Predicting Persistent Postconcussive Problems in Pediatrics (5P): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Predicting Persistent Postconcussive Problems in Pediatrics (5P) (NCT01873287), a registered phase not stated study. The primary outcome is Count of Children Who Have Persistent Post-concussive Symptoms (PCS) at One-month Follow-up..

2. Clinical Question

The study addresses prognosis after concussion, not whether one treatment is better than another. Its stated aims include determining one-month PCS incidence, deriving a prediction rule, and assessing that rule's accuracy, reliability, and acceptability in a subsequent cohort. These aims make the intended use of the study different from estimating a treatment effect: the focus is on identifying patterns of patient information associated with a later outcome.

The brief summary describes PCS as persistent somatic, cognitive, physical, psychological, or behavioural changes lasting more than one month following injury. The primary endpoint operationalizes the outcome at the one-month assessment with a specific PCSI change-from-baseline criterion. The record does not report how many participants met that criterion, the rule's predictors, or its predictive performance. Consequently, the research question is clear, but the posted registry information does not establish how well the rule performed.

3. Trial Design

This is a prospective, multicentre clinical prediction-rule derivation and validation study, not a randomized treatment trial. Its stated objective is to develop a rule for predicting PCS in children and adolescents presenting to an emergency department after acute head injury, then assess the rule in a subsequent cohort. The brief summary describes recruitment across nine large Canadian pediatric sites. The registry record identifies Children's Hospital of Eastern Ontario as the sponsor.

The age range in the study description is 5-17 years. The registered enrollment is 3063. The study began in 2013-08, had primary completion in 2015-06, and is marked COMPLETED. Those dates describe the registry timeline; they do not provide the duration of follow-up for each participant or the timing of any validation analysis.

Because the objective is prediction, the central methodological task is to relate information recorded around clinical presentation to a later PCS outcome. The planned use of clinical history and physical examination is stated in the brief summary. However, the registry record does not identify the candidate predictors, the rule-building procedure, how data were divided between derivation and validation, or the performance criteria used to assess accuracy and reliability. Those details should not be inferred from the stated objective alone.

The record names a primary outcome and two secondary outcomes: the PedsQL Total Score at 4-weeks Post-injury and Neuropsychological Evaluation. It lists three outcome measures overall, but the registry record does not report numerical results for these outcomes or explain the measurement procedures for the secondary outcomes. Allocation is not stated, and no randomized comparison is described. This design therefore supports an account of the intended prediction research, but the registry entry alone is insufficient to reconstruct the full analytic design or evaluate the rule's validation.

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: Completed 1-month followup.

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 Count of Children Who Have Persistent Post-concussive Symptoms (PCS) at One-month Follow-up., measured over 1 month.

The primary endpoint is the count of children with persistent postconcussive symptoms at one-month follow-up. The case definition requires an increase from pre-concussion baseline of three or more symptoms on the validated PCSI at one-month. The registry description says this criterion is consistent with the ICD-10 definition of PCS. The recorded population is children aged 5-17 years who completed 1-month followup.

A clearly defined endpoint is important in prediction research because the rule is intended to forecast a particular outcome. A symptom-based threshold turns a broad clinical concept into an operational outcome that can be applied to participants. The change-from-baseline element also makes the definition depend on symptoms relative to a pre-concussion reference, rather than simply on the presence of symptoms at follow-up.

The endpoint is a participant count, not a mean score or a time-to-event measure. The registry record does not report the number of participants who met the definition, the denominator completing follow-up, or the resulting incidence. It also does not report how missing baseline or follow-up PCSI information was handled. Thus, the endpoint definition is specified, but the registry entry does not provide the observed primary result needed to describe how frequently PCS occurred in the study population.

The named secondary outcomes concern quality of life and neuropsychological evaluation. Their presence indicates that the study registered outcomes beyond the primary symptom-based count, but the registry record does not provide their numerical results or enough detail to interpret their scales or assessment procedures.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Prediction studies and treatment trials answer different statistical questions. A treatment trial commonly compares outcomes under different assigned interventions. A prediction study instead asks whether information about a participant can help forecast a later outcome. The 5P study is described as deriving and validating a clinical prediction rule from information in history and physical examination. Allocation is not stated, and the registry record does not describe a randomized treatment comparison.

What does derivation mean?

Derivation is the process of building a prediction rule from observed participant information and outcomes. In general, a rule uses candidate predictors to estimate or categorize the chance of a defined outcome. Statistical modelling must account for how predictors are selected, how the outcome is coded, and how overly optimistic performance is avoided. These are general methodological considerations, not reported details of this study. The registry record does not state which predictors were considered, what modelling method was used, or how the rule was selected.

Why is validation separate from derivation?

A rule can appear to perform well in the data used to build it simply because it has adapted to patterns specific to those participants. Validation evaluates whether performance is retained when the rule is applied to data not used for its derivation. The brief summary says the investigators intended to assess the rule in a subsequent cohort. The registry record does not explain how that cohort was assembled, whether the validation was internal or external, or whether the rule was fixed before evaluation. Without those details and the performance results, the validity of the rule cannot be assessed from this entry.

What would prediction performance describe?

In general, discrimination describes how well a rule distinguishes participants who experience the outcome from those who do not. Calibration describes agreement between predicted risks and observed outcomes. A rule may separate higher-risk from lower-risk participants while still systematically overestimating or underestimating risk; these are distinct properties. A confidence interval, in general, conveys uncertainty around an estimate under the assumptions of the analysis. Wider intervals indicate less precision, while narrower intervals indicate greater precision. The registry record does not report discrimination, calibration, confidence intervals, or any other numerical measure of prediction performance.

Accuracy, reliability, and acceptability are named as aims, but these terms do not by themselves specify statistical methods. Accuracy could refer to different performance measures, and reliability may depend on the repeatability or consistency assessment chosen. Acceptability may require a defined measure or a qualitative assessment. The registry record does not state how any of these properties were operationalized. They therefore cannot be interpreted as achieved results merely because they appear in the study objectives.

How should the endpoint count be understood?

The primary outcome is a count of participants meeting a binary case definition: the participant either meets the PCSI criterion at one-month or does not. The criterion is an increase from pre-concussion baseline of three or more symptoms. In general, a count requires a denominator to be interpretable as a proportion or incidence. The registry record identifies the population as those completing 1-month followup, but does not report the number completing it or the number meeting the case definition. No incidence calculation can therefore be described from the entry.

Binary-outcome analyses can use methods that estimate the probability of an event or relate that probability to predictors. The choice depends on the research question, data structure, and assumptions. General considerations include whether observations are independent, whether predictor relationships are modelled appropriately, and whether there are enough outcome events for the planned model. Over-dispersion, in general, means that the variability in observed counts is greater than assumed by a chosen model; if ignored, uncertainty estimates can be too small. The registry record does not report the model family, model assumptions, event count, or diagnostic checks, so none of these issues can be evaluated for this study from the posted entry.

Sample-size planning is also distinct from observed power. A planning calculation uses assumptions about the outcome, effect to detect, significance level, allocation, and statistical method to illustrate how much information may be needed. The registry record does not report its sample-size justification or its own design assumptions. Any planning illustration based on an observed effect is not the trial's original design rationale and is not evidence of observed power. The stated assumptions for such illustrations include a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; however, the study is not described as a randomized two-arm comparison. These assumptions should not be mistaken for the actual prediction-rule analysis plan.

When a registry record contains no analysis results, there is no sponsor analysis to compare with a recalculation. No recalculated primary result is reported in this entry. In general, a recalculation from summary data can differ from an adjusted analysis because it may omit covariates, account for clustering differently, or use a different model. If a discrepancy between recalculation and sponsor analysis is identified, the sponsor's own analysis takes precedence. The registry record does not report the information needed to perform or assess such a recalculation for this outcome.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("PPCS in Derivation", "PPCS in Validation")
n      <- c(1701, 883)
events <- c(510, 291)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  PPCS in Derivation  1701  510  30.0  27.8  32.2  27.9  32.2  27.9  32.2
ARM  PPCS in Validation  883  291  33.0  29.9  36.2  29.9  36.1  29.9  36.1

11. Multiplicity, Missing Data and Analysis Populations

The record lists three outcome measures: the primary PCS count, PedsQL Total Score at 4-weeks Post-injury, and Neuropsychological Evaluation. Multiple outcomes can raise multiplicity concerns because examining many outcomes can increase the chance of at least one apparently notable finding by chance. Statistical plans may address this by identifying a primary endpoint, adjusting inference, or treating secondary outcomes as exploratory. These are general approaches. The registry record does not report a multiplicity strategy, a hierarchy among secondary outcomes, or an analysis plan for them.

The registry entry lists 0 analyses. It therefore does not give numerical analysis results for the primary or secondary outcomes. The registry record does not report whether analyses were adjusted for baseline characteristics, how predictors were selected, or what criteria were used to evaluate the prediction rule.

Missing outcome data can affect both estimated event frequency and prediction performance. If participants without follow-up differ from participants with complete assessments, an analysis restricted to completers may not represent the full enrolled group. General methods can include complete-case analysis, imputation, or sensitivity analyses, but their appropriateness depends on why data are missing and on the model. The registry record does not report the number missing at one-month, reasons for missingness, the handling of missing PCSI or predictor information, or any sensitivity analysis.

The record identifies completed 1-month followup as the primary outcome population, but does not provide a participant flow from enrollment to analysis. It also does not define analysis populations for derivation and validation. These omissions limit interpretation of both the endpoint denominator and the generalizability of any results, which are not posted in the registry entry.

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 safety entry lists 3063 participants at risk for serious adverse events, with 0 affected. For other adverse events, it also lists 3063 at risk and 0 affected. The death count is 0. These are the counts recorded in the registry; they are not a comparative estimate between treatment groups, because the study is not described as a randomized treatment comparison.

A count of 0 affected means no events were entered in that registry category for the recorded at-risk group. It does not, by itself, explain how events were sought, which event definitions were used, whether follow-up was complete, or whether all events were reported. The serious adverse event terms list is empty, so no specific serious event types are named. The registry record does not report further safety methods or context.

These counts should not be treated as a formal demonstration of absence of risk or as evidence about a treatment effect. The study's stated purpose is prediction of PCS, not evaluation of a treatment. The safety figures are best read as registry-recorded counts with the limitations of the available reporting.

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

The record supports a description of what the study set out to do: develop and validate a prediction rule for PCS in children and adolescents after acute head injury. It defines the primary case as an increase from pre-concussion baseline of three or more symptoms on the validated PCSI at one-month. It also records enrollment of 3063 and identifies secondary outcomes concerning quality of life and neuropsychological evaluation.

Those details do not show how often PCS occurred in the enrolled population. The primary outcome count and the number completing 1-month followup are not reported in the registry record. The entry also does not report coefficients, a final risk rule, predicted risks, performance measures, or validation findings. It is therefore not possible to judge from this record whether the rule accurately predicts PCS, whether its predictions are well calibrated, or whether it performed consistently in a subsequent cohort.

The distinction between an objective and a result is important. Stating that investigators intended to assess accuracy, reliability, and acceptability does not establish that any particular level of performance was observed. Likewise, the study's COMPLETED status does not substitute for numerical findings. The posted registry information describes the planned research and selected administrative and safety information, but it does not report the evidence needed to assess the prediction model.

The entry also does not support a treatment conclusion. No randomized treatment comparison is described, and the clinical question is prognostic rather than therapeutic. The adverse-event counts are registry entries, not comparative safety estimates. They do not establish treatment benefit or harm.

The appropriate interpretation is consequently limited: the registry record documents a large prospective, multicentre prediction-rule study with a specified symptom-based endpoint, but does not post its primary outcome or predictive results. Conclusions about incidence, accuracy, or clinical performance would require information not reported in the registry record.

14. Why This Trial Matters Statistically

This study is statistically notable for focusing on prognosis rather than a treatment contrast. Clinical prediction research requires a clearly defined outcome, candidate predictors measured in a relevant population, and evaluation of performance beyond the data used to build the rule. The 5P record specifies a one-month PCSI-based case definition and states an intention to derive a rule from history and physical examination, then assess it in a subsequent cohort. Those are central ingredients of a prediction question.

The distinction between derivation and validation is especially important. A rule's performance in its development data may overstate how well it will work elsewhere. A subsequent cohort can provide a stronger assessment if it is genuinely independent and the rule is evaluated as specified. The registry record does not give the cohort details or validation estimates, so the value of this planned step cannot be judged from the posted entry.

Statistical interpretation also depends on transparent reporting of denominators and uncertainty. The registry identifies 3063 participants enrolled, but does not report the number completing the primary follow-up or the number meeting the PCS definition. Without these figures, readers cannot assess the outcome frequency or precision. Without prediction-performance measures, readers cannot evaluate whether the rule meaningfully separates or estimates risk.

The record's omissions illustrate why study status and enrollment alone are not substitutes for results. A completed study can have an important objective and a carefully defined endpoint, yet an incomplete public record may still leave its main statistical conclusions unavailable. Here, the registry communicates the intended prediction research, but leaves the central empirical questions unresolved.

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 main limitation for interpreting this specific registry entry is that it does not report the primary PCS count, the number of participants who completed 1-month followup, or numerical results for the secondary outcomes. This prevents assessment of outcome frequency and precision. The entry lists 3063 as enrollment, but enrollment does not identify the denominator analyzed for the primary endpoint.

The registry record does not describe the candidate predictors, the modelling procedure, the handling of continuous or categorical variables, or the rule ultimately derived. It also does not explain how the subsequent validation cohort was assembled or whether the rule was evaluated without modification. These details are essential for evaluating overfitting and transportability, but should not be inferred from the study title or stated aims.

There is no reported analysis plan, multiplicity approach, missing-data procedure, or description of derivation and validation analysis populations. The record also does not provide confidence intervals or other precision estimates. Consequently, even if a point estimate were available, the uncertainty and assumptions behind it could not be assessed from the registry entry alone.

The brief summary is truncated, and the registered description does not give full operational details for recruitment or outcome assessment. The record names secondary outcomes but does not report their results or measurement details. It also gives safety counts without reporting ascertainment procedures or additional context. These limitations concern what the registry record documents; they do not establish that the underlying study lacked methods or analyses, only that those details are not reported in the registry record.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Predicting Persistent Postconcussive Problems in Pediatrics (5P) — NCT01873287.
• Registry record last updated 2019-11-18; status last verified 2019-10.
• Results first posted 2019-11-18.
• 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

NCT01873287, titled Predicting Persistent Postconcussive Problems in Pediatrics (5P), is a completed prospective, multicentre prediction-rule derivation and validation study sponsored by Children's Hospital of Eastern Ontario. The registered enrollment is 3063; the study began in 2013-08 and had primary completion in 2015-06.

The study concerns children and adolescents aged 5-17 years presenting after acute head injury. Its primary outcome is the count meeting a PCS definition at one-month: an increase from pre-concussion baseline of three or more symptoms on the validated PCSI. Secondary outcomes include PedsQL Total Score at 4-weeks Post-injury and Neuropsychological Evaluation.

The registry record does not report the primary outcome count, follow-up denominator, prediction-rule details, or performance results. Safety entries list 0 serious adverse events, 0 other adverse events, and 0 deaths among the recorded at-risk group of 3063 participants.