Trial of KEEP-P, a Preventive Intervention for Foster Preschoolers (KEEP-P): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Trial of KEEP-P, a Preventive Intervention for Foster Preschoolers (KEEP-P) (NCT03106636), a randomized na study. The primary outcome is Parenting Stress Via Self-report Measure Parenting Stress Index IV (PSI-IV) Short Form.

2. Clinical Question

3. Trial Design

This was a randomized, parallel-group trial with single masking, according to the registry record. The record does not identify who was masked, so the masking designation cannot be interpreted more specifically. Enrollment was 442. The sponsor was the University of Oregon, and the listed condition was child behavior. The trial status is completed.

The two listed behavioral interventions were KEEP-P and KEEP-P+. The KEEP-P arm used the basic curriculum, delivered in a 2-hour weekly group format for 12 weeks. KEEP-P+ was described as an augmented curriculum that added information about recent findings in early brain development and a video coaching component. Details beyond that description are not reported in the registry record.

The registry's brief summary says the investigators planned to examine parenting, disrupted placements, and child outcomes. The primary outcome listed for this analysis is parenting stress, assessed using the Parenting Stress Index IV Short Form. The stated timeframe is change from baseline at endpoint, defined as 3-4 months post-baseline. The record also lists 16 outcome measures, but does not provide a single reduced comparison for this primary outcome.

Although the study is marked completed, the registry record does not report outcome estimates, measures of uncertainty, or a statistical test for the primary comparison. It also does not report a detailed statistical analysis plan, the rationale for the enrollment target, or the role of the person or people masked. The endpoint and design can therefore be described, but the trial's estimated comparative effect cannot be reconstructed from the information 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

Analysis population for the primary outcome, as defined in the registry: The number analyzed in the endpoint and change rows is different from the baseline row because we were not able to retain every participant for the post-intervention data collection.

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 Parenting Stress Via Self-report Measure Parenting Stress Index IV (PSI-IV) Short Form, measured over Change from baseline at endpoint (3-4 months post-baseline).

The listed primary endpoint is parenting stress measured with the Parenting Stress Index IV (PSI-IV) Short Form. The measure is based on the parent's perception of the child's characteristics, the parent's personal characteristics, and the interaction between parent and child. It contains 36 items using a 5-point Likert scale from “strongly agree” to “strongly disagree.” Item 32 is reverse coded; for the other items, higher scores indicate higher stress. The total score is calculated by summing the item scores and ranges from 36 to 180.

The registry describes the target as change from baseline at endpoint, with endpoint occurring 3-4 months post-baseline. Change is useful when the question concerns how a measured characteristic shifts from its starting level over a defined follow-up period. A comparison of changes between randomized groups can describe whether their observed trajectories differ. This is a general explanation of the role of a change score, not a statement about the analysis actually used in this trial.

The measure focuses on perceived stress in the parent-child relationship, rather than directly measuring placement disruption or child outcomes. The registry record does not report group-specific baseline or endpoint scores, change estimates, or a between-group result. Consequently, the scale's measurement description is available, but the observed stress outcome is not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry record describes random allocation, two parallel intervention groups, and single masking. Randomization is generally intended to make the groups comparable at the time of assignment, on average, so that an outcome difference can be interpreted as a comparison of assigned interventions. That general rationale does not establish that all potential sources of bias were eliminated in this particular trial. The record does not give the randomization procedure, allocation concealment details, or the identity of the person or people masked.

What does a mean and standard deviation describe?

The primary outcome is continuous: it is a total score on a scale. A mean describes the average score in a group, while a standard deviation describes the spread of individual scores around that average. Together they summarize central tendency and variation, but they do not show the full score distribution. These are general statistical definitions. The registry record does not provide the groups' outcome means or standard deviations for the primary comparison.

For a change outcome, each participant's follow-up score is considered in relation to baseline. A group mean change describes the average within-group shift, while a comparison of mean changes describes the difference between groups in those shifts. A within-group change alone does not answer the randomized comparative question: both groups could change over time, and the question is whether their changes differ. This is a general distinction. The registry record specifies a change-from-baseline timeframe but does not identify the exact model or contrast used to analyze it.

Why does uncertainty matter?

An estimate is subject to sampling variability: another sample from the same target population could yield a different value. A confidence interval is a way to express uncertainty around an estimate under the assumptions of the analysis. A narrow interval indicates greater precision than a wide interval, but precision is not the same as practical importance. A p-value is a measure tied to a specified null hypothesis and statistical model; it is not the probability that a treatment is effective, nor does it communicate the size of an effect. These explanations are general. No confidence interval or p-value for the primary comparison is reported in the registry record.

The PSI-IV Short Form is bounded, with total scores from 36 to 180. In general, the scale's limits and score distribution can matter when selecting and interpreting a statistical model. A mean-based summary can be useful, but the distribution, outliers, and concentration near a scale boundary may affect how well a model's assumptions fit. The record does not report distributional diagnostics or a particular model, so no claim can be made about those features in this trial.

What does over-dispersion mean?

For count outcomes, over-dispersion means the observed variability exceeds what a chosen count model expects. If it is ignored, standard errors and resulting uncertainty estimates may be too small. This general concept is not a description of the primary endpoint here, which is a continuous score, and the registry record does not report an over-dispersion analysis. It is relevant only as a general example of why model assumptions should match the outcome type.

How should planning calculations be read?

A sample-size calculation is a planning exercise: it asks what sample size might be needed under specified assumptions about the target effect, significance level, allocation, and model. The record's planning assumptions describe an illustration based on the observed effect, with a two-sided significance level of 0.05, equal allocation between arms, a normal approximation, and effects expressed as a stated fraction of the observed effect. These assumptions are not evidence that the original trial was designed using that calculation. The registry record does not report the trial's own sample-size justification.

Any sample-size illustration based on an observed effect is conditional on that effect estimate and the calculation's assumptions. It does not describe the trial's observed power. Power is a design concept tied to a specified effect and assumptions; calculating power after observing results does not establish whether a finding is true or important. The registry record does not provide numerical outputs for such planning illustrations in the registry record for this analysis, so no sample-size result can be stated.

The registry record identifies 16 outcome measures and reports no analysis count. It also indicates that the primary outcome has several categories or time points and was not reduced to a single comparison. As a result, the available information does not establish which contrast, model, or inferential procedure should be treated as the primary statistical result. Any more specific account of covariate adjustment, repeated-measures modeling, handling of clustering, or interim monitoring is not reported in the registry record.

10. Multiplicity, Missing Data and Analysis Populations

The registry record lists 16 outcome measures and indicates that the primary outcome involves several categories or time points rather than one reduced comparison. When a study examines multiple outcomes, time points, or contrasts, the chance of obtaining at least one apparently unusual result can increase if each is tested separately. General approaches to address multiplicity include designating a primary contrast in advance, adjusting inferential thresholds, or interpreting secondary analyses as exploratory. Which approach was used in this trial is not reported in the registry record. No single primary-outcome comparison is reported, so the multiplicity strategy and its effect on interpretation cannot be assessed.

For the parenting-stress outcome, the registry says that the number analyzed in endpoint and change rows differs from the baseline row because not every participant could be retained for post-intervention data collection. This is a documented limitation in follow-up. Missing outcome measurements can reduce precision and can also introduce bias if the probability of missingness is related to participant characteristics or outcomes. That is a general concern, not a claim about the reason for missing data in this trial.

The registry record does not report the number missing for the primary endpoint, reasons for missingness, imputation methods, sensitivity analyses, or the analysis population used. It therefore does not show whether the reported analyses, if any, included all randomized participants or only those with observed follow-up data. These details are important because conclusions can depend on how missing measurements and analysis populations are handled. The registry record also does not report a protocol-defined hierarchy for outcomes or a correction for multiple comparisons.

11. 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 registry record lists serious adverse events and other adverse events by intervention group. For KEEP-P, 226 participants were at risk in each of the listed serious and other event summaries, and 0 were affected in either category. For KEEP-P+, 190 participants were at risk in each summary, and 0 were affected in either category. The record lists 0 deaths in both groups. No serious-event terms are listed.

These figures describe what was recorded for the stated groups and denominators. They do not establish that no adverse events of any kind occurred beyond the categories and reporting captured in the registry record. A zero count is not proof of zero underlying risk; it means that no affected participants are listed in these entries. The registry record does not provide further event definitions, follow-up details for safety reporting, or a discussion of ascertainment.

The safety denominators should not be treated as the number analyzed for the parenting-stress endpoint. The record separately notes that endpoint and change analyses involved fewer participants than the baseline row because some participants were not retained for post-intervention data collection. The relationship between the safety populations and the endpoint analysis population is not reported in the registry record.

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

The registry record supports a description of a randomized trial comparing two behavioral programs for foster preschoolers and their caregivers. It identifies the primary measure as the Parenting Stress Index IV Short Form and specifies a change-from-baseline endpoint at 3-4 months post-baseline. It also describes the basic KEEP-P curriculum and the additional components of KEEP-P+.

It does not report a single primary-outcome comparison or numerical results for parenting stress. As a result, there is no reported effect estimate to describe, no uncertainty interval to evaluate, and no statistical test result to interpret. It is not possible from the registry record to determine whether the groups differed in mean stress change, how large any difference was, or how precise an estimate might have been. The lack of a reported result should not be interpreted as evidence that the programs were equivalent or that there was no difference.

The registry's brief summary describes intended areas of investigation, including parenting, disrupted placements, and child outcomes. Stated study aims are not findings. The record does not provide numerical results for those broader aims in the registry record for this analysis. Likewise, the safety entries listing zero affected participants do not answer the primary parenting-stress question and do not establish that all possible harms were absent.

In general, a trial's findings are interpreted by considering the estimate, its uncertainty, the analysis population, missing data, outcome measurement, and the prespecified analysis plan. Here, important analysis details and the primary comparative result are not reported in the registry record. The record therefore describes what the trial set out to compare and how the primary construct was measured, but it does not support a conclusion about the comparative outcome.

13. Why This Trial Matters Statistically

This trial is statistically relevant because it compares two versions of an intervention rather than an intervention with no-treatment control. KEEP-P+ is described as adding components to the basic KEEP-P curriculum, so the design addresses a contrast between the basic and augmented versions. The interpretation of such a comparison depends on having a clearly specified outcome contrast and a reported estimate with uncertainty.

The primary measure is a continuous score with a defined range and a specified follow-up timeframe. A change-from-baseline question asks about the difference in score trajectories between randomized groups, not simply whether each group changed from its own baseline. That distinction is important in randomized trials because within-group change alone does not isolate a comparative effect.

The registry also identifies a follow-up limitation: the number analyzed at endpoint and for change differs from the baseline count because not every participant was retained for post-intervention data collection. Missing follow-up can affect both precision and validity, depending on its extent and pattern. Without details on how missing measurements were handled, readers cannot fully evaluate this issue.

Finally, the primary outcome is not reduced to a single comparison in the registry record, despite the listing of 16 outcome measures. That makes it difficult to identify which contrast should carry the primary inferential weight. The statistical importance of this record lies in the contrast between a clearly described trial framework and the absence of a reported primary comparison. The design and measurement details help define the research question, but they cannot substitute for the outcome estimate needed to evaluate it.

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 most important limitation for interpreting the listed primary outcome is that the registry record does not report a single comparison or numerical result for parenting stress. It supplies the outcome definition and timeframe but not group scores, changes, an effect estimate, or uncertainty. The comparative result therefore cannot be evaluated from the registry record.

The record states that the endpoint and change rows include fewer participants than the baseline row because some participants were not retained for post-intervention data collection. It does not give the number missing for this outcome, reasons for missingness, or methods used to address it. Missing data may affect interpretation, but the degree and direction of any effect cannot be determined from the information reported.

The primary outcome has several categories or time points and is not reduced to a single comparison. The registry lists 16 outcome measures, but the statistical plan for selecting or adjusting analyses is not reported in the registry record. The analysis model, covariate handling, and analysis population are also not reported.

Some design details are limited. The trial is identified as single-masked, but the masked role is not specified. The KEEP-P+ description names additional early brain development information and a video coaching component, but further details are not reported in the registry record. The record also does not report the trial's sample-size justification. These gaps limit evaluation of the intended analysis and the intervention contrast, independently of the missing primary result.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Trial of KEEP-P, a Preventive Intervention for Foster Preschoolers (KEEP-P) — NCT03106636.
• Registry record last updated 2020-07-14; status last verified 2020-07.
• Results first posted 2020-07-14.
• 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 a completed randomized, parallel-group trial with single masking, sponsored by the University of Oregon and enrolling 442 participants. It compares the basic KEEP-P curriculum with the augmented KEEP-P+ program for foster preschoolers and their caregivers. KEEP-P was delivered in 2-hour weekly group sessions over 12 weeks; KEEP-P+ added early brain development information and a video coaching component.

The listed primary outcome is parenting stress measured by the PSI-IV Short Form, with change from baseline assessed at 3-4 months post-baseline. The record explains the measure and notes reduced retention for endpoint data collection. It lists 16 outcome measures but does not report a single primary-outcome comparison or its numerical results. Consequently, the comparative parenting-stress finding and its precision are not reported in the registry record.