Custom-made vs 3D-printed spinal braces: which choice does the evidence support?

Scoliosis is a three-dimensional deformity, so every good brace must correct in three planes. The word “3D” in advertising usually describes manufacturing, not a treatment principle. Three concepts need to be told apart:

When we talk about braces, “2D” and “3D” refer to the number of planes corrected, not to the machine used. Scoliosis is a deformity in three planes: the frontal plane (viewed from behind, a C or S curve measured by the Cobb angle), the sagittal plane (viewed from the side, changes of the natural curves) and the transverse plane (viewed from above, rotation of the vertebrae producing the rib hump).

“2D” braces are the classic single-principle designs. Each came from a particular school and rests on one main corrective principle, acting mostly in the frontal plane with part of the sagittal plane, so they are usually grouped as two-plane correction.

The common mistake is to merge the two meanings: any 3D-printed brace is called “3D correction”, while a plaster-moulded brace is called a “2D brace”. That is wrong, because the way a brace is manufactured does not decide how many planes it corrects.

A custom-made brace does not follow a fixed template. The practitioner examines, reads the X-rays, assesses spinal flexibility, then chooses and combines principles for each curve pattern so that all three planes are corrected:

These forces are created during the plaster casting itself and refined when the model is modified and the brace is trialled. This is the technology widely used in developed countries such as the US, UK, France, Germany and across Europe.

A caution when quoting these: the comparative studies are mostly retrospective with small samples (the Minsk Rigo-Chêneau group had only 13 cases). There is not yet a large randomised trial directly comparing multi-principle custom braces with classic braces. We state this limit openly so you can weigh the evidence at its true level.

Both methods are the same in examination and assessment; they differ in how the shape is captured, the model corrected and the brace formed.

Wong and Cheng's group (Hong Kong Polytechnic University) showed that CAD/CAM shortened model modification time by about 63.5% compared with manual work. So the proven benefit of digital technology is mainly saving labour time, not improving treatment outcomes.

A plaster cast does not merely copy the body's shape. While the plaster sets, the practitioner uses their hands to apply forces directly to the trunk: pushing at the curve apex, lengthening, counter-rotating the pelvis and rib cage. The curve is therefore already corrected and reduced on the mould.

With 3D scanning the patient simply stands still with arms raised. The scanner records the trunk exactly as deformed, with no corrective force applied. All correction is therefore pushed into the computer-modification stage, with three consequences:

With plaster casting the correction is made from the mould itself, so little modification is needed and the brace fits the body more closely and accurately.

A balanced word on the evidence is needed: direct comparisons of the two ways of capturing the shape are few and do not yet answer the question above. A study on rigid mannequins (Kumagai et al., Neurospine 2026) found that braces made from a 3D scan fitted the model more closely (smaller gaps, especially in curved areas) and were faster to make (about 55 versus 100 minutes); a rigid mannequin has no curve to correct, so it cannot measure the benefit of hands-on correction while casting. A retrospective French study of 101 Garches braces for 32 children (Viallard et al., Prosthet Orthot Int 2026) found no significant difference in Cobb-angle change, number of fittings or skin problems between plaster casting and 3D scanning, and staff preferred scanning for the child’s comfort; that group was children with neuromuscular disease, not adolescent idiopathic scoliosis. The reason for preferring a hands-on corrective mould here is therefore a mechanistic argument based on clinical practice, not a conclusion from a direct trial in AIS; good results still depend on the orthotist’s skill whichever method is used. Corrective plaster casting remains widely used in spinal orthotic practice in many developed countries, including Europe and the USA, while 3D-printed braces are the newer approach.

The effectiveness of bracing in general has strong evidence. But that evidence was built on custom vacuum-formed braces; 3D-printed braces have only a few small, short-term studies.

Overall: 3D-printed braces give comparable short-term results, sometimes slightly better or worse, but with no clear difference. Most studies have fewer than 50 patients, measure only immediate correction and do not follow patients to skeletal maturity. The one large randomised trial registered in China (ChiCTR1900027379) has not yet published results.

The 2011 SOSORT guidelines also noted that studies comparing CAD/CAM with plaster moulding are converging on one conclusion: how a brace is made does not affect the final outcome or the patient's perception. In other words, digital technology does not make the brace better, so there is no scientific basis to pay extra for a “3D brace” expecting higher effectiveness. The guidelines also advise each team to use the type of brace it understands and masters best, since no type is currently recommended over others.

The thermoplastics used in vacuum-formed braces (polyethylene, polypropylene) have been used for decades on millions of patients. Their durability, tolerance of 18–23 hours of wear a day and skin compatibility are established. By contrast, printed materials (such as nylon PA12) still raise open questions:

For this reason the orthotics profession still regards printed braces as a technology under validation, not yet reliable enough to replace vacuum-formed braces.

The two large studies below show that custom rigid braces are effective and safe in the long term. Both were done with conventional vacuum-formed braces; there is no equivalent data for 3D-printed braces.

Systematic review and meta-analysis by Costa et al. (Journal of Clinical Medicine, 2021): 33 studies, of which 16 at low or moderate risk of bias were pooled. Success was defined as curve progression of no more than 5°.

Lange et al. (Scoliosis, 2011) followed 272 patients in Oslo, Norway, for an average of 24.7 years (16–32 years); mean age at review was 40.4.

Limitations to state: the Lange study had no control group. Costa et al. also note that most bracing studies carry a considerable risk of bias; even so, success rates did not depend on study quality.

In the United States, since February 2024 CMS guidance states that a 3D-printed brace may be paid under the custom-made brace code only when it meets all the quality criteria of a custom-made brace: made for one patient, based on a clinically modified mould, measurements or image. The same authority makes it clear that 3D printing is only a fabrication technique and is not by itself enough for a product to count as “custom-made”. This has three implications:

In other words, a patient who pays extra for a “3D brace” is paying for technology, not for a proven treatment benefit. If printed braces were truly superior, developed countries would have a very strong incentive to replace plaster and vacuum forming entirely: machines run day and night, labour is freed, training costs fall. In reality plaster-moulded vacuum-formed braces remain the common choice and printed braces are accepted only as “equivalent”.

The economic advantage of 3D printing depends on the balance between machine cost and labour cost, and in Vietnam that balance is the reverse of developed countries.

Even in Canada the value depends on the model: a printed brace is 37% cheaper when printed in-house but 27% more expensive when outsourced (University of Alberta). The core advantage of 3D printing is therefore optimising production and finance where labour is expensive, not a treatment benefit to the patient. In Vietnam, patients usually pay more for a product with no proven superior effect.

Custom vacuum-formed braces are strong on evidence, durability and adjustability; 3D-printed braces are strong on production speed, weight, thinness and appearance, but have not shown a treatment advantage.

Developed countries such as the US, UK and Europe still mainly use custom vacuum-formed braces; digital technology is introduced chiefly to save labour in some steps.

The literature identifies four factors that influence bracing outcomes: risk of progression by age and curve size, the amount of in-brace correction, hours of wear, and quality of wear. Design quality decides correction and comfort, and so indirectly whether the child is willing to wear the brace. No technology replaces these four steps:

SOSORT sets very specific requirements for the prosthetist-orthotist (CPO) who makes spinal braces: at least 2 years of continuous work with the same experienced physician; at least 2 years of continuous practice in spinal bracing; and at least 2 braces a week, about 100 a year, over the last 2 years. None of these requirements mentions the type of machine used. SOSORT stresses that experience and preparation are the only way to avoid risk to patients and achieve good results. Other recommendations in the guidelines:

A 14-year-old boy with an S-curve of 41° and 33° Cobb responded very well to a custom-made brace. After 8 months without wearing it, the curve rose to 48° and 37°. The family switched to a 3D-printed brace hoping their son would be more comfortable and cooperative; some time later the lower curve improved slightly but the upper curve kept increasing, so the family chose to return to a custom-made brace using the traditional method.

The case shows that whatever the technology, patient adherence and the skill of the person making the brace remain decisive. One case cannot settle the question of a technology; we present it as a practical illustration, not statistical evidence.

The custom-made brace, moulded by plaster and vacuum-formed from thermoplastic, remains the validated and widely used standard in developed countries. The 3D-printed brace is a promising direction for weight and appearance, but has no evidence of superior effectiveness, and the major insurers treat it only as a manufacturing technique. For any individual case, the quality of the brace is decided by the expertise of the person who examines, designs and makes it.

In short, there are several things parents should know about 3D-printed braces before choosing:

The main advantages of 3D-printed braces are fast production, continuous machine operation and a lighter, thinner brace. These are manufacturing benefits, not treatment benefits that have been proven.