A Study of Flurbiprofen 8.75 mg Lozenge in Patients With Pharyngitis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of Flurbiprofen 8.75 mg Lozenge in Patients With Pharyngitis (NCT01048866), a randomized phase 3 study. The primary outcome is Time Weighted Sum of Pain Intensity Differences (SPID) in Sore Throat Pain Intensity Scale (STPIS) Over the 24 Hours Post-baseline (STPIS SPID24).

2. Clinical Question

The study addressed painful pharyngitis and the registry describes its aims as assessing analgesic efficacy relative to the vehicle lozenge and safety during treatment. The primary measure drew on participants’ repeated pain-intensity ratings over the period from baseline to 24 hours after dosing. The reported comparison is between group averages on that measure; it is not a direct count of participants whose symptoms improved.

The registry’s primary analysis population was restricted to participants who took the first full dose and had the required follow-up pain scores. Four participants, two in each group, did not have sufficient 24-hour data for inclusion in that primary analysis. The registry describes an intent-to-treat population with these additional endpoint-specific eligibility conditions. The reasons for the short follow-up beyond the stated evaluability rule are not reported in the registry record.

The central statistical question is whether the observed difference in average pain-intensity summaries is compatible with no group difference under the reported analysis, and how uncertain the estimated contrast is. This comparison alone does not establish the clinical importance of a difference or answer every question about pharyngitis care.

3. Trial Design

The registry describes a Phase 3, randomized, double-blind, parallel-group, placebo-controlled study. Its official title also characterizes it as repeat-dose and multi-centre. The study was designed to compare flurbiprofen 8.75 mg lozenges with their vehicle control lozenges in patients with painful pharyngitis. The registry lists an enrollment of 198 and marks the study as completed.

The listed study period began in 2009-11, with primary completion in 2011-03. Reckitt Benckiser LLC is named as sponsor. These record details describe the registered study, rather than supplying additional evidence about how participants were recruited or how the study was run at each site.

Random allocation is intended, in general, to reduce systematic differences between treatment groups at baseline. Double masking is intended, in general, to limit the influence of treatment knowledge on participants’ reports and study assessments. A parallel design assigns participants to a group for the comparison rather than comparing each participant under both study conditions. These are general explanations of the design terms; the registry record does not report the sequence-generation or concealment procedures, how masking was maintained, or whether masking was assessed.

The intervention list includes placebo, flurbiprofen, and acetaminophen 650mg. The primary results identify only the flurbiprofen lozenge and placebo lozenge groups. The role of acetaminophen in the study, including whether it was used as rescue medication or in another way, is not reported in the registry record. The incomplete arm descriptions do not provide a basis for reconstructing any further treatment instructions.

Enrollment and the primary analysis population are not identical concepts. The primary endpoint description specifies that participants needed the first full dose and adequate 24-hour pain-intensity data to be included in that analysis. The results therefore concern the registry-defined evaluable population for this endpoint, not automatically everyone enrolled.

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: Three hundred and thirty-six patients were screened.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Intent to treat population of participants who took the first full dose of medication, and had STPIS values recorded \>= 21 hours post initial dose. Four participants (2 Flurbiprofen, 2 placebo) did not have sufficient 24 hour data to be included in the primary analysis of the primary endpoint.

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 Time Weighted Sum of Pain Intensity Differences (SPID) in Sore Throat Pain Intensity Scale (STPIS) Over the 24 Hours Post-baseline (STPIS SPID24), measured over baseline (pre-dose), post-dose - hourly up to 24 hours.

The primary endpoint is the time-weighted sum of pain-intensity differences over 24 hours after baseline, reported as STPIS SPID24. Participants completed the Sore Throat Pain Intensity Scale on a 100-mm visual analog scale: 0-mm indicates no pain upon swallowing and 100-mm indicates severe pain. SPID24 combines changes from baseline over the scheduled follow-up, with the time between assessments reflected in the sum. It is therefore a summary of the pain-intensity trajectory over the period, rather than a single rating at the end.

Repeated measurements can capture how symptoms change over time and can distinguish a sustained change from a rating taken at only one moment. That is the general statistical reason a time-weighted summary may be used for a symptom endpoint. Its interpretation depends on the scale’s coding and the calculation used. The registry reports negative group means and a negative contrast; these values should be read on the posted SPID24 scale, not mistaken for raw pain ratings on the 0-mm to 100-mm scale.

The record gives the theoretical range as -104412 to 27588, based on the mean baseline STPIS and specified patterns of pain. It also describes handling of rescue medication and missing intermediate ratings. A last recorded time point before 21 hours made a participant not evaluable for this primary analysis. After rescue medication, subsequent 24-hour STPIS values were assigned the baseline value; some missing scores were interpolated linearly when earlier and later scores were available.

7. Results Posted in the Registry

• Flurbiprofen 8.75 mg Lozenge: mean -522.90 versus -326.50 on control; difference -196.40 (95% CI -318.31 to -74.49), standardized effect d = -0.45; z-test p = 0.002.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Time Weighted Summed Differences in Difficulty Swallowing Scale (DSS) During the Initial 2 Hours From Baseline

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Flurbiprofen 8.75 mg Lozenge: mean -37.10 versus -18.10 on control; difference -19.00 (95% CI -28.69 to -9.31), standardized effect d = -0.54; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Flurbiprofen 8.75 mg Lozenge than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Time Weighted Summed Differences in Difficulty Swallowing Scale (DSS) During the Initial 24 Hours From Baseline

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Flurbiprofen 8.75 mg Lozenge: mean -555.50 versus -379.40 on control; difference -176.10 (95% CI -302.06 to -50.14), standardized effect d = -0.39; z-test p = 0.006.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Flurbiprofen 8.75 mg Lozenge than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Time Weighted Summed Differences in Swollen Throat Scale (SwoTS) During the Initial 2 Hours From Baseline

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Flurbiprofen 8.75 mg Lozenge: mean -35.50 versus -16.90 on control; difference -18.60 (95% CI -28.15 to -9.05), standardized effect d = -0.54; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Flurbiprofen 8.75 mg Lozenge than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The posted secondary measures span related aspects of sore-throat symptoms and assessment: difficulty swallowing, swollen throat, pain intensity, participant-reported relief, and investigator assessment. The registry lists 18 secondary outcomes. The available detailed secondary comparisons show a broadly similar direction in the reported numerical contrasts for the symptom-summary measures described, with group differences on their own scales. These outcomes complement the primary pain-intensity summary by addressing other time windows or related symptoms.

Secondary outcomes can help assess whether the primary pattern is accompanied by changes in related measures, but they do not answer the same question as the primary endpoint. Different scales, time windows, and methods of assessment can produce results that are not directly comparable. An outcome covering the initial hours, for example, summarizes a different period from one covering 24 hours. The full details of all listed secondary measures are not reported in the registry record, so a complete assessment of their estimates and precision is not possible from those details alone.

Unadjusted secondary p-values carry less weight when many outcomes are examined. In general, testing multiple hypotheses increases the chance that some results appear notable by chance, even if no adjustment is made. A multiplicity adjustment can address that issue, but the registry record does not report whether a correction or hierarchy was used for the secondary analyses. Therefore, secondary p-values should be treated as supportive or exploratory evidence rather than as independent confirmation of the primary finding.

The relationship between primary and secondary results is best judged by considering the outcome definitions, uncertainty intervals, and analysis plan together. A consistent direction across related measures may be informative, but it does not remove concerns about multiple testing or turn secondary outcomes into a substitute for the primary endpoint. No conclusions about the unreported secondary estimates should be inferred from the broad pattern alone.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Time Weighted Summed Differences in Difficulty Swallowing Scale (DSS) During the Initial 2 Hours From Baseline
• Time Weighted Summed Differences in Difficulty Swallowing Scale (DSS) During the Initial 24 Hours From Baseline
• Time Weighted Summed Differences in Swollen Throat Scale (SwoTS) During the Initial 2 Hours From Baseline
• Time Weighted Summed Differences in Swollen Throat Scale (SwoTS) During the Initial 24 Hours From Baseline
• Time Weighted Sum of Pain Intensity Differences (SPID) in Sore Throat Pain Intensity Scale (STPIS) Over the 2 Hours Post-baseline (STPIS SPID2)
• Time Weighted Sum of Pain Intensity Differences (SPID) in Sore Throat Pain Intensity Scale Over 24 Hours Post-baseline (STPIS SPID24) For Participants With Baseline Practitioner's Assessment of Pharyngeal Inflammation (PAIN) of Moderate or Severe
• Sore Throat Relief As Reported by Participants 2 Hours After Initial Dose
• Investigators' Clinical Assessment (CLIN) Of Study Medication as a Treatment for Sore Throat at 24 Hours After Initial Dose

9. Statistical Interpretation

10. Statistical Methods Explained

The primary endpoint is continuous: it is a numerical summary of pain-intensity differences over time. For continuous outcomes, comparing group means is one way to summarize the average contrast. The posted group means and standard deviations describe each group’s average and the spread of individual values around that average. A mean difference then expresses the flurbiprofen group average relative to the placebo group average on the same endpoint scale.

The recalculated primary contrast is an unadjusted mean difference of -196.4. “Unadjusted” means this calculation uses the posted summary statistics and does not account for covariates or other features of an adjusted model. It is not necessarily the same quantity as an analysis that accounts for model terms or produces least-square means. The registry’s own analysis is reported separately as an ANOVA least-square means difference.

In general, standard deviation describes variability among observations, whereas a standard error describes uncertainty in an estimated quantity such as a mean. Greater variation among participants tends to make a group mean less precise when other things are equal. The confidence interval for a mean difference combines information about the estimated contrast and its sampling uncertainty. The posted standard deviations and group sizes are sufficient for the simple recalculation presented, but they do not reproduce every potential detail of the registry’s ANOVA model.

In general, analysis of variance (ANOVA) compares group means through a statistical model. Least-square means are model-based group averages, and their difference can reflect the model’s specification. The registry reports ANOVA and a least-square means difference but does not report the model covariates, any transformations, or other model details in the record. Accordingly, the reported estimate should be distinguished from the simple unadjusted recalculation rather than treated as a duplicate calculation.

A p-value is calculated under a statistical model and a null hypothesis, usually a hypothesis of no group difference for the contrast. In general, a small p-value indicates that data at least as discrepant as those observed would be less expected under that null model. It does not state the probability that a result is due to chance, show the probability that one treatment is better, or determine whether a difference matters to patients. The primary registry analysis reports p = 0.002; the unadjusted recalculation reports p = 0.002. Both are tied to the primary contrast, but they arise from different approaches.

The endpoint is constructed from repeated measurements. General repeated-measures summaries can use information across scheduled times, but their interpretation depends on which time points are included and how unobserved values are handled. The record specifies hourly assessment through 24 hours, excludes participants whose last recorded time was before 21 hours from primary evaluability, assigns baseline values to post-rescue scores, and uses linear interpolation for certain missing intermediate scores. These rules are part of the endpoint definition and affect the data represented by the summary.

Linear interpolation is a general method that fills a missing value by assuming a straight-line change between an earlier and a later observed value. This is a simplifying assumption, not a direct observation of the missing score. The registry describes its use when a missing STPIS score had non-missing scores at earlier and later assessments, with the missing assessment treated at its nominal time. The record does not report sensitivity analyses assessing alternative missing-data assumptions.

The record identifies the primary population as participants who took the first full dose and had STPIS values recorded at least 21 hours after the initial dose. This is an endpoint-specific analysis population. Although the record uses the term intent to treat, the stated eligibility conditions mean the primary result should be understood according to that explicit definition. The reasons that four participants did not meet the 24-hour data requirement are not reported in the registry record.

Confidence intervals are useful because they show both the direction and the precision of an estimate on the outcome scale. A narrower interval indicates less uncertainty in the estimate under the model assumptions; a wider one indicates more. The primary recalculated interval is from -318.3 to -74.49. It does not account for every uncertainty in study design, missing-data handling, outcome measurement, or generalizability. Statistical uncertainty and those broader design limitations are related but not interchangeable.

Standard deviation pooling is used in the recalculation to summarize group variability for the standardized effect estimate and planning illustrations. In general, a pooled standard deviation combines variability across groups, and standardized effects express a mean contrast relative to that variability. The recalculated standardized effect for the primary comparison is -0.452. Such a standardized statistic can help describe the contrast relative to observed spread, but it does not preserve the original endpoint units and is not itself a measure of clinical importance.

The planning illustrations use a normal approximation and equal allocation assumptions, as well as a two-sided significance level of 0.05. They scale the observed contrast to consider smaller fractions of that effect. This is a general planning exercise based on the observed result, not information about how the completed trial was designed. The registry’s original sample-size justification and statistical plan are not reported in the registry record.

Over-dispersion is a general concern when data vary more than a model assumes, because standard errors can be too small and intervals too narrow if the extra variation is ignored. For this continuous endpoint, the posted standard deviations describe observed spread, but the record does not report an over-dispersion assessment. The concept is useful when judging models generally; it should not be taken as evidence that over-dispersion occurred in this study.

11. Confidence Intervals and Precision

The unadjusted 95% confidence interval for the primary mean difference runs from -318.3 to -74.49. It describes a range of contrasts compatible with the observed data and the calculation’s assumptions. Since both endpoints are negative, the interval does not include zero. The reported ANOVA interval, from -321.0 to -72.2, also lies on the negative side of zero.

Precision is about uncertainty in the estimated group contrast, not the consistency of individual responses. The interval is not a prediction interval for participant outcomes, and it does not imply that every participant experienced a change resembling the group average. Nor does its 95% label mean there is a 95% probability that this particular fixed interval contains the true effect. In general, the confidence level describes the long-run behavior of the interval procedure under its assumptions.

The endpoints of the interval indicate that the estimated contrast is not exact: a range of magnitudes remains compatible with the analysis. The estimate’s direction is more clearly determined than its precise size under this interval, but the record does not identify a threshold for clinical importance against which to judge the range. Statistical precision alone cannot decide whether the magnitude is meaningful to patients.

The reported ANOVA estimate and confidence interval are close to the unadjusted recalculation, which gives a useful consistency check on the summary-statistic comparison. The two intervals should still be understood as arising from different analyses. Both depend on statistical assumptions and the endpoint data included; neither expresses uncertainty from every possible source, such as missing-data assumptions or generalizability to other settings.

Confidence intervals are generally more informative than a p-value alone because they display a range of plausible effect sizes on the outcome scale. The registry’s p-values describe evidence against a no-difference hypothesis under the analyses, while the intervals help convey the estimated direction and precision. Neither statistic establishes clinical importance or a causal explanation by itself.

12. 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 for 80% power and participants per arm for 90% power. They are not total event counts: the event-count convention applies to time-to-event outcomes, whereas this endpoint is continuous. The figures are illustrations calculated from the observed effect, not the original trial’s sample-size design. The trial’s own design assumptions and sample-size justification are not reported in the registry record.

The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each scenario considers an effect that is a stated fraction of the observed effect. The 80% power column represents the planning sample size intended, under those assumptions, to give an 80% chance of detecting the specified effect if that effect were the true one. The 90% power column represents the corresponding planning target for a 90% chance.

As the assumed effect becomes smaller, more participants are needed to detect it at the same power and significance level. That relationship is a general property of statistical planning: a smaller contrast is harder to distinguish from random variation. The calculations do not establish what effect would occur in a future study, and their assumptions may not match a future study’s design or analysis.

These sample-size figures are planning illustrations, not observed power. Observed power is not an appropriate way to interpret the completed result; the observed estimate and its confidence interval provide the relevant description of effect and precision. The planning figures should be read as conditional scenarios based on the observed effect and stated assumptions, not as evidence that the completed trial was adequately or inadequately designed.

13. 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 = -196.400; pooled standard deviation σ = 434.938.
2. n = 2 × (1.96 + 0.842)² × 434.938² / -196.400² = 77.0, rounded up to 77 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 <- -196.4                      # observed difference in means
sigma <- 434.937594981309                     # 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(1048866)
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 = 77 (80% power), 104 (90% power)
 75% of observed effect: n per arm = 137 (80% power), 184 (90% power)
 50% of observed effect: n per arm = 308 (80% power), 413 (90% power)
power.t.test: 78 
Simulated power at n = 77 per arm: 0.792 (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.

14. Analyses Reported by the Sponsor

The registry reports the sponsor’s analysis as an ANOVA comparison of least-square means. For the primary endpoint, that analysis estimated a difference of -196.6, with a 95% confidence interval from -321.0 to -72.2 and p = 0.002. The simple unadjusted recalculation from the posted group means gives a difference of -196.4, with a 95% confidence interval from -318.3 to -74.49 and an unadjusted p-value of 0.002.

The estimates are close, but they are not identical. The recalculation is based on posted summary statistics and is unadjusted; the reported result is an ANOVA least-square means difference. The record does not describe the model details needed to reproduce that reported analysis fully. The close estimates indicate that the simple comparison broadly tracks the reported result, while the sponsor’s analysis remains the registry’s stated analysis.

The registry does not flag a discrepancy between the reported and recalculated results. The distinction remains important: a summary-statistic recalculation is an independent check, not a replacement for the analysis reported in the record. The p-values should also be read in the context of their respective analyses and not as measures of clinical importance.

15. Multiplicity, Missing Data and Analysis Populations

The registry identifies one primary endpoint and lists 18 secondary outcomes. It reports 18 analyses overall. When a study evaluates multiple outcomes or multiple comparisons, the chance of obtaining at least one apparently unusual result can increase. In general, statistical methods can account for multiplicity by adjusting significance thresholds or by specifying a testing hierarchy in advance. Whether such a method was used for the secondary outcomes is not reported in the registry record.

The record does not report the original statistical analysis plan, an outcome hierarchy, interim analyses, or a multiplicity strategy. Those details should not be inferred from the number of posted outcomes or from the reported p-values. Without them, the primary result is the clearest basis for interpretation, while secondary findings require caution because their relationship to any prespecified testing strategy is unclear.

Missing data are addressed in part through endpoint-specific rules. Participants whose last recorded time point was before 21 hours were considered not evaluable for the primary endpoint. The record states that four participants, two in each group, did not have sufficient 24-hour data for inclusion. It also says that after rescue medication, all subsequent STPIS values in the 24-hour interval were assigned the baseline value. Certain missing intermediate scores were imputed by linear interpolation when scores at earlier and later assessments were available.

These rules make the analysis more explicit, but they do not show what would have happened under different assumptions about missing observations or rescue medication. The reasons for insufficient follow-up and the results of sensitivity analyses are not reported in the registry record. As a general matter, conclusions can be sensitive to missing-data mechanisms and imputation choices, especially when the unobserved values may differ systematically from observed ones.

The population is described as intent to treat, but the stated primary-analysis definition also requires taking the first full dose and having adequate recorded follow-up. The exact effect of these eligibility conditions cannot be determined from the posted summary. The record does not report additional analyses using a different analysis population to assess the impact of these restrictions.

16. 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.

Most frequent non-serious adverse event terms

The registry reports serious adverse events for the flurbiprofen group among 101 participants at risk and for the placebo group among 97 participants at risk. It records 0 serious adverse events affecting participants in either group. This means the posted record contains no serious adverse events for these groups during the recorded safety assessment; it does not establish that serious events are impossible or that the groups have identical safety profiles.

The record also lists other, non-serious adverse events affecting 34 participants in the flurbiprofen group and 28 in the placebo group. These are separate from the serious-event figures and should not be combined with them. The registry does not give the serious-event terms because the listed top serious terms are empty. Death figures are not reported in the registry record.

Adverse-event counts are descriptive. Without further detail on event definitions, timing, severity, relatedness, and follow-up completeness, they cannot by themselves establish a causal relationship or provide a comprehensive safety comparison. A zero count in a study does not rule out uncommon events or events occurring outside the recorded observation period.

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

The primary result describes a difference in average 24-hour time-weighted sore-throat pain-intensity summaries between the flurbiprofen and placebo groups in the registry-defined primary analysis population. The recalculated contrast is -196.4, and the reported ANOVA contrast is -196.6. Their confidence intervals lie on the negative side of zero. These statistics describe the observed comparison and its uncertainty under the respective analyses.

The result does not, by itself, establish how important the difference is to patients. The registry does not report a minimal clinically important difference for SPID24, and the time-weighted units are not the same as millimeters on the original pain-intensity scale. Statistical evidence against a no-difference hypothesis is not equivalent to evidence that an effect is large, noticeable, or worthwhile in every clinical context.

Nor does the result show that every participant had a similar response. Group means summarize averages, while participant experiences can vary. The posted standard deviations indicate spread in the outcome, but the registry does not provide individual-level distributions in the registry record used for this analysis. The mean contrast should not be read as a prediction for any one person.

The findings also do not answer all questions about safety, repeat dosing beyond the instructions described, or comparisons with other treatments. The primary posted comparison is flurbiprofen against placebo. Although acetaminophen 650mg is listed among interventions, its role and any corresponding result are not reported in the registry record. The serious adverse event counts are limited to the posted record and do not establish that no risk exists.

Finally, this is one completed Phase 3 study in patients with pharyngitis. Its results do not automatically generalize to other populations, settings, or treatment regimens. The record does not provide enough information to judge how representative the participants were or to determine the influence of all protocol and analysis choices. The appropriate conclusion is limited to what the registered comparison and posted endpoint results describe.

18. Why This Trial Matters Statistically

This study illustrates why a clinical endpoint’s construction matters. The primary outcome is not a single symptom score but a time-weighted summary of changes from baseline across repeated pain assessments. Such a summary incorporates the timing of measurements, so its units and sign need to be interpreted according to the endpoint definition rather than as raw scale points.

It also demonstrates the difference between a simple contrast calculated from group summaries and a model-based contrast reported by a sponsor. The unadjusted recalculation and ANOVA least-square means difference are close, but they are distinct estimands and methods. Reporting both helps readers see whether a basic summary-statistic comparison is broadly consistent with the registry’s analysis without pretending that it reproduces the model.

Precision adds information that a p-value alone cannot supply. The confidence interval describes a range of plausible contrasts under the analysis assumptions, while the p-value addresses compatibility with a no-difference hypothesis. Neither tells readers whether the difference is clinically important; that judgment needs an interpretable outcome scale and a meaningful threshold, which is not reported for this endpoint.

The record also highlights how analysis populations and missing-data rules shape a result. Eligibility for the primary analysis depended on having adequate follow-up, and the registry specified how rescue medication and some missing intermediate scores were handled. Those details are not administrative footnotes: they define the data summarized and the assumptions behind the estimate.

Finally, the study lists many secondary outcomes, making multiplicity relevant. A broad family of outcomes can provide context, but unadjusted tests should not be treated as many independent confirmations. Clear reporting of the testing strategy, missing-data sensitivity analyses, and clinical-importance thresholds would help readers assess the evidence beyond the primary statistical contrast.

19. 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 the primary endpoint definition and several data-handling rules, but does not report the full statistical analysis plan, detailed ANOVA specification, or original sample-size justification. The reported least-square means analysis therefore cannot be reproduced completely from the record details. The recalculation is necessarily an unadjusted comparison based on posted group summaries.

Primary-endpoint evaluability required recorded STPIS data at least 21 hours after initial dosing. Four participants were excluded from that endpoint analysis because they lacked sufficient 24-hour data. The reasons for missing follow-up are not reported in the registry record. No sensitivity analyses for alternative missing-data assumptions are reported in the registry record for this analysis.

The record assigns baseline values to post-rescue pain scores and uses linear interpolation for certain missing intermediate assessments. These methods make the calculation operational, but they embed assumptions about symptom values after rescue medication and between observed assessments. Their influence on the result cannot be evaluated without additional analyses.

The registry lists 18 secondary outcomes, but does not report a multiplicity strategy. This limits the strength of any interpretation that relies on multiple secondary p-values. In addition, the role of acetaminophen 650mg is not reported in the registry record, despite its appearance in the intervention list. No further comparison involving it should be inferred.

The clinical meaning of the primary contrast is also difficult to judge because the record does not report a threshold for a clinically important SPID24 difference. The summarized group results do not describe individual response distributions, and the safety counts alone cannot rule out uncommon events. Further limitations specific to representativeness, site conduct, or masking cannot be assessed from the registry entries.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Study of Flurbiprofen 8.75 mg Lozenge in Patients With Pharyngitis — NCT01048866.
• Registry record last updated 2017-09-21; status last verified 2017-08.
• Results first posted 2017-09-21.
• 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.

23. Record Summary

NCT01048866 is a completed Phase 3, randomized, double-blind, parallel-group, placebo-controlled study of flurbiprofen 8.75 mg lozenges in patients with painful pharyngitis. Reckitt Benckiser LLC is listed as sponsor. The registry records enrollment of 198, a study start of 2009-11, and primary completion in 2011-03.

The primary endpoint was the time-weighted sum of sore-throat pain-intensity differences over 24 hours. The unadjusted recalculated mean difference was -196.4, with a 95% confidence interval from -318.3 to -74.49. The registry’s reported ANOVA estimate was -196.6, with a 95% confidence interval from -321.0 to -72.2. The analyses are close but use different methods.

The primary analysis required adequate 24-hour pain data, and the registry specifies rules for rescue medication and some missing scores. The record lists 18 secondary outcomes but does not report a multiplicity strategy. The posted results should be interpreted as evidence about this registered comparison and endpoint, not as a complete account of clinical importance, individual responses, or all safety questions.