Predict the Best Level of Care Placement for Each Child's Behavioral Health Needs - Effectiveness Study: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Predict the Best Level of Care Placement for Each Child's Behavioral Health Needs - Effectiveness Study (NCT06834763), a registered phase not stated study. The primary outcome is Mean Difference in Average Domain Z-scores Across Raters Within Two Weeks on the Clinical Scale of the Treatment Outcome Package (TOP-CS) Between a) Baseline and b) Follow-up.

2. Clinical Question

The study asks whether children and adolescents have better behavioral-health outcomes when they are in a placement that PSP identifies as near the placement with the highest predicted likelihood of success for them. PSP is described as a site-customized machine-learning tool that provides placement-specific predictions. The primary hypothesis concerns whether outcomes differ between clients whose follow-up placement is within one standard deviation of their highest-predicted placement and clients whose placement is not within that range. A secondary hypothesis concerns whether improvement over time is positively associated with the number of days spent in at least one placement within that range.

The registry record does not report the primary outcome estimates or a statistical comparison between the groups. The clinical question is therefore clear, but the posted information does not answer it with an observed treatment-effect estimate.

3. Trial Design

This is an effectiveness study of clinical decision support in a naturalistic setting. The intervention is access by the clinical team to Placement Success Predictor results. PSP is described as using site-customized machine-learning models to predict the likelihood of a good outcome for a youth in each placement type. The study compares clients according to whether their placement at follow-up is within one standard deviation of the placement with the highest predicted likelihood of success for that client.

The registry describes two placement categories: “In PSP-Recommended Placement” and “Not In PSP-Recommended Placement.” The definitions refer to the client’s placement and PSP results at follow-up. The record does not describe randomized assignment; allocation is not stated, and the groups are not a like-for-like control comparison. Because group membership is based on placement status at follow-up, the categories represent observed placement circumstances, not documented randomized treatment assignments.

The listed enrollment is 403. The study status is completed, with a start date of 2025-02-03 and a primary completion date of 2026-02-26. The primary outcome is an average of domain z-scores across raters on the Clinical Scale of the Treatment Outcome Package, assessed at baseline and follow-up. The stated timing is baseline within 2 weeks of study start and follow-up approximately 60-120 days later.

The registry lists a secondary outcome comparing a risk-adjusted predicted TOP-CS total score at baseline with the actual average total score across raters at follow-up. It also describes a secondary hypothesis relating improvement over time to the number of days in a placement within the PSP-defined range. The record lists two outcome measures but zero analyses. Specific statistical procedures, covariate adjustments, sample-size justification, and details about how the study handled follow-up timing 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 Mean Difference in Average Domain Z-scores Across Raters Within Two Weeks on the Clinical Scale of the Treatment Outcome Package (TOP-CS) Between a) Baseline and b) Follow-up, measured over At baseline (within 2 weeks of study start) and approximately 60-120 days later.

The primary endpoint is the mean difference in average domain z-scores across raters on the Clinical Scale of the Treatment Outcome Package (TOP-CS), between baseline and follow-up. Baseline is defined as within 2 weeks of study start; follow-up is approximately 60-120 days later. The outcome description says the reported value is the mean difference in the average z-score between those assessments. A negative mean difference indicates improvement, meaning reduced symptom severity.

TOP-CS is intended to measure behavioral-health symptoms and well-being over the past 2 weeks. The Child Treatment Outcome Package has 48 items and assesses 13 domains for children ages 3 - 18. The Adolescent TOP-CS has 58 items and assesses 12 domains for adolescents ages 11 - 21. The named domains include Depression, Violence, and Suicidality. Responses range from “All” to “None of the Time” on a 6-point Likert scale. The TOP-Case Mix companion form assesses stressful life events and comorbidity, and the outcome description says scores are risk-adjusted for case-mix variables assessed using 37 items on that form.

A domain z-score of 0 represents the general population mean. The domain scores are averaged to form a summary score per participant; higher, more positive average z-scores indicate greater symptom severity and lower behavioral well-being. An average across domains provides a broad summary, while necessarily combining information from distinct areas of functioning. The registry record does not report the observed endpoint values, the mean difference, or its uncertainty, so the size and precision of any change cannot be assessed from the posted results.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is continuous: it summarizes TOP-CS z-scores rather than counting whether a participant did or did not experience an event. For a continuous outcome, a group mean describes the average score in that group, while a mean difference expresses the contrast between two averages on the outcome’s scale. The registry identifies the primary measure as a mean and its dispersion as a standard deviation. It does not report the actual means, standard deviations, or mean difference.

A standard deviation describes how spread out individual measurements are around their group mean. It is not the same as the uncertainty in the estimated mean. A confidence interval, if reported, would describe the range of effect estimates compatible with the data under the analysis assumptions. A narrow interval would indicate greater precision than a wide interval, but neither its width nor its location is available in the registry record. A p-value, if reported, would address compatibility with a specified null model; it would not by itself describe the size or practical importance of an effect. No p-value is reported.

The TOP-CS outcome is constructed from domain z-scores. A z-score expresses a domain score relative to a reference distribution, and the registry defines 0 as the general population mean. The domain z-scores are averaged into a participant-level summary. The registry states that more positive average values indicate worse symptom severity and lower behavioral well-being. It also states that a negative baseline-to-follow-up mean difference indicates improvement. The score’s direction matters: a reader should interpret any numerical estimate using the registry’s stated convention rather than assume that a positive change is favorable.

The Child TOP-CS and Adolescent TOP-CS differ in item count and number of assessed domains. The child version has 48 items and 13 domains for ages 3 - 18; the adolescent version has 58 items and 12 domains for ages 11 - 21. Both use a 6-point response scale from “All” to “None of the Time” and assess the past 2-week experience. The outcome description says case-mix risk adjustment uses information from 37 items on the companion TOP-Case Mix form, including stressful life events and comorbidity. The exact score-construction and adjustment procedures beyond that description are not reported in the registry record.

Repeated measurement at baseline and follow-up can help describe change over time, but the analysis must account appropriately for the relationship between measurements from the same participant. General statistical practice distinguishes a comparison of follow-up values from an analysis of within-participant change; those approaches answer related but not identical questions. The registry identifies a baseline-to-follow-up mean difference, but the detailed statistical model, treatment of repeated measures, and handling of measurements from multiple raters are not reported. The endpoint title says scores are averaged across raters, but the number of raters and the rules for combining their assessments are not reported.

The study’s placement groups are defined from observed follow-up placement relative to the highest PSP-predicted likelihood of success. In general, an observational group comparison can be affected by differences in participant characteristics or care circumstances between groups. Random assignment can reduce some forms of such imbalance, but allocation is not stated, and the record explicitly cautions that this is not a randomized comparison. No adjusted group-effect estimate or adjustment variables are reported. The registry also does not describe how the one-standard-deviation placement range was calculated beyond the group definitions.

Over-dispersion is a general concern when count outcomes vary more than a simple count model assumes; methods that account for over-dispersion can prevent uncertainty from being understated. That issue is not directly the primary endpoint’s stated form, which is a continuous mean score rather than a count. The registry does not report the model used for the secondary hypothesis about days in a placement, so it is not possible to assess how that association was analyzed or how its uncertainty was estimated.

The record describes planning illustrations based on the observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. Such figures are planning illustrations, not observed power and not trial outcomes. The stated effect to detect is a fraction of the observed effect; the record notes that time-to-event planning figures would be total events and other figures participants per arm. No specific planning counts or sample-size justification are reported in the registry record. The record does not state that these illustrations were the basis of the trial’s design.

10. Confidence Intervals for Each Arm

No two-arm comparison is shown for this trial, but each arm's mean can be given with a t-based 95% confidence interval computed from the posted mean, spread and number of participants.

R code and output

arm <- c("In PSP-Recommended Placement, Child Sample", "Not In PSP-Recommended Placement, Child Sample", "In PSP-Recommended Placement, Adolescent Sample", "Not In PSP-Recommended Placement, Adolescent Sampl")
n   <- c(66, 129, 39, 169)
m   <- c(-0.09, -0.15, 0.09, -0.16)
sp  <- c(0.54, 0.51, 0.49, 0.63)
sd  <- sp      # posted standard deviations
for (i in seq_along(n)) {
  se <- sd[i] / sqrt(n[i]); h <- qt(0.975, n[i] - 1) * se
  cat(sprintf("CARM|%s|%d|%.3f|%.3f|%.3f|%.3f\n", arm[i], n[i], m[i], sd[i], m[i] - h, m[i] + h))
}

CARM  In PSP-Recommended Placement, Child Sample  66  -0.090  0.540  -0.223  0.043
CARM  Not In PSP-Recommended Placement, Child Sample  129  -0.150  0.510  -0.239  -0.061
CARM  In PSP-Recommended Placement, Adolescent Sample  39  0.090  0.490  -0.069  0.249
CARM  Not In PSP-Recommended Placement, Adolescent Sampl  169  -0.160  0.630  -0.256  -0.064

11. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary outcome and one secondary outcome, and the brief summary describes a primary and a secondary hypothesis. The primary outcome is the mean difference in average TOP-CS domain z-scores across raters from baseline to follow-up. The listed secondary outcome compares a risk-adjusted predicted TOP-CS total score at baseline with an actual average total score at follow-up. A further secondary hypothesis concerns the association between improvement over time and days spent in a placement within the PSP-defined range.

When several outcomes or hypotheses are examined, the chance of obtaining at least one apparently notable finding can increase. General methods for handling multiplicity include prespecifying a hierarchy of tests or adjusting inference across a defined family of comparisons. The registry record does not report a multiplicity strategy, a testing hierarchy, or which analyses were considered confirmatory. It also lists zero analyses, so no outcome-specific statistical findings are available to interpret.

Missing-data methods are not reported in the registry record. The posted information does not state how incomplete TOP-CS forms, missed follow-up assessments, or participants without an evaluable baseline and follow-up were handled. It does not report the number of participants with endpoint data, reasons for missingness, or any sensitivity analyses. The enrollment of 403 should not be read as the number analyzed.

Analysis populations are also not reported in the registry record. It is not stated whether analyses included all enrolled clients, only clients with follow-up measurements, or another defined group. The record does not specify how participants were assigned to analysis groups if placement status changed, or how the number of days in PSP-defined placements was determined for the secondary hypothesis. These omissions prevent assessment of the analysis population and of how missingness or changes in placement could affect the estimates.

12. 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 lists serious adverse event figures separately for child and adolescent samples and for clients in and not in PSP-recommended placement. The at-risk counts are 66 and 129 for the two child groups, and 39 and 169 for the two adolescent groups. Serious adverse events are recorded as 0 in each of the four groups. The listed counts for other adverse events are also 0 in all four groups, and deaths are recorded as 0 in all four groups.

These figures describe the adverse-event entries in the registry record. They do not establish that no event occurred outside the reported collection or recording process, and they do not provide details about monitoring or event definitions. The record does not list serious adverse-event terms. The group counts should not be interpreted as randomized safety comparisons: the placement categories are not a randomized comparison, and allocation is not stated. Differences in group size also make raw counts unsuitable as a stand-alone comparison of risk.

The safety figures do not address whether PSP predictions were accurate or whether TOP-CS outcomes changed. Those are separate questions. The registry record does not report the primary outcome estimates, so the absence of recorded adverse events cannot fill that gap or establish effectiveness.

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

The registry record describes an evaluation of PSP in a naturalistic behavioral-health setting. Clinical teams have access to placement-specific predictions, and the study asks whether clients whose follow-up placement is within one standard deviation of the highest predicted likelihood of success have better outcomes than clients outside that range. The primary outcome is a mean difference in average TOP-CS domain z-scores across raters between baseline and follow-up.

The record does not report the primary endpoint estimate, group-specific results, uncertainty intervals, or hypothesis-test results. It therefore does not show the direction, magnitude, or precision of any difference in TOP-CS outcomes. The record’s completed status and enrollment of 403 do not provide the missing outcome analysis. Nor does the listed secondary outcome have a posted result in the registry information.

Even with an outcome estimate, the comparison would need to be interpreted in light of the nonrandomized design. Allocation is not stated, and clients are grouped based on their placement status at follow-up relative to PSP’s predictions. The record does not report adjustment methods or demonstrate that the groups were comparable on factors that could influence both placement and outcomes. A group association would not, by itself, establish that PSP access or recommended placement caused an outcome difference.

PSP’s predictions and participants’ observed TOP-CS scores are different kinds of information. The study is intended to examine how placement relative to the predictions relates to outcomes, but the record does not report predictive performance, model calibration, or validation results. The posted information therefore describes a research question and measures, not evidence that the predictions were accurate or that a particular placement improved outcomes.

The safety entries show zero recorded serious adverse events, other adverse events, and deaths in each listed group. They do not substitute for the missing effectiveness analysis. Overall, the record establishes what was planned to be evaluated and how the primary outcome is defined, while leaving the central comparative result and its statistical precision unreported.

14. Why This Trial Matters Statistically

This study raises a statistical question about individualized prediction in a setting where placement decisions are made for children and adolescents with behavioral-health needs. PSP is intended to estimate the likelihood of a good outcome across placement types for each youth, rather than provide only a single prediction for all clients. Evaluating such a tool requires attention to both the predictions and the outcomes observed after care decisions.

The registry frames the primary comparison around whether a client is in a placement within one standard deviation of the placement with the highest predicted likelihood of success. This makes the meaning of the comparison depend on how predictions are generated and how placement status is classified. The record describes the broad rule but does not report model-performance statistics or the detailed prediction procedures. Without those details and outcome estimates, readers cannot assess how well predicted likelihoods corresponded to observed TOP-CS outcomes.

The study also illustrates the distinction between prediction and causal evidence. A model may identify patterns associated with outcomes, but a comparison between clients who do and do not receive a predicted placement is not automatically a causal test. Here the registry states that the comparison is not randomized and that allocation is not stated. Differences in client needs or care circumstances could be related to both placement and outcome, and the analysis methods needed to evaluate those possibilities are not reported.

Finally, clear reporting of effect estimates, uncertainty, analysis populations, and missing-data methods is essential for evaluating a completed study. The record provides the endpoint definition and enrollment but lists zero analyses. Consequently, it does not permit assessment of the primary finding or its precision. This distinction matters statistically: a clearly specified question is not the same as a reported answer, and the presence of a completed study status does not replace the information needed to evaluate the evidence.

15. 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 principal limitation of the posted record is that it contains no primary outcome analysis. No group-specific averages, mean difference, confidence interval, or p-value are reported. It also lists zero analyses. As a result, the direction, size, and precision of the primary comparison cannot be evaluated from the registry information.

The study is not described as randomized, and allocation is not stated. The groups are based on whether clients’ follow-up placements fall within the PSP-defined range. This can make the groups systematically different, and the record does not report an adjustment strategy or variables used to address possible differences. The comparison should not be treated as a like-for-like control comparison or as establishing causation.

The enrollment is 403, but the registry does not report the number with complete baseline and follow-up TOP-CS assessments, the number included in each outcome analysis, or the analysis population. Missing-data procedures, sensitivity analyses, and reasons for missing observations are not reported. The number and handling of raters, detailed score-combination procedures, and the exact statistical model are also not reported in the registry record.

The record does not provide details of the site-customized models, their predictive performance, or calibration. It also does not report the protocol’s sample-size justification, interim analyses, or a statistical analysis plan. A planning illustration is not a substitute for those design details. The completed status and the safety entries do not resolve these limitations, and the zero recorded adverse events do not establish how completely adverse events were ascertained.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Predict the Best Level of Care Placement for Each Child's Behavioral Health Needs - Effectiveness Study — NCT06834763.
• Registry record last updated 2026-08-07; status last verified 2026-08.
• Results first posted 2026-07-09.
• 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.

19. Record Summary

The registry record describes a completed naturalistic effectiveness study of clinical-team access to Placement Success Predictor results. PSP is intended to make placement-specific predictions using site-customized machine-learning models. The listed enrollment is 403, and the study compares clients in a PSP-defined placement range at follow-up with clients outside that range. Allocation is not stated, and the record cautions that this is not a randomized comparison.

The primary outcome is the mean difference in average TOP-CS domain z-scores across raters between baseline and follow-up, with a negative mean difference defined as improvement. The record lists one secondary outcome and two outcome measures overall, but zero analyses. It does not report primary outcome estimates, uncertainty, missing-data procedures, or analysis populations.

Serious adverse events, other adverse events, and deaths are recorded as 0 in each of the four listed child and adolescent placement groups. These entries do not answer the primary effectiveness question. The posted information describes the study question and outcome definitions but does not provide the comparative results needed to assess them.