A malunited fracture or a developmental deformity is seldom confined to one plane. Angulation in the coronal and sagittal planes usually coexists with rotation and translation, and the true deformity lies in an oblique plane that no standard radiograph shows directly. The principles for analysing such deformities from radiographs are well established (Paley, 2002), but carrying out a multiplanar correction freehand, with wires and an image intensifier, depends heavily on the surgeon's spatial judgement.

Three-dimensional planning from computed tomography (CT) offers a different route. The deformed bone is modelled, compared with a template (usually the mirrored opposite side), and the osteotomy is simulated on screen. A patient specific guide then transfers the simulated cut and the planned position of the fragments to the patient. The approach was developed for malunions of the forearm and has since been applied around the knee and elsewhere (Murase et al., 2008; Victor and Premanathan, 2013). Systematic reviews of three-dimensional printing in surgery list surgical guides among its commonest uses and report shorter operating time and better accuracy as the usual claimed benefits, with cost and production time as the usual drawbacks (Tack et al., 2016; Martelli et al., 2016).

Producing the guides inside the hospital addresses those two drawbacks and brings the surgeon closer to the design. It also makes the hospital a manufacturer. This article proposes a procedure by which an orthopaedic department can do this in an orderly way. It describes a method; it does not report results, and the procedure has not been validated here.

A cutting guide is a negative of a chosen area of bone surface, extended into a body that carries slots for the saw blade and sleeves for drills or wires. Its accuracy depends on a chain of steps, each of which adds error: the resolution of the scan, the fidelity of the segmentation, the quality of the plan, the dimensional accuracy of the printer, and the surgeon's ability to seat the guide on a bone surface that has been cleared of soft tissue. The radiological literature on medical printing describes this chain and the points at which it commonly fails (Mitsouras et al., 2015), and guidance from the Radiological Society of North America sets out expectations for image quality, segmentation and quality control of printed models (Chepelev et al., 2018).

The regulatory position differs between jurisdictions and should be established locally before any guide is used. In the European Union, the Medical Devices Regulation permits health institutions to manufacture and use devices in house under the conditions of its Article 5(5), which include an appropriate quality management system, documentation of the design and manufacturing process, and a justification that the need cannot be met by an equivalent device on the market (European Union, 2017). In the United States, the Food and Drug Administration has issued guidance on technical considerations for additively manufactured devices (FDA, 2017). A cutting guide is in contact with tissue for a short period and must be sterile, so biological evaluation of the material and validation of the sterilisation cycle are unavoidable parts of the task (Christensen and Rybicki, 2017).

The procedure has seven steps, shown in sequence in Figure 1. The paragraphs that follow describe each step; the checks and records attached to them are gathered in a table further on.

The procedure needs people before it needs machines. A surgeon who plans, an engineer or technologist who segments and designs, and a named person responsible for quality are the minimum, and the roles should be separated so that no one checks their own work. A quality management system is required under the European in-house exemption, and ISO 13485 is the recognised model for it (ISO, 2016). In practice this means written procedures for each step, controlled versions of software and design files, traceability from the patient to the material lot, training records, and a route for reporting and investigating failures.

The economic case should be made honestly. Published analyses suggest that guides and anatomical models can save operating time, and that the value of that time can offset production costs (Ballard et al., 2020), but a department printing a small number of guides a year carries fixed costs in staff time, software, validation and maintenance that a simple cost per print conceals. Collaboration with a university engineering department or a regional hub may suit low-volume units better than independent production. Departments should also decide in advance which cases justify a guide. Simple uniplanar corrections gain little; multiplanar malunions and osteotomies close to joints are where the planning effort is most likely to be repaid (Hoekstra et al., 2016).

What is described here is a proposed procedure, assembled from the literature and from standards. It has not been tested in this article. A department adopting it would need to establish at least four things in its own setting before clinical use and during it. The first is dimensional accuracy: that printed and sterilised guides match their design within a stated tolerance. The second is transfer accuracy: that the correction achieved, measured on postoperative CT or calibrated radiographs, agrees with the plan. Sawbone or cadaveric trials are the appropriate first test. The third is safety, including sterility assurance and the absence of material-related reactions. The fourth is benefit to the patient, which is the hardest to show, because better radiographic accuracy does not by itself prove better function.

The published evidence for guides in deformity surgery consists mainly of case series and technical reports, and comparative studies against conventional technique are few (Tack et al., 2016). The procedure also depends on judgement at several points that no checklist replaces: the choice of template, the decision that a segmentation is good enough, and the decision in theatre to trust or discard a guide. It is confined to single-use polymer cutting and drilling guides and does not extend to printed implants, which raise different and stricter requirements.

Designing and printing patient specific cutting guides within a hospital is technically within reach of an orthopaedic department with access to CT, suitable software and a qualified printer. What makes it safe is the discipline around the technology: a fixed imaging protocol, independent checking of the segmentation, a surgeon-owned plan, a design that seats in one position, a validated route through printing and sterilisation, and records that tie every guide to a patient. The seven-step procedure proposed here is intended as a starting framework for departments to adapt, test on models and audit against their own results before and after they bring it into clinical use.

To be declared by the author(s) before publication.

To be declared by the author(s) before publication.