Clear Aligner Accuracy: Planned vs Achieved

Simplified Orthodontics Academy · Draft for owner review · 30 October 2026

Key points

In research, “accuracy” often means how closely a measured tooth position matches a digital plan. It is not the same as treatment success or patient satisfaction.
Studies differ in which teeth and movements they include, how they align digital models, and how they calculate the gap between planned and achieved movement.
Published percentages describe groups or movement observations. They are not the probability that one patient will achieve a movement.
A useful paper reports its sample, movement type, measurement method, timing and limits together with each result.

How clear aligner accuracy is measured

In clear aligner accuracy studies, a digital setup shows a planned position for each tooth. A follow-up record, often a digital model made from an intraoral scan, shows the tooth position measured at a later time. A study can compare those positions after aligning the models to a chosen reference. The difference is a study measurement—not proof that the original plan was biologically guaranteed or that the measured result is clinically adequate.

“Clear aligner accuracy” is therefore not one universal number. Authors may report a distance or angle between positions, a percentage of planned movement achieved, or an average across many teeth and movement observations. A percentage calculated in one study may not be directly comparable with a percentage from another if their definitions, teeth, or registration methods differ.

Figure 1. A study comparison begins with a baseline record, defines a planned position and measures a later position. The result depends partly on how the models are aligned and which surfaces are measured.

How studies measure the comparison

One approach is to register the initial and follow-up digital models on a reference region, then compare the planned and measured positions of each tooth. The reference matters: aligning a whole arch is not identical to aligning selected teeth or a region assumed to be stable. Authors may also use different software, surface-matching settings and landmarks. These choices can change the measured difference even when the clinical records are the same.

For example, Kravitz and colleagues compared predicted and achieved virtual models and used stationary posterior teeth as the reference for their superimposition. They evaluated anterior teeth and several movement types. Haouili and colleagues compared digital plans with post-treatment intraoral scans using best-fit superimpositions. They assessed teeth from the central incisors through the second molars, but did not assess torque because it could not be confirmed from the surface scans without radiographs. These methods answer related questions, but their samples and measurement scope are not interchangeable. [1] [3]

A surface scan records visible crown shape. It does not directly show the root. If a paper reports crown movement from surface models, do not silently read it as a measurement of root position, torque or the full three-dimensional movement of the tooth. The paper’s methods should state exactly what was captured.

A plan may also include an intended overcorrection. If a study does not account for that choice, its planned-versus-achieved difference can reflect how the target was set as well as how much movement was expressed. Kravitz and colleagues identified unaccounted-for overcorrection as a limitation of their analysis. This is another reason to inspect the plan definition before interpreting an accuracy percentage. [1]

Three movements can produce different measurements

“Tooth movement” is not a single outcome. Tipping changes crown and root orientation around a center of rotation. Translation describes a tooth shifting in roughly the same direction as a whole. Rotation describes turning around the tooth’s long axis. A study may score these separately because a crown surface can show them differently.

Figure 2. A simplified teaching diagram. Tipping, translation and long-axis rotation describe different spatial changes; these panels are not targets, instructions or clinical thresholds.

What published studies show—and what they do not

In a prospective study of 37 patients and 401 anterior teeth, Kravitz and colleagues reported mean movement accuracy of 41%. Their study reported 47.1% for lingual constriction and 29.6% for extrusion. Those percentages summarize agreement between planned and measured movement in that study. They do not mean that 41% of patients achieved treatment, or that the remaining patients failed. The study focused on anterior teeth, did not assess posterior movement, and did not include radiographs to evaluate root-related factors. [1]

A systematic review by Rossini and colleagues included 11 studies: two randomized trials, five prospective non-randomized studies and four retrospective studies. The review described extrusion as difficult to control and reported 30% accuracy for that movement. It also judged the evidence as limited by study number, quality and heterogeneity. Since the underlying studies did not use one shared measurement definition, the review’s percentage should be read in its reported context—not as a universal property of aligners. [2]

Haouili and colleagues studied 38 patients and included teeth from central incisors through second molars. Their reported mean accuracy across movements was 50%; buccolingual crown tipping had 56% accuracy and rotation 46%. The paper also reported 28% for mesial rotation of mandibular first molars, 37% for distal rotation of maxillary canines and 35% for intrusion of mandibular incisors. These are group-level movement results, not patient probabilities. The study was conducted in one practice with one experienced orthodontist, which limits how broadly its results can be generalized. [3]

It can be tempting to compare the 41% reported in one study with the 50% reported in another and call the difference an improvement. That is not a head-to-head comparison: the studies differed in sample, teeth assessed, reference method and treatment context. Haouili and colleagues described improvement in their discussion, but the two headline averages alone cannot isolate what caused a difference. [1] [3]

A later systematic review by Robertson and colleagues included seven articles, with six assessed as having moderate risk of bias and one high risk. The authors judged certainty about specific movement efficiency to be low to moderate and noted that not all clinical scenarios had been assessed. This is a reminder that a neat percentage can sit on top of a small, uneven evidence base. [4]

Why results vary between studies

Variation can begin before the first aligner is made. Studies may include different malocclusions, teeth, movement types and amounts. They may differ in how a plan stages movement, what attachments or other auxiliaries are used, how adherence is captured, and whether the outcome scan is taken before or after finishing procedures. A study that measures only the front teeth answers a narrower question than one that measures the full dentition.

Measurement choices matter too. A scan-to-scan comparison depends on the registration region, model quality and operator decisions. Some papers report millimetres or degrees; others report a percentage. A percentage can be especially hard to interpret when the planned movement is small, because a small absolute discrepancy may become a large relative difference. These values should be read alongside the paper’s definition and method-error assessment.

Review design adds another layer. Retrospective studies can depend on which records were available and which cases reached the analysis stage. Reviews may combine studies that use different aligner systems, populations, movements and measurement methods. If the authors do not pool results, that may be a sensible response to incompatible evidence rather than a missing calculation. [2] [4]

A practical way to read an aligner-accuracy paper

Find the unit behind the result. Is the denominator patients, teeth, arches or movement observations? A tooth-level average is not a patient-level success rate.
Identify the two records. What exactly counts as “planned,” and when and how was “achieved” measured?
Check the registration method. Which teeth or surfaces were used to align the models? Were they assumed to be stable, and was that assumption discussed?
Read the outcome definition. Is the result a distance, angle, percentage, or a rating? Does the paper define its formula and direction of error?
Look for measurement reliability. Did the study repeat measurements, report error or use blinded assessors? Were missing records or exclusions described?
Keep the conclusion inside the sample. Compare studies only when their teeth, movements, records and protocols are sufficiently alike.

Limits of the evidence

A model comparison does not capture every dimension of orthodontic outcome. A crown surface scan cannot directly establish root position. Accuracy relative to a digital plan does not, by itself, describe occlusal quality, refinement needs, comfort, satisfaction or the appropriateness of that plan. Results from one system, protocol or clinician cannot be assumed to describe another.

Most importantly, a study average does not predict what will happen for an individual. The authors’ numbers describe selected samples under stated methods. They are useful for understanding how researchers define and measure movement, but they cannot be converted into a forecast for a particular patient.

This is educational and research content only; it does not diagnose, prescribe, recommend a device, or provide patient-specific guidance.

References

Kravitz ND, Kusnoto B, BeGole E, Obrez A, Agran B. How well does Invisalign work? A prospective clinical study evaluating the efficacy of tooth movement with Invisalign. Am J Orthod Dentofacial Orthop. 2009;135(1):27–35. doi:10.1016/j.ajodo.2007.05.018. PMID: 19121497. Author-hosted full text.
Rossini G, Parrini S, Castroflorio T, Deregibus A, Debernardi CL. Efficacy of clear aligners in controlling orthodontic tooth movement: a systematic review. Angle Orthod. 2015;85(5):881–889. doi:10.2319/061614-436.1. PMID: 25412265. Publisher full text.
Haouili N, Kravitz ND, Vaid NR, Ferguson DJ, Makki L. Has Invisalign improved? A prospective follow-up study on the efficacy of tooth movement with Invisalign. Am J Orthod Dentofacial Orthop. 2020;158(3):420–425. doi:10.1016/j.ajodo.2019.12.015. PMID: 32620479.
Robertson L, Kaur H, Fagundes NCF, Romanyk D, Major P, Flores Mir C. Effectiveness of clear aligner therapy for orthodontic treatment: a systematic review. Orthod Craniofac Res. 2020;23(2):133–142. doi:10.1111/ocr.12353. PMID: 31651082.