Fractures 101


2024 - Donna Magid, M.D., M.Ed., FAUR

PLEASE NOTE 3 points where I do not agree with many conventional texts

a) I tend to discourage using "Complete/incomplete" as descriptor-- at microscopic level even nondisplaced or 'minmal' fracture likely truly disrupts osseous integrity, and will also show up as more extensive at 10-14 days.  

b) Eponyms (previously, Political Incorrect “DeadWhiteGuys"-- not a fan of useless nondescriptive nomenclature.AND most wildly over- or mis-used (Colles, Jones, etc—all drift into mis-use for anything in the general neighborhood of the original definition). DESCRIBE it. As a newcomer, would you be more inclined to understand “Lisfranc injury”, or “tarsal-metatarsal fracture subluxation”?

c) Salter I drawings—Many illustrations show the epiphysis displaced from the metaphysis. if the epiphysis were that displaced off the physis the (radiographically invisible) physis would be totally destroyed; = Salter V.  My concept of Salter I is a force travelling through but not visibly/grossly disturbing metaphyseal/epiphyseal relationship. Since vascular supply and physeal chondrocyte architecture are NOT markedly disrupted/destroyed, Salter I fractures are difficult to see/dx. and heal well.  Most clinicians consider it a de facto Salter I if a child is point-tender over a physis, and radiograph negative.


FRACTURES occur when loaded bone material fails. The most subtle or challenging fractures may be absolutely radiographically normal at first imaging, but if mechanism of injury, clinical factors, or your instinct says ‘Worry”, advisable to treat presumptively 'if clinically indicated' (decision up to clinician); Radiologist should suggest “…follow-up area of interest 10-14 days if clinically indicated”. Remember that an image remaining negative before that may be false negative—subtle fractures may need the 10-14 days for the osteoblasts (new matrix, callus), osteoclasts (debris removal), mineralization or new matrix and/or periosteal reaction, and hyperemia, to induce sufficient change for a fracture line, periosteal reaction, early new matrix, or trabecular distortion to become diagnosable. Certain fractures—such as scaphoid—may take even longer to become visible due to tenuous blood supply. Other fractures, which may be life threatening - C spine, proximal femur—may need other imaging modalities more urgently. Appropriate phrasing needs wiggle room--'No definite acute abnormality, if clinically indicated...";   NOT "No evidence of fx" and NOT 'Normal". The patient had enough pain to present, that is not 'normal' and it is evidence of something going on.

WOLFF’S LAW: "Bone is laid down according to the forces acting on it; skeletal architectural transformation reflects its environment". Bone therefore ‘talks’ to a certain extent, helping distinguish acute from chronic or healing, or acquired in childhood from acquired in adulthood. Gradual resumption of use (partial weight-bearing) of an immobilized injury is intended, besides rehabbing muscle and tissues and vascular supply, to start initiating more normal forces through healing bone to enhance its healing and direct remodeling. (Julius Wolff is a D-W-G German anatomist, 1836-1902. So he lived, presumably, to see radiographs confirm his tissue observations).

Contributing factors (which also determine type of fracture expected):

  • Material properties ie cortical vs cancellous, normal vs osteopenic; peds vs. adult
  • Geometry eg stress risers, contours, bowing which direct force distribution
  • Loading conditions, rate of load eg fell 30 feet vs. torsion in stance vs. fatigue/chronic

CHILDREN are different from adults!

  • Materially: bone tissue has different histology, elasticity, and other mechanical properties
  • Structurally: size, shape, density, periosteum characteristics, blood supply,, and mineral mass differ
  • Growth and remodeling: open physes, rapid bone turnover, greater capacity to remodel


ONE VIEW IS NO VIEW! Need at least 2 views at 90 degrees to each other, sometimes more (eg, standard ankle, hand, shoulder, knee, exams all include +3 views) or you will err. Any 2D representation of a 3D structure can only show 2 of the 3 (X,Y,Z) axes; the 2d view provides the missing data and determines what superimposes vs what truly lies within

Figure 1. Fracture Types


Fracture lines described relative to longitudinal (spinal) axis; may imply mechanism of injury:

  • Transverse (A)
  • Oblique (B)
  • Spiral (often implies torsion or twisting force) (C)
  • Intraarticular(J)                                                                                                                                              


  • ‘Greenstick’ (bowing or plastic deformity) (D,H)
  • Incomplete (rarely used- vague; and don't describe other fxs. as 'complete')
  • Torus or Buckle (often axial load) (E)

Salter I-II-III-IV-V if physis involved (See Fig. 3 KQ below)


  • Closed (simple): skin intact over injury
  • Compound (open): skin compartment violated ie internal/outside environment communication
  • Comminuted: 3 or more pieces, no matter how small or how few (F)
  • Butterfly’ (F): significant wedge- or triangular-shaped fragment of diaphyseal bone in comminuted fracture, usually measured/described           
  • *Avulsion (I): fragment pulled off as result of muscle contraction or ligamentous/tendinous resistance
  • Pathologic (G): fracture through abnormal bone. Classically transverse (‘banana fracture”)
  • Stress: fracture occurring via fatigue (abnormal repetitive forces on normal bone, eg military recruit or new physical activity) or insufficiency (normal, often repetitive, forces on abnormal bone such as osteopenic new exerciser)
  • "Chip"-- NO

POSITION of fragments: assume starting point of normal, or ANATOMIC. CHANGES reference where DISTAL twent relative to PROXIMAL. Be very specific and clear. Redundancy (“with 25 degrees medial angulation of the DISTAL FRAGMENT relative to the PROXIMAL FRAGMENT”) is better than miscommunication (“…25 degrees medial angulation” may suffice in communicating with Orthopaedics, but many Emergency Medicine or other house officers will be unfamiliar with the distal-relative-to-proximal Orthopaedic convention, hence the redundancy of "...of the distal portion")). Once fractured, bone may be ‘near” or ‘apparently anatomic’ if truly NOT displaced—but is unlikely to be 100% anatomic; use word sparingly.I virtually always say ‘Near anatomic” or “Grossly anatomic”. See ‘Alignment” below.

Figure 2. Fractures


ANGULATION (*U): again, reference is DISTAL to PROXIMAL; referring to angling of the longitudinal (spinal, diaphyseal) axis of the distal fragment relative to the proximal fragment’s axis. For clarity be redundant. “(Medial/lateral, anterior/posterior, varus/valgus) of the DISTAL fragment relative to the PROXIMAL fragment” would all describe changes in the longitudinal axis of the distal fragment relative to the proximal; measurements always welcome (some ‘operate/don’t operate’ decisions may be triaged by degrees of angulation; for example, the 5th MC or ‘Boxer’s” fx.)

DISPLACEMENT (V) refers to motion of distal fragment perpendicular to the long (diaphyseal, spinal) axis; i.e. distal fragment moved   (medial, lateral, anterior, posterior) without also angulating long axis

DISTRACTION (T): ‘distracted xxx mm longitudinally” means increased gap between fragments (osteoblasts don’t like distance challenges)

IMPACTED, OVERRIDING, TELESCOPED (S) implies compaction or shortening (usually of a long bone) along that same longitudinal axis

ROTATION (*W) is motion around the long axis, i.e. spinning in the axial plane like the Equator over 24 hrs (imagine a dowel passed through a paper towel tube). This usually needs reduction for best functional outcome, child or adult. Remember to get BOTH ENDS of a long bone on ONE view eventually and every so often during healing (mostly a problem in tall adults, femur or tibia-fibula)

SUBLUXATION (*X) and DISLOCATION (*Y) are descriptors for ARTICULAR SURFACE relationships, not fracture fragments. SUBLUXATION is partial loss of articular contact of apposed bones; DISLOCATION is total loss of articular surfaces contact. “One View Is No View”; a seemingly normal articulation may be totally dislocated on a second view. Describe and measure directions of distal articular surface’s change relative to the proximal articular surface. Surfaces may change relationships as non-osseous restraints (tendon, ligament, soft tissues, etc.) fail.

Figure 3. Subluxation and Dislocation


SUBLUXATION (*X) implies early or partial loss of contact of one articular surface from another; partial contact is maintained. Restraining structures (ligaments, tendons, capsule) typically stretched but at least partially intact.

DISLOCATION (*Y) is total loss of contact of two apposed, normally articulating surfaces. Use the worst-case scenario view (i.e., may look subluxed on AP but is clkearly dislocated on lateral). One View Is No View!!

ALIGNMENT implies reference to the anatomic starting baseline or (implied and expected) normal. Be precise and descriptive (degrees, directions, mm or cm) and objective (‘improved’, ‘atypical’, ‘near-anatomic’, ‘decreasing displacement/angulation/rotation than initial image” and even ‘now apparently near-anatomic’ would be ok but “poor’, ‘good’, ‘failed’, ‘unsuccessful’, would NOT be. One can instead say ‘post-reduction images show no significant change in initial (angulation, displacement, whatever)” or “Views 10:46 hrs are unchanged from 09”17 hrs”, or ‘There continues to be 1.6 cm medial displacement and 35 degrees varus angulation of the distal fragment”. LAWYERS love subjective opinions.

PHYSEAL INJURIES: Salter (or Salter-Harris) Classification (*K-Q) (one of the few DeadWhiteGuy, DWG, names still acceptable to most); physeal injuries assigned IV rating implying severity and prognosis. The worst consequence of a Salter injury is either partial or complete physeal destruction, leading to complete longitudinal growth arrest, or where physis is partially compromised and heals with a bar (bony point-fusion across physis), creating a bonsai-tree-like imbalance of forces as bone attempts to grow longitudinally. Consequences depend on child's age (chronological and bone age) at time of physeal injury (infant triples length, toddler doubles length, injured teenager's bone may already be at 95% potential adult length), severity of physeal injury (IV), which bone (upper extremity may be less serious than lower extremity; distal phalanx less noticeable than proximal phalanx; etc.),; and degree of angular distortion if physeal bar is inducing curved growth.

The physeal component, involving radiolucent cartilage, is often overlooked. If a fracture line approximates the physis, the fracturing force can be assumed to have continued to travel through the physis (path of least resistance). The less disturbed or displaced the physeal tissue is, the less likely growth will be disrupted or distorted. Significant physeal trauma may result in growth arrest, either in one spot (bony bar), or across the entire physis. The former can lead to angular deformity as the remaining physis grows; the latter, total loss of longitudinal growth (length).

Figure 4. Salter-Harris Classification of Physeal Injuries

  • Salter I: no apparent radiographic fx; possible focal soft tissue swelling; point-tender. One may catch a bit of focal soft tissue swelling or bone reaction at some point first few days but we do not know how many Salter I fxs. there really are:  a) not all come in to be seen/radiographed; b) those that do may feel fine rapidly and skip follow up visit, and c) many are completely radiographically invisible. Always safe to say, when confronted with apparently normal child’s bone image: “No definite abnormality identified. Physes are open. If clinically indicated f/u 10-14 days following injury may be helpful to r/o subtle Salter injury”. PHYSEAL COMPONENT ASSUMED, THO' OFTEN INVISIBLE, ALL SALTER FXS.



  • Salter II: fx. Through physis and metaphyseal side of physis (‘corner”). Most common; ~65%=II
  • Salter III: through physis and epiphysis ie intraarticular (hence potentially more severe than a II)
  • Salter IV: through physis, metaphysic and epiphysis. (New math: “Salter II plus III =s IV”)
  • Salter V: complete destruction physis either through total disruption/displacement of metaphysic from epiphysis, or from impaction/axial load/crush essentially destroying physis. Will lead to growth arrest. Occasionally it is unclear initially whether an injury is a Salter V or not, particularly when there is displacement along the physis and fracture fragments: “Salter II, cannot r/o Salter V, fracture distal femur.” Final classification may not be possible in some cases until retrospectively (physis undamaged and resumed normal growth—or not).

BONE LENGTHENING: harnessing normal healing and Wolff’s Law to lengthen bones and restore symmetry of the extremities, most commonly the LE (UE or arm asymmetries are less noticeable cosmetically and functionally). Reasons for asymmetric bone length: most common is trauma, such as serious Salter injury in childhood with growth arrest or disruption,  bone loss, infection, or non-union.  Developmental or congenital deformities (e.g., proximal focal femoral deficiency, amniotic bands, fibular hemimelia, craniofacial deformities), childhood events such as septic arthritis or Perthes disease, neurofibromatosis, and idiopathic leg length discrepancies may all benefit from bone lengthening (distraction osteogenesis). Gains may approach 6 inches. Calculations based on child’s age, gender, bone age, health, predicted future growth of bone of interest, and biologic parental heights help determine the final (post-pubertal) length desired, to achieve symmetry and render the lengthening goal more precise.

There are several variations, but essentially an iatrogenic fracture (osteotomy) is introduced in a segment of bone with optimal healing potential (ie, not through the abnormal segment of bone, nor through poorly vascularized regions; but one with excellent vascular supply and anticipated healing potential such as the proximal tibia or distal femur). Longitudinal fixation of a long pin parallel to the diaphysis, with transverse pins through proximal and distal bone, allow very slow distraction of the osteotomy using an external magnet . As the bone begins its normal healing response to osteotomy or fracture—bleeding, clot contraction, activity of osteoblasts, matrix deposition, neovascularity—the dynamic lengthening frame is extended very slowly, ~ mm/day increments, ‘tricking’ the bone cells into healing a slowly-expanding distraction gap. The rate is balanced to avoid premature bone consolidation (expanded too slowly) or over-distraction of the osteoblasts, which leads to fibrous, rather than bony, union and may injure stretching nerves and muscles. It may take up to several months (about 4-5 weeks per cm, roughly) to establish desired length of matrix and new bone (and of associated nerves, tendons, ligaments, muscles); and to rehab the extremity and coalescing new bone for full weight-bearing. Older children make excellent candidates—with the healing capacities of children but with sufficient motivation, comprehension, compliance and cooperation to tolerate the complex care schedules, wound and daily hygiene, physical therapy, discomfort, extended care and severe mobility limitations involved. Social/family/psychological factors are also weighed in assessing candidates. In the US bone-lengthening is NOT performed as a strictly cosmetic procedure in those desiring augmented height.


    Fractures need to be immobilized to heal; motion at a fracture site causes fibrosis and failure to heal. Immobilizers may be external—cast, splint, brace, etc; or internal—which is instrumentation, not ‘hardware’. This includes rods, pins, nails, screws, plate and screws, etc. Don’t get too specific—“Since last exam instrumentation has appeared…”..plate and screws appearing…”intramedullary fixation…”ACL surgery has appeared…” since the more specific one attempts to be the more likely to make an error (even the Op Note may mis-name something. And unless you show the good taste to elect MSK Fellowship/career, you do not need to know Cottrell-Duboset from Luque rods (and as patents expires, many look-alikes emerge), nor lag from cortical from cancellous screws)

   Once a fracture is instrumented, subsequent radiographs should show no significant evidence of loosening (<1 MM metal-bone interface allowed), nor obvious change in position/angulation). Compare old and new images carefully for positional or exposure differences before concluding the instrumentation has moved.


One NEED NOT BE CERTAIN. “Possible…”Cannot r/o…”Suboptimally seen…”Positionally different from previous exam…”Possible trabecular irregularity vs superimposed contour…”. DESCRIBE first, titrate your level of certainty.


EXAM: XR (side) (part) (# views)

INDICATION: Fracture/Post-reduction/Open Reduction Internal Fixation/instrumentation removal/


FINDINGS: If in cast or splint: ‘Cast material markedly limits detail’

“Again seen is (describe precisely where in bone, comminution, displacement, angulation, impaction, surgical fixation, associated foreign matter, etc), with radiographic progression of healing/reactive change’…or ‘with no definite change, since last exam.’

IMPRESSION: Unremarkable (side, anatomy) fx/post-operative follow up. Or, Increasing displacement/angulation….Follow up as clinically indicated; or (if inappropriate image obtained, eg humerus when fracture is humeral neck, or forearm when fracture is elbow or wrist) Follow-up coned and centered to any area of interest would give better detail , if clinically indicated.

Describe any special/unusual concerns—infection/inflammation, hardware migration/change, positional or exposure limitations, etc.


Long bone views (humerus, tibia, forearm, femur) are always ‘Limited exam of joint above/joint below’ as opening sentence.

Hand, foot when area of interest known to be fingers/toes: “Coned and centered exam of the ring finger/of the toes would give better detail at follow up if clinically indicated”.

If true--particularly of toes, when entire foot requested instead of coned views: “Limited exam of the area of interest, exam coned to the first toe would give better detail at follow up if clinically indicated”

Caveat: no generic template anticipates every possible finding/scenario; add details, descriptors, potential concerns, relevant history, as judged appropriate