Rate of Complete Symptom Relief, Prevention of Symptom Relapse: Grades A and B Esophagitis of Esomeprazole Therapy: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Rate of Complete Symptom Relief, Prevention of Symptom Relapse: Grades A and B Esophagitis of Esomeprazole Therapy (NCT01874535), a randomized phase 4 study. The primary outcome is The Rates of Complete Symptom Relief.

2. Clinical Question

The outcome is presented as counts: 152 of 207 participants in the four-week group and 157 of 201 in the eight-week group. The registry record also includes an analysis labeled as a hazard ratio and a chi-squared method, but the posted information does not explain how this time-to-event label relates to the stated count-based endpoint. The unadjusted comparison expresses the difference in the proportion with the outcome as a risk difference. These descriptions address related but not necessarily identical statistical summaries, so their interpretations should remain distinct.

3. Trial Design

This was a Phase 4, randomized, parallel-group trial with no masking. The registry record lists 408 enrolled participants and names Chang Gung Memorial Hospital as sponsor. The condition was erosive esophagitis. The arm descriptions specify participants with Los Angeles grade A or B esophagitis receiving daily esomeprazole 40 mg for either four weeks or eight weeks.

Random allocation is intended to create groups that are comparable on average at the start of a trial, although the record does not provide information about the randomization sequence or allocation concealment. With no masking, participants and study personnel were not masked to group assignment, according to the registry’s masking designation. The record does not report additional details about the conduct of treatment or outcome assessment.

The registry lists a start date of 2010-06, primary completion in 2016-06, and status as completed. These dates describe the trial record, but do not establish the timing of each participant’s follow-up or the duration of the entire study for individual participants. The record lists one outcome measure and one analysis, and no posted secondary outcomes. The primary outcome is titled “The Rates of Complete Symptom Relief,” but the description says the rate was assessed at the end of initial treatment, whereas the stated timeframe is at 20 weeks after the end of initial treatment. This discrepancy in wording leaves the exact timing unclear.

The registry record does not report the trial’s original sample-size justification, its detailed statistical analysis plan, interim analyses, or analysis populations. The planning illustrations associated with the posted effect are separate from the trial’s original design assumptions.

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: Baseline characteristics of the enrolled patients

6. Primary Endpoint

The registry lists the primary outcome as The Rates of Complete Symptom Relief, measured over at the 20 weeks after the end of initial treatment..

The primary outcome is the number of participants recorded as having complete symptom relief. The registry identifies the unit as participants and the parameter as a number, so the posted arm results are counts rather than an individual-level symptom score or a change from baseline. The record gives 152 participants with the outcome in the four-week group and 157 in the eight-week group.

A binary outcome distinguishes whether each participant did or did not meet a defined condition. Counting participants who meet the condition can make a clinical outcome straightforward to interpret, but interpretation depends on how the condition was defined and when it was assessed. The registry record does not provide a detailed operational definition of complete symptom relief, such as the symptom criteria used to classify a participant.

The endpoint’s timing is also unclear in the record: its description refers to the end of initial treatment, while its timeframe states at the 20 weeks after the end of initial treatment. This matters because symptom status may be considered differently depending on the assessment point. The record’s brief summary discusses symptom relapse and sustained healing, but the posted primary outcome measure is complete symptom relief. No secondary outcomes are posted. The outcome should therefore be interpreted as the reported count for this named symptom endpoint, not as a direct result for relapse, sustained healing, or another unposted measure.

7. Results Posted in the Registry

• 8 Week Group: 78.1% (157/201) versus 73.4% (152/207) on control; risk difference 4.7 percentage points (95% CI -3.6 to 13.0), risk ratio 1.06 (0.95–1.19); unadjusted two-proportion z-test p = 0.270.

8. Statistical Interpretation

9. Statistical Methods Explained

Two kinds of statistical information appear in the registry record: arm-level counts and a between-group comparison recalculated from those counts. Keeping these distinct helps avoid treating every reported statistic as if it answered the same question.

What does a count-based outcome represent?

The primary result records how many participants in each group had complete symptom relief. For a binary outcome, each participant is classified according to whether the outcome occurred. An arm-level count needs its group size for context, which is why the posted results include both the number with relief and the number in the group. The counts are 152 of 207 for four weeks and 157 of 201 for eight weeks. The registry does not provide further detail about the outcome’s classification criteria.

The posted comparison expresses the difference in the two group proportions as a risk difference. This is an absolute measure: it describes the difference between the proportions meeting the outcome in the two groups. Its estimate is 0.047. A confidence interval accompanies the estimate to show a range of values consistent with the data under the calculation’s assumptions. The interval, -0.036 to 0.130, crosses zero, which represents no absolute difference. The unadjusted p-value is 0.27. The p-value describes how unusual results at least this discordant with a no-difference model would be, under the assumptions of that test; it is not the probability that the null hypothesis is true.

How do relative measures differ?

The same comparison lists a risk ratio and an odds ratio. A risk ratio compares the proportions with the outcome, while an odds ratio compares the odds of the outcome. Both are relative measures rather than absolute differences. A value of one is the no-difference reference for these measures. The posted risk ratio is 1.064, with a confidence interval from 0.953 to 1.187. The posted odds ratio is 1.291, with a confidence interval from 0.819 to 2.035. The intervals include one, indicating that the summaries are not precise enough to exclude the no-difference reference on those scales. Relative measures can appear numerically different from the risk difference because they use different scales and definitions.

What does “unadjusted” mean?

The unadjusted re-calculation uses posted summary statistics to compare the groups without adjustment for other variables. It should not be described as the sponsor’s original analysis. An adjusted analysis may account for additional variables or use a different model, but the registry record does not give enough detail about the sponsor’s statistical procedure to determine what was done beyond the posted method and label. The record identifies the sponsor’s method as chi-squared and its parameter as a hazard ratio, yet does not explain how a hazard ratio was obtained or how it connects to the count-based primary result.

General statistical context: a chi-squared test is commonly used to assess whether categorical outcome distributions differ between groups. A hazard ratio, in contrast, is generally associated with time-to-event analysis. These terms refer to different kinds of summaries. Their joint appearance in the record is not enough to establish the exact model, data structure, or analysis used. The sponsor’s value of 0.009 and p <0.05 should therefore be described as posted, not reinterpreted as a specific effect estimate beyond the record’s own label.

Why can analysis labels matter?

Statistical estimates depend on the outcome definition, assessment timing, analysis population, and model. The record calls the primary outcome a number of participants, but its timeframe and description differ about when relief was assessed. The sponsor result is labeled as a hazard ratio even though the posted outcome is count-based. These features make it important to report the analysis as the registry states it and separately explain the unadjusted re-calculation. The record does not report an analysis plan that would resolve these questions.

What can a non-significant result establish?

General statistical context: a p-value above a conventional threshold does not demonstrate that groups are equivalent. It can reflect a small or absent difference, imprecision, or both. Establishing equivalence or non-inferiority ordinarily requires a prespecified margin and a design and analysis suited to that question. The registry record does not report such a margin or an equivalence or non-inferiority objective. The risk-difference interval is therefore best read as an estimate with uncertainty, not as proof of equality.

The estimates and intervals describe statistical uncertainty conditional on the methods and data used. They do not resolve uncertainty created by unclear endpoint timing, missing-data handling, or the limited analysis detail in the registry record. Statistical interpretation should remain within those boundaries.

10. Confidence Intervals and Precision

The risk-difference estimate for eight-week versus four-week treatment is 0.047, with a confidence interval from -0.036 to 0.130. A confidence interval is a measure of statistical precision: it gives a range of effect values compatible with the observed data and the method’s assumptions. It is not a guarantee that the true effect lies in the interval, nor does it describe the range of individual participants’ outcomes.

Because this interval includes zero, the posted unadjusted data are compatible with no absolute difference as well as with a difference in either direction. The interval extends from a negative estimate to a positive one, so it does not establish a single direction with precision. Its width reflects uncertainty in the estimated contrast given the group sizes and outcome counts. The record does not provide an adjusted confidence interval for the sponsor’s analysis.

The relative measures also have intervals that include their no-difference reference: the risk ratio is 1.064, with an interval from 0.953 to 1.187, and the odds ratio is 1.291, with an interval from 0.819 to 2.035. Those intervals are on different scales from the risk-difference interval and should not be compared by their raw widths.

Precision is not the same as clinical importance. The registry record does not specify a threshold for a clinically meaningful difference, and the unclear endpoint timing further limits interpretation. The intervals describe uncertainty in the posted summaries; they do not correct for limitations in outcome definition, missing-data reporting, or analysis documentation.

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 participant counts per arm for the binary outcome, not total event counts. They are illustrations based on the observed effect, not the original trial’s design calculations. The planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is defined as the stated fraction of the observed effect.

For planning at 100% of the observed effect, the figures are 1,315 participants per arm for 80% power and 1,760 participants per arm for 90% power. Planning for 75% of the observed effect gives 2,377 participants per arm for 80% power and 3,182 for 90% power. Planning for 50% of the observed effect gives 5,433 participants per arm for 80% power and 7,272 for 90% power.

Power is the probability that a study’s statistical test will meet its chosen significance criterion if a particular effect is truly present and the planning assumptions hold. The 80% and 90% columns describe two target probabilities of detecting the specified effect. They are not probabilities that the treatment is effective, and they do not describe the chance that the current trial’s result is correct. These figures are not observed power.

The smaller the effect chosen for planning, the more participants are needed to distinguish it from random variation under the stated assumptions. The registry record does not report whether these assumptions were used for the original trial, nor does it report the original sample-size justification. The listed figures should therefore be read only as planning illustrations tied to the posted observed effect, not as evidence about the adequacy of the trial’s original enrollment target.

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 = 0.734; treatment proportion p₁ = 0.781; average p̄ = 0.758.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6060 = 1.1877.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6051 = 0.5092.
4. n = (1.1877 + 0.5092)² / (0.0468)² = 1,314.9, rounded up to 1,315 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 <- 0.734299516908213                          # control proportion
p1 <- 0.781094527363184                          # 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(1874535)
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 = 1,315 (80% power), 1,760 (90% power)
 75% of observed effect: n per arm = 2,377 (80% power), 3,182 (90% power)
 50% of observed effect: n per arm = 5,433 (80% power), 7,272 (90% power)
power.prop.test: 1314.9 
Simulated power at n = 1,315 per arm: 0.805 (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’s sponsor-reported analysis lists a chi-squared method, a parameter labeled “Hazard Ratio (HR),” a value of 0.009, and p <0.05. No confidence interval is posted for this result. The record does not explain the analysis in enough detail to identify its model, the outcome timing it used, or how the hazard-ratio label relates to the count-based primary outcome.

The unadjusted re-calculation based on the posted group counts gives a risk difference of 0.047, with a confidence interval from -0.036 to 0.130 and an unadjusted p-value of 0.27. This is an absolute comparison of the recorded outcome proportions; it is not the same statistic as a hazard ratio. The registry’s sponsor result and the re-calculation should therefore be reported separately rather than treated as competing estimates on the same scale.

The record does not provide enough methodological information to reconcile the results. The sponsor’s result remains the analysis reported by the registry, while the unadjusted calculation is a distinct summary derived from the posted counts. Neither description supplies missing information about endpoint timing or analysis populations.

14. Multiplicity, Missing Data and Analysis Populations

The registry record lists one outcome measure and one analysis, with no secondary outcomes posted. It does not report whether additional outcomes or analyses were specified in a protocol, whether multiple statistical tests were conducted, or whether any adjustment for multiple comparisons was planned. The absence of posted secondary outcomes does not establish that no other analyses were performed.

Multiplicity arises when several outcomes, subgroups, time points, or statistical tests are examined. General statistical context: conducting more tests can increase the chance of a statistically significant result arising by chance, unless the analysis plan accounts for those tests. The record does not report enough information to assess whether this issue affected the sponsor’s reported p-value or the selection of the primary result.

Missing-data handling is not reported in the registry record. The posted arm counts provide the group sizes and the number recorded with complete symptom relief, but do not explain whether every enrolled participant had an outcome assessment, how any unavailable outcomes were treated, or whether the analysis used all randomized participants. The record also does not describe exclusions, withdrawals, adherence, or imputation procedures. These omissions prevent assessment of the sensitivity of the result to missing outcomes.

The analysis populations are not reported in the registry record. It is therefore not possible to establish whether the posted analysis included all randomized participants, only those who completed treatment, or another subset. General statistical context: different analysis populations can answer different questions, and excluding participants after randomization may affect the comparability created by random allocation. The registry record does not provide the detail needed to determine whether or how that concern applies.

The registry also does not report the original statistical analysis plan, interim analyses, or a prespecified threshold for interpreting the result. These are documentation gaps, not evidence that a particular procedure was or was not used.

15. 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.

The serious adverse event figures list zero affected participants in the four-week group, among 207 at risk, and zero in the eight-week group, among 201 at risk. The record also lists zero other adverse events in each group and zero deaths in each group. No serious adverse event terms are listed.

These figures describe what was posted for the stated groups. A zero count means no events of that category were recorded in the registry’s posted results; it does not establish that the risk is absent or that no event occurred outside the reporting framework. The record does not give definitions, observation windows, event-collection procedures, or detail about how safety events were attributed to treatment.

The group counts shown for safety match the arm sizes listed for the primary outcome, but the record does not explain the safety analysis population or whether every participant was followed for the same duration. Nor does it provide information about withdrawals or completeness of safety reporting. The absence of listed event terms should therefore be interpreted as limited posted information, not as a comprehensive characterization of safety.

These statistics are descriptive. They do not by themselves establish a comparative safety difference between treatment durations, and they do not support medical advice.

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

The posted primary outcome counts show 152 of 207 participants in the four-week group and 157 of 201 in the eight-week group recorded as having complete symptom relief. The unadjusted risk difference is 0.047, with a confidence interval from -0.036 to 0.130 and an unadjusted p-value of 0.27. These results describe the registry’s count-based comparison and its uncertainty.

The interval includes no absolute difference and values on either side of it. Consequently, the unadjusted comparison does not determine a precise direction or size of difference. It also does not establish equivalence: no equivalence margin or non-inferiority framework is reported. The sponsor’s analysis is labeled as a hazard ratio and has a reported value of 0.009 with p <0.05, but the registry does not provide enough methodological detail to reconcile it with the count-based comparison.

The endpoint’s assessment time is ambiguous because its description says relief was assessed at the end of initial treatment, while its timeframe says at 20 weeks after the end of initial treatment. The registry does not provide a detailed symptom-relief definition or describe missing-data handling and analysis populations. These limits matter when deciding what the counts represent and how broadly they can be interpreted.

The brief summary refers to symptom relapse and sustained healing, but the posted primary endpoint is complete symptom relief, and no secondary outcome is posted. The primary result should not be read as a direct estimate of relapse prevention or sustained healing. The findings are statistical descriptions of the posted record, not treatment recommendations or a basis for individual medical decisions.

17. Why This Trial Matters Statistically

This trial illustrates why the outcome definition, timing, and statistical scale need to be read together. Its primary result is a count of participants with complete symptom relief, while the registry’s separate sponsor analysis is labeled as a hazard ratio. Because a count-based risk difference and a hazard ratio summarize outcomes differently, their estimates cannot be compared as if they were the same quantity.

The unadjusted risk difference and its confidence interval make uncertainty visible. The point estimate alone gives an incomplete account: the interval extends across zero and includes both negative and positive values. Reporting the estimate together with its precision helps distinguish the observed contrast from the range of values still compatible with the data and method.

The record also demonstrates how small wording differences can affect interpretation. The outcome description refers to the end of initial treatment, but the timeframe specifies 20 weeks after that treatment ended. Without clarification, readers cannot confidently identify which assessment point the counts represent. Similarly, the brief summary discusses relapse and sustained healing, while the posted primary outcome is complete symptom relief. Keeping these outcome concepts separate avoids overreading the result.

Finally, planning illustrations based on an observed effect are not substitutes for a trial’s original sample-size rationale. Their assumptions and intended use need to be distinguished from the trial design. The registry record does not report the original justification or detailed analysis plan, underscoring the importance of transparent reporting for interpreting clinical trial statistics.

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.

Several limitations are specific to the posted record. First, the primary endpoint’s timing is inconsistent: its description places assessment at the end of the initial treatment phase, whereas its timeframe states at the 20 weeks after that phase ended. The record does not resolve this discrepancy.

Second, the registry does not provide a detailed definition of complete symptom relief. It also does not explain how symptom information was collected or how participants were classified. Without those details, the endpoint’s operational meaning cannot be fully assessed.

Third, the sponsor analysis is labeled as a chi-squared method with a hazard-ratio parameter, but no confidence interval or further model description is posted. The record does not explain how that analysis relates to the binary count endpoint. The unadjusted risk-difference calculation is a separate summary and cannot fill in the missing details of the sponsor’s analysis.

Fourth, missing-data procedures and analysis populations are not reported. The registry does not explain whether the posted counts include all randomized participants or how participants without an outcome assessment were handled. The record also does not report adherence, withdrawals, or exclusions.

Finally, the original sample-size justification, detailed statistical analysis plan, interim analyses, and any multiplicity procedures are not reported in the registry record. These are limits of the available documentation and should not be taken to imply that a particular method was or was not used. No secondary outcomes are posted, so the results do not provide direct estimates for the relapse or sustained-healing questions named in the brief summary.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Rate of Complete Symptom Relief, Prevention of Symptom Relapse: Grades A and B Esophagitis of Esomeprazole Therapy — NCT01874535.
• Registry record last updated 2018-09-04; status last verified 2017-06.
• Results first posted 2018-09-04.
• 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

This completed Phase 4 randomized, unmasked, parallel-group trial enrolled 408 participants with erosive esophagitis. It compared daily esomeprazole 40 mg for four weeks with the same daily dose for eight weeks among participants described as having Los Angeles grade A or B esophagitis.

The primary outcome was complete symptom relief. The posted counts were 152 of 207 in the four-week group and 157 of 201 in the eight-week group. The unadjusted risk difference was 0.047, with a confidence interval from -0.036 to 0.130 and p = 0.27. The sponsor’s posted analysis was labeled as a chi-squared method and hazard ratio, with value 0.009 and p <0.05; the registry does not explain how it relates to the count-based endpoint.

The endpoint timing is worded inconsistently, and missing-data handling, analysis populations, and the original statistical plan are not reported in the registry record.