Study to Investigate the Potential of Probiotics in Yogurt: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study to Investigate the Potential of Probiotics in Yogurt (NCT00492583), a randomized phase 3 study. The primary outcome is Number of Days Children Are Out of School Sick.

2. Clinical Question

3. Trial Design

The registry record describes a randomized, quadruple-masked, parallel-group phase 3 study sponsored by Georgetown University. The study status is completed. Enrollment is listed as 182. The record lists 95 participants in the placebo arm and 87 in the Bifidobacterium lactis (BB-12) arm. These are the reported arm counts; the record does not describe participant flow after enrollment or explain whether every participant contributed data to the primary outcome.

The intervention was a daily oral drink. The placebo and active arms are both described as receiving 4 fluid ounces (112 grams) per day. The registry record does not give further information about treatment duration, adherence monitoring, or whether intake was supervised. The outcome timeframe is 90 days, but the record does not specify whether this was the entire treatment period or how the assessment period related to exposure.

Randomized allocation is reported, but the method used to generate the allocation sequence and whether allocation was concealed are not reported in the registry record. Masking is listed as quadruple; which groups of people were masked, and how masking was maintained, are not reported. The model is listed as parallel, indicating separate intervention groups rather than a crossover structure. The record does not report whether randomization or analysis accounted for daycare-center grouping.

The study start is listed as 2007-07 and primary completion as 2008-08. These dates describe the registry record and do not, by themselves, establish when recruitment began or when all outcome data were collected. Details of the protocol, changes to planned methods, and study conduct beyond the listed design features are not reported in the registry record.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Overall Study

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 Number of Days Children Are Out of School Sick, measured over 90 days.

The primary endpoint is titled “Number of Days Children Are Out of School Sick.” Its unit is days per 100 person days, and its timeframe is 90 days. In practical terms, it is intended to describe sick-related absence in relation to the person-time represented by children being observed. Expressing a count relative to person-time can make observation periods more comparable when participants contribute different amounts of time, although the registry record does not explain the calculation used for this endpoint.

Sick-day absence is relevant to the stated study question because the summary focuses on illness-related absences from daycare. It is a count-based outcome rather than a continuous measure such as a laboratory value. Count outcomes can have a skewed distribution and can include many observations with no events; statistical methods therefore often account for their discrete nature and possible variation beyond a simple count model. Those are general considerations, not methods confirmed for this study.

The outcome description says the measure was assessed for the entire population, while the posted results give a value for each arm. The record does not report a more detailed endpoint definition, such as how partial days were handled, how illness-related absence was determined, or how person-time was calculated. Nor does it report an analysis comparing the groups. The endpoint values alone therefore describe the posted arm results without establishing the estimated treatment contrast or its uncertainty.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry identifies the endpoint as a count-other outcome and gives its unit as days per 100 person days. It does not state the statistical model used, the method for estimating the arm values, or the method intended for a comparison. The explanations below describe general statistical considerations for outcomes of this kind; they should not be taken as a description of the analysis conducted in this trial.

Why does the outcome count as count data?

In general, count data record how many events or units occur. Sick days are discrete observations rather than measurements that can take any value along a continuous scale. Count outcomes often have a concentration of low values, including zero, and may have a right-skewed distribution. These properties can make methods that assume normally distributed measurements a poor fit unless the data and modeling approach justify that assumption.

Common general approaches for counts include Poisson models and extensions that allow more variability than the basic Poisson model. A Poisson model assumes that the mean and variance are linked in a particular way. When the observed variability exceeds that assumption, over-dispersion is present. Ignoring over-dispersion can make standard errors too small and uncertainty intervals too narrow. Negative binomial models or other approaches may accommodate excess variability, but the registry record does not say whether any such model was used in this study.

What does the person-time unit contribute?

In general, a rate expresses an event count relative to time at risk or time observed. Using person-time can be useful when participants contribute unequal observation periods, because the denominator reflects how much follow-up contributes to the rate. A rate analysis may use the logarithm of person-time as an offset in a count model. That approach relies on a clearly defined denominator and suitable model assumptions. Although this endpoint is reported per 100 person days, the registry record does not explain how person-time was accumulated, whether follow-up differed across participants, or how the rate was estimated.

It also does not describe how sick days were counted when a child had more than one illness, whether absence days were recorded by caregivers or centers, or how incomplete follow-up was handled. Such details affect the interpretation of both the numerator and denominator. Without them, the unit can be stated as recorded, but the mechanics underlying the rate cannot be reconstructed.

What would an arm comparison need to communicate?

In general, a comparison should identify the effect measure, its uncertainty, and the analysis population. For a count rate, an analysis might report a rate difference or rate ratio with a confidence interval, depending on the study question and model. A confidence interval communicates the statistical precision of the estimate under the model's assumptions; it is not a guarantee that the true effect lies within the interval. A hypothesis test may also be reported, but a p-value alone does not describe effect magnitude or precision.

No two-arm comparison, confidence interval, p-value, or model estimate is reported for this primary endpoint. The registry record explicitly indicates that values were missing or not estimable for a two-arm comparison and lists zero analyses. The reason for that status and any alternative analysis are not reported in the registry record. The two arm-level values therefore cannot be used to infer a test result or its precision.

How should randomization and masking be understood?

Random allocation is intended, in general, to reduce systematic differences between groups at assignment. Its value depends on the allocation process, follow-up, outcome measurement, and analysis. The record identifies the study as randomized and quadruple-masked, but the sequence-generation method, concealment procedure, and identities of those masked are not reported. Those missing details prevent a fuller assessment of how allocation and masking were implemented.

The study is listed as parallel-group. That means the record describes separate arms receiving the active drink or placebo, rather than each participant crossing between interventions. Whether allocation occurred at the child or daycare-center level is not reported. Because children were recruited in a daycare context, correlation among children within a center could be a consideration in general if allocation or outcomes were clustered by center. The registry record does not establish that such clustering occurred or explain whether any analysis accounted for it.

What can planning calculations tell readers?

In general, sample-size planning asks how much information a future study would need under specified assumptions about an effect, variability, significance level, allocation, and analysis. Such calculations are prospective planning illustrations, not evidence that a completed study had a particular probability of detecting an effect. They also do not replace an estimate and uncertainty interval from observed data.

The record describes planning assumptions involving a two-sided significance level of 0.05, equal allocation, a normal approximation, and effects expressed as fractions of an observed effect. It does not report a planning estimate in the registry record for this interpretation. More importantly, the primary outcome lacks an estimable two-arm comparison and the record supplies no reported variance or model estimate for it. Any planning illustration should therefore be distinguished from observed trial evidence, and it cannot fill the gap in the missing inferential result.

10. Multiplicity, Missing Data and Analysis Populations

The registry record lists one outcome measure and no posted secondary results. It also lists zero analyses. The record does not report whether multiple outcomes or multiple analyses were planned in the protocol, whether any interim analyses occurred, or whether the primary endpoint was selected from a broader set of measures. Those aspects of the statistical plan are not reported in the registry record.

Multiplicity is a general concern when several outcomes, time points, or statistical comparisons are tested. Repeated testing can increase the chance of a false-positive result unless the analysis plan addresses it. Here, no inferential primary comparison is posted, and no secondary outcome results are listed. The information available does not support an assessment of multiplicity adjustments or the overall testing strategy.

The registry record does not describe missing outcome data, reasons for missingness, or how missing observations were handled. It does not report the number of participants who completed follow-up, the number contributing to the endpoint analysis, or an imputation method. The arm counts are listed, but they should not be taken as evidence that every participant had complete outcome data.

Analysis populations are also not specified. The record does not state whether results were analyzed by assigned group, among participants who adhered to the intervention, or in another defined population. Nor does it report a prespecified statistical analysis plan or deviations from one. These omissions matter because the interpretation of a randomized comparison depends partly on who was included, how missing outcomes were addressed, and whether analysis decisions followed a plan established before results were known.

11. Safety Results

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

Other (non-serious) adverse events

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

Most frequent non-serious adverse event terms

The registry record reports zero serious adverse events in the placebo group, for which 95 participants are listed as at risk, and zero serious adverse events in the Bifidobacterium lactis (BB-12) group, for which 87 participants are listed as at risk. These are the serious-event counts recorded for the two arms. The record does not list serious adverse-event terms.

Other adverse events are reported for 3 participants in the placebo group and 3 in the active group. The record does not give event descriptions or indicate whether these were events, participants with events, or a particular event category beyond the “other” classification. The death fields are not reported in the registry record.

Zero reported serious adverse events should be read as a registry count, not as proof that no serious events occurred under all possible definitions or that the intervention is free of risk. The record does not provide details about adverse-event collection, definitions, monitoring period, severity grading, or causal assessment. Without those details, the safety figures are limited to the counts and at-risk group sizes recorded.

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

The trial record presents a randomized comparison of a probiotic yogurt drink with a placebo drink in daycare children ages 1 to 3. It reports arm-specific values for the primary sick-day endpoint over 90 days: 2.51 days per 100 person days for placebo and 2.82 days per 100 person days for Bifidobacterium lactis (BB-12). It also reports the arm sizes and a limited set of adverse-event counts. These details describe what is recorded, but they do not provide a formal estimate of the difference between groups.

The registry does not report a confidence interval, p-value, adjusted comparison, or statistical model for the primary outcome. It indicates that values were missing or not estimable for a two-arm comparison. As a result, the posted arm values cannot establish whether the groups differed beyond what could plausibly arise from sampling variability. They also do not quantify the precision of the comparison.

That limitation is not the same as evidence that the groups were equivalent. Establishing equivalence or non-inferiority generally requires an appropriate design, a prespecified margin, and an analysis that estimates uncertainty relative to that margin. Such information is not reported in the registry record. Conversely, the record also does not support a claim that the active drink reduced illness-related absences. The descriptive values alone do not settle either claim.

The record also leaves important interpretive details unclear: how the rate was calculated, how missing observations were managed, which analysis population was used, and what statistical method was planned or applied. The study is listed as randomized and quadruple-masked, but implementation details are absent. These gaps affect how confidently the endpoint values can be connected to the intervention.

The defensible conclusion is narrow: the registry records two arm-level values and does not provide an estimable inferential comparison for the primary outcome. The safety fields record zero serious adverse events in each arm and 3 other adverse events in each arm, but collection and classification details are not reported. The record therefore supplies descriptive information rather than a complete statistical account of the trial's primary question.

13. Why This Trial Matters Statistically

This study illustrates why a randomized design and arm-level outcome values are not enough to interpret a treatment comparison statistically. Randomization is an important design feature, but a reader also needs a clearly reported estimate of the between-group contrast and its uncertainty. Without those, descriptive results cannot distinguish a potentially meaningful difference from variation that may occur by chance.

The endpoint also highlights the importance of aligning the statistical method with the outcome's structure. Sick days are count data, and the reported unit uses person-time. In general, the analysis needs to account for the discrete outcome, the amount of observation time, and any extra variability or clustering that the design entails. The registry record does not report how these issues were handled. The absence of that information is particularly relevant because a two-arm comparison was not estimable in the posted results.

Registry reporting is useful not only for positive findings but also for identifying where interpretation is limited. This record names the primary outcome, timeframe, intervention, comparator, design features, and arm results. At the same time, it does not provide a model-based contrast, precision estimate, or details about missing data and analysis populations. Readers can therefore separate what is recorded from what remains unanswered, rather than treating arm-level values as if they were a complete efficacy analysis.

The study also shows why safety counts require context. Counts of serious and other adverse events are informative as recorded, but they do not explain how events were defined, sought, or attributed. A zero count should not be interpreted as a comprehensive demonstration of safety when the monitoring and classification procedures are not described.

14. 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 primary limitation for interpreting the trial is that the registry record does not report an estimable two-arm comparison for its primary outcome. The arm-specific values are available, but no confidence interval, p-value, model, or adjusted effect estimate is reported. Without a comparison estimate and its uncertainty, the primary research question cannot be answered from the posted outcome data alone.

The endpoint's calculation is not described in enough detail to assess its construction. The unit is days per 100 person days, but the registry record does not explain how sick days or person-time were defined, how incomplete follow-up was treated, or how the reported values were estimated. It also does not report a dispersion measure, which limits assessment of variability for this count-based outcome.

Information about missing data, participant flow, analysis populations, adherence, allocation procedures, masking implementation, and daycare-center clustering is not reported in the registry record. These are relevant to understanding whether the analysis reflects the randomized groups and whether the outcome data were measured and analyzed consistently. The record does not establish that any particular problem occurred; it simply lacks the detail needed to evaluate these aspects.

Safety interpretation is limited by the absence of event definitions, descriptions of the other adverse events, collection methods, follow-up details, and death counts. The registry lists the counts and arm sizes but not the context needed to assess completeness or attribution. The record also lists no secondary outcome results, leaving no reported secondary measures to help contextualize the primary endpoint.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Study to Investigate the Potential of Probiotics in Yogurt — NCT00492583.
• Registry record last updated 2011-08-17; status last verified 2011-06.
• Results first posted 2011-08-17.
• 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.

18. Record Summary

The registry record describes NCT00492583, “Study to Investigate the Potential of Probiotics in Yogurt,” as a completed, randomized, quadruple-masked, parallel-group phase 3 study sponsored by Georgetown University. It lists enrollment of 182 and a healthy population; the study summary describes daycare children ages 1 to 3. The active intervention was Bifidobacterium lactis (BB-12), compared with placebo, with both drinks described as 4 fluid ounces (112 grams) orally per day.

The primary endpoint was sick days per 100 person days over 90 days. Posted arm values are 2.51 for placebo and 2.82 for the active arm. The registry reports no analyses and indicates that a two-arm comparison was not estimable. Serious adverse-event counts are zero in each arm; other adverse events are reported for 3 participants in each arm. Further details about the statistical analysis, missing data, analysis populations, and adverse-event collection are not reported in the registry record.