Reopening Schools Safely and Educating Youth Research Study (ROSSEY): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Reopening Schools Safely and Educating Youth Research Study (ROSSEY) (NCT04859699), a randomized na study. The primary outcome is School Attendance.

2. Clinical Question

3. Trial Design

ROSSEY is a randomized, parallel-group study with schools as the unit of randomization. The registry describes the design as a cluster-randomized controlled trial involving K-5 schools. In a cluster design, groups rather than individual participants are assigned to study arms. Here, the clusters are schools, while the attendance endpoint is measured for index children. This distinction matters when assessing how much information the trial provides: outcomes from people attending the same school may be more alike than outcomes from people in different schools.

The registry lists seven schools in the control arm and seven in the intervention arm. Control schools and enrolled families were not to receive the comic books and videos; intervention schools and families were assigned the COVID-19 health education materials. The study is recorded as unmasked, so masking is not described as part of the design. The registry lists enrollment as 1095 and status as completed. The recorded study start is 2022-01-17 and primary completion is 2024-05-08.

The record describes three aims. One is a qualitative assessment of rural Latino community social, ethical, and behavioral needs and resources related to returning students to school and maintaining onsite learning. Another is the school-based cluster-randomized trial. The remaining aim is a qualitative assessment of implementation outcomes for the Yakima School District testing program and risk communication intervention, involving school stakeholders, parents, and children. These aims encompass more than the primary attendance comparison, so the trial record should not be read as describing only a single outcome or activity.

The phase is recorded as NA. The registry record does not report additional details of allocation procedures, a sample-size justification, or the statistical analysis plan. It also does not provide a numerical result from the sponsor's analysis of the cluster-randomized comparison. Those absences limit what can be concluded from the posted attendance information.

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.

Before assignment: Parent-child dyads were counted individually and the number of participants reflects each individual in the study. Some parents had more than one child (siblings) included in the study.

Overall Study

Analysis population for the primary outcome, as defined in the registry: Total number of school days missed by students (index child only) due to symptoms of respiratory illness at Post Survey 1 (Control = 191; Intervention = 217)

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.

Population: Consented kindergarten through 5th grade students and parents enrolled in participating school district; children includes index and siblings. Children in the intervention arm = 325 (246 index and 79 siblings); Children in the control group = 307 (217 index and 90 siblings).

6. Primary Endpoint

The registry lists the primary outcome as School Attendance, measured over Up to nine months.

The primary endpoint is school attendance expressed as the mean number of school days missed because of respiratory illness symptoms. The measurement is limited to the index child and is assessed at Post Survey 1 (T2), within a timeframe of up to nine months. The stated outcome therefore focuses on absence associated with a specified reason, rather than all absences or a general measure of attendance.

A mean is a summary of a continuous measure: it describes the average number of missed days in a group. In general, this kind of endpoint can make group-level attendance patterns easier to describe than a list of individual values. The standard deviation, if reported, describes how spread out individual observations are around the group mean. Neither arm-specific mean nor standard deviation is reported in the registry record.

The registry's primary outcome population description gives total school days missed as Control = 191 and Intervention = 217. Those are reported as totals, not as arm means. They do not, on their own, indicate how the missed days were distributed among index children or how many children contributed data. The record does not report the denominators used for the primary outcome, an estimated difference between group means, a confidence interval, or a test result. Consequently, the endpoint definition is clear, but the posted information is insufficient to quantify the comparative result or its precision.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is a continuous measure summarized by a mean, with standard deviation identified as its dispersion measure. The registry does not report the arm means or standard deviations, and it does not report a sponsor analysis estimate for the primary comparison. The methods below explain general statistical principles relevant to this type of trial; they are not claims about analyses performed by ROSSEY unless a method is explicitly described in the registry record.

Why does the cluster design matter?

General statistical explanation: when assignment occurs by group, observations within a group can be correlated. Children attending the same school may share features of the school environment, policies, calendar, or local context. If an analysis assumes that all child-level observations are independent, it may understate uncertainty because it effectively treats related observations as more independent information than they are. Cluster-aware methods account for the school-level assignment and the within-school relationship when estimating uncertainty.

The registry explicitly identifies schools as the randomization unit and cautions that a simple individual-level re-analysis would be misleading. The number of schools recorded in each arm is seven. The number of individual participants does not replace the number of randomized clusters when considering how much independent randomized information is available. The exact cluster-aware model or variance method used by the sponsor is not reported in the registry record.

What does a mean and standard deviation describe?

General statistical explanation: the mean is the arithmetic average of observed values. The standard deviation describes the spread of those values around the mean; it is not the uncertainty in the estimated mean. A standard error or confidence interval addresses uncertainty in an estimate and depends on the study design and analysis, including clustering when assignment is by school. A larger spread among individuals and stronger similarity within schools can affect the precision of a group comparison.

For this endpoint, the individual value is the number of school days missed by the index child because of respiratory illness symptoms. The registry gives arm totals of missed days, not the distribution of individual values. A total does not reveal whether missed days were concentrated among a small number of children or spread more evenly. It therefore cannot substitute for the group means and variability needed to describe the outcome distribution or estimate a comparative mean difference.

How should uncertainty be assessed?

General statistical explanation: a confidence interval gives a range of values compatible with the data and the statistical model at the stated confidence level. Its width communicates precision; a narrow interval indicates greater precision than a wide interval, all else equal. A confidence interval for a treatment contrast can also show which effect sizes remain compatible with the observed data. The registry record does not report a confidence interval for the primary outcome.

A p-value, when used, assesses how compatible the data are with a specified null hypothesis under an analysis model. It does not measure the size or importance of an effect, and it does not establish that the null hypothesis is true when it is not small. No primary-outcome p-value or other test result is reported in the registry record. Therefore, the posted attendance totals do not permit an assessment of statistical evidence under a specified test.

What do the planning illustrations mean?

General statistical explanation: sample-size calculations describe the amount of information that a future or planned study might require under stated assumptions. They are planning illustrations, not observed power for a completed trial. The registry describes illustrations based on a fraction of the observed effect, with a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation. It states that non-time-to-event illustrations are expressed as participants per arm. These assumptions do not establish what assumptions were used to design ROSSEY.

The record does not report the numeric planning results or a numeric observed effect in the primary outcome description. Consequently, those illustrations cannot be interpreted from the registry record as a justification for the original enrollment or as evidence that the trial had adequate precision. In general, planning calculations depend on the effect size and variability assumptions as well as design features. For a cluster-randomized study, the role of clustering also needs to be considered; the particular assumptions used for any planning calculation are not reported.

Why should unadjusted recalculation be treated cautiously?

General statistical explanation: a recalculation from summary statistics can sometimes reproduce a simple unadjusted comparison, but it may not match an analysis that accounts for the study design or adjusts for prespecified factors. In a cluster-randomized study, a calculation that treats children as independently assigned would not reflect school-level randomization. The registry specifically notes that an individual-level re-analysis would be misleading and that only the sponsor's own analysis is shown. The numerical sponsor analysis is not reported in the registry record, so no recalculated estimate or comparison can be described.

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("Control Group", "COVID-19 Comic Books and Videos")
n   <- c(191, 217)
m   <- c(6.57, 7.34)
sp  <- c(7.72, 8.45)
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  Control Group  191  6.570  7.720  5.468  7.672
CARM  COVID-19 Comic Books and Videos  217  7.340  8.450  6.209  8.471

11. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary attendance outcome and four secondary outcomes: emotional regulation of index children at baseline (T1), perceived stress of parents at baseline (T1), emotional regulation of index children at Post Survey 1 (T2), and perceived stress of parents at Post Survey 1 (T2). These outcomes address different measures and time points. The record also describes qualitative aims, which are not represented by these listed quantitative outcomes alone.

When a study evaluates multiple outcomes or multiple comparisons, the chance of at least one apparently notable result can increase if each is assessed separately without accounting for multiplicity. General approaches can include prioritizing a primary outcome, prespecifying a testing sequence, or adjusting inference across a family of tests. The registry record does not report whether any multiplicity adjustment, testing hierarchy, or other strategy was planned or used. The number of analyses is recorded as zero, and numerical inferential results are not reported for the listed outcomes.

The registry record also does not report how missing outcome data were handled, whether any participants or schools were excluded from a particular analysis, or which analysis population was used. The number contributing to the primary endpoint is not stated. These omissions matter because the analyzed population and missingness can affect both the estimated group comparison and its uncertainty. No assumptions about missing-data mechanisms, imputation, or analysis inclusion should be inferred from the completion status alone. The corresponding details are not reported in the registry record.

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 events, other adverse events, and deaths for parent and child groups. For Control Group -Parent, the record lists 217 at risk, 0 serious events, 0 other events, and 0 deaths. For Health Communication Strategies Parent, it lists 246 at risk, 0 serious events, 0 other events, and 0 deaths.

For Control Group - Child, the record lists 307 at risk, 0 serious events, 0 other events, and 0 deaths. For Health Communication Strategies - Child, it lists 325 at risk, 0 serious events, 0 other events, and 0 deaths. No top serious-event terms are listed.

These are registry-reported counts for the listed groups. They indicate that no events were recorded in those categories in this record; they do not establish that no events occurred outside the recorded scope, or that the intervention has no safety concerns. The registry record does not provide further details about event collection, definitions, observation periods, or how the listed parent and child groups relate to the primary attendance analysis. These safety counts should therefore be read separately from the attendance outcome and its unreported comparative analysis.

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

The registry record establishes the study's broad design and primary measurement: K-5 schools were randomized, and the primary outcome was the index child's mean number of days missed from school because of respiratory illness symptoms at Post Survey 1 (T2). The comparison was between schools assigned to receive COVID-19 health education comic books and videos and control schools that were not to receive those materials. The outcome population description reports total missed days for each arm, but not the group means or a statistical comparison.

What the posted record does not establish is whether the intervention changed attendance. It does not report a mean difference, an uncertainty interval, a test result, or an analyzed sample size for the primary outcome. The two totals cannot be interpreted as a comparative effect without information about the number of children contributing and the distribution of outcomes. The school-level randomization also means that a valid estimate of uncertainty must reflect clustering.

Likewise, the registry record does not establish that the intervention had no effect. A missing estimate is not the same as an estimate close to zero, and a completed status does not supply the missing numerical analysis. The appropriate conclusion from the record is limited: the study was designed to assess an attendance outcome in a cluster-randomized setting, and the recorded totals are available, but the comparative result and its precision are not reported in the registry record.

The listed adverse-event counts are also limited to the categories and groups recorded. They should not be combined with the attendance results or treated as a comprehensive account of all possible safety outcomes. The study's qualitative aims address community needs and implementation experiences; the registry description identifies those aims but does not provide their findings. As a result, neither the effectiveness comparison nor the qualitative findings can be characterized from the information recorded.

14. Why This Trial Matters Statistically

This trial is statistically notable because the unit assigned to study arms is the school, while the primary outcome is measured for children. That structure is common in studies where an intervention is delivered through an institution or community setting. It requires the analysis to respect the level of randomization rather than treating every individual outcome as if it came from an independent assignment.

The registry directly signals this issue: a simple individual-level re-analysis would be misleading, and only the sponsor's own analysis is appropriate to represent the clustered comparison. That caution is consequential because the number of children enrolled and the number of schools randomized describe different aspects of the study. For inference about a school-assigned intervention, the clustering can influence the uncertainty around an estimated effect.

The outcome also illustrates why the form of a reported statistic matters. The primary endpoint is defined as a mean, while the registry's outcome description gives totals of missed days by arm. Totals do not provide the same information as averages and dispersion, and they cannot alone support an estimate of the comparative effect. Clear reporting of the estimand, contributing population, group summaries, and cluster-aware uncertainty would be needed to interpret the trial's attendance comparison.

Finally, the study combines a randomized evaluation with qualitative aims concerning community needs and implementation. Those components can answer different kinds of questions: a randomized comparison assesses outcomes by assigned arm, while qualitative assessment can describe perspectives and experiences. The registry record identifies these aims but does not report the corresponding findings. The distinction helps readers avoid treating the trial description or outcome counts as a complete account of either effectiveness or implementation.

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 for interpreting the primary outcome is that the registry record does not report the arm means, standard deviations, denominators, estimated comparison, confidence interval, or test result. It gives total missed-day counts of Control = 191 and Intervention = 217, but these totals do not show the distribution of absences or the average for an index child. Without a numerical analysis, the size and precision of any group difference cannot be assessed.

The cluster-randomized structure further limits what can be recovered from the posted summary. Schools, rather than individual children, were randomized, and the record lists seven schools in each arm. The registry cautions against simple individual-level re-analysis. The specific sponsor model and its handling of within-school correlation are not reported in the registry record, nor is a numerical sponsor result available to interpret.

Other key analysis details are also absent. The registry record does not report the number contributing to the primary endpoint, the handling of missing data, the analysis population, or any multiplicity strategy for the primary and secondary outcomes. It does not state how the follow-up observations were selected for the endpoint beyond the stated Post Survey 1 (T2) timing and timeframe of up to nine months. These details could affect the interpretation and precision of an outcome estimate, but their impact cannot be determined from the record.

Finally, the registry description of the qualitative aims does not include their findings, and the adverse-event listings do not include additional event-collection details. These limitations concern the completeness of the available record, not proof of a particular result or a flaw in the conduct of the study. The registry record does not report whether additional analyses or findings are available elsewhere.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Reopening Schools Safely and Educating Youth Research Study (ROSSEY) — NCT04859699.
• Registry record last updated 2025-10-08; status last verified 2025-09.
• Results first posted 2025-10-08.
• 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

ROSSEY is a completed randomized, parallel-group study recorded as unmasked, with K-5 schools as the unit of randomization. The registry lists enrollment as 1095 and identifies COVID-19 health education comic books and videos as the behavioral intervention. Control schools were not to receive those materials. The study includes qualitative aims as well as a school-based cluster-randomized trial.

The primary outcome is the index child's mean number of school days missed because of respiratory illness symptoms at Post Survey 1 (T2), within a timeframe of up to nine months. The record reports totals of Control = 191 and Intervention = 217, but does not report arm means or a comparative statistical analysis. Because the assignment was clustered by school, the record cautions against a simple individual-level re-analysis. The numerical sponsor analysis, analysis populations, missing-data methods, and multiplicity approach are not reported in the registry record.