMA lifting the graft off the bed, INFECTION, or SHEAR/movement - so meticulous haemostasis, a bolster/tie-over dressing, and IMMOBILISATION are essential; the graft fits on the RECONSTRUCTIVE LADDER above primary closure/secondary intention and below local/regional and free flaps.
- “Graft take = imbibition (0-48h) -> inosculation (48-72h) -> revascularisation (~day 4-7); needs a vascular bed.
- “Won't take on bare bone/tendon/cartilage (no periosteum/paratenon/perichondrium), infected or irradiated beds -> use a FLAP.
- “FTSG (less contracture, durable, sensate) for palm/fingertips/flexion creases; STSG (reliable, large) for dorsum/large wounds. Failure = haematoma/infection/shear.
- “Ladder, 8 rungs: secondary intention -> primary -> delayed primary -> STSG -> FTSG -> local flap -> regional/pedicled -> free. It is an ELEVATOR, not a staircase - pick the rung that gives the best function.
- “Bare bone/tendon with a CLEAN bed: dermal substitute now, graft at a second sitting (two-stage). INFECTION contraindicates the matrix and the graft alike.
A well-vascularised bed: muscle, fascia, granulation, paratenon-covered tendon, periosteum-covered bone, perichondrium-covered cartilage.
Bare cortical bone, bare tendon, bare cartilage, infected/heavily contaminated or irradiated beds, or exposed hardware/joint - cover with a flap.
Graft Types & Take Physiology
A skin graft carries no blood supply of its own, so it survives only if the recipient bed can nourish and revascularise it. Take occurs in three phases: plasmatic imbibition in the first ~48 hours, when the graft absorbs nutrients by diffusion from the bed; inosculation at ~48-72 hours, when graft and bed capillaries align and connect; and revascularisation from about day 4-7, when new vessels grow into the graft. A split-thickness graft (epidermis + part of dermis) is thinner, so it imbibes and revascularises more readily and takes more reliably, but it contracts more and is less durable; a full-thickness graft (epidermis + full dermis) needs a better bed but contracts less and is more durable and sensate.

- Split-thickness (STSG)
- Epidermis + part of dermis
- Full-thickness (FTSG)
- Epidermis + full dermis
- Split-thickness (STSG)
- More reliable (thinner)
- Full-thickness (FTSG)
- Needs a better bed
- Split-thickness (STSG)
- MORE
- Full-thickness (FTSG)
- LESS
- Split-thickness (STSG)
- Poorer
- Full-thickness (FTSG)
- Better
- Split-thickness (STSG)
- Heals by re-epithelialisation (e.g. thigh)
- Full-thickness (FTSG)
- Closed primarily (limited size)
- Split-thickness (STSG)
- Large/dorsal defects, granulating wounds, temporary
- Full-thickness (FTSG)
- Palm, fingertips, across flexion creases
Hand-Specific Principles & Donor Sites
- Palm and fingertips: use a FTSG, ideally from GLABROUS (non-hairy) skin - the hypothenar eminence or the instep - to match the thick, hairless, durable palmar skin and provide better sensation; an FTSG/glabrous graft is reliable for fingertip and donor-site defects.
- Across flexion creases: prefer FTSG (or a flap) because STSG secondary contraction can cause a flexion contracture - a major functional problem in the hand.
- Dorsum and large defects: STSG (e.g. from the thigh) gives reliable cover; meshing expands area and lets exudate escape but worsens cosmesis/contraction.
- Donor sites for FTSG: wrist/hypothenar crease, antecubital fossa, medial arm, groin (hairless areas preferred for the hand).
- Exposed bone/tendon/cartilage without periosteum/paratenon/perichondrium, or infected/irradiated beds: a graft will fail - choose a flap (local, regional or free).
The commonest reasons a graft fails are a HAEMATOMA or SEROMA lifting it off the bed, INFECTION, and SHEAR/movement. Prevent these with meticulous haemostasis, a bolster (tie-over) dressing or negative-pressure dressing to maintain contact, and immobilisation/splinting of the hand for the first ~5 days; treat any infection and avoid grafting onto a clearly contaminated wound (delayed grafting after the bed is clean - 'delayed primary' grafting - is an option). Counsel about graft hyperpigmentation and reduced bulk.
Primary vs Secondary Contraction (Why FTSG Across Creases)
- Primary contraction. The immediate elastic recoil a graft shows the moment it is harvested, due to the elastin in the dermis - so it is greater in a full-thickness graft (more dermis/elastin) than a split graft. Practically, harvest an FTSG slightly larger than the defect to allow for it.
- Secondary contraction. The contraction of the grafted wound as it heals, driven by myofibroblasts in the recipient bed - this is greater with a split graft (little dermis to restrain it) and less with a full-thickness graft, because the more dermis a graft carries, the more it inhibits the myofibroblast-driven contraction. This is the functionally important one in the hand.
- Why it dictates hand choices. Across a flexion crease or on the palm/fingertip, the marked secondary contraction of an STSG would pull the digit into a flexion contracture - so a full-thickness (ideally glabrous) graft is preferred there; on the dorsum or large wounds, where contracture matters less, the more-reliable STSG is acceptable.

Primary contraction = immediate recoil at harvest, worse in FTSG (more elastin) - so cut an FTSG a bit large. Secondary contraction = wound-bed myofibroblast contraction during healing, worse in STSG and inhibited by dermis - which is exactly why FTSG (less secondary contraction) is used across flexion creases and on the palm/fingertips to avoid a disabling contracture.
Operative Technique: Harvest, Inset and the Donor Site
- STSG harvest. Taken at a set, calibrated thickness with a powered dermatome or a Watson/Humby knife (the guard sets the depth), usually from the thigh; the donor heals by re-epithelialisation from retained dermal adnexa under a semi-occlusive/alginate dressing.
- Meshing. The STSG can be passed through a mesher (ratios such as one-and-a-half to one) to expand its area and let exudate/haematoma escape - at the cost of a cobblestoned appearance and more contraction; small hand grafts are often fenestrated/pie-crusted by hand instead.
- FTSG harvest. Cut to a template of the defect, taken full-thickness, then defatted (all subcutaneous fat trimmed off the dermis so the bed can nourish it), and the donor site closed primarily - which limits the size available.
- Inset and securing. The graft is trimmed to fit, sutured/secured without tension, fenestrated to drain, and held in firm contact by a tie-over (bolster) dressing or negative-pressure dressing; the hand is splinted/immobilised and the graft typically first inspected at about day five.
STSG: calibrated dermatome/Humby harvest (often thigh), donor re-epithelialises; mesh (about 1.5:1) or fenestrate to let fluid out. FTSG: template, full-thickness, defat, close donor primarily. Always secure with a tie-over bolster or NPWT and immobilise, inspecting at about day five - the contact, haemostasis and immobilisation are what give take.
The Reconstructive Ladder - and the Rung Between Graft and Flap
The Ladder in Full
The ladder orders the options for closing a defect by increasing complexity, from the simplest upward:
- Option
- Healing by secondary intention
- When it is the right answer in the hand
- Small defects with no exposed bone - and at the fingertip this often gives the best sensation and padding of any option
- Option
- Primary closure
- When it is the right answer in the hand
- Clean wound with skin that comes together without tension
- Option
- Delayed primary closure
- When it is the right answer in the hand
- Contaminated wound closed once it is clean, typically after a few days
- Option
- Split-thickness skin graft
- When it is the right answer in the hand
- Large or dorsal defects, granulating beds, temporary cover
- Option
- Full-thickness skin graft
- When it is the right answer in the hand
- Palm, fingertips, across flexion creases - wherever secondary contraction would disable
- Option
- Local flap
- When it is the right answer in the hand
- Exposed bone, tendon or joint where a graft cannot take - V-Y, Moberg, cross-finger, homodigital island
- Option
- Regional / pedicled flap
- When it is the right answer in the hand
- Larger defects beyond the reach of a local flap - radial forearm, posterior interosseous, groin
- Option
- Free tissue transfer
- When it is the right answer in the hand
- Large composite defects, or where local and regional tissue is unavailable or inadequate
The ladder is a hierarchy of complexity, not an obligation to climb one rung at a time. The reconstructive elevator is the more accurate modern framing: choose the rung that gives the best functional result, which in the hand often means going straight to a flap rather than accepting a graft that will contract across a crease. The digit-specific version of these rungs is set out in Fingertip Injuries, and closure technique more broadly in Wound Closure Techniques.
Exposed Bone or Tendon Does Not Automatically Mean a Flap
The rule that a graft will not take on bare bone, tendon or cartilage is sound, but there is a step between that finding and committing to a flap. A dermal substitute (an acellular collagen-based matrix) can be laid over the exposed structure to generate a vascularised neodermis, which is then grafted at a second sitting - converting an ungraftable bed into a graftable one without committing the patient to flap surgery.
In a systematic review of 190 studies of dermal substitutes, wound depth was the primary indication, and the recommendation was explicitly staged: a one-stage approach for deep dermal to full-thickness wounds larger than 10 cm, and a two-stage approach for wounds of that depth with limited donor sites or exposed bone or tendon. The relevant contraindications were wound infection and allergy to the matrix components, with limited data in diabetes, chronic vascular disease and immunocompromise, and no definitive age or anatomical-site thresholds. In the hand specifically, a dermal matrix has been used to cover bone exposed after tumour excision, with a split-thickness graft applied subsequently and complete healing.
A dermal substitute widens what can be grafted, but it does not rescue an infected wound - infection is a contraindication to the matrix just as it is to the graft. So the sequence of questions is: is the bed infected? If so, neither is an option until the infection is cleared. If the bed is clean but a structure is bare, a dermal substitute followed by delayed grafting is a legitimate alternative to a flap. If the bed is clean and vascularised, graft it directly - and let the site, not the availability of the technique, decide between split and full thickness.
Know the ladder as eight rungs from secondary intention to free tissue transfer, and know that it is really an elevator - in the hand the functionally correct answer is often a flap rather than the next rung up. And do not stop at "bare tendon means flap": a dermal substitute with delayed grafting is the recognised two-stage route for a clean bed with exposed bone or tendon, with infection the contraindication to both.
Evidence & Key Studies
Comparison of reading man flap and skin grafting for donor defects in homodigital reverse-flow flaps
- For donor-site defects of homodigital fingertip flaps, full-thickness skin grafting and the reading-man flap gave SIMILAR flap survival, complications, range of motion, two-point discrimination and aesthetic outcomes.
- No patients developed cold intolerance or hypersensitivity, supporting FTSG as a reliable option for finger donor-site resurfacing.
- Demonstrates the role of skin grafting within the reconstructive options for fingertip/donor defects.
Delayed primary full-thickness skin graft for a distal interphalangeal fingertip injury: a case report
- A fingertip injury with bone exposed distal to the DIP but with INTACT PERIOSTEUM (and no tendon over the bone) was reconstructed with a full-thickness skin graft after wound preparation.
- The literature review emphasises that FTSGs are unreliable over poorly vascularised beds and only work without serious blood-supply issues - i.e. an adequate vascular bed is essential.
- Outcomes were good function and aesthetics with normal touch/vibration; disadvantages were limited soft-tissue volume and graft hyperpigmentation.
Indications for the use of dermal substitutes in acute burns and reconstructive surgery: a systematic review
- Across 190 studies, wound depth emerged as the primary indication for using a dermal substitute alongside split-thickness grafting.
- A one-stage approach is recommended for deep dermal to full-thickness wounds larger than 10 cm; a two-stage approach is advised for wounds of that depth with limited donor sites or with exposed bone or tendon.
- Contraindications are wound infection and allergy to the matrix components; data remain limited in diabetes, chronic vascular disease and immunocompromise, and no definitive age or anatomical-site thresholds were identified.
The reliability of FTSG for fingertip/donor-site resurfacing comes from the cited Qin comparison, and the requirement for an adequately vascularised bed (FTSG over periosteum-covered bone, not bare bone) and the FTSG advantages/disadvantages from the cited Soraya case report. The indications for a dermal substitute - wound depth as the primary determinant, the one-stage versus two-stage split, the explicit inclusion of exposed bone or tendon in the two-stage indication, and infection as a contraindication - come from the cited van den Bosch systematic review, which is drawn largely from burn and reconstructive practice rather than from hand-specific trials; the hand application over bone exposed after tumour excision is a single reported case.
Graft thickness is described here as calibrated on the dermatome rather than in millimetres, because no verified thickness ranges were identified; the practical trade-off is stated instead - a thicker graft carries more dermis and so contracts less but takes less reliably and leaves a donor site that heals more slowly. The graft-take phases, the STSG/FTSG properties, the eight rungs of the reconstructive ladder and the reconstructive elevator concept are standard, well-established plastic and hand-surgery teaching. (See also Fingertip Injuries and Wound Closure Techniques; no dedicated flap or soft-tissue-coverage topic exists in the library, which is why the flap rungs are named here rather than deferred.)
Clinical Decision Scenarios
Practise clinical reasoning and management decisions out loud
“What is the difference between a split- and full-thickness skin graft, how does a graft take, and how would you choose between them in the hand?”
“A graft you placed on the hand has failed. What are the likely reasons and how do you prevent them?”
Mnemonics & Memory Aids
III (take)
Hook:Graft take = the three I's: Imbibition, Inosculation, Ingrowth.
HIS (failure)
Hook:Grafts fail for HIS reasons: Haematoma, Infection, Shear.
Types
- STSG: epidermis + part dermis - reliable, large, donor re-epithelialises, but MORE contraction
- FTSG: epidermis + full dermis - less contraction, durable, sensate; donor closed primarily
- Glabrous FTSG (hypothenar/instep) for palmar/fingertip match
Take physiology
- Imbibition (0-48h) -> inosculation (48-72h) -> revascularisation (~day 4-7)
- Needs a WELL-VASCULARISED bed (no own blood supply)
- Won't take: bare bone/tendon/cartilage, infected/irradiated -> FLAP
Hand choices
- Palm/fingertips/flexion creases: FTSG (avoid STSG contracture)
- Dorsum/large/temporary: STSG (+/- mesh)
- FTSG donors: wrist/hypothenar, antecubital, medial arm, groin
Failure & prevention
- Causes: haematoma/seroma, infection, shear, poor bed
- Prevent: haemostasis, mesh/fenestrate, bolster/tie-over or NPWT, immobilise
- Inspect at about day 5
Reconstructive ladder (8 rungs)
- Secondary intention -> primary closure -> delayed primary closure
- STSG -> FTSG
- Local flap -> regional/pedicled flap -> free tissue transfer
- It is an ELEVATOR, not a staircase - choose the rung that gives best function
- Clean bed with bare bone/tendon: dermal substitute, then graft at a second sitting
- Infection contraindicates the dermal matrix and the graft alike