Two records exist in almost every trauma fixation failure, and they degrade on completely different timescales. Serial radiographs are archived, dated and durable, and read in sequence they document how a construct was loaded month by month. The removed hardware carries the physical evidence of the same history, and it is routinely cleaned, cut or discarded within hours of coming out. Understanding what each record supports, and which one is about to disappear, largely determines what an investigation can say later.
Imaging is a time series, not a snapshot
A single radiograph showing broken hardware establishes very little beyond the fact of the break. The value sits in the sequence: the immediate post-operative films, the routine follow-up series, and any imaging taken when symptoms changed. Placed in order, those studies show the construct drifting or holding steady, and the drift carries the meaning.
This is why an imaging request should be for the complete study set with original metadata rather than for selected images. Acquisition dates, projections and technique all matter, and a report is not a substitute for the images.
What the first post-operative films establish
The earliest images fix the starting conditions: reduction quality, the residual gap at the fracture, implant position and length, which holes carry screws and which were left empty, screw purchase and trajectory, and whether hardware was contoured or cut at implantation. Every later comparison is measured against that baseline.
They also fix what the construct was at time zero, the only reliable way to separate features present from the outset from changes that developed under load.
Progression is the signal
The findings that matter are almost always changes rather than states. Callus that forms and then stops, a fracture line that stays sharply defined long after it should have blurred, screws that migrate or back out, a lucent halo widening around a screw shaft, or a gap that opens rather than closes all indicate that load was not transferring to bone.
Read as a series, those observations amount to a loading history. They indicate roughly when the construct stopped sharing load and began carrying it alone, the interval in which fatigue damage accumulates fastest.
Radiographic changes that often precede hardware fracture
Certain patterns are recognised precursors rather than incidental findings: progressive angulation at the fracture site, loss of reduction, hardware bending visible as a change in contour between studies, and lucency concentrated at the screws nearest the fracture. None of them identify a cause on their own.
What they do is establish sequence and timing, and sequence is usually the contested point. A construct that was visibly deforming across several studies before it broke tells a different story from one that was radiographically stable until it was not.
Cross-sectional imaging and what metal does to it
CT resolves union and lucency far better than plain films, but metal produces streak and beam-hardening artefact that can obscure exactly the bone-implant interface the study was ordered to assess. Metal artefact reduction techniques help, and knowing whether they were used affects how much weight a reported finding will bear.
MRI has its own constraints around implanted metal. For fixation hardware questions the practical consequence is that absence of a finding on cross-sectional imaging is weaker evidence than presence of one.
The explant usually leaves theatre before anyone considers it
Removed hardware has no clinical value once it is out, so hospital routine treats it as waste or as a specimen. It may be discarded, returned to a manufacturer under a complaint process, decontaminated and stored loose in a bag, or handed to the patient. By the time a matter is framed the hardware may be several custody steps away.
A short written preservation request placed with the operating team, naming the retained material and instructing that it not be cleaned or cut, is the single most effective intervention available. It has to precede the procedure to be useful.
What cleaning, cutting and packaging destroy
Ultrasonic cleaning, enzymatic detergents and autoclaving remove or alter surface deposits, corrosion products and adherent tissue, and can obscure the fracture initiation region. Cutting hardware to ease extraction sometimes passes directly through the zone of interest. Loose packaging lets fragments abrade one another in transit, superimposing fresh damage on the original surfaces.
Fragments matter individually. Two mating halves of a broken plate, the screws adjacent to the fracture and any debris recovered from the surrounding tissue each carry different information, and their relationship to one another is lost once they are pooled in one container.
Retrieval practice and chain of custody
ASTM F561 provides the standard practice for retrieval, handling and analysis of medical implants removed from patients, covering documentation, decontamination that preserves surfaces, and staged examination from non-destructive inspection through to destructive testing. Working to a recognised practice rather than an improvised one is what makes the findings defensible.
The sequencing rule is simple: everything non-destructive first, fully documented, before anything irreversible. Sectioning, cleaning for microscopy and mechanical testing all foreclose later options, so they follow photography, dimensional inspection and surface examination, and follow notice to the other side.
Reading the two records against each other
Neither record is complete alone. Imaging shows position, timing and healing but not the surface of the metal. The explant shows mechanism and origin but carries no dates. Correlating them — matching a fracture origin to the screw hole that sat over the gap on the films — is where a conclusion becomes reasonably supported.
Where the two disagree, the physical artefact generally governs on mechanism and the imaging generally governs on timing. Saying which record supports which part of the opinion is what keeps the analysis reviewable.
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.