Pediatric Fractures (Salter-Harris)
- Kids have a soft cartilage growth plate (the physis) near the ends of long bones — it's the weakest link, so it fails before the ligaments do.
- The Salter-Harris (SH) system grades growth-plate fractures I–V by how the fracture line travels through the physis, the metaphysis, and the epiphysis.
- Mnemonic SALTR: I = Slipped, II = Above, III = beLow, IV = Through, V = cRushed.
- The grade matters because higher numbers (especially III, IV, and the sneaky V) more often disturb growth — meaning crooked or stunted bones down the road.
- A normal growth plate is a lucent line that can look like a fracture; comparison with the other side and knowing the normal anatomy keeps you out of trouble.
A child's bone is not just a small adult bone — it's a work in progress. Near each end sits a band of growth cartilage, and because cartilage is softer than the bone and the surrounding ligaments, it becomes the part that snaps when a kid lands wrong off the monkey bars. The whole Salter-Harris system is just a tidy way of saying which way the crack ran through that soft band.
Meet the growth plate
The physis is the growth plate: a disc of cartilage sitting between the epiphysis (the bone cap at the very end, near the joint) and the metaphysis (the flared shaft just below it). Think of a long bone as a pencil with an eraser stuck on the end. The eraser is the epiphysis, the metal band crimping it on is the physis, and the wooden pencil body is the metaphysis. New bone gets laid down at that band, which is how kids get taller — and it's also exactly where things break.
On an X-ray the physis shows up as a clean lucent (dark) line because cartilage doesn't stop X-rays the way bone does. That's a beautiful normal finding, and also the first trap: it can be mistaken for a fracture by anyone who forgot it's supposed to be there.
The five types (just follow the crack)
Here is the entire classification, which I promise is less scary than the Roman numerals make it look. Picture the physis as a horizontal line, the epiphysis above it, the metaphysis below.
| Type | Where the fracture goes | The SALTR word |
|---|---|---|
| I | Straight through the physis only; epiphysis "slips" off the shaft | Slipped |
| II | Through the physis, then up into the metaphysis (a metaphyseal corner breaks off) | Above |
| III | Through the physis, then down into the epiphysis (into the joint) | beLow |
| IV | A single line crossing metaphysis, physis, and epiphysis | Through |
| V | The physis is crushed/compressed — no displaced fragment to see | cRushed |
Type II is the one you'll meet most often. Type IV crosses every layer like a kebab skewer through all the ingredients, which is why it tends to need surgery to line everything back up. And type V is the villain — it's a compression injury, so there's often nothing dramatic on the first film. You find out it happened later, when the bone grows crooked.
The "above" and "below" in the mnemonic are from the perspective of the physis, with the metaphysis sitting above and the epiphysis below — which feels upside-down until you accept that mnemonics were not designed by anatomists.
Why the number actually matters
This isn't classification for its own sake. The grade is a rough guess at the risk to future growth. Types I and II usually do well because the fracture spares the layer of the physis responsible for growth. Types III and IV cross the joint surface and the growth plate, so they carry a higher chance of growth arrest and joint problems — and they more often need precise surgical alignment. Type V quietly damages the growth engine itself, which is why its reputation is worse than its X-ray.
When a growth plate gets injured and stops working in one spot, the bone can keep growing everywhere else and end up angled, or stop growing and end up short. That's the long-term stakes hiding behind a single Roman numeral.
When you describe one of these, don't just say "Salter-Harris II." Say which bone, which physis, and whether the fragments are displaced or angulated — that's what actually drives whether the child goes to the cast room or the operating room. The same plain fracture description habits apply here.
The traps that catch everyone
A Salter-Harris I can have a completely normal-looking X-ray if nothing is displaced — the only clue may be soft-tissue swelling or point tenderness over the physis. Cartilage doesn't show, so a non-displaced slip can hide in plain sight. Believe the exam, and get comparison views when in doubt.
Two more things keep me honest. First, the normal physis is lucent and irregular at certain ages, so the best friend you have is a comparison view of the other side — symmetry is reassuring, asymmetry is suspicious. Second, multiple fractures of differing ages, or fractures that don't fit the story, should make you think about non-accidental trauma — a pattern you never want to miss.
If you remember nothing else: in a child, the soft growth plate breaks first, Salter-Harris just tells you which way the crack wandered, and the bigger the number the more you should worry about the bone's future shape.