Conventional Prophylactic Oral Dexamethasone vs Short-course IV Dexamethasone in Paclitaxel Hypersensitivity: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Conventional Prophylactic Oral Dexamethasone vs Short-course IV Dexamethasone in Paclitaxel Hypersensitivity (NCT03598426), a randomized phase 3 study. The primary outcome is Number Participants With Any-grade Paclitaxel-associated Hypersensitivity Reactions During the First Paclitaxel Infusion.

2. Clinical Question

3. Trial Design

This is a phase 3, randomized, open-label, parallel-group trial registered as NCT03598426. The registry lists enrollment of 90 and describes the study as prospective and single-center. Participants were assigned to conventional, short-course, or combined dexamethasone premedication approaches. The brief summary says block randomization was used. Further details on sequence generation, allocation concealment, and implementation of randomization are not reported in the registry record.

The conventional approach involved oral dexamethasone at home 12 hours and 6 hours before paclitaxel infusion, with intravenous diphenhydramine and famotidine at the clinic 30 minutes before infusion. The short-course approach used intravenous dexamethasone with those two antagonists 30 minutes before infusion. The combined approach included oral dexamethasone 12 hours before treatment and additional intravenous dexamethasone with the antagonists 30 minutes before infusion. The registry's combined-arm description ends mid-sentence, although the preceding regimen details are listed.

Masking is listed as none, so the trial was open-label. The registry lists Loma Linda University as sponsor, a start date of 2018-08-08, primary completion on 2025-01-07, and status as completed. The posted primary outcome is a count of participants with any-grade paclitaxel-associated hypersensitivity during the first infusion, assessed from infusion start through at least three hours after infusion. The registry describes a secondary outcome concerning participants who discontinued treatment during the first infusion, but results for that outcome are not reported in the registry record. Details of the original sample-size justification and the trial's full 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.

Population: All females patients with gynecologic malignancy over the age of 18 years seen at Loma Linda University Cancer Center.

6. Primary Endpoint

The registry lists the primary outcome as Number Participants With Any-grade Paclitaxel-associated Hypersensitivity Reactions During the First Paclitaxel Infusion, measured over Change between start of the paclitaxel infusion and at least three hours post-infusion at the first paclitaxel treatment.

The primary endpoint counts participants with any-grade paclitaxel-associated hypersensitivity reactions during the first paclitaxel infusion. The registry sets the observation window from the start of infusion to at least three hours after infusion. This is a participant-level event outcome: it records whether a participant experienced a reaction within the specified period, rather than measuring the number of reactions per participant or the time until a reaction.

The registry reports 2 participants with reactions in the conventional group, 8 in the short-course group, and 6 in the combined group. The corresponding group sizes are 29, 25, and 32 participants. The named primary contrast is short-course versus conventional; the combined-versus-conventional contrast is also posted. Those contrasts help express the observed difference between regimens while retaining the event counts as the underlying data.

A defined endpoint and observation window make the outcome comparable across the randomized groups. The any-grade definition includes reactions across grades rather than restricting the endpoint to severe reactions. The registry does not report the grade-by-grade breakdown, criteria used to classify reactions, or the number of reactions among participants who experienced an event. The endpoint therefore describes occurrence within the stated window, not severity, recurrence, or outcomes beyond that window.

7. Results Posted in the Registry

• Short-Course: 32.0% (8/25) versus 6.9% (2/29) on control; risk difference 25.1 percentage points (95% CI 4.6 to 45.6), risk ratio 4.64 (1.08–19.86); unadjusted two-proportion z-test p = 0.018.
• Combined: 18.8% (6/32) versus 6.9% (2/29) on control; risk difference 11.9 percentage points (95% CI -4.5 to 28.2), risk ratio 2.72 (0.60–12.42); unadjusted two-proportion z-test p = 0.171.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is binary at the participant level: each participant either had an any-grade hypersensitivity reaction during the specified observation window or did not. A count and group size describe that outcome directly. For example, the registry reports 2 events among 29 participants in the conventional group and 8 among 25 in the short-course group. The combined group has 6 events among 32 participants. These are observed counts, not estimates of reaction severity or of repeated-event frequency.

What does the risk difference describe?

A risk difference subtracts the event proportion in one group from the event proportion in another. It expresses the contrast on the absolute probability scale. For short-course versus conventional, the posted estimate is 0.251; for combined versus conventional, it is 0.118. A positive estimate in these stated comparisons means the observed event proportion was higher in the named treatment group than in the conventional group. The size of the estimate is uncertain, which is why the confidence interval matters alongside the point estimate.

What does a confidence interval add?

A confidence interval gives a range of values compatible with the data and the method used to construct it. It is not a guarantee that the true effect lies within the particular interval, nor does it express the range of individual participants' outcomes. In general, wider intervals indicate less precision. The primary contrast's posted 95% confidence interval is 0.046 to 0.456. The combined contrast's interval is -0.045 to 0.282. The former lies above zero; the latter includes zero. These patterns describe the intervals and should be considered together with the modest group counts.

How should the p-values be read?

A p-value describes how unusual data at least as discrepant as the observed data would be under a specified null model and its assumptions. It is not the probability that the null hypothesis is true, and it does not measure the clinical importance of the contrast. The registry posts unadjusted p-values of 0.018 for short-course versus conventional and 0.171 for combined versus conventional. The brief summary says p-values less than 0.05 would be considered statistically significant. The record does not report adjusted p-values, so these should be treated as unadjusted comparisons.

Why are ratio measures also listed?

Risk ratios and odds ratios express contrasts on relative scales rather than the absolute scale used by the risk difference. The posted short-course risk ratio is 4.64, with a confidence interval from 1.084 to 19.86; its odds ratio is 6.353, with a confidence interval from 1.203 to 33.55. For combined versus conventional, the risk ratio is 2.719, with a confidence interval from 0.595 to 12.42; the odds ratio is 3.115, with a confidence interval from 0.576 to 16.86. Relative measures can appear large when event proportions are low, and their intervals show substantial uncertainty. They do not replace the underlying counts or the absolute risk-difference interpretation.

What analysis details are available?

The brief summary says one-way analysis of variance would be used to assess differences among strategies and states a threshold of 0.05. However, the primary outcome is a participant count with an event, and the registry does not report the detailed statistical procedure used for the posted contrasts. It lists zero analyses, and no adjusted or model-based results are posted. It is therefore not possible to determine from the record whether the proposed analysis was used as written, whether any covariates were included, or how assumptions were assessed. Those details are not reported in the registry record.

In general, analysis methods should match the outcome type and account for the design and data structure. For a binary endpoint, methods can model event probabilities or compare event proportions, with uncertainty intervals that reflect the sample size. If event counts are sparse, estimates can be unstable and intervals can be wide; that is relevant when interpreting the posted counts without assuming a particular unreported method. The risk differences and intervals are available as posted summaries, but the record does not give enough detail to reconstruct the sponsor's full analysis.

10. Confidence Intervals and Precision

The point estimate alone does not show how precisely a group contrast is known. The posted 95% confidence interval for short-course versus conventional runs from 0.046 to 0.456 around a risk difference of 0.251. The interval is entirely above zero, but it covers a broad range of positive differences. The result is therefore not a precise estimate of the size of the contrast, even though the interval does not include zero.

For combined versus conventional, the risk difference is 0.118, and its 95% confidence interval runs from -0.045 to 0.282. This interval includes zero and also includes positive values. It is compatible with no difference as well as with differences favoring either group. That uncertainty should not be mistaken for evidence that the two approaches are equivalent.

The group counts are 29, 25, and 32, with 2, 8, and 6 participants experiencing the endpoint, respectively. In general, estimates based on limited event counts tend to be less precise than estimates supported by more information. The registry also posts broad intervals for the risk ratios and odds ratios. These intervals reinforce the need to focus on the range of plausible effects rather than treating any single point estimate as exact.

A confidence interval concerns uncertainty in the estimated group contrast under the analysis method; it does not represent the range of outcomes expected for an individual patient. The record does not report adjustment for multiple comparisons or additional analyses that might affect precision. The posted intervals should therefore be read as the intervals reported for the individual, unadjusted comparisons.

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 values are participants per arm, not total events. The endpoint is not a time-to-event outcome, so the planning figures are not event totals. The registry's original design assumptions and sample-size justification are not reported in the registry record. The planning illustration instead uses the observed effect and assumes a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. It examines effects stated as a fraction of the observed effect.

For short-course versus conventional, planning for 100% of the observed effect gives 38 participants per arm for 80% power and 51 participants per arm for 90% power. Planning for 75% of the observed effect gives 60 participants per arm for 80% power and 79 participants per arm for 90% power. Planning for 50% of the observed effect gives 113 participants per arm for 80% power and 151 participants per arm for 90% power.

For combined versus conventional, planning for 100% of the observed effect gives 124 participants per arm for 80% power and 166 participants per arm for 90% power. Planning for 75% of the observed effect gives 199 participants per arm for 80% power and 266 participants per arm for 90% power. Planning for 50% of the observed effect gives 397 participants per arm for 80% power and 530 participants per arm for 90% power.

In plain terms, the 80% column describes a design intended to detect the stated effect with 80% power under the listed assumptions; the 90% column uses a higher power target. Smaller effects require more participants per arm under the same planning framework. These values are planning illustrations, not observed power, and they should not be read as the trial's original target or as a guarantee of a future result.

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, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.069; treatment proportion p₁ = 0.320; average p̄ = 0.194.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.5597 = 1.0971.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5309 = 0.4468.
4. n = (1.0971 + 0.4468)² / (0.2510)² = 37.8, rounded up to 38 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
p0 <- 0.0689655172413793                          # control proportion
p1 <- 0.32                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^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 (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(3598426)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 38 (80% power), 51 (90% power)
 75% of observed effect: n per arm = 60 (80% power), 79 (90% power)
 50% of observed effect: n per arm = 113 (80% power), 151 (90% power)
power.prop.test: 37.8 
Simulated power at n = 38 per arm: 0.832 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-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. Multiplicity, Missing Data and Analysis Populations

The registry posts two contrasts against the conventional group: short-course versus conventional and combined versus conventional. Each has an unadjusted p-value. When several comparisons are considered, the chance of at least one apparently unusual result can increase. In general, multiplicity procedures address that issue by adjusting the interpretation or testing threshold. No multiplicity adjustment is reported in the registry record, and no adjusted p-values are posted. The p-values should therefore be read as the stated unadjusted comparisons, not as family-wise adjusted evidence.

The brief summary describes one-way analysis of variance and a p-value threshold of 0.05, but the detailed statistical analysis plan is not reported in the registry record. The registry lists zero analyses. It does not report whether the planned method was changed, how the binary outcome was analyzed, whether any covariates were used, or whether prespecified subgroup analyses were conducted. The analysis population definitions are also not reported in the registry record.

Missing outcome data can affect estimates if participants lack endpoint information, particularly when missingness differs between groups or relates to the outcome. The registry does not report the amount or handling of missing primary-outcome data, participant exclusions after randomization, protocol deviations, or an intention-to-treat analysis. It is not possible to determine from the posted record whether all randomized participants contributed to the primary endpoint or whether any imputation or sensitivity analyses were used. These omissions limit assessment of the robustness of the posted contrasts.

The secondary outcome concerning discontinuation during the first paclitaxel infusion is listed, but its results and analysis are not reported in the registry record. Definitions for the analysis populations and the handling of participants who discontinued or did not complete the observation period are likewise not reported. These gaps do not change the posted event counts, but they constrain evaluation of how completely the results reflect the randomized groups.

14. Safety Results

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

Deaths

Other (non-serious) adverse events

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

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: grade 3 dexamethasone related hypersensitivity reaction (3.1%).

The registry lists serious adverse events separately from other adverse events. For serious adverse events, the number affected is 0 in the conventional group, 0 in the short-course group, and 1 in the combined group. The corresponding numbers at risk are 29, 25, and 32. Deaths are listed as 0 in each group. The listed top serious term is “Grade 3 dexamethasone related hypersensitivity reaction,” with 3.1% shown.

These figures describe what the registry records during the reported safety collection; they do not by themselves establish whether a regimen caused an event. The record does not report the serious-adverse-event definition, the complete event narratives, or the timing and duration of each event. The listed serious-event figures should not be combined with the separate any-grade primary endpoint counts: they refer to different outcome categories. In general, zero recorded events in a group does not establish that an event is impossible or that there is no safety risk. The small group counts also limit how precisely uncommon events can be characterized.

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

The posted primary comparison shows a higher observed proportion of participants with any-grade hypersensitivity in the short-course group than in the conventional group during the first infusion observation window. Its risk-difference estimate is 0.251, with a 95% confidence interval from 0.046 to 0.456 and an unadjusted p-value of 0.018. The combined-versus-conventional estimate is 0.118, with a 95% confidence interval from -0.045 to 0.282 and an unadjusted p-value of 0.171. The latter interval includes zero and is compatible with a range of effects.

These results describe the participants and outcome recorded in this trial. They do not establish the effect of the regimens in every oncology setting, for all patients receiving paclitaxel, or beyond the first infusion. The trial is described as single-center and open-label, and the posted endpoint covers reactions during the infusion and at least three hours afterward. Outcomes outside that interval are not addressed by this primary measure.

The results do not establish that any regimen is clinically preferable, nor do they prove that a regimen prevents reactions or causes them. The study compares premedication approaches, but the posted record does not include a direct comparison of combined versus short-course treatment. The any-grade endpoint also does not show the severity distribution of reactions, and the secondary discontinuation outcome has no posted result.

Confidence intervals express uncertainty in the estimated contrasts, not certainty about an individual participant's outcome. The short-course interval is entirely above zero but remains broad; the combined interval includes zero and spans effects in either direction. The unadjusted p-values do not answer questions about multiplicity, unreported missing data, or the trial's full statistical plan. A careful interpretation keeps these limitations alongside the observed counts.

16. Why This Trial Matters Statistically

This trial illustrates why a clinical comparison should be read through both its event counts and its uncertainty estimates. A count of participants with an event is direct and interpretable, while a risk difference places the group contrast on an absolute scale. A confidence interval then indicates how much uncertainty remains around that contrast. Reporting only a point estimate or p-value would obscure important features of these results.

The short-course comparison has a posted risk difference of 0.251 and a 95% confidence interval from 0.046 to 0.456. The combined comparison has a risk difference of 0.118 and an interval from -0.045 to 0.282. Their different interval positions matter: one is entirely above zero and the other includes zero. Yet both intervals are broad, emphasizing that statistical evidence about a contrast and precision about its size are related but distinct questions.

The record also illustrates how multiple comparisons and analysis-plan transparency affect interpretation. Two active approaches are compared with the conventional group, but only unadjusted p-values are posted. The brief summary mentions one-way analysis of variance, while detailed methods for the binary endpoint are not reported. Without analysis-population definitions, missing-data handling, or a multiplicity plan, readers cannot fully evaluate the analysis pathway from randomized assignment to reported contrast.

Finally, the registry distinguishes a primary any-grade hypersensitivity endpoint from serious adverse events and a listed discontinuation outcome. Those endpoints answer different questions and should not be conflated. Keeping outcome definition, observation window, comparison, and uncertainty in view is central to statistical interpretation of a randomized clinical trial.

17. 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 describes a single-center, open-label trial. Lack of masking can affect outcome assessment or participant reporting when outcomes require clinical judgment, although the extent of any such influence is not reported. The record does not provide participant demographic or clinical characteristics, detailed eligibility criteria, or information about how representative the enrolled population is of other paclitaxel-treated settings. The official title names breast and gynecologic oncology patients, while the brief summary describes gynecologic oncology patients.

The primary outcome is limited to any-grade hypersensitivity during the first infusion through at least three hours afterward. The registry does not post reaction grades by group, the criteria used to classify an event, or outcomes from later infusions. As a result, the endpoint does not describe the full course of treatment, recurrence, or severity distribution. The listed secondary discontinuation outcome has no posted result.

Although the brief summary mentions block randomization, allocation concealment and implementation details are not reported in the registry record. The registry does not report the original sample-size justification, detailed analysis plan, analysis populations, missing-data handling, exclusions, or sensitivity analyses. It lists zero analyses. The brief summary's reference to one-way analysis of variance does not explain how the binary participant-level outcome was analyzed.

The posted groups have limited numbers of participants and events, and the risk-difference intervals are broad. This limits precision, especially for the combined comparison, whose interval includes zero. Only unadjusted p-values are posted for the two comparisons against conventional treatment; no multiplicity adjustment is reported. Serious adverse-event figures are also sparse, and additional safety definitions and event details are not reported in the registry record.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Conventional Prophylactic Oral Dexamethasone vs Short-course IV Dexamethasone in Paclitaxel Hypersensitivity — NCT03598426.
• Registry record last updated 2026-06-09; status last verified 2026-06.
• Results first posted 2026-06-09.
• 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.

21. Record Summary

NCT03598426 is a completed phase 3 randomized, open-label, parallel-group trial sponsored by Loma Linda University. The registry lists enrollment of 90, a start date of 2018-08-08, and primary completion on 2025-01-07. The trial compared conventional oral dexamethasone premedication, short-course intravenous dexamethasone, and a combined regimen before paclitaxel infusion.

The primary outcome counts participants with any-grade paclitaxel-associated hypersensitivity during the first infusion, from infusion start through at least three hours afterward. The primary posted contrast is short-course versus conventional; combined versus conventional is also reported. The record lists one secondary outcome concerning discontinuation during the first infusion, but its results are not reported in the registry record.