Non-Pharmacological Interventions on Sleep in Post-Acute Rehabilitation: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Non-Pharmacological Interventions on Sleep in Post-Acute Rehabilitation (NCT00333619), a randomized na study. The primary outcome is Pittsburgh Sleep Quality Index.

2. Clinical Question

This is a comparison of two behavioral approaches, not a comparison with no contact: the control group received structured social visits. That distinction matters when interpreting what the trial contrast represents. Any group difference would reflect the assigned multicomponent sleep program relative to the active-control experience, rather than the effect of receiving attention or structured visits versus receiving nothing. The registry summary describes the broader project as testing sleep/wake patterns and functional recovery as well as sleep, but the posted primary endpoint is the Pittsburgh Sleep Quality Index at 3-month follow-up.

3. Trial Design

The registry describes a randomized, triple-masked, parallel-group trial. Participants were allocated to the nonpharmacological sleep intervention or an active control. The record lists enrollment of 219. The intervention arm had 110 participants in the posted primary outcome results, and the active-control arm had 109. The registry does not identify which people or roles were masked, or how masking was maintained; those details are not reported in the registry record.

The intervention had several coordinated components: sleep assessment, environmental changes intended to support sleep and wakefulness, a structured bedtime routine, and cognitive-behavioral elements. The active-control condition offered daily 15-minute social visits with structured social activities. This comparison therefore evaluates the intervention package as a whole. The registry does not report separate randomized comparisons of the individual intervention components, so their distinct contributions cannot be separated from these results.

The trial status is recorded as completed. The recorded start is 2007-01 and primary completion is 2010-12. The sponsor is the US Department of Veterans Affairs. The registry lists the phase as not applicable. Further design details, including the trial sites, eligibility criteria, allocation sequence procedures, masking roles, and the original sample-size justification, are not reported in the registry record.

The primary endpoint was the mean Pittsburgh Sleep Quality Index score at 3-month follow-up, summarized separately for each arm. The posted arm summaries give a mean of 8.6 with a standard deviation of 3.8 for the intervention and a mean of 9.4 with a standard deviation of 3.8 for the active control. These are outcome summaries, not estimates adjusted for baseline values or other covariates. Whether the sponsor's ANOVA included adjustments beyond the group comparison is 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

Reasons for not completing this period:

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 Pittsburgh Sleep Quality Index, measured over 3-month follow-up.

The primary endpoint was the Pittsburgh Sleep Quality Index (PSQI) at 3-month follow-up. It is an 18-item questionnaire measuring subjective sleep quality and sleep disturbances. The registry describes a total score ranging from 0 to 21 and states that a score > 8 indicates poor sleep quality. A lower score corresponds to fewer reported sleep problems on this scale; a group mean, however, describes an average and does not show how many individuals fell above or below a threshold.

A continuous score can preserve information across the scale rather than reducing each participant's experience to a simple category. Comparing arm means addresses whether the average reported PSQI score differed between randomized groups at the specified follow-up. The standard deviation helps describe how spread out individual scores were around each arm's mean. Neither summary alone describes the full distribution, the number of participants meeting a threshold, or changes from baseline.

The PSQI is used to operationalize subjective sleep quality and sleep disturbance, outcomes aligned with the intervention's stated aim. It is a self-reported measure, so the endpoint represents participants' questionnaire responses rather than an objective sleep recording. The registry record does not report a separate objective sleep measure in the primary result, a baseline PSQI result, or an interpretation of a clinically important difference for this score. The observed arm means should therefore be read as group-level questionnaire summaries at 3-month follow-up, not as a complete account of each participant's sleep or functional recovery.

7. Results Posted in the Registry

• Nonpharmacological Sleep Intervention: mean 8.60 versus 9.40 on control; difference -0.80 (95% CI -1.81 to 0.21), standardized effect d = -0.21; z-test p = 0.119.

8. Statistical Interpretation

9. Statistical Methods Explained

The comparison has two related forms: the sponsor's posted analysis and a simple unadjusted re-calculation from the arm-level summary statistics. The registry reports an ANOVA analysis with p = 0.10. The re-calculation estimates the difference in group means from the reported means and standard deviations, yielding -0.8, with an unadjusted confidence interval and p-value. These should not be treated as the same analysis: the re-calculation uses only posted summary information and does not reproduce any unreported adjustments or modeling choices in the sponsor's analysis.

What does a mean difference represent?

For a continuous outcome, the mean is the arithmetic average of participants' scores in an arm. A mean difference compares those averages. The contrast is expressed as intervention minus active control: a negative value indicates a lower average in the intervention arm, and a positive value indicates a higher average there. In this trial, the reported contrast is -0.8 PSQI units. This is a group-average comparison at 3-month follow-up; it is not the average change from baseline, because baseline scores are not included in the posted result.

Means can be useful because they summarize the location of a group distribution in a single interpretable scale unit. They can also conceal variation in individual responses or unusual values. The standard deviation describes the spread of scores around the mean, but it does not tell whether scores were symmetric, clustered, or affected by outliers. The record does not report distribution plots, medians, ranges, or other summaries that would help assess those features.

What does the standard deviation contribute?

The standard deviation is a measure of variability among participant scores, not a measure of how precisely the mean has been estimated. The posted standard deviation is 3.8 in each arm. Precision of a mean, and of the difference between means, also depends on the number of observed participants and the analysis method. The arm counts in the posted outcome results are 110 and 109. The re-calculation uses these group summaries to estimate uncertainty around the mean difference. It does not account for unreported features such as covariate adjustment, repeated measurements, or missing-data procedures.

Because the outcome is continuous, an analysis of means is a natural way to express the average group contrast. The sponsor identifies ANOVA as its method. In general, analysis of variance compares average outcomes across groups within a statistical model. With two groups, a basic unadjusted ANOVA addresses a group mean contrast, though the registry's brief method label alone does not establish whether the model included covariates or other features. Those specifics are not reported in the registry record.

How should the p-values be understood?

In general, a p-value measures how inconsistent observed data are with a specified null hypothesis under the assumptions of the analysis. It is not the probability that the null is correct and does not indicate the probability that the intervention will help a particular person. The registry reports p = 0.10 for the ANOVA. The summary-statistic re-calculation reports p = 0.119. The latter is labeled unadjusted and is calculated from the posted group summaries. The two values differ because they represent distinct reported analyses or calculation procedures; the registry does not give enough model detail to explain the difference. The record does not report a prespecified decision threshold or whether the reported p-value was adjusted for multiple outcomes.

A conventional threshold such as 0.05 is sometimes used to organize decisions about statistical evidence, but crossing or not crossing that threshold does not measure clinical importance. The uncertainty interval adds information about plausible effect sizes, while the point estimate gives one estimated contrast. A p-value by itself cannot show whether the likely effect is small or large, whether estimates are precise, or whether the scale difference matters to participants.

Why can over-dispersion matter in general?

For outcomes modeled as counts, over-dispersion means the observed variability exceeds what a model such as a simple Poisson model expects. If ignored, uncertainty can be understated and tests can appear more decisive than warranted. That issue is not the central feature of this primary endpoint: PSQI is a continuous score analyzed as a mean, not a count outcome. The registry does not report an over-dispersion assessment or a distributional diagnostic. More generally, assumptions about the outcome distribution, independence of observations, and model fit matter for valid standard errors and inference, but the record does not provide diagnostic results.

What does randomization contribute?

In general, random assignment supports a causal comparison by making treatment groups comparable in expectation, including with respect to measured and unmeasured baseline factors. It does not guarantee identical groups in a particular trial, especially in finite samples. Baseline PSQI and other baseline characteristics are not reported in the posted results, so balance cannot be assessed from this record. The randomized design is a strength for interpreting the assigned-group contrast, but conclusions also depend on adherence, follow-up, missingness, and the analysis population. Those details are not reported in the registry record.

The study is listed as triple-masked, which indicates masking at three roles or levels, but the registry does not state which people were masked. In general, masking can reduce the influence of expectations on care, assessment, or reporting. Since the PSQI is subjective, who knew assignments and whether that could affect questionnaire responses are relevant questions. The label alone does not answer them.

Finally, the planning illustrations use the observed effect as a reference and vary the effect to be detected. Their stated assumptions are a two-sided significance level of 0.05, equal allocation, and a normal approximation; the selected effect is the stated fraction of the observed effect. For this continuous endpoint, the results are expressed as participants per arm. Such planning calculations describe hypothetical designs under their assumptions. The registry's own original sample-size assumptions and justification are not reported.

10. Confidence Intervals and Precision

The re-calculated mean difference is -0.8 PSQI units, with a 95% confidence interval from -1.807 to slightly above zero. A confidence interval describes a range of effect estimates compatible with the observed data and the calculation's assumptions. In repeated use of the same procedure under its assumptions, intervals constructed this way would cover the fixed underlying effect at the stated rate. It is not a range containing a specified proportion of individual participant scores, nor a probability statement that the true effect lies within this one realized interval.

The interval includes zero, which corresponds to no average difference between groups. It also spans values on either side of zero, so the posted data do not pin down the direction with high precision. Its lower limit permits a larger reduction in the intervention group than the point estimate, while the upper limit permits a small difference in the other direction. The result therefore supports a range of interpretations rather than a single precise effect size.

The standard deviations are 3.8 in both arms, and the endpoint sample sizes are 110 and 109. These quantities contribute to uncertainty in the mean difference. The record does not report a confidence interval for the sponsor's ANOVA result, and the re-calculated interval is unadjusted. Any difference between unadjusted and adjusted estimates, if an adjusted analysis was performed, cannot be evaluated from the registry record.

Statistical precision and clinical importance are separate. Even a narrow interval would not establish that an effect matters to patients without an appropriate clinical interpretation; a wider interval can leave substantial uncertainty even when the point estimate is small. The record does not state a minimally important PSQI difference or an equivalence margin. Consequently, the interval cannot be used to conclude that the groups were equivalent. It indicates uncertainty around the average contrast, not the full range of individual experiences.

11. 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 participants across both arms and not total events. The calculation assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each planning scenario uses the stated fraction of the observed effect as the effect to detect. The registry's original design assumptions and sample-size justification are not reported in the registry record.

The 80% power column gives the planned number of participants per arm for a hypothetical design with an 80% probability of rejecting the null hypothesis if the assumed effect is the true effect, under the calculation's assumptions. The 90% power column gives the corresponding per-arm number for a 90% probability. Power is a property of a planned study under specified assumptions; it is not a statement about whether this trial's result is true or whether the intervention is clinically important.

At 100% of the observed effect, the planning illustration gives 355 participants per arm for 80% power and 475 for 90% power. Planning for 75% of the observed effect gives 630 participants per arm for 80% power and 843 for 90% power. Planning for 50% of the observed effect gives 1,417 participants per arm for 80% power and 1,897 for 90% power. These figures illustrate that smaller effects require more participants to detect under the same assumptions. The estimates are based on the observed effect, which is uncertain; using it as a planning reference does not make it a known underlying effect.

These numbers are a planning illustration, not observed power. They do not show the probability that the trial's finding is correct, and they do not replace the confidence interval as a description of uncertainty in the observed contrast. The figures also do not establish what sample size the investigators originally planned or why they chose it.

12. 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 = -0.800; pooled standard deviation σ = 3.800.
2. n = 2 × (1.96 + 0.842)² × 3.800² / -0.800² = 354.2, rounded up to 355 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 <- -0.800000000000001                      # observed difference in means
sigma <- 3.8                     # 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(333619)
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 = 355 (80% power), 475 (90% power)
 75% of observed effect: n per arm = 630 (80% power), 843 (90% power)
 50% of observed effect: n per arm = 1,417 (80% power), 1,897 (90% power)
power.t.test: 355.1 
Simulated power at n = 355 per arm: 0.799 (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.

13. Analyses Reported by the Sponsor

The registry reports the sponsor's analysis as an ANOVA with p = 0.10 for the comparison of the nonpharmacological sleep intervention and active control. It does not report the ANOVA's estimated mean difference, confidence interval, model covariates, or other model details. The reported result therefore indicates a test result but does not, by itself, provide the magnitude or precision of the sponsor's estimated contrast.

The unadjusted re-calculation from arm-level means and standard deviations gives a mean difference of -0.8, a 95% confidence interval from -1.807 to slightly above zero, and p = 0.119. This is a transparent calculation based on the posted summaries, not a reproduction of unreported features of the sponsor's model. The sponsor's p-value and the re-calculated p-value are close but not identical; the record does not explain the difference.

The posted record does not flag a discrepancy between these results. The two analyses should still be read according to what each reports: the sponsor gives an ANOVA p-value, while the re-calculation provides an unadjusted estimate and interval based on summary statistics. Neither value should be interpreted as a measure of clinical importance. Since the sponsor's effect estimate and analysis details are not reported, the record does not allow a full comparison of the underlying methods or estimates.

14. Multiplicity, Missing Data and Analysis Populations

The registry record does not report how multiplicity was handled. The project summary describes interests in sleep/wake patterns and functional recovery, and the record lists two outcome measures, but no secondary outcome results are posted in the record. The available primary result concerns PSQI at 3-month follow-up. It is not reported in the registry record whether the investigators designated one outcome and time point as primary in a prespecified analysis plan, adjusted statistical tests for multiple comparisons, or treated the ANOVA p-value as one test among several.

In general, testing multiple outcomes, time points, or subgroup contrasts can increase the chance of obtaining at least one apparently unusual result under a null model. Adjustment procedures can control some forms of this error, while a prespecified hierarchy can define which tests receive confirmatory interpretation. Without details about the planned outcomes and testing strategy, the impact of multiplicity on the reported p-value cannot be determined. No multiplicity-adjusted p-value is reported in the registry record.

Missing-data handling is also not reported in the registry record. The arm counts attached to the PSQI summaries are 110 and 109, but these counts alone do not explain whether every randomized participant completed the questionnaire, whether participants were excluded, or how missing scores were treated. The number and reasons for missing follow-up outcomes, imputation methods, and sensitivity analyses are not reported. If participants with unavailable outcomes differed from those with observed outcomes, the observed group means might not represent all randomized participants equally well.

The analysis population is not reported in the registry record. It is therefore unclear whether the analysis followed the intention-to-treat principle, included only participants with observed outcomes, or used another population definition. The record does not describe treatment adherence or crossover, either. Random allocation identifies the assigned-group comparison, but how the analysis handles participants after assignment affects the interpretation of that comparison.

Subgroup analyses, baseline adjustment, and interim analyses are not reported in the registry record. No assumptions about them should be inferred from the brief ANOVA label or from the posted group summaries. These omissions limit assessment of the analysis plan and robustness, but do not change the arithmetic of the unadjusted comparison based on the reported means and standard deviations.

15. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

The registry records zero serious adverse events among 110 participants in the nonpharmacological sleep intervention arm and zero among 109 participants in the active-control arm. It also records zero other adverse events among 110 participants in the intervention arm and zero among 109 in the active-control arm. These are the event counts and at-risk denominators posted for the two categories.

Zero recorded events in these fields means that no events were entered in those categories for the listed groups; it does not establish that no unwanted experiences occurred, that all participants had complete safety follow-up, or that the groups have identical safety profiles. The registry does not provide serious adverse event terms, and deaths are not reported in the registry record. Definitions, collection procedures, event timing, and duration of safety monitoring are also not reported.

These counts should be read as descriptive registry entries rather than as a detailed safety analysis. With no event counts in either group, the record offers no comparative estimate for these categories. It also does not report information that would allow assessment of event severity beyond the serious-versus-other classification, relatedness to the intervention, or completeness of ascertainment.

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

The posted primary outcome is a comparison of average PSQI scores at 3-month follow-up. The intervention arm mean was 8.6 and the active-control arm mean was 9.4. The unadjusted mean difference was -0.8, indicating a lower average score in the intervention arm. Because higher PSQI scores indicate poorer subjective sleep quality and sleep disturbance, the observed direction is toward a lower score in the intervention group. This is a description of the posted data, not a conclusion that the intervention caused improvement.

The interval estimate includes zero and extends from -1.807 to slightly above zero. That uncertainty means the summary data do not distinguish a nonzero average difference from no difference at the usual confidence-interval interpretation. The sponsor's reported ANOVA p-value is 0.10, while the unadjusted re-calculation gives p = 0.119. Neither statistic alone answers whether the effect is important to participants. The registry does not report a minimally important difference or an equivalence margin, so the size and clinical meaning of the contrast cannot be classified against such a benchmark.

The result is specific to the assigned sleep program compared with daily structured social visits, and to PSQI at 3-month follow-up. It does not isolate the contribution of light exposure, reduced daytime time in bed, bedtime routine, sleep assessment, or cognitive-behavioral elements. It does not establish effects on objective sleep measures, longer-term outcomes, or functional recovery. Those outcomes are not reported as results in the registry record.

The registry describes the study as randomized and triple-masked, but it does not identify the masked roles or give enough information to assess adherence, baseline balance, missing outcomes, or the analysis population. Those details matter when moving from an observed group contrast to an interpretation about the effect of assignment. The record also does not report a multiplicity strategy. These gaps leave uncertainty about how robustly the analysis represents all randomized participants and how the primary test fits into the broader outcome plan.

The most defensible reading is that the observed group means differ in the direction of lower PSQI scores in the intervention arm, while the estimated difference is imprecise and compatible with no average difference. The result does not prove the groups are equivalent, and it does not establish a clinically meaningful effect. It describes what the posted endpoint summaries show, with the limits imposed by the uncertainty interval and the details absent from the registry record.

17. Why This Trial Matters Statistically

This trial illustrates why a clinical result should be read through its estimate and uncertainty, rather than through a p-value alone. The arm means suggest a lower PSQI score for the intervention group, but the unadjusted confidence interval includes zero. The point estimate gives the direction and size of one estimated group contrast; the interval conveys that a range of effects remains compatible with the posted data and calculation assumptions.

The active control is statistically important to the interpretation. Participants assigned to control received daily structured social visits, so the randomized comparison concerns a multicomponent sleep intervention relative to an active social-contact condition. The contrast is not equivalent to treatment versus no attention. This helps define the estimand—the particular group comparison the trial addresses—and cautions against describing the result as a comparison with usual care or no intervention.

The endpoint also demonstrates the difference between an average scale score and a clinically interpretable outcome. PSQI provides a continuous subjective measure, and the registry gives a threshold for poor sleep quality, but it does not report how many individuals crossed that threshold. Nor does it provide a minimally important difference. Without those details, the mean contrast cannot be translated into a claim about the proportion of participants with a meaningful change.

Finally, the record shows the value and limits of a simple re-calculation. Posted means, standard deviations, and arm counts support an unadjusted estimate and interval, but they cannot reconstruct a sponsor model whose details are absent. Reporting the analysis method, estimate, confidence interval, missing-data approach, and multiplicity plan would make the result easier to evaluate and reproduce. Statistical interpretation depends not only on the result but also on understanding what comparison was planned and how participants and outcomes were handled.

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 provides a concise set of design and outcome details, but several limitations constrain interpretation. The primary result is a single follow-up comparison at 3-month follow-up; baseline PSQI scores and changes over time are not reported. As a result, the posted means do not show within-participant change or allow assessment of baseline balance on the primary measure.

The registry reports the sponsor's ANOVA p-value but not its effect estimate, confidence interval, covariates, or other model specifications. The unadjusted re-calculation uses arm-level summaries and cannot establish whether the sponsor's model handled baseline scores or other variables. The record does not provide diagnostic information about distributional assumptions, nor does it describe missing-data methods or the analysis population. These omissions make it difficult to judge whether the reported analysis incorporates all randomized participants and how sensitive the findings are to analytic choices.

Although the study is labeled triple-masked, the identities of the masked roles are not reported. Masking may be especially relevant for a subjective questionnaire, but the record does not explain how it was implemented. Adherence to the multicomponent intervention, fidelity of delivery, and possible contamination between groups are not reported. The individual contributions of its components cannot be separated because the trial compares the package with an active control rather than randomizing components separately.

The record does not report participant characteristics in detail, eligibility criteria, study sites, a clinical-importance threshold for PSQI, or a multiplicity plan. It lists two outcome measures but posts no secondary outcome results. Generalizability beyond older people undergoing post-acute rehabilitation cannot be assessed from the limited participant description. The safety entries show zero events in the posted categories, but event definitions, monitoring duration, and completeness are not reported. These limits do not negate the reported summaries; they define what those summaries can support.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Non-Pharmacological Interventions on Sleep in Post-Acute Rehabilitation — NCT00333619.
• Registry record last updated 2015-04-24; status last verified 2014-10.
• Results first posted 2015-04-06.
• 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

The registry record describes a completed, randomized, triple-masked, parallel-group trial of a multicomponent nonpharmacological sleep intervention versus an active social-visit control in older people undergoing post-acute rehabilitation. Enrollment was 219. The primary endpoint was mean Pittsburgh Sleep Quality Index score at 3-month follow-up.

The posted outcome summaries give a mean of 8.6 (standard deviation 3.8; 110 participants) for the intervention and 9.4 (standard deviation 3.8; 109 participants) for active control. The sponsor reports an ANOVA p-value of 0.10. An unadjusted re-calculation gives a mean difference of -0.8, with a confidence interval from -1.807 to slightly above zero and p = 0.119. The interval includes zero, so the estimate remains uncertain about the average group difference.

The registry records zero serious and zero other adverse events in both arms. Details of missing-data handling, analysis populations, multiplicity, the sponsor model, and the original sample-size justification are not reported in the registry record. The posted results describe the PSQI group comparison; they do not establish clinical importance or report results for the project's other stated aims.