RESEARCH DISCIPLINE
Statistical Analysis for Dental Research
Consulting for Oral Health & Clinical Dental Studies
Dental research statistical analysis for split-mouth and parallel trials, periodontal and caries cohorts, implant survival and diagnostic studies, with the analysis built on the level where treatment was applied and results reported at the level of the clinical decision.
No Commitment Required
Strictly Confidential
In-House PhD-Level Statisticians
Fixed Project Fee
Same-Day Response
Statistical Analysis for Dental Research
Dental research statistical analysis is the statistical design, analysis and reporting of oral health studies, in which the units measured are teeth, surfaces, implants or sites nested within patients, while the treatment may be applied to the patient, one side of the mouth or a single tooth. Counting every tooth as a separate participant overstates the information in a study, a split-mouth trial analyzed as two independent groups ignores its pairing, and teeth lost during follow-up leave later examinations incomplete. OnlineSPSS, a specialist statistical consultancy for academic, clinical and professional research, provides dentistry data analysis through in-house PhD-level statisticians, who decide which level carries the treatment, how lost teeth are handled and at which level results are reported with you before the models are fitted.
WHEN TO ENGAGE
Statistical Support at Different Stages of Dental Research
Dental researchers contact us from the choice of the unit to randomize to a reviewer's question about whether teeth were treated as independent.
Designing a Clinical Dental Study You are planning a randomized or observational study and need to decide whether treatment goes to patients, sides of the mouth or teeth, how many of each and which outcome is primary.
Analyzing Tooth- or Site-Level Data Your measurements are recorded per tooth, surface or site, and the model, the level of summary and the handling of teeth missing at follow-up must be chosen and checked.
Evaluating a Diagnostic Test You are comparing a radiographic, optical or clinical test with a reference standard and need accuracy reported at patient and tooth level with intervals that respect the clustering.
Revising After Review A reviewer or advisor has asked whether teeth were treated as independent, whether a split-mouth design could have carried the effect across to the other side, or at what level accuracy was summarized.
RESEARCH CONTEXT
Research Designs and Data We Work With in Dentistry
Dental datasets are recorded per patient, per tooth and per site, and the level at which treatment and outcome sit determines which analyses are valid.
01
Split-Mouth Trials
Treatments allocated to opposite sides or quadrants of the same mouth, so each patient gives a paired comparison and carry-across between sides has to be ruled out.
02
Parallel Randomized Dental Trials
Patients randomized to restorative, periodontal or preventive treatments, each with several teeth or implants measured, so the unit of randomization is smaller than the number of teeth.
03
Periodontal and Caries Cohorts
Follow-up of probing depth, attachment loss or caries increment across sites and visits, where teeth are lost or extracted between examinations.
04
Implant and Restoration Survival
Time to failure of implants, crowns or fillings, several per patient, with removals, recalls and lost contact that leave follow-up incomplete.
05
Imaging and Detection Studies
Radiographs, scans or optical devices read by examiners against a reference standard such as histology or clinical examination, with several teeth per patient.
06
Orthodontic and Measurement Studies
Cephalometric, scan and model measurements taken repeatedly by one or more examiners, where measurement error and agreement must be quantified before groups are compared.
METHODOLOGICAL CONSIDERATIONS
Statistical Issues We Consider in Dental Research
The level of treatment, the lost teeth and the summary level are checked before results are trusted, since each can change a conclusion more than the model does.
Which Level Carries the Treatment A treatment given to the patient but measured on many teeth has far fewer independent units than teeth. The analysis and any sample size are built on the unit that was randomized.
Pairing in Split-Mouth Designs Opposite sides are paired, so the analysis uses the within-patient difference. Carry-across to the other side cannot be estimated from the data, so it is judged clinically before the design is chosen.
Teeth That Are Lost Teeth extracted or lost during follow-up are often the most diseased ones, so analyzing survivors alone can flatter a treatment. Loss is an outcome or competing event where it relates to treatment.
Patients With More Teeth Patients with many teeth contribute more observations, and when their oral health differs from that of patients with few, a tooth-weighted estimate answers a different question from a patient-weighted one.
Bounded Scores and Many Zeros Caries scores, bleeding percentages and surface counts are bounded and often zero. They are modeled with a count, binomial or ordinal model, not as a normal outcome, and the level of summary is stated.
Accuracy at the Level of Decision Sensitivity and specificity pooled over teeth in one mouth look more precise than they are. Accuracy is reported where the decision is made, with cluster-aware intervals and a defined reference standard.
STATISTICAL METHODS
Methods Selected for the Research Question and Study Design
The method families below follow the level of treatment and the outcome of your study, and each line says why a family suits dental data.
Longitudinal & Multilevel Analysis
Teeth, sites and visits nested within patients enter as levels, and split-mouth differences are within-patient contrasts, so inference reaches the patient from tooth-level data.
Clinical, Survival & Epidemiological
Implants and restorations fail at different times in the same mouth, so time to failure is analyzed with censoring and with variance estimates that allow for clustering within patients.
Sensitivity & Specificity (Diagnostic Accuracy)
Caries and lesion detection is judged against histology or clinical examination, with accuracy computed at patient and tooth level and agreement measured between examiners.
Regression & Predictive Modeling
Caries increment, bleeding and graded outcomes are modeled with the distribution they follow, with patient-level covariates such as smoking beside tooth-level ones such as tooth type.
Comparative & Hypothesis Testing
Left-right comparisons in a split-mouth study use paired methods on the within-patient difference, McNemar's test for paired binary outcomes and rank-based versions for ordinal scores.
Bayesian & Computational Methods
Missed examinations are imputed under stated assumptions with patient as a level, and bootstrap intervals for patient-level summaries resample whole patients, not teeth.
DELIVERABLES
Statistical Outputs Prepared for Your Research
The outputs follow the structure of a dental study, from the review of tooth coding to the files that reproduce every table.
Data & Study Review Tooth numbering, site codes, patient identifiers and the allocation level are reviewed with the teeth missing at each visit, so every record is tied to a tooth, a mouth and a time point.
Statistical Analysis The agreed mixed, GEE, survival or paired analyses, with the clustering structure checked, the split-mouth pairing and side allocation verified and sensitivity analyses for lost teeth.
Effect Estimates & Uncertainty Within-patient and between-group differences, ratios or hazards with confidence intervals, and patient-level and tooth-level accuracy with intervals that allow for clustering.
Tables & Figures Patient and tooth flow diagrams, baseline tables at each level, survival curves for implants or restorations, agreement plots and plots of split-mouth differences.
Statistical Interpretation Results are explained in clinical terms at the level of the decision, patient or tooth, with the limits of the design stated, such as examiner calibration or loss of teeth.
Analysis Files Syntax or scripts in SPSS, R, Stata or SAS, with the data documented in long format by patient, tooth and site, so that every table can be regenerated.
REPORTING
Reporting Dental Research
Dental studies report to the guideline for their design, and the within-mouth and clustered structure adds requirements of its own.
Parallel randomized trials in dentistry report to CONSORT 2025, and within-person designs such as split-mouth trials add the CONSORT extension for within-person trials, which asks how the pairing was handled in the analysis. Cohort, case-control and cross-sectional oral health studies follow STROBE, and studies of an imaging or clinical test report to STARD 2015 with the unit of analysis for accuracy stated. Flow diagrams give patient and tooth counts separately, each result is stated at the level it was analyzed, and the instructions of your target journal are checked at the end.
ILLUSTRATIVE EXAMPLE
How the Analysis May Look in Practice
An illustration of how a split-mouth question moves from a protocol to a result that is analyzed at the right level.
ILLUSTRATIVE EXAMPLE · PEDIATRIC DENTISTRY RESEARCH · UNITED STATES
A pediatric dentistry group planned a split-mouth comparison of two fissure sealant materials in children's permanent molars, with retention assessed at 12 months.
Which material went on which side was randomized within each child, but some children lacked a qualifying molar on one side, so not every child gave a complete pair. Before outcomes were examined, we fixed a mixed logistic model with a random intercept for child, which keeps children with an incomplete pair, and a conditional analysis of complete pairs as a sensitivity check. Retention was assessed by examiners calibrated beforehand, and the results were reported per tooth and summarized per child.
Outcome: The group received the odds ratio for retention with its 95% confidence interval, the retention proportion by material, CONSORT 2025 and within-person extension flow counts, a methods paragraph and the scripts.
This example does not describe a real client or project. We never share details of actual client work, in line with the confidentiality we promise every researcher. This illustrative example is included to show the type, scope, and standard of work involved in this kind of analysis.
RESEARCHER VOICE
What Researchers Say
Feedback from a dental researcher on clustered data that could not be treated as independent rows, and on tables and interpretation written for the research context.
“Our project had a very field-specific problem around the clustered structure of the data rather than treating every row as independent. The team understood why it mattered in Dentistry and chose the analysis accordingly. The final tables and interpretation felt written for our research context rather than copied from a generic statistics template.”
HOW WE ASSESS YOUR PROJECT
We Review Your Actual Research
Your assessment is based on a review of your actual research materials and analytical requirements, not a generic project category.
You Submit Your Project Details
Share whatever is available at the current stage, such as the research aim, protocol or proposal, dataset, variable list, manuscript or statistical requirements.
We Review Your Research and Define the Scope
Your statistically trained Research Coordinator reviews the materials to understand the study design, research questions, data structure, analytical requirements and expected deliverables. If clarification is needed, additional documents or information can be requested before the assessment is finalized.
Your Analysis Is Completed and Delivered
Once the scope and fixed project fee are agreed, the in-house statistician completes the work and the Research Coordinator remains the communication point through delivery, with the analysis, deliverables and timeline as agreed.
SOFTWARE
Statistical Software for Dental Research
Dental analyses are run in the software your clinic or school already uses, with syntax or scripts provided where applicable.
Statistical Support Within Your Research OnlineSPSS provides the agreed methodological, analytical, and statistical reporting support. The researcher or research team retains the scientific question, substantive interpretation, conclusions, and overall responsibility for the research.
FREQUENTLY ASKED QUESTIONS
Questions About Statistical Analysis for Dental Research
Answers to common questions from dental researchers, on teeth within a mouth, split-mouth trials, accuracy, sample size, confidentiality and the assessment.
How should several teeth from one patient be analyzed?
The analysis should match the level at which treatment was applied. Teeth from one mouth share the patient’s habits, oral hygiene and biology, so a model with the patient as a level, or generalized estimating equations with patient clusters, gives intervals that are honest, and the results can be stated per patient or per tooth. Our Statistical Methods pages describe the families involved.
How is a split-mouth trial analyzed?
Split-mouth analysis works on the within-patient difference between the two sides, because each patient serves as their own control. Side allocation is checked, and treatments that can spread to the other side, such as rinses or treatments with a systemic effect, are not suited to the design. Incomplete pairs are handled with a model that keeps those patients.
Should sensitivity and specificity be calculated per tooth or per patient?
Accuracy is reported at the level where the clinical decision is made, and the level is stated. A patient-level figure answers whether a person has disease, and a tooth-level figure answers which tooth is affected, and the second needs intervals that allow for several teeth per patient. Both rest on a clearly defined reference standard.
Can you estimate sample size for a dental trial with clustered teeth?
Yes. The sample size depends on the unit that was randomized, on the number of teeth per patient and on how alike teeth in one mouth are, so a large count of teeth does not by itself give a large sample. We calculate it for the primary outcome and the planned analysis, and write the assumptions into the analysis plan.
How is confidential clinical dental data handled?
Study datasets, protocols and analysis files are handled confidentially, and project information is not published. An NDA is available where required and can be arranged before sensitive materials are shared. Read more about confidentiality and data security.
What happens when I request an assessment, and how long does it take?
You share your research question, design and relevant materials, and a statistically trained Research Coordinator reviews the request and responds the same day. The analysis, deliverables, timeline and one fixed project fee are then confirmed before any work begins, so the timeline follows from the agreed scope.
Still have a question? Find more answers in our full FAQ.
NEXT STEP
Discuss Your
Dental Research With Us
Share your research question, study design and available data, even if only a protocol or a tooth coding sheet exists. We review the structure of your measurements and confirm the approach before any work begins.
No Commitment Required
Strictly Confidential
Fixed Project Fee
In-House PhD-Level Statisticians
Same-Day Response
Serving Researchers Since 2014