Safety and Efficacy Study of Axid Use in Infants Suffering From Gastroesophageal Reflux Disease (GERD): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Safety and Efficacy Study of Axid Use in Infants Suffering From Gastroesophageal Reflux Disease (GERD) (NCT00373334), a randomized phase 3 study. The primary outcome is Infant Gastroesophageal Reflux Questionnaire Revised (I-GERQ-R) Success.

2. Clinical Question

3. Trial Design

This was a Phase 3, randomized, parallel, multicenter study. The official title describes it as double-blind, while the registry’s masking field records quadruple masking. The record does not explain which trial roles were masked. Braintree Laboratories is listed as the sponsor. The study status is completed; the recorded start is 2006-08 and primary completion is 2008-02.

The study included two nizatidine dose arms and a control arm labeled Conservative Measures Only. The registry lists 42 primary-outcome participants in the 2.5 mg/kg b.i.d. arm, 49 in the 5.0 mg/kg b.i.d. arm, and 42 in the control arm. Registry enrollment is 138, but that figure should not be confused with the endpoint analysis denominators or adverse-event risk sets. The record does not describe allocation ratios or explain the different analysis denominators.

The stated objective was to evaluate efficacy, acceptability, and safety of Axid Oral Solution versus placebo for GERD symptoms in infants. The posted primary endpoint concerns I-GERQ-R success at 8 weeks. Two secondary outcomes are listed: investigator assessments of GERD relief and GERD severity. The arm listing’s control label and the objective’s placebo wording are not fully reconciled in the registry record, so the comparator is described using the listed arm name.

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

Analysis population for the primary outcome, as defined in the registry: The analysis population includes all patients that took drug and had any efficacy data reported. 5 patients that were lost to follow up (1 nizatidine 2.5 group, 1 nizatidine 5.0 group, 3 placebo group) were not included because they had no efficacy data and had not reported any adverse events.

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 Infant Gastroesophageal Reflux Questionnaire Revised (I-GERQ-R) Success, measured over 8 weeks.

The primary endpoint counts participants who meet the registry’s definition of success on the Infant Gastroesophageal Reflux Questionnaire Revised (I-GERQ-R) at 8 weeks. The questionnaire has 12 questions assessing GERD frequency and severity. Its score ranges from a minimum of 0, indicating minimal symptoms, to a maximum of 42, indicating more frequent and/or severe symptoms.

Success is not simply any score improvement. It requires a reduction of at least 5 points from baseline, at least 4 weeks of treatment, and no discontinuation because of lack of efficacy or an adverse event. The resulting endpoint is binary: each participant either meets the success definition or does not. The posted arm results therefore report counts of participants with success, rather than changes in average questionnaire score.

A threshold-based outcome can make a result easier to interpret as the share meeting a prespecified clinical response definition. It also discards information about how far scores moved for participants who fall on either side of the threshold. The registry record does not explain why this particular threshold and these treatment-duration and discontinuation conditions were selected. It also does not post the underlying score distributions, so the counts alone cannot describe the full range of symptom changes.

7. Results Posted in the Registry

• Nizatidine 5.0 mg/kg b.i.d.: 65.3% (32/49) versus 64.3% (27/42) on control; risk difference 1.0 percentage points (95% CI -18.7 to 20.7), risk ratio 1.02 (0.75–1.38); unadjusted two-proportion z-test p = 0.919.
• Conservative Measures Only: 73.8% (31/42) versus 64.3% (27/42) on control; risk difference 9.5 percentage points (95% CI -10.1 to 29.2), risk ratio 1.15 (0.86–1.53); unadjusted two-proportion z-test p = 0.345.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is binary because each participant is classified as either an I-GERQ-R success or not a success under the registry definition. For this kind of outcome, a basic comparison can describe the difference between the proportions meeting the endpoint in two arms. A risk difference expresses that contrast in absolute terms. A value of zero corresponds to equal proportions; positive and negative values indicate direction according to the chosen arm order. The registry’s sponsor results do not include risk-difference estimates.

What does a confidence interval describe?

A confidence interval gives a range of values compatible with the data under the calculation’s assumptions. For a risk difference, an interval that includes zero means the data are compatible with no difference as well as with differences in either direction. It does not establish that the true effect is zero. The recalculated intervals for both control comparisons include zero, and their breadth indicates that the posted counts do not pin down a precise difference. These are unadjusted intervals based on summary counts, not confidence intervals from the sponsor’s reported analysis.

How should p-values be read?

A p-value measures how unusual results at least as discordant with a null hypothesis would be if that null model and the analysis assumptions held. It is not the probability that the null hypothesis is true, nor does it measure clinical importance. The registry reports p-values of 0.528 and 0.341 from Cochran-Mantel-Haenszel analyses. The recalculated unadjusted p-values are 0.919 and 0.345. The difference between those sets of values is consistent with the fact that they come from different analyses; the registry does not report enough detail to reproduce the sponsor’s procedure.

The Cochran-Mantel-Haenszel method is generally used to compare categorical outcomes while accounting for stratification. The registry identifies this method but does not report the strata or other details of its implementation. No further assumptions about the sponsor’s analysis should be inferred from the method name alone. In particular, an unadjusted calculation from the overall counts is not equivalent to an analysis that incorporates stratification or other features that are not represented in those margins.

Why can over-dispersion matter?

In general, over-dispersion means that observed variation is greater than expected under a model’s assumed distribution. When present, it can make standard errors too small and confidence intervals too narrow if the model does not account for it. The primary endpoint is a binary participant count, and the registry does not report an over-dispersion assessment or a model-based approach that would allow one to evaluate it. This is a general statistical consideration, not a claim that over-dispersion occurred in this trial.

Counts, denominators, and analysis populations must also be kept distinct. The primary endpoint denominators are 42, 49, and 42 across the listed arms. Safety reporting uses separate risk sets. The registry says its primary analysis population includes patients who took drug and had efficacy data, and that 5 lost-to-follow-up patients without efficacy data were excluded. It does not describe an intention-to-treat analysis or provide a complete account of handling missing endpoint values beyond that statement.

Finally, multiple outcomes and comparisons can raise multiplicity concerns. Testing more than one contrast or outcome can increase the chance of at least one apparently unusual result if no adjustment or prespecified hierarchy is used. The record lists two secondary outcomes and two sponsor comparisons for the primary endpoint, but does not report a multiplicity strategy. This general concern does not establish that the reported analyses were or were not adjusted.

10. Confidence Intervals and Precision

The unadjusted recalculated risk-difference intervals for each comparison extend across zero, so the posted counts are compatible with no difference and with effects in either direction. The intervals are broad, which means the recalculated comparisons estimate the size of a difference with limited precision. This is important alongside the point estimates: an uncertain estimate should not be read as evidence that the groups are exactly alike.

The registry’s own Cochran-Mantel-Haenszel results report p-values but no effect estimates or confidence intervals. Without those estimates, the magnitude and precision of the sponsor’s adjusted comparisons cannot be assessed from the registry record. The sponsor p-values alone cannot show whether the plausible effects are small or large.

Precision depends on the amount of information in the data and on the observed event proportions. The primary endpoint counts come from relatively limited arm denominators, and the endpoint is dichotomized using a success threshold. Both features constrain the detail that can be recovered from the posted summaries. The original trial’s sample-size justification and precision targets are not reported in the registry record. The planning illustrations use the observed effects for a hypothetical design exercise; they do not reveal the trial’s original planning assumptions.

Confidence intervals describe uncertainty under a calculation’s statistical assumptions. They do not address every possible source of uncertainty, such as missing outcomes, differences in analysis populations, measurement limitations, or incomplete reporting. The registry record notes five excluded participants without efficacy data, but does not give enough information to assess how their unobserved outcomes might affect the comparison.

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 sample-size estimates are participants per arm for 80% power and participants per arm for 90% power. They are not total event counts. The planning illustration assumes a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. It uses the stated fraction of the observed effect as the effect to detect. The trial’s own original design assumptions and sample-size justification are not reported in the registry record.

For the 5.0 mg/kg b.i.d. versus Conservative Measures Only comparison, planning for 100% of the observed effect gives 34,389 participants per arm for 80% power and 46,037 for 90% power. Planning for 75% of the observed effect gives 61,238 per arm for 80% power and 81,979 for 90% power. Planning for 50% of the observed effect gives 138,010 per arm for 80% power and 184,756 for 90% power.

For 2.5 mg/kg b.i.d. versus Conservative Measures Only, the corresponding estimates are 369 and 494 participants per arm when planning for 100% of the observed effect; 670 and 897 when planning for 75%; and 1,538 and 2,058 when planning for 50%. In each pair, the first value is for 80% power and the second for 90% power. The 90% column represents a higher chance of detecting the specified effect if it is real, under the planning assumptions, and therefore requires more participants in these illustrations. Smaller effects need more participants to detect than larger effects.

These estimates are planning illustrations, not observed power. They should not be interpreted as evidence about the probability that this completed trial detected an effect.

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 = 0.643; treatment proportion p₁ = 0.653; average p̄ = 0.648.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6754 = 1.3238.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6754 = 0.5684.
4. n = (1.3238 + 0.5684)² / (0.0102)² = 34,388.7, rounded up to 34,389 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 <- 0.642857142857143                          # control proportion
p1 <- 0.653061224489796                          # 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(373334)
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 = 34,389 (80% power), 46,037 (90% power)
 75% of observed effect: n per arm = 61,238 (80% power), 81,979 (90% power)
 50% of observed effect: n per arm = 138,010 (80% power), 184,756 (90% power)
power.prop.test: 34388.7 
Simulated power at n = 34,389 per arm: 0.794 (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 two sponsor analyses using the Cochran-Mantel-Haenszel method: 5.0 mg/kg b.i.d. versus Conservative Measures Only has p = 0.528, and 2.5 mg/kg b.i.d. versus Conservative Measures Only has p = 0.341. It does not report an effect estimate or confidence interval for either analysis. Those reported results are the sponsor’s analyses and take precedence over simple recalculations when describing the sponsor’s conclusions.

The recalculated comparisons use only the posted arm counts and denominators. They are unadjusted: the p-values are 0.919 for the higher-dose comparison and 0.345 for the lower-dose comparison. They are not attempts to reproduce the Cochran-Mantel-Haenszel method. The registry does not report the strata or other implementation details needed for that reproduction. The difference between the sponsor and recalculated p-values therefore reflects distinct analytical procedures, not a basis for replacing the sponsor results.

No discrepancy is flagged in the registry record. The recalculations are best read as a transparent, simple analysis of the marginal counts, while the sponsor’s reported method and p-values remain the primary record of its own analysis. Neither p-value alone describes the size or clinical importance of a possible difference.

14. Multiplicity, Missing Data and Analysis Populations

The registry lists one primary outcome, I-GERQ-R success at 8 weeks, and two secondary outcomes: investigator assessments of GERD relief and GERD severity. It also reports two sponsor comparisons against the control arm for the primary outcome. Testing multiple comparisons or outcomes raises a general multiplicity issue because the chance of at least one apparently notable result can increase across a family of tests.

The registry record does not report whether a multiplicity adjustment, testing hierarchy, or other strategy was prespecified or applied. It also does not report the results for the two secondary outcomes. The record lists three outcome measures and six analyses, but does not explain the relationship among all analyses or identify which tests, if any, formed a prespecified family. It is therefore not possible to determine from the record whether reported p-values account for multiple testing.

For missing data, the registry states that the primary analysis population includes patients who took drug and had any efficacy data reported. It says that 5 patients lost to follow up were not included because they had no efficacy data and had not reported adverse events. Their arm assignments are reported as 1 in the 2.5 mg/kg arm, 1 in the 5.0 mg/kg arm, and 3 in the control arm. The record does not provide their unobserved endpoint outcomes, explain additional missing-data handling, or state whether sensitivity analyses were conducted.

The primary endpoint denominators are reported by arm, while the safety risk sets differ. The registry does not characterize the primary population as intention-to-treat, and it does not provide a complete analysis-population flow. The record also does not report a protocol-level statistical analysis plan, interim analyses, or a sample-size justification. These omissions limit assessment of prespecification, multiplicity, and the potential effect of missing data.

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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: respiratory distress (2.2%).

The registry reports serious adverse events among separate safety risk sets: 0 among 43 participants in the 2.5 mg/kg b.i.d. arm, 0 among 50 in the 5.0 mg/kg b.i.d. arm, and 1 among 45 in the Conservative Measures Only arm. These denominators differ from the primary endpoint denominators, so the safety counts should be read using their own risk sets rather than compared as if they came from the same analysis population.

The registry lists respiratory distress at 2.2% among the top serious adverse-event terms. It does not provide a group-specific breakdown for that term in the registry record reported. The death counts are not reported in the registry record. A count of zero serious events in an arm does not establish that the risk is absent; uncommon events can be difficult to characterize from limited participant counts. These figures describe the posted safety data and do not by themselves settle comparative safety.

The registry also reports other adverse events affecting 27 participants in the lower-dose arm, 31 in the higher-dose arm, and 31 in the control arm, each using the respective safety risk set. Those are not serious adverse-event counts. The registry record does not provide additional detail in the listed summary to interpret event types or severity beyond the serious-event term noted above.

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

The posted primary endpoint results describe how many participants met a specific I-GERQ-R success definition at 8 weeks in each arm. The counts are 27 of 42 for nizatidine 2.5 mg/kg b.i.d., 32 of 49 for nizatidine 5.0 mg/kg b.i.d., and 31 of 42 for Conservative Measures Only. The sponsor’s reported comparisons have p-values of 0.341 and 0.528. The registry does not provide sponsor effect estimates or confidence intervals, so those analyses do not convey the size or precision of the adjusted contrasts.

The unadjusted recalculations from the arm counts yield p-values of 0.345 and 0.919, with intervals that include zero. These analyses are simpler than the reported Cochran-Mantel-Haenszel analyses and should not be substituted for them. Taken together, the posted information does not establish a clear difference between either nizatidine arm and the control on the primary endpoint. It also does not establish that the groups are identical or rule out effects in either direction.

The endpoint is a thresholded measure. It captures whether a participant met the stated response definition, not the full pattern of questionnaire scores or the size of every participant’s change. The registry record does not report the secondary outcome results, a detailed missing-data analysis, or the sponsor’s effect estimates and confidence intervals. Those gaps restrict what can be concluded about the wider set of outcomes and the precision of the sponsor comparisons.

These results are not, by themselves, a general conclusion about all infants with GERD or all uses of nizatidine. The registry concerns a defined trial population, doses, comparator label, endpoint, and follow-up period. Statistical findings describe the posted data under their analysis assumptions; they are not medical advice or a guarantee about an individual outcome.

17. Why This Trial Matters Statistically

This trial illustrates why an endpoint definition matters as much as the arm counts. I-GERQ-R success requires a specified score reduction, a minimum treatment duration, and no discontinuation for lack of efficacy or an adverse event. The binary outcome is straightforward to count, but it compresses a symptom score into a success-or-not classification. Readers should therefore distinguish the reported response count from a complete account of symptom change.

It also illustrates the importance of separating an analysis result from its uncertainty. A p-value is not an effect size, and the registry’s sponsor analyses do not include effect estimates or confidence intervals. The unadjusted recalculations do include intervals, but those use the marginal counts and do not reproduce the sponsor’s reported method. The two types of results answer related but not identical statistical questions.

The record further shows why analysis populations need careful attention. Primary outcome denominators differ from safety risk sets, and the registry notes five participants without efficacy data who were excluded from the primary analysis. Those details affect how counts should be read and which participants are represented in each analysis. The record’s limited missing-data and multiplicity detail also constrains a full audit of the reported comparisons.

Finally, the sample-size illustrations demonstrate how planning depends on the effect one aims to detect: smaller assumed effects require more participants. They are not evidence of observed power and do not reveal the trial’s original planning rationale. Together, these issues make transparent reporting of estimands, analysis populations, intervals, and analysis plans central to interpreting clinical trial results.

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 trial-specific limitations arise from what the registry does and does not report. The primary endpoint is a dichotomized questionnaire response, so the posted counts do not show the underlying score distributions or the range of changes within each arm. The registry does not report the rationale for the success threshold or the associated treatment-duration and discontinuation conditions.

The sponsor’s Cochran-Mantel-Haenszel analyses are reported only as p-values. Their effect estimates, confidence intervals, strata, and implementation details are not reported in the registry record. Consequently, the adjusted comparisons’ magnitude and precision cannot be evaluated, and the simple unadjusted recalculations cannot reproduce them. The two secondary outcomes are listed, but their outcome results are not reported.

The registry reports that 5 lost-to-follow-up participants lacked efficacy data and were excluded, but it does not provide their potential outcomes or report sensitivity analyses for missing data. It does not characterize the primary analysis as intention-to-treat. The primary endpoint denominators and safety risk sets differ, and a complete account of analysis-population flow is not reported.

The arm listing calls the control Conservative Measures Only, while the study objective describes a placebo comparison. The registry record does not clarify this difference or give detailed descriptions of the arms. It also does not report a multiplicity plan, original sample-size justification, interim analyses, or the trial’s own statistical analysis plan. These gaps limit independent assessment of design choices and the full analysis strategy.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Safety and Efficacy Study of Axid Use in Infants Suffering From Gastroesophageal Reflux Disease (GERD) — NCT00373334.
• Registry record last updated 2009-11-18; status last verified 2009-11.
• Results first posted 2009-11-17.
• 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

NCT00373334 is a completed, randomized, parallel, multicenter Phase 3 study sponsored by Braintree Laboratories. It evaluated nizatidine oral solution at 2.5 mg/kg b.i.d. and 5.0 mg/kg b.i.d. against an arm listed as Conservative Measures Only in infants age 30 days up to 1 year with GERD symptoms. The primary endpoint was I-GERQ-R success at 8 weeks.

Success was defined as a reduction of at least 5 points from baseline, with at least 4 weeks of treatment and no discontinuation for lack of efficacy or an adverse event. The registry reports 27 of 42, 32 of 49, and 31 of 42 successes across the listed arms. Sponsor comparisons report p-values of 0.341 and 0.528. Effect estimates and confidence intervals for those sponsor analyses are not reported in the registry record.

The record lists two secondary outcomes but does not report their results. It also notes five participants without efficacy data who were excluded from the primary analysis and provides separate risk sets for safety reporting.