Sleep in Pediatric HSCT: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Sleep in Pediatric HSCT (NCT04106089), a randomized na study. The primary outcome is Patient Acceptability Questionnaire.

2. Clinical Question

3. Trial Design

The registry lists this study as randomized, single-masked, and crossover, with phase recorded as NA. It is titled Sleep in Pediatric HSCT and is sponsored by Children's Hospital of Philadelphia. The study began on 2019-11-18, had primary completion on 2022-12-01, and is listed as completed. Enrollment is recorded as 50.

The study description characterizes the design as an aggregated N=1 randomized controlled design: each patient serves as their own control, and randomization determines the 5-day intervention period. The intended comparison is a protected nighttime sleep window, created by extending the interval between vital-sign checks, against observation periods with regular checks. The two listed sequences are Observation-Intervention-Observation and Observation-Observation-Intervention. Each sequence contains three 5-day periods, but the registry record does not provide further detail on how periods were scheduled for individual patients or how the crossover was handled in the analysis.

The primary acceptability outcome was reported by 40 patients who completed the full study. The posted sequence-specific questionnaire groups contain 22 and 18 patients. For the actigraphy outcomes, the corresponding posted group sizes are 10 and 17. The reasons for these different numbers of observations are not reported in the registry record.

The record lists seven outcome measures, including the primary questionnaire and three posted actigraphy outcomes: sleep efficiency, total sleep time, and awakenings-minutes per day. It reports no analyses for the primary or secondary outcomes. The registry record does not report the original sample-size justification, a detailed statistical analysis plan, or how randomization and masking were implemented. Those details cannot be inferred from the crossover label alone.

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: 40 patients completed the full study and reported on acceptability at day 15

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 Patient Acceptability Questionnaire, measured over 15 days after transplant.

The primary endpoint was the Patient Acceptability Questionnaire, assessed 15 days after transplant. It was created for this study to evaluate patients' acceptability of adjusting vital-sign checks. Participants rated agreement with each statement on a 5-point Likert scale, from not at all (1) to extremely (5); higher values mean greater agreement. The posted outcome is a mean measured in units on a scale, accompanied by a standard deviation.

Acceptability is relevant to a behavioral and care-delivery intervention because changing the timing of routine checks depends in part on whether patients find the change acceptable. A questionnaire can capture patients' reported experience directly, whereas actigraphy measures aspects of sleep. The study description also identifies feasibility and supportive-care engagement as interests, but the primary endpoint reported in the registry is acceptability, not a combined measure of those concepts.

The registry record does not report the questionnaire's individual statements, how responses were combined into a score, or evidence about its measurement properties. It also does not report a threshold that would define an acceptable score or a meaningful difference. Accordingly, the means describe the posted ratings on the study's scale; they do not by themselves establish a validated cutoff or show how patients would interpret a particular difference.

7. Results Posted in the Registry

• Observation-Observation-Intervention: mean 3.55 versus 3.63 on control; difference -0.08 (95% CI -1.03 to 0.87), standardized effect d = -0.05; z-test p = 0.869.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Actigraphy Sleep Efficiency

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Observation-Observation-Intervention: mean 80.56 versus 74.63 on control; difference 5.93 (95% CI -8.27 to 20.13), standardized effect d = 0.34; z-test p = 0.413.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Observation-Observation-Intervention and the control. The interval gives the range of effects that remain compatible with the data.

Actigraphy Total Sleep Time

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Observation-Observation-Intervention: mean 474.19 versus 374.43 on control; difference 99.76 (95% CI -15.04 to 214.56), standardized effect d = 0.69; z-test p = 0.089.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Observation-Observation-Intervention and the control. The interval gives the range of effects that remain compatible with the data.

Actigraphy Awakenings-minutes Per Day

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Observation-Observation-Intervention: mean 108.94 versus 122.44 on control; difference -13.50 (95% CI -90.76 to 63.76), standardized effect d = -0.14; z-test p = 0.732.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Observation-Observation-Intervention and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes cover three related but distinct aspects of sleep: the proportion of time in bed spent asleep, minutes slept per day, and minutes of awakenings per day. Their posted group means point in different directions across measures: the Observation-Observation-Intervention group has higher mean sleep efficiency and total sleep time, and lower mean awakenings-minutes. The confidence intervals for all three comparisons include zero, leaving uncertainty about whether the differences reflect a consistent underlying pattern.

These actigraphy results complement, rather than replace, the primary acceptability outcome. Acceptability is a report about patients' agreement with questionnaire statements, while actigraphy provides sleep measurements. A favorable numerical direction on one measure does not imply a change in acceptability, nor does the primary questionnaire comparison settle the sleep findings. The registry does not report a joint analysis connecting acceptability with actigraphy or supportive-care engagement.

The actigraphy group sizes are 10 and 17 for each posted comparison, below the primary outcome group sizes. The reason for the smaller actigraphy samples and the handling of unavailable measurements are not reported in the registry record. The estimates may consequently be less precise, as their broad confidence intervals also indicate.

These secondary results carry less evidential weight than a prespecified primary analysis because multiple outcomes create more opportunities for chance patterns. In general, evaluating several outcomes without adjustment can increase the chance that at least one appears notable by chance alone. The registry record reports no analyses and does not say whether multiplicity adjustments or a hierarchy among secondary outcomes were planned. The secondary figures are therefore best treated as descriptive and hypothesis-generating, not as independent confirmation of the primary result or as definitive evidence of a consistent sleep effect.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Actigraphy Sleep Efficiency
• Actigraphy Total Sleep Time
• Actigraphy Awakenings-minutes Per Day

9. Statistical Interpretation

10. Statistical Methods Explained

The posted outcomes are continuous summaries expressed as means and standard deviations. A mean describes the average score or measurement in a group. A standard deviation describes how spread out individual values are around that average. It is not the uncertainty in the mean: uncertainty also depends on the amount of data and, in a crossover study, on how measurements from the same participant are related.

The primary outcome uses a questionnaire scale with responses from 1 to 5. The registry presents the result in units on a scale and reports a mean and standard deviation for each randomized sequence. The registry record does not describe how the questionnaire statements were combined, so the method used to produce a participant-level summary cannot be assessed from the record. Nor does it report whether the scale was analyzed as a continuous score in a prespecified statistical model.

A mean difference compares the two sequence-group means in the stated order. For the primary result, the comparison is Observation-Observation-Intervention relative to Observation-Intervention-Observation. A negative value means the posted mean in the first-listed comparison group is lower; a positive value would mean it is higher. This describes the arithmetic contrast, but does not by itself account for the crossover structure. Because the study description says each patient serves as their own control, a within-person analysis could use information from repeated periods. The registry reports no analyses, so it is not possible to determine whether such an analysis was conducted or how it was specified.

What does a confidence interval convey?

A confidence interval gives a range of estimates compatible with the data under the method used to construct it. In general, a wider interval indicates less precision, while a narrower interval indicates greater precision. An interval that crosses zero for a mean difference includes no difference among the values compatible with that interval; it can also include differences in either direction. This is not proof that the true difference is zero. The registry record does not identify the interval method or report an adjusted interval for the crossover design.

How should a p-value be read?

In general, a p-value measures how inconsistent the observed data are with a stated null hypothesis under a statistical model. It does not state the probability that the null hypothesis is true, and it does not quantify clinical importance. The primary comparison has an unadjusted p-value of 0.869. Since no registry analyses or detailed analysis plan are reported, the assumptions behind an official inferential test are not available for review. The posted p-value should be understood as an unadjusted summary rather than evidence that a particular crossover model was used.

Why does the crossover structure matter?

In general, crossover studies collect data from the same participants under more than one condition. Repeated measurements from one person tend to be related, and the timing or order of conditions can also matter. An analysis that treats all observations as independent may not represent that structure. Period effects, carryover, and within-person correlation are general issues to consider when evaluating crossover results; whether and how they were addressed in this study is not reported in the registry record. The two randomized sequences make timing explicit, but sequence-specific means alone do not show the within-person contrasts.

What can the posted secondary summaries add?

The actigraphy outcomes are means and standard deviations for sleep efficiency, total sleep time, and awakenings-minutes per day. These measures describe distinct aspects of sleep, in their stated units, over the period from the day of transplant through 15 days after transplant. Their group sizes are smaller than those for the primary questionnaire. The registry lists no reported analyses for these outcomes, so the posted comparisons do not establish whether an adjusted analysis was performed or how the multiple outcomes were handled.

Standard deviations also help explain why a group mean should not be mistaken for a typical individual response. A group can have a mean near the other group's mean while individual measurements vary substantially. In general, over-dispersion means that data vary more than a chosen model anticipates; if that occurs, a model's standard errors and intervals may be too optimistic unless it accommodates the extra variability. The registry does not report a model or a dispersion assessment, so there is no basis for asserting that over-dispersion occurred. These general considerations explain why means, variability, confidence intervals, and design-aware analyses all contribute different information.

11. Confidence Intervals and Precision

The primary mean difference is -0.08 scale units, with a reported confidence interval that extends from -1.033 to an upper bound recorded as 0.873. Because the recorded upper bound has more than three decimal places, it is not repeated; the important interpretation is that the interval crosses zero and includes values on both sides. This indicates that the point estimate is not precise enough to rule out differences in either direction under the interval method used. The interval's breadth should be considered alongside the scale and the posted standard deviations.

The group standard deviations for acceptability were 1.61 and 1.46, while the group sizes were 22 and 18. These describe variation and the amount of posted data contributing to the respective means. The registry gives no confidence intervals for the individual group means and no reported analysis that accounts for repeated measurements. Thus, the precision attached to the posted contrast cannot be assumed to reflect all the design features of the randomized crossover study.

For each secondary comparison, the reported interval also includes zero: sleep efficiency has an interval from -8.266 to 20.13 percentage points; total sleep time from -15.04 to 214.6 minutes per day; and awakenings-minutes per day from -90.76 to 63.76 minutes per day. These intervals are broad relative to their point estimates and include both directions of difference. They therefore leave substantial uncertainty about the size and direction of each contrast.

Confidence intervals describe statistical uncertainty, not the range of individual patient experiences or a guaranteed range for the true effect. In general, interval interpretation depends on the analysis method and its assumptions. The registry record does not report the methods used to calculate these intervals, nor does it identify intervals adjusted for multiple outcomes or crossover features. The figures are most useful as indicators of uncertainty around the posted contrasts, not as definitive boundaries on an intervention effect.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning figures are participants per arm, not total events. They are illustrations based on the observed effect in the posted comparison, not the original trial's design justification; that original justification is not reported in the registry record. The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set in turn to 100%, 75%, or 50% of the observed effect.

The 80% power column gives the estimated participants per arm for a design with an 80% chance of detecting the selected effect under those planning assumptions, if the effect and model assumptions hold. The 90% power column gives the corresponding participants per arm for a design with a 90% chance. These are prospective planning illustrations, not observed power for the completed study. They do not show the probability that the observed result is true or that a finding will replicate.

Smaller target effects require more participants because a smaller difference is harder to distinguish from random variation. That pattern is reflected in the rows for a smaller fraction of the observed effect. The figures are based on a two-arm equal-allocation illustration and should not be mistaken for a tailored calculation of this randomized crossover design. The registry record does not report the assumptions used to plan the actual study, and the planning figures do not account for details that are not specified, such as the crossover analysis method or missing outcome data.

The comparison-specific planning numbers should be interpreted cautiously when the observed estimate is uncertain. In particular, using a noisy observed effect as the basis for planning can yield unstable projections. These figures communicate how the required group size changes under the stated assumptions; they do not validate the original enrollment or establish that the study was designed to detect a particular clinically important difference.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = -0.080; pooled standard deviation σ = 1.545.
2. n = 2 × (1.96 + 0.842)² × 1.545² / -0.080² = 5,852.5, rounded up to 5,853 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- -0.0800000000000001                      # observed difference in means
sigma <- 1.54469635919337                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(4106089)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 5,853 (80% power), 7,835 (90% power)
 75% of observed effect: n per arm = 10,405 (80% power), 13,929 (90% power)
 50% of observed effect: n per arm = 23,411 (80% power), 31,340 (90% power)
power.t.test: 5853.5 
Simulated power at n = 5,853 per arm: 0.803 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Multiplicity, Missing Data and Analysis Populations

The registry lists seven outcome measures, while three secondary actigraphy outcomes are posted in addition to the primary questionnaire. It reports zero analyses. The record therefore does not state which outcomes were included in formal hypothesis tests, whether any analysis plan ranked outcomes, or whether the multiplicity from assessing several outcomes was addressed. In general, testing multiple outcomes without adjustment can increase the likelihood of chance-positive findings. The unadjusted secondary comparisons should be interpreted with that concern in mind.

The record does not report a prespecified analysis plan, a primary model, covariate adjustments, missing-data methods, or rules for including participants in each analysis. It also does not report whether results were analyzed according to randomized sequence, treatment period, or within-person contrast. Those omissions prevent readers from checking whether the posted summaries correspond to the intended crossover analysis or whether analytical choices were made before results were known.

For the primary outcome, 40 patients are described as completing the full study and reporting acceptability at day 15. The sequence-specific groups have 22 and 18 observations. The actigraphy outcomes have 10 and 17 observations in the two groups. The registry does not explain why outcome counts differ, whether missingness was related to participant characteristics or measurements, or how missing data were handled. It also does not specify whether the primary analysis population was all enrolled participants, all randomized participants, completers, or another group.

The total enrollment is 50, but enrollment alone does not establish how many participants were randomized, began the intervention, or contributed usable data to each endpoint. The record does not report those analysis-population details. Without them, the possibility of attrition or outcome-specific missingness cannot be assessed, and no missing-data adjustment can be assumed.

15. 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 adverse-event record lists zero serious adverse events in each sequence group and zero deaths in each. It gives 29 participants at risk for serious adverse events in the Observation-Intervention-Observation group and 21 at risk in the Observation-Observation-Intervention group. The record also lists zero other adverse events in the groups, with the same respective at-risk counts. No serious adverse-event terms are listed.

These figures mean that the registry reports no events in the listed categories for the stated at-risk groups. They should not be interpreted as proof that no adverse events could occur, or as a complete account of every possible safety outcome. The registry record does not give further detail about the event-collection process, event definitions, observation window, or the relationship between the at-risk counts and the primary outcome groups. The counts are safety reporting summaries, not a statistical comparison designed to establish equivalence between sequences.

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

The posted primary result describes similar average acceptability ratings across the two randomized sequences, with a mean difference of -0.08 and an unadjusted p-value of 0.869. Its confidence interval crosses zero and allows values in either direction. This is an uncertain estimate, not evidence that the sequences are identical. Nor does the p-value establish that a difference is absent or that the intervention is acceptable to every patient.

The result addresses a study-created questionnaire at 15 days after transplant. It does not directly measure sleep, and it does not establish effects on all aspects of supportive care. The actigraphy outcomes provide separate sleep summaries, but their intervals also include zero and the registry reports no analyses. Taken together, the posted figures do not settle whether protecting a nighttime window changes sleep or patient experience.

The study's crossover design is important to the meaning of the comparisons. Although the description says each patient serves as their own control, the reported means are presented by randomized sequence. The registry does not report a within-person analysis or explain how periods and repeated observations were modeled. Consequently, the posted group comparison should not be treated as a definitive estimate of an individual-level intervention effect.

These findings also do not demonstrate that the study was sufficiently large to detect a particular effect. The planning illustrations use the observed effects and stated assumptions; the original sample-size justification is not reported. Similarly, the absence of recorded adverse events in the listed categories is not proof that the intervention has no safety risks. The registry summaries answer narrower questions: what mean ratings and actigraphy measurements were posted for the recorded groups, and what adverse-event counts were entered.

In general, a clinical result is best interpreted by considering the outcome's meaning, the size and precision of the estimate, the design, missing data, and analysis methods together. Here, important analysis and missing-data details are not reported in the registry record. The posted numbers are informative descriptions of this completed study, but they do not support a stronger conclusion than their uncertainty and the available documentation allow.

17. Why This Trial Matters Statistically

This study raises a statistical question that is easy to overlook: the randomized schedule is not simply two independent groups receiving different treatments. The registry describes a crossover in which each patient serves as their own control, with intervention timing randomized across sequences. That structure can make within-person comparisons informative, but it requires analysis that recognizes repeated observations and the order of periods. The registry's sequence-specific means do not reveal whether that structure was incorporated into an inferential model.

The primary outcome also illustrates the difference between a meaningful research question and a precise estimate. Patient acceptability is central to whether a change in routine care is workable, yet the posted questionnaire comparison has an interval spanning both directions. A nonsignificant unadjusted p-value cannot resolve whether the underlying difference is negligible, meaningful, or simply estimated imprecisely. Scale interpretation and prespecified thresholds would help readers judge practical importance, but neither is reported in the registry record.

The contrast between enrollment, primary outcome reporting, and actigraphy group sizes makes outcome availability another statistical issue. Different outcomes have different numbers of observations, while reasons and missing-data methods are not reported. Readers therefore cannot determine whether the groups contributing to each outcome are comparable or how missing measurements may influence the results.

Finally, several outcomes are relevant to sleep and care engagement, but no registry analyses are reported and multiplicity handling is not described. Reporting descriptive estimates with uncertainty is valuable, but the absence of an analysis plan limits how strongly the results can be interpreted. These design and reporting features make the study a useful example of why estimates, intervals, analysis populations, and model alignment all matter alongside a point estimate.

18. 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 registry record leaves several trial-specific questions unresolved. Although it identifies a randomized, single-masked crossover design and gives two 5-day sequences for each phase, it does not explain how masking was implemented, who was masked, or how sequence allocation was generated and concealed. It also does not report the actual sample-size justification or how the study's aggregated N=1 design was reflected in the analysis.

There are no analyses listed for the primary or secondary outcomes. The record does not describe a crossover model, within-person estimates, handling of period or sequence effects, covariate adjustment, or the method used to calculate the reported intervals. The comparisons therefore cannot be evaluated as adjusted estimates of the randomized intervention effect from the registry information alone.

Outcome availability differs: 40 patients completed the full study and reported acceptability, while actigraphy summaries are based on smaller groups. The record does not give reasons for these differences or describe missing-data handling or analysis populations. It also does not report the individual questionnaire items, scoring procedure, or a threshold for interpreting acceptability.

Other limitations include the lack of reported multiplicity procedures despite several outcomes, and the absence of a reported link between the acceptability and sleep measures. The record includes adverse-event counts but not detailed definitions or collection methods. These limitations concern what can be established from the registry record; they do not show that undocumented methods were or were not used.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Sleep in Pediatric HSCT — NCT04106089.
• Registry record last updated 2024-10-18; status last verified 2024-10.
• Results first posted 2024-10-18.
• 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.

22. Record Summary

NCT04106089, Sleep in Pediatric HSCT, is listed as a completed randomized, single-masked crossover study sponsored by Children's Hospital of Philadelphia. It enrolled 50 participants and compared sequences involving regular vital-sign checks with a 5-day period of extended time between checks intended to protect nighttime sleep.

The primary posted outcome is a study-created acceptability questionnaire at 15 days after transplant. The mean ratings were 3.63 and 3.55 across the two sequences, with a mean difference of -0.08 and an unadjusted p-value of 0.869. Three actigraphy outcomes are also posted. The registry reports no analyses and does not provide the original sample-size justification or detailed analysis methods.