Salter-Harris Classification Interpreter
Salter-Harris Classification Interpreter
Type a physeal fracture I to V from which structures the line crosses — with the published growth-disturbance rates by type, the zone of the physis each type passes through, and why the mnemonic teaches the anatomy and not the prognosis.
Salter-Harris type from the structures involved
Physis + metaphysis + epiphysis + crushA physeal fracture of the distal radius with a triangular metaphyseal fragment and no extension into the epiphysis; no crush and no established arrest
The five types, and the zone each one crosses
- the germinal layer, not the type number
- the OTA’s teaching set gives the zones: type I through the zone of hypertrophy, type II through endochondral ossification and hypertrophy, type III through the “hypertrophic, proliferative, and germinal zones”, type IV through all four. Types III and IV reach the germinal layer, which generates new cartilage, and that is where the growth-arrest risk comes from
- the mnemonic, in its two incompatible versions
- one source gives SALTER — Slipped, Above, Lower, Through Everything, Rammed — and another gives SALTR, with “Through/Transverse” for type IV. Both are in circulation and both are printed here. Either way the mnemonic encodes WHERE the line runs and says nothing about the germinal layer, which is the part that decides the outcome
- the risks are not monotonic
- in a meta-analysis of 564 distal femoral physeal fractures, growth disturbance occurred in 36% of type I, 58% of type II, 49% of type III and 64% of type IV. Type III is LOWER than type II. “Higher type, worse outcome” is folklore
- displacement beat type
- in the same analysis, “The odds of a displaced fracture having growth arrest was 4 times greater than that of a nondisplaced fracture” — a larger effect than any difference between the types
- site, before anything else
- type I carries 36% growth disturbance at the distal femur and 3 to 5% at the distal tibia, and an unselected cohort of 1,585 children with tibial or distal femoral physeal fractures found clinically significant growth disturbance in 5.0% overall. A type without a bone is not a risk
- type V is diagnosed backwards
- no line, minimal displacement, and “usually diagnosed retrospectively” once deformity appears. It was not in the original description, and no reliable arrest rate for it was found
- derivation
- Salter RB and Harris WR, “Injuries Involving the Epiphyseal Plate”, J Bone Joint Surg Am 1963;45(3):587–622. The paper was not reachable for this page; the type definitions and the zone anatomy here are quoted from the Orthopaedic Trauma Association’s own teaching set and from RCEM Learning
Worked example
A physeal fracture of the distal radius with a triangular metaphyseal fragment and no extension into the epiphysis; no crush and no established arrest
No crush and no established arrest, so the type V rule does not fire
The physis is involved, so the injury is inside the classification
A Thurston-Holland fragment is present and the epiphysis is not involved, so this is type II — the commonest type, 75% of physeal injuries in the distribution quoted here
Add epiphyseal extension to the same fracture and it becomes type IV, crossing all four zones of the physis including the germinal layer, with 64% growth disturbance in the 564-fracture distal femoral meta-analysis
Remove the metaphyseal fragment and keep the epiphyseal extension and it is type III — and in that same analysis type III carried 49%, LESS than type II's 58%. The ordering is not monotonic
Remove both and it is type I, the physis alone — 36% at the distal femur and 3 to 5% at the distal tibia, which is the whole argument for naming the bone before quoting a risk
Growth disturbance by type, with the cohort each figure came from
| Type | Distal femur, 564-fracture meta-analysis | Distal tibia, OTA teaching set | Zone crossed |
|---|---|---|---|
| I | 36% | 3 to 5% | zone of hypertrophy |
| II | 58% | 17 to 36% | endochondral ossification and hypertrophy |
| III | 49% | 13 to 50% | hypertrophic, proliferative and germinal |
| IV | 64% (the OTA set renders the same figure as 65%) | 13 to 50% | all four zones |
| V | not separately reported | not separately reported | crush of the whole physis |
| Any type, unselected | 5.0% (95% CI 3.8 to 6.6%) clinically significant growth disturbance in 1,585 children with tibial or distal femoral physeal fractures at one centre, 2008–2018 | same cohort | — |
The mnemonic, both versions, and what neither one tells you
| Type | SALTER version | SALTR version | What decides the prognosis |
|---|---|---|---|
| I | S = Slip (separated or straight across) | S = Slipped | germinal layer spared |
| II | A = Above the physis, or Away from the joint | A = Above | germinal layer spared |
| III | L = Lower, below the physis in the epiphysis | L = Lower | germinal layer involved, and intra-articular |
| IV | TE = Through Everything | T = Through / Transverse | all four zones including the germinal layer, and intra-articular |
| V | R = Rammed (crushed) | R = Rammed | epiphyseal blood supply, and usually found late |
Where the line runs, which zone it crosses, and why the type number is not a ranking
Salter and Harris typed injuries involving the epiphyseal plate by which structures the fracture line crosses. Type I runs through the physis alone. Type II takes a triangular corner of metaphysis with it — the Thurston-Holland fragment — and exits away from the joint. Type III crosses the physis and the epiphysis into the joint. Type IV crosses metaphysis, physis and epiphysis together. Type V is a crush of the plate, with no line to see.
The mnemonic every trainee learns teaches that geography accurately and the prognosis not at all. What decides whether a child grows normally is the zone of the physis the line passes through. The Orthopaedic Trauma Association’s own teaching set sets it out: type I goes through the zone of hypertrophy, type II through endochondral ossification and hypertrophy, type III through the hypertrophic, proliferative and germinal zones, type IV through all four. The germinal layer is where new cartilage is generated, and types III and IV are the ones that reach it. That is the reason RCEM Learning can give types I and II an excellent prognosis while describing growth arrest and angular deformity as common problems in III and IV.
Then the numbers complicate it. A meta-analysis of 564 distal femoral physeal fractures found growth disturbance in 36% of type I, 58% of type II, 49% of type III and 64% of type IV. Type III is below type II: the risks are not monotonic in the type number, so “higher type, worse outcome” is not supportable. In the same analysis displacement mattered more than type at all — a displaced fracture carried four times the odds of arrest. And the site dominates everything: type I is 36% at the distal femur and 3 to 5% at the distal tibia, while an unselected cohort of 1,585 children with tibial or distal femoral physeal fractures found clinically significant growth disturbance in 5.0%. A Salter-Harris type quoted without a bone is not a risk estimate.
Type V deserves its own sentence, because it is not found the way the others are. There is no line; there is a narrowed plate and, usually, a deformity that appears months later. It is diagnosed retrospectively, it was not in Salter and Harris’s original description, and no dependable arrest rate for it was found for this page — so none is given.
A classification is not a score. Garden IV is not one unit worse than Garden III, a Salter-Harris IV is not twice a II, and nothing on this page adds up — these are named categories, not an ordinal measurement, and a mean or a difference taken over them has no meaning. This page states which published category the entered findings fall in and what the literature reports for that category in the cohorts it was measured in. It renders no clinical or surgical decision: not whether to operate, not which implant, not whether to take or withhold an image, and not whether a limb can take weight. Those are the treating clinician’s, and the classification is one input among many. Agreement between observers is part of what a classification means. A category two experienced readers assign differently on the same film is not a measurement of the fracture, and the published kappa values for these systems run from almost nothing to almost perfect. Each page here prints them with the number of observers and the number of radiographs they came from.
Frequently asked questions
How are Salter-Harris fractures classified?
By which structures the fracture line crosses. Type I is through the physis only; type II through the physis and out through the metaphysis, leaving a Thurston-Holland fragment; type III through the physis and into the epiphysis; type IV through metaphysis, physis and epiphysis together; type V a crush of the physis with no fracture line.
Which Salter-Harris types risk growth arrest?
The ones that cross the germinal layer of the physis — types III and IV — carry the risk by mechanism, and type V by disruption of the epiphyseal blood supply. But the measured rates do not follow the type numbers. In a meta-analysis of 564 distal femoral physeal fractures the figures were 36% for type I, 58% for type II, 49% for type III and 64% for type IV, and displacement carried four times the odds of arrest whatever the type.
Why does the same type carry such different risks?
Because the site matters more than the type. Type I is reported at 36% growth disturbance at the distal femur and 3 to 5% at the distal tibia. The distal femur is the worst physis in the body and the series meta-analysed for it were selected ones. An unselected cohort of 1,585 children with tibial or distal femoral physeal fractures found clinically significant growth disturbance in 5.0%, 95% CI 3.8 to 6.6%.
What does the SALTR mnemonic actually teach?
Where the fracture line runs, and nothing about the prognosis. It also comes in two incompatible versions — SALTER, with “Through Everything” for type IV, and SALTR with “Through/Transverse” — both of which are in current use. Neither mentions the germinal layer of the physis, which is the structure that decides whether growth continues.
Why can type V not be diagnosed on the first radiograph?
Because there is nothing to see except a narrowed growth plate, and displacement is minimal or absent. RCEM Learning describes type V as usually diagnosed retrospectively, when limb deformity becomes apparent; the OTA teaching set calls it a late diagnosis after physeal arrest and deformity have already developed. It was also not part of Salter and Harris’s original description.
Related calculators
References
- Salter RB, Harris WR. Injuries Involving the Epiphyseal Plate. J Bone Joint Surg Am. 1963;45(3):587–622. Citation read in the JBJS Classics commentary (orthobuzz.jbjs.org, 30 March 2015), which describes the paper as setting out “the famous five types of physeal injury”, and in Wikipedia’s reference list; the paper itself was not reachable. Every criterion below is quoted from a named reproduction.
- Orthopaedic Trauma Association. Growth Plate Injuries, teaching slide set P14, ota.org/sites/files/2018-06. Type I “Physis only injured”, “Fracture through zone of hypertrophy”; type II “Physis +metaphysis” with a “Thurston-Holland metaphyseal fragment”; type III “Physis+Epiphysis Injured” involving the “hypertrophic, proliferative, and germinal zones”; type IV “Epiphysis, physis, metaphysis injured”, “All four zones of physis involved”; type V “Crush injury to entire physis”, “Very difficult initial diagnosis as minimal displacement”, “Late diagnosis after complication of physeal arrest and deformity has occured”. Distal tibia growth-arrest risks (attributed there to Leary et al 2009, cohort size not stated on the slide): I 3 to 5%, II 17 to 36%, III and IV 13 to 50%, triplane 7 to 21%, Tillaux “low risk”.
- Growth disturbance after distal femoral growth plate fractures in children: a meta-analysis. J Orthop Trauma. October 2009. Named as Basener et al in the OTA slide set above; abstract read at qxmd.com/r/19897989 and the full text was not reachable, so the author list is not reproduced here. “Of the 564 fractures, 291 (52%) had a growth disturbance.” “Growth disturbance occurred in 36% of SH 1 fractures, 58% in SH 2, 49% in SH 3, and 64% in SH 4 fractures.” “Twenty-two percent (112/506)” developed a leg-length discrepancy of at least 1.5 cm. “The odds of a displaced fracture having growth arrest was 4 times greater than that of a nondisplaced fracture.” The OTA slide set renders the type IV figure as 65% where the abstract says 64% — both are printed on this page.
- Epidemiology of physeal fractures and clinically significant growth disturbances affecting the distal tibia, proximal tibia and distal femur: a retrospective cohort study. DOI 10.5435/JAAOS-D-22-00303; record read at research.luriechildrens.org. 1,585 patients aged 0.5 to 18.9 years treated at one level I paediatric trauma centre 2008–2018. “The incidence of CSGD was 5.0% (95% confidence interval, 3.8% to 6.6%).” Distal femoral and proximal tibial fractures were associated with increased risk. No breakdown by Salter-Harris type is given. This is the unselected denominator against which the distal-femoral meta-analysis’s 52% has to be read.
- Effectiveness of Surgical versus Conservative Treatment for Distal Femoral Growth Plate Fractures: A Systematic Review. Open Orthop J. 2019;13:117. 466 patients from 15 retrospective case series, whose own Salter-Harris counts (70 I, 276 II, 58 III, 45 IV, 3 V, 8 VI) sum to 460 — a discrepancy the paper does not explain. “it was not possible to assess the primary outcomes” by type. Reports Basener’s “incidence of 52% in growth disturbance, with 22% of the growth disturbance greater than 1.5cm”; Arkader et al 2007’s cohort of 73 patients (18 SH I, 43 SH II, 4 SH III, 7 SH IV, 1 SH V); Plánka et al 2008’s 6 of 31 patients with leg-length discrepancy or angular deformity (19.3%); and Caterini et al 1991’s “complication rate of 71.4%” over 12 to 36 years.
- Tzavellas AN, Kenanidis E, Potoupnis M, Pellios S, Tsiridis E, Sayegh F. Interobserver and intraobserver reliability of Salter-Harris classification of physeal injuries. Hippokratia. 2016;20(3):222–226. “Twenty-eight independent raters” and “50 randomly selected radiographs of physeal injuries”. “Overall kappa was 0.45”; “The kappa value for specialists was 0.53” and “only 0.39 for the residents”; “Kappa value for junior residents was 0.3 and for senior residents 0.44”. “The mean kappa for intraobserver reliability was 0.52”, specialists 0.55 against residents 0.49 (p = 0.346).
- Royal College of Emergency Medicine Learning. Salter-Harris classification summary, in the module Soft tissue and bony injuries in children. “type I and II have an excellent prognosis, although complete or partial growth arrest may occur in displaced fractures”, attributed to the germinal layer usually remaining intact; “Types III and IV have a worse prognosis, as growth arrest and angular deformity are common problems”; “Type V fractures have poor prognosis”, were not in Salter and Harris’s original description, and are usually diagnosed retrospectively “when limb deformity becomes apparent”.
- Wikipedia, Salter–Harris fracture, and Life in the Fast Lane, Salter-Harris Classification (litfl.com). The two give the mnemonic differently — “SALTER” with “TE = Through Everything” for type IV against “SALTR” with “Through/Transverse” — and both are printed on this page rather than one being picked. Wikipedia also gives the type distribution (I 6%, II 75%, III 8%, IV 10%, V 1%) and that physeal injuries are “15% of childhood long bone fractures”, with no cohort named for either, which is how they are presented here.
Not medical advice. For healthcare professionals and education. Reference intervals vary by laboratory and assay — always use your own laboratory's. Never base a dose or a treatment decision on this page alone. Full disclaimer at calcengines.com/disclaimer/