Effects of Probiotic and Prebiotic Combinations on Premature Infants: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Effects of Probiotic and Prebiotic Combinations on Premature Infants (NCT00282113), a randomized na study. The primary outcome is Weight Gain.

2. Clinical Question

The study description states that its purpose was to assess growth, intestinal bacterial colonization, and fecal short chain fatty acid content. Its stated hypothesis anticipated improved short-term growth, more healthy bacteria in stool, and increased fecal short chain fatty acid content in the combination groups relative to control groups. These statements describe the rationale and hypothesis, not the findings.

The primary outcome was weight gain rather than a measure of bacterial colonization or stool chemistry. The registry record reports group means and standard deviations for weight gain, along with an overall sponsor analysis and comparisons with placebo. The outcome counts show that fewer infants were measured at the 5-week time point than the enrollment figure. The registry record attributes this limited availability to infants leaving the study when discharged from the neonatal intensive care unit.

3. Trial Design

The registry record describes a completed, randomized, triple-masked, parallel trial sponsored by the University of California, Davis. Enrollment was 90. The listed study start was 2004-10, and primary completion was 2006-08. The phase is listed as NA. The three intervention arms were ProBioPlus, Culturelle, and placebo; the first two are categorized as dietary supplements, while placebo is categorized as other.

The primary outcome was weight gain in grams, defined as weight at five weeks minus birth weight. The recorded primary outcome counts were 19 for ProBioPlus, 15 for Culturelle, and 17 for placebo. The registry record explains that infants left the study when discharged from the neonatal intensive care unit, so many were unavailable for measurement at five weeks. These outcome counts therefore describe the infants remaining at that measurement time, not the full enrollment.

Two secondary outcomes were recorded at 4 weeks: stool colonization with bifidobacteria and stool short chain butyric acid content. For the colonization outcome, the listed counts were 11 in each arm. For butyric acid content, the counts were 11 for ProBioPlus, 12 for Culturelle, and 10 for placebo. The record lists three outcome measures and three analyses.

Further details about the randomization procedure, masking roles, allocation concealment, protocol deviations, and analysis populations are not reported in the registry record. The original sample-size justification and planned statistical methods are also not reported in the registry record. The posted design information identifies the broad trial structure but does not establish how every operational or analytic decision was made.

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: Infants left the study as they were discharged from the NICU, therefore many of the infants were not available for measurement at five weeks. The number reported is the number of infants remaining in the study at 5 weeks. Weight in grams is reported.

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 Weight Gain, measured over 5 weeks.

The primary endpoint is weight gain in grams, defined as the infant’s weight at five weeks minus birth weight. A mean is used because weight gain is a continuous measurement: it can take a range of values rather than falling into a small set of categories. Reporting a mean for each arm gives a summary of the observed weight change in the infants measured at the specified time point.

Weight gain was selected as the primary outcome because the study description identifies growth as one of the central questions for the probiotic and prebiotic combinations. It is a direct measure of short-term growth within the study’s chosen timeframe. The reported means and standard deviations describe both the average gain and the spread of measurements in each arm.

The outcome is specifically tied to five weeks, not to a common follow-up time for every infant. The registry record states that infants left the study when discharged from the neonatal intensive care unit, leaving many unavailable for measurement at five weeks. Thus, the available measurement counts differ across arms and are smaller than enrollment. The reported group means describe observed participants with a five-week measurement; the registry record does not report a method for incorporating infants without that measurement into the primary outcome analysis.

Weight gain alone does not capture the secondary questions about stool colonization or stool chemistry. Those outcomes were assessed separately at 4 weeks. Interpreting the primary result therefore requires keeping the growth measure distinct from the secondary biological measures, and recognizing that the endpoint is limited to the infants measured at the stated time.

7. Results Posted in the Registry

• ProBioPlus: mean 618.00 versus 682.00 on control; difference -64.00 (95% CI -197.96 to 69.96), standardized effect d = -0.31; z-test p = 0.349.
• Culturelle: mean 590.00 versus 682.00 on control; difference -92.00 (95% CI -228.51 to 44.51), standardized effect d = -0.46; z-test p = 0.187.

8. Secondary Efficacy Endpoints

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

Stool Colonization With Bifidobacteria

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • ProBioPlus: 63.6% (7/11) versus 27.3% (3/11) on control; risk difference 36.4 percentage points (95% CI -2.4 to 75.1), risk ratio 2.33 (0.81–6.76); unadjusted two-proportion z-test p = 0.087.
  • Culturelle: 18.2% (2/11) versus 27.3% (3/11) on control; risk difference -9.1 percentage points (95% CI -43.9 to 25.7), risk ratio 0.67 (0.14–3.24); unadjusted two-proportion z-test p = 0.611.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between ProBioPlus and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Culturelle and the control. The interval gives the range of effects that remain compatible with the data.

Stool Short Chain Butyric Acid Content

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • ProBioPlus: mean 5.10 versus 7.70 on control; difference -2.60 (95% CI -7.03 to 1.83), standardized effect d = -0.51; z-test p = 0.251.
  • Culturelle: mean 8.30 versus 7.70 on control; difference 0.60 (95% CI -4.05 to 5.25), standardized effect d = 0.11; z-test p = 0.801.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between ProBioPlus and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Culturelle and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes address stool colonization with bifidobacteria and stool short chain butyric acid content, rather than growth. The posted results show varying patterns across the two outcomes and the two supplement comparisons. Their estimates are imprecise, and the sponsor-reported overall tests do not provide detailed effect estimates or confidence intervals. Taken together, the secondary findings do not form a uniformly precise pattern that resolves the primary growth question.

These outcomes relate to the study’s rationale because the study description anticipated changes in intestinal bacteria and fecal short chain fatty acid content. They remain distinct from the primary weight-gain endpoint. A result for colonization or butyric acid cannot by itself establish a difference in growth, and the primary result does not establish what happened on either secondary measure. The outcomes should be interpreted as separate pieces of evidence rather than combined into a single conclusion.

The secondary measures were available for substantially fewer participants than the enrollment figure, with counts varying by arm and endpoint. The registry record does not explain why the secondary measurement counts are limited, or report how missing secondary measurements were handled. This constrains how confidently the observed patterns can be generalized to all enrolled infants.

All recalculated secondary comparison p-values are unadjusted. In general, when several endpoints and contrasts are examined, the chance of finding an apparently notable result by chance increases. Without a reported multiplicity plan, an isolated unadjusted result carries less evidential weight than a primary result designated in advance and interpreted in the context of all tests. The registry record does not report whether the secondary outcomes were confirmatory or exploratory. They are best treated as supportive, hypothesis-generating information rather than independent confirmation of a treatment effect.

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

• Stool Colonization With Bifidobacteria
• Stool Short Chain Butyric Acid Content

9. Statistical Interpretation

10. Statistical Methods Explained

The registry record reports a mixed-effects regression for the primary outcome and an overall p-value of p >0.5. It does not report the model specification, the estimated effects, standard errors, or confidence limits for that analysis. The pairwise mean differences and confidence intervals discussed alongside the posted group summaries are recalculated unadjusted comparisons of each supplement arm with placebo. They are not a replacement for the sponsor’s reported analysis.

What does a mean difference describe?

For a continuous outcome, a mean difference expresses the separation between the average measurements in two groups, in the outcome’s original units. The posted comparisons use placebo as the reference. A negative estimate means the observed mean in the named supplement group was lower than the placebo mean; a positive estimate means it was higher. This direction is a description of the group summaries, not a judgment about the importance of the difference.

For weight gain, the recalculated differences were -64.0 grams for ProBioPlus versus placebo and -92.0 grams for Culturelle versus placebo. The corresponding standard deviations describe variation among the observed infants in each arm. Standard deviation is not the uncertainty in the estimated mean difference; that uncertainty is reflected by the confidence interval.

What does a confidence interval add?

In general, a confidence interval gives a range of effect estimates compatible with the data and the statistical procedure used. It is not a guarantee that the true value lies inside the displayed interval, and it does not assign equal probability to every value in the interval. A narrower interval indicates greater precision than a wider interval, assuming comparable methods and outcomes. An interval spanning zero for a mean difference includes no difference as one value compatible with the data, while also including other possible differences.

The posted unadjusted comparisons also include p-values. In general, a p-value describes how unusual data at least as incompatible with a specified null hypothesis would be under the assumptions of the test. It is not the probability that the null hypothesis is true, nor a measure of clinical importance. The unadjusted p-values are not reported as adjusted for making multiple comparisons or testing multiple outcomes.

Why is there an overall model and a pairwise recalculation?

The mixed-effects regression is the overall analysis reported by the registry. A mixed-effects model can represent group differences while accounting for structure in the data through fixed and random effects; the exact structure depends on the model specification. In general, repeated measurements or clustered observations may motivate such a model, but the registry record does not report the model terms or the reason this method was selected. No model details should be inferred from the method name alone.

The recalculated comparisons use the posted arm-level means, standard deviations, and outcome counts to estimate simple, unadjusted differences versus placebo. In general, such calculations do not account for covariates, missing data patterns, or other design features unless those are incorporated explicitly. They also do not recreate a regression analysis from incomplete model information. The registry record does not report a discrepancy between the sponsor analysis and the recalculations; however, the two approaches answer related but not necessarily identical analytic questions.

How should variation and missing measurements be considered?

The standard deviations show substantial variability in weight gain within each arm relative to the estimated differences. In general, greater outcome variability makes a group contrast harder to estimate precisely when participant numbers are limited. The primary outcome counts differ among arms, and the registry explains that discharge from the neonatal intensive care unit left many infants unavailable at five weeks. In general, missing outcomes can affect interpretation, especially if the availability of a measurement is related to the outcome or to group assignment. The registry record does not report an analysis of that possibility or a description of missing-data assumptions.

For the secondary binary outcome, the posted comparisons use risk differences, which express the difference between the observed proportions in the compared groups. For the continuous butyric acid outcome, they use mean differences. In general, risk differences are expressed on an absolute scale, whereas ratios compare relative levels of risk. The registry also reports overall chi-squared and ANOVA tests for the secondary outcomes, but does not provide effect estimates or confidence limits from those reported analyses.

All pairwise p-values displayed for these recalculations are unadjusted. In general, testing several groups or outcomes raises the chance of at least one apparently unusual result even when there is no underlying effect. This is why isolated secondary comparisons deserve caution, particularly when no multiplicity strategy is reported. The registry record does not report a prespecified multiplicity adjustment or a detailed statistical analysis plan.

11. Confidence Intervals and Precision

Precision concerns how narrowly the data locate an effect estimate. The primary unadjusted mean-difference interval for ProBioPlus versus placebo ranges from -198.0 to 69.96 grams; for Culturelle versus placebo it ranges from -228.5 to 44.51 grams. Each interval includes zero and spans differences in both directions. That breadth means the posted summaries do not pinpoint a narrow range of plausible mean differences.

The estimated differences themselves are -64.0 grams and -92.0 grams. Those values are point estimates, not exact measures of the underlying group differences. The confidence intervals communicate uncertainty around them. In general, a point estimate should not be read alone when the interval is broad, because multiple materially different values may remain compatible with the data.

The observed primary outcome counts were 19, 15, and 17 across the three arms. The registry record explains that many infants were unavailable at five weeks after discharge from the neonatal intensive care unit. Limited outcome counts and variability in weight gain contribute to the uncertainty seen in the unadjusted comparisons. The registry record does not report confidence intervals for the sponsor’s mixed-effects regression, so its precision cannot be assessed from a model-based interval in the posted record.

In general, confidence intervals describe statistical uncertainty under the analysis assumptions; they do not address every source of uncertainty. They do not resolve concerns about missing measurements, the choice of analysis, or whether the observed population represents other infants. The secondary confidence intervals are also broad, and should be read as uncertain estimates rather than definitive evidence of no difference. Precision is therefore a central consideration across the outcomes, not just a technical supplement to the p-values.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning estimates are expressed as participants per arm, not total events. They are illustrations based on the observed effect, not the trial’s original design calculations. The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The planning effect is the stated fraction of the observed effect. The trial’s own original sample-size assumptions are not reported in the registry record.

The 80% power column describes the number of participants per arm the illustration estimates would be needed for an 80% chance of detecting the specified effect under those assumptions. The 90% power column gives the corresponding number for a 90% chance. These are prospective planning quantities under a hypothetical design; they do not describe the probability that the completed trial detected an effect.

Rows that use 100% of the observed effect use the observed estimate as the target effect; the rows using 75% or 50% use smaller target effects. Smaller effects are harder to distinguish from random variation, so the planning estimates require more participants per arm as the target effect gets smaller. The displayed estimates vary considerably by outcome and comparison because the observed effects and variability differ.

These estimates are not observed power. Observed power is not a useful substitute for reading the effect estimates and their uncertainty after a trial has been completed. The planning illustrations can instead help readers understand how sample-size needs change under specified assumptions. They should not be taken as evidence that the original trial was designed around these effects, since that original design justification is not reported.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = -64.000; pooled standard deviation σ = 204.324.
2. n = 2 × (1.96 + 0.842)² × 204.324² / -64.000² = 160.0, rounded up to 160 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- -64                      # observed difference in means
sigma <- 204.323993005342                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(282113)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 160 (80% power), 215 (90% power)
 75% of observed effect: n per arm = 285 (80% power), 381 (90% power)
 50% of observed effect: n per arm = 640 (80% power), 857 (90% power)
power.t.test: 161 
Simulated power at n = 160 per arm: 0.797 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Analyses Reported by the Sponsor

The registry record reports a mixed-effects regression across the three primary-outcome groups, with p >0.5. It does not report an effect estimate, confidence interval, or further model details for this analysis. The reported result is therefore an overall test without a posted estimate of the size and precision of the group differences under that model.

The pairwise mean differences and confidence intervals are recalculated from the posted group means, standard deviations, and counts. They are unadjusted comparisons of ProBioPlus versus placebo and Culturelle versus placebo. Their p-values are also unadjusted. They are useful for describing the observed arm summaries, but they should not be presented as the sponsor’s own analysis or as a reproduction of the mixed-effects model.

The registry record does not flag a discrepancy between the posted overall result and these recalculations. Even so, an overall test and pairwise contrasts are not identical questions: the overall analysis addresses differences among all three groups, while each recalculation contrasts one supplement with placebo. The posted information does not allow reconstruction of the model or direct comparison of model-based estimates with the unadjusted estimates.

The sponsor’s reported analysis takes precedence as the trial’s own analysis. Its p-value should be interpreted as evidence that the reported model did not identify an overall group difference under its method, not as proof that the groups were equivalent. The unadjusted intervals add a view of the uncertainty around simple pairwise estimates, but their limitations and the missing model details remain important.

15. Multiplicity, Missing Data and Analysis Populations

The registry record identifies one primary outcome and two secondary outcomes, and reports an analysis for each. It does not report a prespecified strategy for controlling multiplicity across the outcomes or pairwise comparisons. The recalculated comparison p-values are unadjusted. A p-value without multiplicity adjustment addresses its individual comparison, not the overall chance of at least one false-positive finding across a family of tests.

In general, testing multiple outcomes, arms, and contrasts can increase the probability of obtaining a small p-value by chance. A prespecified hierarchy or adjustment can help define how results should be interpreted across tests. Because no such strategy is reported in the registry record, the secondary comparison results warrant caution and should not be treated as though they independently confirm a finding. The record does not report whether any multiplicity approach was used in the original analysis.

Missing outcome information is an important issue for the primary endpoint. The record states that infants left the study when discharged from the neonatal intensive care unit, and many were unavailable for measurement at five weeks. It gives the number remaining at that time for each arm. It does not report whether the mixed-effects model included incomplete observations, the assumptions used for missing measurements, or any sensitivity analysis examining departures from those assumptions. The missing-data methods are not reported in the registry record.

The registry record also does not report an intention-to-treat population definition, a per-protocol population, or how exclusions and protocol deviations were handled. The outcome counts are available, but the analysis population and participant flow beyond the stated discharge-related unavailability are not described in detail. The specific randomization procedure, allocation concealment, and full statistical analysis plan are not reported in the registry record. These omissions limit assessment of whether the analyses followed a prespecified plan and how robust the estimates are to analytic choices.

16. Safety Results

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

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

The posted serious adverse event figures report zero affected participants in each arm. The listed numbers at risk were 31 for ProBioPlus, 30 for Culturelle, and 29 for placebo. The adverse-event section does not list serious-event terms. These figures describe the serious adverse events recorded in the registry for the stated groups; they do not by themselves establish that no serious events occurred outside the reporting period or definitions used.

The number at risk in the adverse-event listing is distinct from the primary outcome measurement count. The registry record does not explain the difference between the adverse-event group counts and the weight-gain counts, nor does it report exposure duration or detailed ascertainment methods for adverse events. No deaths are listed as a numerical result.

Zero reported serious adverse events should be read as a count, not as proof that an intervention is free of risk. In general, a lack of observed events in a limited sample cannot rule out uncommon events. The record also reports zero other affected participants in the listed groups, but provides no event terms. Interpretation is therefore constrained by the limited safety detail available in the registry record.

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

The posted primary data describe average weight gain among the infants measured at five weeks. The placebo group mean was 682.0 grams, compared with 618.0 grams for ProBioPlus and 590.0 grams for Culturelle. The recalculated unadjusted contrasts have confidence intervals that include zero and extend across a range of negative and positive differences. The sponsor’s reported overall mixed-effects analysis has p >0.5, with no model-based estimate or interval posted.

These results do not establish that the groups are the same. A failure to identify a statistically clear difference is not evidence of equivalence unless equivalence is assessed with an appropriate design and prespecified margins, which are not reported in the registry record. Nor do the results establish a precise effect in either direction. The confidence intervals indicate substantial uncertainty around the unadjusted pairwise estimates.

The observed results are limited to infants with available measurements at the outcome time. The registry record says many infants were unavailable at five weeks after discharge from the neonatal intensive care unit. Since missing-data methods are not reported, the extent to which those observed summaries represent all randomized infants cannot be determined from the record.

The secondary outcomes concern stool colonization and butyric acid, not weight gain. Their patterns are variable and the comparisons are unadjusted, with no reported multiplicity strategy. They therefore do not resolve the uncertainty in the primary outcome or establish a broader biological conclusion. The safety listing reports zero serious adverse events in the recorded groups, but its limited detail cannot establish absence of risk.

Overall, the registry results describe a small set of observed measurements with uncertain group contrasts and limited reporting of analytic details. They support a cautious statistical interpretation, not a clinical recommendation.

18. Why This Trial Matters Statistically

This trial illustrates why a posted p-value should not be read without the underlying estimates and their precision. The sponsor’s primary analysis reports an overall p-value, while the arm summaries allow unadjusted pairwise differences and confidence intervals to be described. The latter show the scale and uncertainty of the contrasts, information that a p-value alone does not convey.

It also illustrates the importance of separating enrollment from the number with an outcome measurement. Although enrollment was 90, the primary endpoint counts were 19, 15, and 17. The registry record explains that discharge from the neonatal intensive care unit meant many infants were unavailable at five weeks. A reader needs to consider both the outcome counts and the missing-data reporting when assessing the reach of the findings.

The trial has multiple outcomes with distinct meanings and measurement times. Weight gain was the primary endpoint; stool colonization and butyric acid were secondary measures. Keeping these questions separate prevents secondary patterns from being used to overstate what the primary growth analysis establishes. The absence of a reported multiplicity plan further matters when interpreting unadjusted secondary tests.

Finally, the planning illustrations show how requirements depend on the effect size targeted and the assumptions used. They do not reveal the original design rationale or establish the trial’s observed power. Together, these features make the registry record useful for teaching the distinction between estimates, uncertainty, analysis plans, missing outcomes, and planning calculations.

19. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The registry record provides limited detail about trial conduct and analysis. The randomization procedure, allocation concealment, masking roles, and protocol deviations are not reported. The intervention arm descriptions are blank, so the record does not specify the contents or administration details of the supplements. The original sample-size justification and planned statistical analysis are not reported.

Availability of the primary measurement is a specific limitation. Many infants were unavailable at five weeks because they left the study after discharge from the neonatal intensive care unit. The record gives the numbers remaining for measurement but does not report missing-data assumptions, a sensitivity analysis, or a detailed account of how the mixed-effects analysis addressed missing measurements. The possibility that measured infants differ from those not measured cannot be assessed from the posted information.

The sponsor’s overall primary analysis is reported only as a method and p-value; its effect estimate, confidence interval, and model specification are absent. This makes it difficult to assess the magnitude and precision of the sponsor’s analysis or to compare it directly with the unadjusted recalculations. The reported secondary analyses likewise lack estimates and confidence limits.

Secondary outcome counts are limited, and the record does not report why those measurements were available for fewer participants or whether missingness was handled analytically. It also does not report a multiplicity plan, analysis-population definitions, or a full statistical analysis plan. The serious adverse-event listing has zero affected participants in each group but no event terms or detailed ascertainment information. These limitations constrain interpretation; they do not show that any particular unreported method or event occurred.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Effects of Probiotic and Prebiotic Combinations on Premature Infants — NCT00282113.
• Registry record last updated 2010-01-28; status last verified 2009-12.
• Results first posted 2010-01-28.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

23. Record Summary

NCT00282113 is a completed randomized, triple-masked, parallel trial sponsored by the University of California, Davis, with enrollment of 90. It compared ProBioPlus and Culturelle with placebo in premature infants. The primary outcome was weight gain in grams from birth to five weeks. The record also lists stool colonization with bifidobacteria and stool short chain butyric acid content at 4 weeks.

The reported primary analysis was mixed-effects regression, with p >0.5 and no posted model-based estimate or confidence interval. Unadjusted recalculated pairwise comparisons have broad confidence intervals that include zero. Many infants were unavailable for the primary measurement after discharge from the neonatal intensive care unit. The registry record does not report the original sample-size justification, detailed model specification, missing-data methods, or multiplicity strategy. The results should therefore be read as uncertain estimates based on the reported outcome data, with important analysis details not reported in the registry record.