Oral Health Educational Methods in Adolescence: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Oral Health Educational Methods in Adolescence (NCT03216746), a randomized na study. The primary outcome is Knowledge Score.

2. Clinical Question

3. Trial Design

The registry describes a randomized, single-masked, parallel-group trial of educational methods in adolescent oral health. The record is marked completed. The listed study dates are 2016-12-01 for study start and 2017-11-15 for primary completion. The sponsor is Universidade Federal do Paraná.

The intervention structure includes four groups: oral orientation, oral orientation and app, video orientation, and video orientation and app. The arm descriptions report 71 participants for oral orientation, 66 for oral orientation and app, 63 for video orientation, and 63 for video orientation and app. These are arm-specific values as recorded; the reason they do not reconcile with the registry enrollment figure is not reported in the registry record.

The oral orientation was standardized guidance delivered by a previously trained researcher and addressed general and oral health, particularly periodontal diseases. The video groups received oral health information through audiovisual material made for the research. The registry names a smartphone app as part of two groups, but its full content and use are not reported in the registry record. The study summary describes educational activities and assessments in phases, but its account is incomplete, so details beyond those explicitly recorded cannot be established.

The primary outcome is a knowledge score assessed at 30 days. The registry lists one outcome measure and one reported analysis. Allocation procedures beyond the randomized designation, the masking role, and the statistical analysis plan are not reported in the registry record.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Overall Study

Reasons for not completing this period:

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

6. Primary Endpoint

The registry lists the primary outcome as Knowledge Score, measured over 30 days.

The primary endpoint is a knowledge score concerning oral health. The study summary says that a questionnaire included statements about periodontal diseases and prevention, with responses scored as correct or incorrect to form a knowledge score. The endpoint is intended to represent knowledge rather than a clinical examination finding. Although the study summary also mentions assessment of plaque and gingival bleeding, these are not the posted primary endpoint in the registry results.

The registry specifies a median and full range for the score at 30 days. It describes the score as ranging “from 13 (five) to 0 (zero),” a wording that does not clearly specify the scale’s endpoints. The posted group results themselves give values and ranges from 5 to 13. The registry states that higher scores mean better outcomes.

A knowledge score is used to quantify questionnaire responses in a form that can be summarized and compared across groups. A median describes the middle of the observed scores, while the full range gives the lowest and highest reported values. Neither summary alone shows the distribution of all scores or how precisely the group results have been estimated. No confidence interval is reported for this endpoint.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does the Kruskal–Wallis test assess?

The Kruskal–Wallis test is a rank-based method for comparing outcome distributions across groups. Rather than relying directly on the group means, it ranks observations across the groups and assesses whether those ranks differ more than would be expected if the groups came from the same distribution. It is commonly used when an outcome is summarized by medians or when a conventional analysis based on means and its assumptions is not suitable. This is general statistical context; the registry does not state why this method was selected for this trial.

The registry reports this test across the four educational groups, with p = 0.001. This is an omnibus result. It gives evidence of a difference somewhere among the group rank distributions, but it does not say which groups account for that result. To identify particular contrasts, the analysis would need to report pairwise comparisons, with appropriate attention to multiple testing. Such comparisons are not reported in the registry record.

What do medians and full ranges show?

A median is the central observation after scores are ordered. It is a compact description of a group’s center and is less influenced by extreme values than a mean. The full range reports the lowest and highest observed values, so it indicates the span of scores but not how densely scores are distributed within that span. Two groups can have the same median and range yet differ in the arrangement of observations between those endpoints.

The posted medians are 11, 13, 11, and 13 for oral orientation, oral orientation and app, video orientation, and video orientation and app, respectively. Their reported full ranges are 5 to 13, 9 to 13, 5 to 13, and 10 to 13. These summaries are descriptive; they are not estimates of uncertainty. The registry does not report confidence intervals for the medians or for contrasts between groups.

Why can a rank-based test be useful?

As general statistical context, rank-based methods can be useful for outcomes whose distributions are skewed, whose measurement scale is ordinal, or for which a model based on a particular distribution is not well justified. Ranking reduces dependence on the numerical spacing between outcome values. However, the Kruskal–Wallis test is not automatically a test of medians alone. Interpreting it as a difference in medians requires conditions about the shapes of the group distributions. The registry does not report distribution plots, diagnostic checks, or an explanation of the assumptions used.

What does the p-value say—and not say?

The reported p-value is a measure of how unusual data at least as inconsistent with the test’s null model would be, if that model and the test assumptions held. It is not a measure of the size of a group difference, the probability that an intervention caused the observed pattern, or the probability that a finding will replicate. The p-value should be read alongside estimates of effect and uncertainty. Those estimates are not available in the registry record.

The trial record also contains planning assumptions describing a sample-size illustration based on an observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. Numerical planning results are not reported in the registry record. Such planning illustrations are not observed power and do not replace an effect estimate with its uncertainty. The listed analysis is across four groups, whereas the described planning assumptions refer to two arms; the relation between those assumptions and the reported analysis is not explained in the registry record.

What cannot be inferred from the reported analysis?

The reported test does not establish that adding an app caused a particular change, because the registry does not report a factorial analysis separating app effects from education-format effects. It does not establish the direction of every group contrast, despite the descriptive medians. It also does not provide adjusted estimates, confidence intervals, or a specified analysis population. These details are not reported in the registry record.

10. Analyses Reported by the Sponsor

The registry reports the sponsor’s overall analysis as a Kruskal–Wallis test across the four groups, with p = 0.001. The posted group summaries are medians and full ranges. Together, these show the observed score summaries and the result of the reported omnibus test, but the registry does not give pairwise tests or an estimated contrast with a confidence interval.

A recalculated two-group comparison is not estimable from the registry record’s analysis information. The record explicitly indicates that values are missing or not estimable for a two-arm comparison for this outcome. Therefore, there is no recalculated two-arm result to set against the sponsor’s four-group analysis. These are also different questions: a two-group calculation would not reproduce the overall four-group test.

The registry does not report a discrepancy between the sponsor’s analysis and a recalculation. The sponsor’s reported analysis should therefore be read on its own terms: it indicates an overall rank-distribution difference among the four groups under the test, without specifying which groups differ or how large a difference is. Additional details of the sponsor’s statistical plan, including any follow-up comparisons, are not reported in the registry record.

11. Multiplicity, Missing Data and Analysis Populations

The record lists one primary outcome measure and one reported analysis, but it does not explain whether other analyses were planned or performed. It does not report a multiplicity strategy, such as whether follow-up comparisons were adjusted when examining several group contrasts. With an overall four-group test, pairwise comparisons would answer more specific questions than the omnibus result and would raise the issue of multiple testing. No such pairwise results are reported in the registry record.

Although the registry describes the study as randomized and gives an enrollment figure, it does not report the analysis population, how participants were assigned to that population, or whether results followed the randomized assignment. It does not describe missing questionnaire responses, exclusions, withdrawals, or how missing endpoint data were handled. The arm descriptions give group counts, but the registry does not explain their relationship to the enrollment figure or give an analysis-specific denominator for the reported test.

These omissions limit assessment of how the analysis was carried out and whether the reported summaries represent all randomized participants, participants with an observed score, or another group. They also make it difficult to assess the potential influence of missing data. Missing-data methods can affect estimates when the probability of a missing observation is related to the outcome or participant characteristics; the relevance of those mechanisms cannot be determined from the registry record.

The record does not report a prespecified analysis plan, interim analyses, or sample-size justification. The study’s planning assumptions describe an illustrative calculation, but numerical results and a connection to the reported four-group analysis are not reported in the registry record. These limitations concern reporting and interpretation; they do not, by themselves, show that an unreported procedure was or was not used.

12. What the Result Does, and Does Not, Mean

The posted results describe a 30-day knowledge-score comparison across four educational groups. The app groups had medians of 13, while the groups without an app had medians of 11. The registry reports an overall Kruskal–Wallis p-value of 0.001. The ranges show that scores varied within each group and overlapped across groups. These observations describe the registry’s reported outcome; they do not by themselves determine the size, precision, or practical importance of group differences.

The overall test does not show whether a specific pair of groups differs. Nor does it isolate the contribution of the app from the type of educational material. A design with four combinations can be examined in several ways, but the registry reports only the four-group omnibus test. A separate estimate of the app effect, the format effect, or an interaction between them is not reported in the registry record.

The outcome is a questionnaire-based knowledge score, not a direct measure of lasting knowledge, oral-health behavior, or clinical condition. The record gives a 30-day timeframe for the primary outcome and does not report later assessment results. Although the study summary mentions plaque and gingival bleeding assessments, results for those measures are not posted as outcomes in the record.

The findings should not be treated as a guarantee of what would occur in other settings or populations. The record describes adolescents enrolled in a public school in Curitiba and gives limited detail about the app, missing data, and analysis population. Those details matter when judging how broadly the reported pattern can be interpreted. The data do not support medical advice or a claim about clinical benefit.

13. Why This Trial Matters Statistically

This trial is statistically interesting because its educational approaches form four groups: oral orientation or video orientation, each with or without an app. The question is therefore broader than a simple comparison of two treatments. The overall Kruskal–Wallis test addresses whether score ranks differ across the four groups, while questions about adding an app or choosing an education format would require more specific contrasts and estimates.

The reported medians provide a straightforward description of the group centers, and the ranges show the spread of observed scores. The omnibus p-value adds evidence about whether the group distributions are alike under the test’s null model. These summaries answer different questions: a p-value does not express the magnitude of a difference, and a median does not quantify its uncertainty. The absence of pairwise estimates and confidence intervals leaves those questions unresolved in the registry record.

The record also illustrates why consistent reporting matters. The enrollment figure differs from the initial sample described in the study summary, and the endpoint scale description contains an unclear parenthetical. The arm counts and outcome summaries remain reportable as recorded, but the reasons for these discrepancies or ambiguities are not reported in the registry record.

For biostatistical interpretation, the key distinction is between the observed score pattern and claims about its size, cause, or generality. The posted results provide a four-group descriptive comparison and an overall test. They do not provide the contrast estimates or analysis details needed to answer every question raised by a multi-arm design.

14. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The record does not report confidence intervals, pairwise comparisons, or an estimate of the size of a contrast. The overall Kruskal–Wallis result cannot identify which groups differ. The posted summaries also do not show the full score distributions, so the extent and pattern of within-group variability cannot be assessed from medians and ranges alone.

Some study details are incomplete. The registry enrollment is 288, while the summary describes an initial sample of 291; the reason is not reported in the registry record. Arm descriptions list their own participant counts, but the record does not explain the differences among these figures or give an analysis-specific denominator. The score-scale description, “from 13 (five) to 0 (zero),” is ambiguous. The reported group ranges are clearer, but they do not resolve the wording in the endpoint description.

The app’s content and implementation are not fully described, and the record does not provide enough detail to assess whether exposure was consistent across participants. The masking role is not specified beyond the single-masked designation. Recruitment, allocation procedures, missing observations, exclusions, and the analysis population are not described sufficiently to assess their influence on the result.

Only the knowledge-score result is posted. The study summary mentions clinical assessments, but their results are not reported as outcomes in the registry record. The listed timeframe is 30 days, with no later outcome results reported. The statistical plan, multiplicity handling, and sample-size justification are also not reported in the registry record. These reporting gaps constrain interpretation; they do not establish that any particular method was omitted from the conduct of the study.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Oral Health Educational Methods in Adolescence — NCT03216746.
• Registry record last updated 2020-05-05; status last verified 2020-04.
• Results first posted 2020-04-22.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

18. Record Summary

The registry describes a completed randomized, single-masked, parallel-group study sponsored by Universidade Federal do Paraná. It compares four adolescent oral-health education groups: oral orientation, oral orientation and app, video orientation, and video orientation and app. The primary outcome is a knowledge score at 30 days, summarized by medians and full ranges.

The reported medians are 11 for oral orientation, 13 for oral orientation and app, 11 for video orientation, and 13 for video orientation and app. An overall Kruskal–Wallis test across the groups has p = 0.001. The record does not report pairwise comparisons or confidence intervals, and it does not post results for the clinical assessments mentioned in the study summary.

Enrollment and sample descriptions differ, and the score-scale wording is ambiguous. Details of missing data, analysis populations, and the statistical plan are not reported in the registry record. The results therefore support a description of the posted 30-day score comparison, while leaving the size and precision of specific group contrasts unclear.