A Study to Compare the Number of Patients With Gastro-Oesophageal Reflux Disease (GORD) Achieving Heartburn and Regurgitation Symptom Resolution After Treatment With Either Rabeprazole Sodium 20mg, Esomeprazole 20mg or Esomeprazole 40mg: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study to Compare the Number of Patients With Gastro-Oesophageal Reflux Disease (GORD) Achieving Heartburn and Regurgitation Symptom Resolution After Treatment With Either Rabeprazole Sodium 20mg, Esomeprazole 20mg or Esomeprazole 40mg (NCT00464308), a randomized phase 4 study. The primary outcome is The Number of Patients With Complete Resolution of Heartburn by Week 4.

2. Clinical Question

3. Trial Design

The TREAT registry record describes a phase 4, randomized, triple-masked, parallel-group trial sponsored by Janssen-Cilag Pty Ltd. It lists enrollment of 1392 and a completed status. The recorded start is 2006-11, and primary completion is recorded as 2008-05. The record does not specify which parties were masked, so the meaning of triple masking beyond that label is not reported in the registry record.

Participants were assigned to three daily regimens: rabeprazole 20 mg/day, esomeprazole 20 mg/day, or esomeprazole 40 mg/day. Each arm description specifies treatment for 28days and a daily placebo tablet or capsule alongside an active tablet or capsule. The arm sizes recorded for the primary outcome were 464 for rabeprazole, 469 for esomeprazole 40 mg/day, and 459 for esomeprazole 20 mg/day.

The primary endpoint was the number of patients with complete resolution of heartburn by week 4. The definition required absence of symptoms for any 7 consecutive days during the 4 week assessment period, measured using PAGI-SYM. The registry states that efficacy analyses used the ITT population and that data were reanalysed using a compliance population defined by taking at least 80% of study medication and completing at least 80% of data recording.

The registry lists three secondary outcome descriptions: median time to complete heartburn resolution, median time to complete relief of regurgitation, and mean percentage of participants with 24-hour heartburn symptom-free periods. It records 7 outcome measures and 4 analyses. Further operational details of randomization, the masking procedure, and the trial's original statistical 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:

Analysis population for the primary outcome, as defined in the registry: The intent-to-treat (ITT) population was used for all efficacy analyses. The data were reanalysed using the compliance population (defined as all subjects who consumed at least 80% of study medication and who completed at least 80% of data recording).

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 The Number of Patients With Complete Resolution of Heartburn by Week 4, measured over week 4 of treatment.

The primary endpoint counts participants who achieved complete resolution of heartburn by week 4. The registry defines complete resolution as absence of symptoms for any 7 consecutive days within the 4 week period, assessed using the PAGI-SYM scale. That scale describes symptoms from 0-None through 5 Very Severe, with intermediate levels of Very Mild, Mild, Moderate, and Severe. The endpoint is therefore a binary outcome for each participant: the participant either met the stated resolution definition or did not.

Counting participants with the event gives an understandable summary of how many in each treatment arm met the endpoint. Comparing the event proportions between arms describes the size and direction of the observed difference. This kind of endpoint is useful for a direct comparison of the proportion meeting a defined symptom criterion during a specified period. It does not describe the timing of resolution within that period; the registry lists time to complete resolution separately as a secondary outcome.

Interpretation depends on the definition and assessment process. The endpoint captures a consecutive symptom-free interval, not simply a single symptom assessment at the end of treatment. The record does not report additional details about how incomplete symptom recordings were handled or how the endpoint was derived when data were missing.

7. Results Posted in the Registry

• Esomeprazole 40 mg/Day: 64.4% (302/469) versus 58.6% (272/464) on control; risk difference 5.8 percentage points (95% CI -0.5 to 12.0), risk ratio 1.10 (0.99–1.22); unadjusted two-proportion z-test p = 0.070.
• Esomeprazole 20 mg/Day: 60.6% (278/459) versus 58.6% (272/464) on control; risk difference 1.9 percentage points (95% CI -4.4 to 8.3), risk ratio 1.03 (0.93–1.15); unadjusted two-proportion z-test p = 0.547.

8. Statistical Interpretation

9. Statistical Methods Explained

General: A binary endpoint records whether each participant experienced a defined event. In this trial, the event was complete heartburn resolution by week 4, under the registry's definition of no symptoms for any 7 consecutive days during the 4 week period. The posted counts report the number meeting that definition in each arm, alongside the arm size. An event count is not itself a rate; the arm size provides the denominator needed to describe the proportion with the event.

What does a risk difference describe?

General: A risk difference subtracts the event proportion in one group from the event proportion in the comparison group. It is expressed on the absolute probability scale, so zero represents equal event proportions. A positive or negative value indicates the direction of the contrast according to the order in which groups are compared. The recalculations use rabeprazole 20 mg/day as the comparator: the estimates are 0.058 for esomeprazole 40 mg/day and 0.019 for esomeprazole 20 mg/day. The interval estimates describe uncertainty around those point estimates.

How do relative measures differ?

General: A risk ratio compares the event proportions as a ratio, while an odds ratio compares the odds of the event. These relative measures answer different questions from an absolute risk difference, and their numeric values should not be read as percentage-point changes. For esomeprazole 40 mg/day versus rabeprazole 20 mg/day, the posted risk ratio is 1.098 with an interval from 0.992 to 1.216, and the odds ratio is 1.277 with an interval from 0.98 to 1.663. For esomeprazole 20 mg/day versus rabeprazole 20 mg/day, the risk ratio is 1.033 with an interval from 0.929 to 1.149, and the odds ratio is 1.084 with an interval from 0.833 to 1.41. The registry record does not describe the methods used to construct these intervals.

What do the p values indicate?

General: A p value summarizes how incompatible the observed data are with a specified null hypothesis, under the assumptions of the analysis. It is not the probability that the null hypothesis is true, and it does not measure the size or clinical importance of a difference. The unadjusted recalculated p values are 0.07 for the esomeprazole 40 mg/day comparison and 0.547 for the esomeprazole 20 mg/day comparison. The record does not report the full statistical procedures used for these recalculations or an adjustment for multiple comparisons.

What is non-inferiority analysis?

General: A non-inferiority analysis asks whether an intervention is not worse than a comparator by more than a prespecified margin. That margin and the direction of the contrast are essential to interpreting the estimate and its confidence interval. The sponsor's posted analysis is identified as non-inferiority and gives a proportion difference of -0.11 with p <0.05 for the rabeprazole and esomeprazole 40 mg/day groups. A non-inferiority margin and confidence interval are not reported in the registry record. The sponsor's analysis takes precedence over the simple recalculation because the record flags a discrepancy; the available information does not support a detailed reconstruction of the sponsor's conclusion.

Why can over-dispersion matter?

General: Over-dispersion occurs when observed variability is greater than a statistical model assumes. If a model assumes less variability than is present, standard errors may be too small and intervals or tests may appear more decisive than warranted. This concept is relevant to some count and clustered-data models, but the primary endpoint in this record is a participant-level binary outcome. The registry does not report a model for over-dispersion or identify it as an issue in the primary analysis, so no trial-specific conclusion about it can be drawn.

The trial's analysis populations are described at a high level: efficacy analyses used the ITT population, and the data were also reanalysed in a compliance population defined by at least 80% medication consumption and at least 80% data-recording completion. The record does not state which population underlies each displayed count or provide the corresponding compliance-population results in the primary outcome information. Details of missing-data handling, covariate adjustment, stratification, and multiplicity procedures are not reported in the registry record.

10. Confidence Intervals and Precision

A confidence interval gives a range of values compatible with the data under the interval's statistical assumptions. It communicates uncertainty around an estimate; it is not a guarantee that the true effect lies within the reported limits. The registry record does not specify the method used to construct the recalculated intervals, so their exact inferential assumptions are not reported.

For esomeprazole 40 mg/day versus rabeprazole 20 mg/day, the recalculated risk difference is 0.058 and the interval runs from -0.005 to 0.120. Since the interval crosses zero, the data are compatible with no difference and with effects in either direction within the interval. The estimate's positive direction should therefore be read alongside the interval rather than in isolation.

For esomeprazole 20 mg/day versus rabeprazole 20 mg/day, the risk difference is 0.019, with an interval from -0.044 to 0.083. This interval also includes zero and covers values on both sides of it. The uncertainty is substantial relative to the point estimate, so the estimate alone should not be treated as a precise ranking of the regimens.

General: Wider intervals indicate less precision, while narrower intervals indicate more precision under the same method and assumptions. Precision is distinct from whether a result is clinically important or whether an analysis meets a prespecified decision rule. The record does not report a clinically meaningful threshold for the primary endpoint. The sponsor-reported non-inferiority result has no confidence limits listed, and its margin is not reported in the registry record; its precision and decision boundary cannot be assessed from the posted sponsor result alone.

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 for these binary-outcome comparisons. They are not total events. The planning illustration assumes a two-sided significance level 0.05, equal allocation between the two arms, and a normal approximation; the effect to detect is the stated fraction of the observed effect. These are planning assumptions for the illustration, not the trial's own original design assumptions, which are not reported in the registry record.

For esomeprazole 40 mg/day compared with rabeprazole 20 mg/day, planning for 100% of the observed effect gives 1,115 participants per arm for 80% power and 1,492 participants per arm for 90% power. Planning for 75% of the observed effect gives 1,996 participants per arm for 80% power and 2,672 participants per arm for 90% power. Planning for 50% of the observed effect gives 4,521 participants per arm for 80% power and 6,051 participants per arm for 90% power.

For esomeprazole 20 mg/day compared with rabeprazole 20 mg/day, the corresponding figures are 9,983 and 13,364 participants per arm for 80% and 90% power when planning for 100% of the observed effect; 17,783 and 23,805 when planning for 75%; and 40,087 and 53,664 when planning for 50%.

In plain terms, the 80% power column gives the per-arm planning size under which the stated assumptions yield 80% power to detect the selected effect; the 90% column does the same for 90% power. Smaller effects require more participants, as the scenarios illustrate. These are planning illustrations, not observed power.

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 = 0.586; treatment proportion p₁ = 0.644; average p̄ = 0.615.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6881 = 1.3487.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6869 = 0.5781.
4. n = (1.3487 + 0.5781)² / (0.0577)² = 1,114.5, rounded up to 1,115 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 <- 0.586206896551724                          # control proportion
p1 <- 0.643923240938166                          # 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(464308)
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 = 1,115 (80% power), 1,492 (90% power)
 75% of observed effect: n per arm = 1,996 (80% power), 2,672 (90% power)
 50% of observed effect: n per arm = 4,521 (80% power), 6,051 (90% power)
power.prop.test: 1114.5 
Simulated power at n = 1,115 per arm: 0.803 (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. Analyses Reported by the Sponsor

The registry reports a sponsor analysis comparing rabeprazole 20 mg/day with esomeprazole 40 mg/day. It is labelled non-inferiority and gives a proportion difference of -0.11 with p <0.05. The record does not list confidence limits for that result or report the non-inferiority margin, so the criterion underlying the sponsor's conclusion cannot be independently assessed from the posted details.

The unadjusted recalculation for esomeprazole 40 mg/day versus rabeprazole 20 mg/day instead gives a risk difference of 0.058, an interval from -0.005 to 0.120, and an unadjusted p value of 0.07. The displayed figures are not interchangeable: the sponsor result is identified as a non-inferiority analysis, whereas the recalculation is a simple unadjusted contrast from posted summary statistics. The registry flags a discrepancy, and the sponsor's own analysis takes precedence.

The record does not give enough detail to explain the difference in reported results, including whether the contrast direction, analysis population, endpoint handling, or statistical procedure differs. Those details should not be inferred from the numerical summaries alone.

14. Multiplicity, Missing Data and Analysis Populations

The registry lists 7 outcome measures and 4 analyses. It identifies the primary heartburn-resolution endpoint and describes secondary outcomes concerning time to complete heartburn resolution, time to complete regurgitation relief, and the mean percentage of participants with 24-hour heartburn symptom-free periods. The existence of multiple outcomes and analyses raises the general issue of multiplicity: when several hypotheses are tested, the chance of at least one apparently notable result can increase if no adjustment or hierarchy is used.

The registry record does not report a multiplicity strategy, such as a prespecified hierarchy, adjustment method, or allocation of significance levels across outcomes. It also does not report whether the sponsor's non-inferiority analysis was the sole primary inferential test or how the other analyses related to it. The unadjusted recalculated p values should therefore be read as unadjusted comparisons, not as values corrected for the full set of outcomes or analyses.

The record states that all efficacy analyses used the ITT population and that data were reanalysed using a compliance population, defined as participants who consumed at least 80% of study medication and completed at least 80% of data recording. It does not report the primary endpoint counts for both populations in the posted primary outcome details, nor does it specify how participants with incomplete data were treated. The missing-data method, including any imputation or sensitivity analysis, is not reported in the registry record.

The record also does not describe interim analyses, stopping rules, or the trial's original statistical plan and sample-size justification. These omissions limit assessment of whether the posted analyses followed prespecified procedures. They do not establish that such procedures were absent; they mean their details are not reported in the registry record.

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.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: lower respiratory tract infection (0.2%), alleged assault (0.2%), chest pain (0.2%), fever (0.2%), perianal abscess drainage (0.2%).

The registry lists serious adverse events for 5 of 464 participants in the rabeprazole 20 mg/day arm, 1 of 469 in the esomeprazole 40 mg/day arm, and 9 of 459 in the esomeprazole 20 mg/day arm. Other adverse events are listed for 151, 155, and 158 participants, respectively. These are counts with the arm sizes recorded as the numbers at risk; the figures describe the posted safety summaries and are not adjusted comparisons.

The serious-event terms listed are lower respiratory tract infection, alleged assault, chest pain, fever, perianal abscess drainage, and urinary retention, each reported as 0.2%. The registry does not assign these listed terms to specific treatment arms in the posted summary, so an arm-by-arm interpretation of each term is not supported by the record. Death counts are not reported in the registry record.

These summaries do not by themselves establish whether treatment caused an event or whether the observed counts differ beyond chance. The record does not provide further detail about event severity grading, exposure-adjusted rates, or a comparative safety analysis. The counts should be read as descriptive safety information within the recorded trial arms.

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

The posted primary outcome counts describe how many participants in each arm met the registry's definition of complete heartburn resolution by week 4. The unadjusted recalculations show positive point estimates for both esomeprazole arms relative to rabeprazole, but both confidence intervals include zero. The estimates therefore have uncertainty that includes no difference under these recalculations, and the data do not provide a precise numerical separation between the groups from the recalculated contrasts alone.

The sponsor's own analysis is reported as non-inferiority for the comparison of rabeprazole 20 mg/day and esomeprazole 40 mg/day, with a proportion difference of -0.11 and p <0.05. Because the registry flags a discrepancy, that sponsor analysis takes precedence over the simple recalculation. However, the registry does not report the non-inferiority margin or confidence limits, limiting what can be independently evaluated about the sponsor's result.

The endpoint concerns a specific heartburn symptom criterion over the 4 week treatment period. It does not directly establish results for regurgitation, time to symptom relief, longer-term outcomes, or other aspects of GORD. Although secondary outcomes for some of these topics are listed, their results are not included in the registry record for the posted primary result.

These findings should not be read as proof that one regimen is clinically preferable. The registry does not report a clinically meaningful threshold, full analysis plan, missing-data procedure, or multiplicity strategy. The record supports description of the event counts and the stated statistical results, but not broader conclusions beyond the measured outcomes and reported analyses.

17. Why This Trial Matters Statistically

This trial illustrates why the analysis question and the outcome definition matter as much as the event counts. The primary endpoint is not simply a symptom score at the final visit: it requires a consecutive 7-day symptom-free period at some point within the 4 week treatment period. That operational definition determines which participants count as having achieved resolution and frames what the result can address.

It also demonstrates the distinction between absolute and relative comparisons. A risk difference describes the absolute contrast in event proportions, while risk ratios and odds ratios express the comparison on relative scales. The posted recalculations include all of these measures, and their confidence intervals make clear that point estimates should be interpreted in light of uncertainty.

The record includes a sponsor-reported non-inferiority result as well as recalculated, unadjusted contrasts. A discrepancy between these analyses is important because non-inferiority conclusions depend on a specified margin and analysis choices. The margin and details needed to reconcile the results are not reported in the registry record. That makes transparent reporting of estimands, populations, margins, and methods central to evaluating such findings.

Finally, the planning illustration shows how design choices relate to effect size and target power. Smaller effects require larger per-arm sample sizes under the stated assumptions. These planning figures are distinct from the trial's original design assumptions, which are not reported. Together, these features offer a practical example of interpreting binary outcomes, confidence intervals, non-inferiority analyses, analysis populations, and planning assumptions without treating any single statistic as a complete account of the evidence.

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 the primary outcome counts and several recalculated comparisons, but important analysis details are absent. It does not report the method used for the recalculated confidence intervals, the non-inferiority margin, or the confidence interval for the sponsor's non-inferiority result. The discrepancy between the sponsor result and the recalculation is flagged but not explained, preventing a full comparison of their estimands and procedures.

The primary endpoint is based on symptom recording during a defined treatment period. Although the scale and resolution definition are described, the handling of incomplete symptom records and missing data is not reported in the registry record. The record notes an ITT efficacy population and a compliance-population reanalysis, but does not provide separate primary endpoint results for both populations in the posted outcome details.

Multiplicity procedures, interim analyses, the original statistical plan, and the trial's own sample-size justification are not reported. The three secondary outcome descriptions are listed, but their results are not available in the registry record used for this summary. Safety summaries are descriptive, and the listed serious adverse event terms are not linked to individual arms in the record presented.

These limitations constrain interpretation of the posted evidence; they do not establish that the trial lacked additional analyses or procedures. They indicate which details cannot be assessed from the registry record.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: A Study to Compare the Number of Patients With Gastro-Oesophageal Reflux Disease (GORD) Achieving Heartburn and Regurgitation Symptom Resolution After Treatment With Either Rabeprazole Sodium 20mg, Esomeprazole 20mg or Esomeprazole 40mg — NCT00464308.
• Registry record last updated 2014-04-21; status last verified 2014-04.
• Results first posted 2010-08-30.
• 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

NCT00464308 is the completed phase 4 TREAT trial, sponsored by Janssen-Cilag Pty Ltd. The registry describes a randomized, triple-masked, parallel-group study with enrollment of 1392 and compares once-daily rabeprazole 20 mg/day with esomeprazole 20 mg/day or 40 mg/day for 28days.

The primary outcome is the number of participants achieving complete heartburn resolution by week 4, defined as absence of symptoms for any 7 consecutive days in the 4 week period using PAGI-SYM. The primary arm counts were 272 of 464 for rabeprazole, 302 of 469 for esomeprazole 40 mg/day, and 278 of 459 for esomeprazole 20 mg/day.

The registry reports a sponsor non-inferiority analysis for rabeprazole versus esomeprazole 40 mg/day and flags a discrepancy with an unadjusted recalculation. The sponsor analysis takes precedence. Details needed to reconcile the results, including the non-inferiority margin, are not reported in the registry record.