Collection of Gingival Crevicular Fluid From Periodontitis Patients: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Collection of Gingival Crevicular Fluid From Periodontitis Patients (NCT00763048), a randomized phase 3 study. The primary outcome is Metabolite Associated With Inflammation (Cadaverine).

2. Clinical Question

3. Trial Design

The registry record describes a completed, randomized, single-masked, parallel-group Phase 3 study sponsored by Colgate Palmolive. Enrollment was 39. The study began in 2008-03, and primary completion was recorded in 2008-04. The record does not explain who was masked, how randomization was carried out, or how allocation was concealed.

The two groups differed in toothpaste ingredients beyond fluoride. The control group used fluoride toothpaste, identified as Colgate Great Regular Flavor. The active-comparator group used fluoride/triclosan/copolymer toothpaste, identified as Colgate Total Toothpaste. The record says all study toothpastes contained fluoride and describes the study as evaluating additional ingredients in the other toothpaste.

The registry’s brief summary says the study collected GCF samples from diseased patients suffering only from gingivitis and/or periodontitis. The listed condition is gingival diseases. Further details about eligibility criteria, clinical severity, recruitment setting, or participant characteristics are not reported in the registry record.

The primary outcome was mean cadaverine, a metabolite associated with inflammation, measured as a normalized ion count from a mass-spectral instrument after 6 weeks of treatment use. The record lists 10 outcome measures and 9 analyses, but additional outcome-level results and analysis details are limited. The relationship between the short calendar interval from study start to primary completion and the stated 6-week treatment-use timeframe is not explained 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.

Before assignment: Subjects were screened to meet medical and oral health criteria and then began a 1 week washout period with fluoride toothpaste.

Overall Study

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 Metabolite Associated With Inflammation (Cadaverine), measured over 6 weeks.

The primary endpoint is the mean quantitative, normalized ion count for cadaverine in gingival crevicular fluid. GCF is fluid collected from the area around the gums. Cadaverine is described as one of 10 inflammation biomarkers found in GCF and associated with inflammation that could lead to gingivitis. The measurement comes from a mass-spectral instrument; it is an instrument-based count rather than a direct clinical rating of gum health.

The endpoint was assessed after 6 weeks of study-treatment use. The registry description says that a lower ion count may indicate less inflammation. That wording is appropriately cautious: the value is a biomarker measurement, and the record does not establish that a particular count corresponds to a specific clinical state or patient benefit.

The outcome is summarized as a mean with standard deviation, so the comparison concerns average marker levels and the spread of individual measurements around each group mean. It does not, by itself, describe how many participants experienced a clinical improvement, whether a change was noticeable to participants, or whether the marker predicts longer-term outcomes. Further information about baseline measurements, changes from baseline, and the clinical interpretation of the ion-count scale is not reported in the registry record.

7. Results Posted in the Registry

• B - Active Comparator: mean 0.92 versus 1.12 on control; difference -0.20 (95% CI -0.71 to 0.31), standardized effect d = -0.24; z-test p = 0.445.

8. Statistical Interpretation

9. Statistical Methods Explained

What does a mean comparison summarize?

A mean is the arithmetic average of the measured values in a group. For this endpoint, the values are normalized ion counts from mass spectrometry. The control mean was 1.119, and the active-comparator mean was 0.921. Comparing means gives a concise description of the groups’ average marker levels, but it does not show every participant’s value or the shape of the distributions.

The group standard deviations were 0.835 for control and 0.785 for the active comparator. A standard deviation describes how dispersed measurements are around a group mean. It is not the same as the uncertainty in the mean itself. The standard errors listed for the group means are 0.182 and 0.185. In general, standard errors describe the sampling uncertainty of estimated means; they are influenced by both measurement variability and the amount of information in a group.

How should the mean difference be read?

The unadjusted re-calculation expresses the contrast as active-comparator mean minus control mean. The estimated mean difference is -0.198 ion count. A negative estimate means that the observed active-comparator mean is lower than the observed control mean. The estimate is a difference between group averages, not an individual participant’s change and not a percentage change.

The pooled standard deviation used in the re-calculation is 0.812. Pooling is a way to represent within-group variability with a common estimate for a two-group comparison. The standardized difference, reported as d = -0.244, expresses the estimated mean difference relative to that pooled variability. It is a dimensionless description of the contrast, not a clinical threshold. Its interpretation depends on the measure and context, and the registry record does not report a threshold for a meaningful change in cadaverine count.

What does a t-test assess?

In general, a two-sample t-test evaluates whether the observed separation between two group means is large relative to the expected sampling variation under a null hypothesis of no mean difference. The re-calculated p-value is 0.445. In plain language, this is not strong evidence against that null hypothesis under the assumptions of the re-calculation. A p-value is not the probability that the null hypothesis is true, and it does not measure the size or clinical importance of an effect.

The registry’s own analysis is described as a one-sided t-test, with p <0.05. A one-sided test evaluates evidence in a specified direction, while a two-sided test allows departures in either direction. Which direction was prespecified, and the details of the sponsor’s test, are not reported in the registry record. A p-value alone does not provide an effect estimate or show how precise that estimate is.

In general, t-based comparisons rely on assumptions about independent observations and the behavior of the measurements or the sampling distribution of the mean. The registry record does not report diagnostic checks, distributional summaries beyond mean and standard deviation, or any methods for addressing departures from assumptions. The re-calculation is unadjusted and based on the posted summary statistics; it should not be mistaken for a reconstruction of every feature of the sponsor’s analysis.

What information does the interval add?

The re-calculated mean-difference interval extends from -0.707 to 0.310. A confidence interval is a range of values compatible with the observed data and the method’s assumptions. Its endpoints communicate uncertainty around the estimated difference, which a single point estimate cannot. Because the interval includes zero, it does not exclude no mean difference. The range also includes negative and positive values, so the data are compatible with differences in either direction at the interval’s stated level. The sponsor analysis has no interval bounds reported in the registry record.

What is and is not documented?

The registry record provides the group means, standard deviations, group sizes, a reported sponsor test, and enough summary information for a simple unadjusted re-calculation. It does not report the original statistical analysis plan, the rationale for choosing a one-sided test, model adjustments, missing-data methods, interim analyses, or the analysis population. Those details cannot be inferred from the endpoint summaries.

General statistical caution is especially relevant when an outcome is one among several measures. Multiple analyses can increase the chance of a seemingly notable result if many tests are considered, unless the analysis strategy accounts for that issue. Whether the sponsor used any multiplicity procedure is not reported in the registry record. The number of listed outcomes and analyses alone does not reveal which tests were planned, which were primary, or how results were selected.

10. Confidence Intervals and Precision

The re-calculated mean difference is -0.198 ion count, with a confidence interval from -0.707 to 0.310. The interval is important because it shows a range of plausible values under the calculation’s assumptions rather than presenting the point estimate as exact. It includes zero and extends to both negative and positive values. Thus, the estimate is compatible with a lower mean in the active-comparator group, little or no difference, or a higher mean in that group.

The interval is broad relative to the point estimate, indicating substantial uncertainty about the size and direction of the contrast. This does not mean that every value within the interval is equally likely, nor does it assign probabilities to particular treatment effects. It also does not establish equivalence: an equivalence conclusion would require a prespecified margin and a suitable analysis, neither of which is reported in the registry record.

The interval belongs to the unadjusted re-calculation from the posted group summaries. The sponsor’s reported analysis has no confidence-interval limits in the registry record, so its precision cannot be evaluated from the reported p-value alone. The record also does not provide a clinical interpretation threshold for the normalized ion count. Consequently, even a statistically precise contrast would not on its own indicate whether the difference is clinically important.

Precision is affected by the variability of measurements and the amount of information available, as well as by the analysis method. The group standard deviations indicate variation in measured counts, but the summary data do not show the individual values or their distribution. No baseline-adjusted estimate or other adjusted interval is reported.

11. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning figures are participants per arm, not total events. They use the observed mean difference as a reference and consider planning for 100%, 75%, or 50% of that observed effect. Under each effect assumption, the display gives the number of participants per arm for 80% power and for 90% power. The 80% power column describes a planning target with an 80% chance of detecting the specified effect under the stated model if that effect is present. The 90% column uses a higher planned chance of detection, which requires a larger planned sample under the same assumptions.

The stated assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is the stated fraction of the observed effect. The original trial’s design assumptions and sample-size justification are not reported in the registry record, so these figures should be read as planning illustrations rather than as the basis for the trial’s actual design. This is not observed power.

Smaller effects require more participants to distinguish them from random variation at the same significance and power targets. That is why the planning figures increase when the assumed effect is a smaller fraction of the observed effect. The calculation uses the observed effect as its starting point, but an observed estimate from a small study can be imprecise. The figures therefore do not guarantee that a future study will detect an effect or reproduce the observed estimate.

These sample-size calculations address a hypothetical future comparison under the listed assumptions. They do not alter the uncertainty in the current trial’s estimate, and they do not show what the enrolled study was designed to detect.

12. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = -0.198; pooled standard deviation σ = 0.812.
2. n = 2 × (1.96 + 0.842)² × 0.812² / -0.198² = 263.1, rounded up to 264 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- -0.198416                      # observed difference in means
sigma <- 0.812285481010217                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(763048)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 264 (80% power), 353 (90% power)
 75% of observed effect: n per arm = 468 (80% power), 627 (90% power)
 50% of observed effect: n per arm = 1,053 (80% power), 1,409 (90% power)
power.t.test: 264.1 
Simulated power at n = 264 per arm: 0.796 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

13. Analyses Reported by the Sponsor

The registry reports the sponsor’s analysis as a one-sided t-test comparing the two groups, with p <0.05. It does not report the sponsor’s test statistic, confidence-interval limits, detailed model, or the direction specified for the one-sided test. The sponsor result is therefore not fully reconstructable from the registry record.

The unadjusted re-calculation from the posted summary statistics gives a mean difference of -0.198, a confidence interval from -0.707 to 0.310, and p = 0.445. It is a simple two-sided re-calculation, not a substitute for the sponsor’s complete analysis. The two results should not be treated as interchangeable: the registry describes different test directionality, and the analysis details needed to determine how the results relate are not reported.

The re-calculation is useful for making the posted means and their uncertainty easier to interpret, but it cannot establish what methods the sponsor applied beyond the brief registry description. The registry record does not explain the relation between the reported one-sided result and the re-calculated two-sided result. Nor does it provide enough information to attribute the difference to a particular analytic choice. The reported sponsor result and the re-calculation should therefore be read with their distinct descriptions and limitations in view.

14. Multiplicity, Missing Data and Analysis Populations

The registry lists 10 outcome measures and 9 analyses. It identifies cadaverine as the primary endpoint, but additional outcome-level results are not posted. No secondary results are posted. The record does not provide a complete list of which outcomes were analyzed, which analyses correspond to the listed outcomes, or whether the primary endpoint and analysis were selected before results were examined.

Multiplicity refers, in general, to the statistical issue that testing several outcomes or comparisons can increase the chance of obtaining a nominally notable result by chance alone. Researchers may address this through prespecification, adjustment procedures, or a hierarchy of outcomes and tests. Which, if any, approach was used is not reported in the registry record. The listed outcome and analysis counts are not enough to determine whether or how multiplicity was handled.

Information about missing observations is also not reported. The registry does not state whether all randomized participants contributed endpoint measurements, how unavailable samples were handled, or whether any values were excluded. The group counts for the reported outcome are listed, but no participant-flow account or missing-data analysis is described. It is therefore not possible to assess whether missingness could have affected the comparison.

The analysis population is not reported in sufficient detail. The record does not say whether the analysis followed an intention-to-treat approach, included only participants with usable samples, or used another population definition. The randomization description alone does not identify which participants entered the primary analysis. There is also no reported analysis plan specifying adjustment variables or handling of protocol deviations. These gaps limit assessment of the completeness and reproducibility of the primary result.

15. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

The serious adverse event figures list 0 events in the control group, with 21 at risk, and 0 events in the active-comparator group, with 18 at risk. The figures for other adverse events also list 0 events in both groups, with the same group-specific at-risk counts. These are the registry’s posted event counts; they should not be interpreted as proof that no adverse events of any kind occurred outside the record’s definitions or reporting period.

The registry record does not provide definitions for serious or other adverse events, details about the period during which events were collected, or information on how events were assessed. Death counts are not reported in the record; they should not be read as zero. The listed counts describe what is entered in the registry, not a complete characterization of safety or a comparison capable of establishing differences in risk.

With limited event information and no further safety-analysis details, the figures are descriptive. They do not establish that either toothpaste is safer, and they do not support individual medical decisions.

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

The posted results describe mean cadaverine ion counts after 6 weeks of study-treatment use. The fluoride-control group mean was 1.119, and the fluoride/triclosan/copolymer group mean was 0.921. The unadjusted re-calculation estimates a lower mean in the active-comparator group, with a mean difference of -0.198. Its interval, from -0.707 to 0.310, includes zero and allows values in either direction.

These findings are about one quantitative biomarker, not a direct measure of symptoms, disease progression, or a clinical diagnosis. The endpoint description says a lower count may indicate less inflammation, but it does not give a threshold for meaningful change or establish how the count relates to clinical outcomes. The average also cannot describe whether participants’ individual values changed consistently or whether a subset of measurements shaped the group result.

The registry’s sponsor analysis reports a one-sided t-test with p <0.05, while the unadjusted re-calculation gives p = 0.445. The record does not explain the relationship between these results or provide enough sponsor-analysis detail to reproduce that result. They should be understood as differently described analyses rather than combined into a single conclusion.

The evidence therefore has important limits: the contrast is uncertain in the unadjusted re-calculation, several analysis details are not reported, and the endpoint is a laboratory marker. No conclusion about clinical benefit, equivalence, or absence of an effect follows from these summaries alone. The results characterize the posted data and the limits of what those data can establish.

17. Why This Trial Matters Statistically

This study is statistically interesting because it illustrates the difference between a biomarker measurement and a clinical endpoint. Cadaverine is a normalized ion count from a mass-spectral instrument, and the registry describes it as associated with inflammation. That makes it possible to compare a continuous laboratory measure between randomized groups, but it does not make the measure a direct account of patient experience or clinical disease.

The trial also illustrates why a point estimate should be read alongside its uncertainty. The observed group means differ, and the re-calculated mean difference has a direction and a numerical size. Yet the interval extends across zero and into both directions. Reporting only the direction of the means would leave out the uncertainty that matters for interpretation.

There is also a useful distinction between a sponsor-reported test and an unadjusted re-calculation. The registry reports a one-sided result, while the re-calculation is a two-sided summary-statistic comparison. Without the fuller methods, readers cannot determine exactly how those analyses correspond. This highlights the importance of clear reporting of test direction, analysis population, missing data, and confidence intervals.

Finally, the record lists multiple outcomes and analyses but posts limited results. That makes it difficult to evaluate multiplicity and completeness. These are general statistical concerns rather than evidence that a particular problem occurred. The record demonstrates how a randomized design can coexist with important uncertainty about measurement interpretation, analysis details, and the scope of conclusions.

18. 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 registry record gives limited information about the study population beyond the summary description of patients with gingivitis and/or periodontitis. Detailed eligibility criteria, baseline characteristics, disease severity, and recruitment setting are not reported. This makes it difficult to judge how representative the enrolled participants were or to understand the clinical context of the biomarker values.

The outcome is a normalized mass-spectral ion count, but the record does not explain the normalization procedure, assay reproducibility, measurement limits, or clinical meaning of specific values. It states that a lower count may indicate less inflammation; it does not report a threshold for meaningful change or validate the marker against a clinical outcome in this study. Baseline endpoint measurements and change-from-baseline results are not reported.

The record does not explain who was masked, the randomization process, allocation concealment, treatment adherence, or handling of protocol deviations. It also does not report participant flow, missing endpoint measurements, the primary analysis population, or model adjustments. These are trial-specific reporting gaps that prevent fuller evaluation of the randomized comparison.

The analysis reporting is incomplete. The sponsor result is described only as a one-sided t-test with p <0.05, without an estimate or interval. The unadjusted re-calculation is based on posted group summaries and does not recover unreported analysis details. Although 10 outcome measures and 9 analyses are listed, additional results and multiplicity procedures are not reported. The safety information likewise lacks definitions and event-collection details. Together, these limitations constrain interpretation of both the endpoint comparison and the broader study record.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Collection of Gingival Crevicular Fluid From Periodontitis Patients — NCT00763048.
• Registry record last updated 2011-06-14; status last verified 2011-06.
• Results first posted 2011-03-01.
• 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.

22. Record Summary

The registry record describes a completed, randomized, single-masked, parallel-group Phase 3 study sponsored by Colgate Palmolive, with enrollment of 39. It compared fluoride toothpaste with fluoride/triclosan/copolymer toothpaste. The primary endpoint was the mean normalized cadaverine ion count in gingival crevicular fluid after 6 weeks of treatment use.

The observed means were 1.119 for control and 0.921 for the active comparator. The unadjusted re-calculation reports a mean difference of -0.198, with an interval from -0.707 to 0.310 and p = 0.445. The sponsor analysis is reported as a one-sided t-test with p <0.05, but its fuller details are not reported in the registry record.

Interpretation is limited by missing information about the analysis population, missing data, multiplicity, and the clinical meaning of the biomarker. The posted results describe a laboratory measure and its uncertain group contrast; they do not establish a clinical effect.