Trauma fixation hardware is not designed to last forever. A plate, screw or intramedullary nail carries load only until the bone it spans can carry that load itself, and every loading cycle in the interval consumes a fraction of the implant's finite fatigue life. So when hardware breaks, the question that matters is rarely whether the metal was sound. It is whether the bone healed inside the window the construct could survive. The fracture surface, read properly, distinguishes a device that failed early from one that outlasted its assignment.
Fixation hardware is a temporary load path
A fixation construct is a load-sharing arrangement with an expiry date built into it. Immediately after surgery the implant carries most of the load across the fracture. As callus forms and mineralises, the bone takes that load back and stress on the implant falls away. This transfer is the whole design premise: the plate is a bridge meant to become redundant.
That framing changes the forensic question. Hardware that fractures late in a bone that never united has done roughly what any comparable construct would. Hardware that fractures early, under loading it should comfortably have carried, is a different matter.
What a fatigue fracture looks like under magnification
Fatigue failure leaves a legible record. Under scanning electron microscopy the surface shows an initiation site, a progressive zone carrying beach marks and striations, and a final overload region where the remaining cross-section gave way in one step. The relative size of those zones is informative: a large progressive zone with a small overload region indicates modest stress over many cycles, while the reverse points to higher loads over a shorter life.
A single-event overload fracture looks nothing like this. It shows ductile dimpling or shear lips across the whole section with no progressive zone, implying a discrete traumatic event rather than accumulated walking.
Where the crack starts, and why that matters
Initiation site location is the most useful single observation. Plates fracture overwhelmingly at an empty screw hole over the fracture gap, because the hole removes material exactly where bending stress peaks. Nails tend to fail at a locking hole for the same reason. These are designed-in geometric features, not defects.
An initiation site that does not correspond to a hole or a section change deserves closer attention. A crack starting at a machining mark, a handling scratch, an inclusion or a surface-marking artefact points away from construct geometry, and materials examination becomes the productive line.
Delayed union is the usual explanation for a broken plate
If bone does not bridge, load never transfers, and the implant keeps accumulating cycles at high stress indefinitely. No fixation device has unlimited fatigue life at physiologic bending loads, so a persistent nonunion eventually breaks whatever spans it. In that sequence the hardware fracture is a consequence of the healing failure rather than its cause.
This is why the clinical record often carries more weight than the metallurgy. Fracture pattern, infection, weight-bearing instructions and adherence, and serial radiographs showing callus that never progressed all bear on how long the construct worked alone.
What the standards actually test
ASTM F382 covers static and fatigue testing of metallic bone plates. ASTM F543 addresses metallic bone screws, ASTM F1264 covers intramedullary fixation devices, and ASTM F1717 specifies a corpectomy-model test for spinal implant constructs. ISO 5835 and ISO 5838 give dimensional requirements for screws and external fixation components, and the common alloys are governed by ASTM F138 and F139 for wrought stainless steel and ASTM F136 for titanium alloy.
These are bench standards run on defined specimens in defined fixtures. They establish that a design meets a stated performance level under controlled loading. They do not predict service life in a particular patient with a particular fracture pattern and gait, and treating a bench result as a service-life guarantee is a common error.
Separating a manufacturing question from the rest
Where a manufacturing contribution is genuinely in issue, the examination is bounded: alloy chemistry against the applicable composition limits, microstructure, hardness, inclusion content and surface finish. Cold-worked stainless steel and titanium alloy differ in fatigue behaviour and notch sensitivity, so the comparison has to be against the right specification.
Conformance findings are strongest when tied to the fractography. A nonconformance unrelated to the initiation site explains nothing.
Time in service is a poor proxy for cycles
Cycles matter more than months. A patient kept protected weight-bearing accumulates a small fraction of the loading of a heavier patient who returned to unrestricted activity early. Reconstructing the actual duty cycle — body mass, activity level, restrictions and adherence — places the failure somewhere on the design's expected life rather than leaving it uncalibrated.
Estimates of that kind belong in an opinion as ranges with their assumptions stated. Precision the record cannot support is the first thing an opposing expert tests.
What fractography does not resolve
The fracture surface establishes mechanism, origin and sequence. It does not establish whether a different implant or different post-operative instructions would have changed the outcome, and it says nothing about surgical judgement. Those are separate inquiries on separate evidence, and folding them together weakens the part of the analysis that is well founded.
Preserving what answers the question
The fracture surfaces are the evidence and they are fragile. Mating faces damage each other during extraction, cleaning and decontamination degrade the origin region, and cutting through the fracture zone to ease removal destroys it outright. ASTM F561 sets out standard practice for the retrieval, handling and analysis of implants removed from patients.
The practical instruction is short. Keep every fragment, do not clean, do not cut, package the pieces separately so surfaces cannot contact one another, and put the preservation request in writing before the revision.
This article is general technical orientation, not a failure analysis, an engineering opinion, or advice on any specific matter. Determining the cause of a particular incident requires hands-on examination by a credentialed expert.