Improving Sleep Using Mentored Behavioral and Environmental Restructuring: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Improving Sleep Using Mentored Behavioral and Environmental Restructuring (NCT03327324), a randomized na study. The primary outcome is Sleep Efficiency (SE).

2. Clinical Question

3. Trial Design

This was a randomized study with an unmasked allocation and a sequential model. The registry record lists 110 participants and describes six facility-and-unit sequences. Each sequence began with Baseline, proceeded to a 3 Month SLUMBER Intervention, and then had a sustainability period. The listed sustainability periods were 39 Month, 33 Month, 27 Month, 21 Month, 15 Month, and 9 Month, attached to different facility-and-unit sequences. The record does not describe a concurrent two-arm comparison for the primary outcome.

The intervention was behavioral and aimed to train nursing facility staff in sleep-promoting approaches and environmental changes for residents. The registry describes the study’s goal as testing potential functional and psychosocial benefits of improved sleep. It also gives context about sleep disturbances in skilled nursing facilities and concerns about reliance on sleep medications, but this context is not a result of the study.

The study is marked COMPLETED. Its recorded start was 2018-03-27, and its primary completion was 2021-02-23. These dates describe the study record; they do not indicate when each sequence’s sustainability period ended. The record lists Insomnia as the condition and does not state a study phase beyond NA.

Although the study is identified as randomized, the allocation details needed to understand how facilities, units, or residents were assigned are not reported in the registry record. The method of randomization, allocation concealment, and any stratification are also not reported. Masking is listed as NONE. The registry record does not report a statistical plan or a rationale for the enrollment target. It lists 12 outcome measures and 0 analyses, and notes that the primary-outcome analysis was skipped because there were fewer than two arms.

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: This is a stepped wedge intervention trial (SWT). In this SWT, all units receive the intervention but at different time points. Every unit provides baseline and follow-up data.

Baseline (Month 0-Month 3)

Reasons for not completing this period:

Intervention (Month 3-Month 6)

Reasons for not completing this period:

Sustainability Follow-Up 1 (Month 6-9)

Reasons for not completing this period:

Sustainability Follow-Up 2 (Month 9-12)

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Sleep/wake time was measured using wrist actigraphs and some participants took off the watch, meaning some participants were not included in the analysis of this outcome measure.

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.

Only the first 6 of 7 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Sleep Efficiency (SE), measured over Baseline.

The primary endpoint was Sleep Efficiency (SE), measured at Baseline. Sleep efficiency is defined in the registry record as the ratio of total sleep time to time in bed, expressed as a percentage of time in bed. The measure was based on wrist actigraphy, which records sleep and wake patterns over time. The record specifies a mean as the summary measure and a 95% confidence interval as the dispersion measure.

As a general statistical concept, sleep efficiency expresses how much of the time spent in bed is recorded as sleep. It can therefore describe sleep continuity in a way that accounts for time in bed, rather than reporting sleep duration alone. Actigraphy offers a measurement approach based on recorded activity, but the registry record notes that some participants took off the watch and were not included in the analysis of this outcome.

The endpoint’s stated timeframe is Baseline. A baseline measurement can characterize the starting outcome value, but by itself it does not show whether the intervention changed that value. To estimate change or compare conditions, an analysis needs outcome data linked to the relevant timepoints or groups. The registry record does not report a primary-outcome estimate, confidence interval, or analyzed comparison. Accordingly, the endpoint definition is available, but the posted results do not establish the observed level of sleep efficiency or its change.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry record identifies the study as randomized, unmasked, and sequential. These terms describe different aspects of design. Randomization refers generally to using chance to assign units or participants to study conditions. Masking concerns whether participants or study personnel are kept unaware of assignments; masking is listed as NONE. A sequential design organizes study conditions or periods in an ordered sequence. The record lists six sequences, each with Baseline, a 3 Month intervention, and a sustainability period of a specified duration.

The record does not report the unit of randomization, the assignment procedure, allocation concealment, or the statistical model used for outcome analysis. It also does not report how repeated measurements, facility or unit clustering, or changes over time would be handled. These details matter in general because measurements from the same person or setting may be related rather than independent. A suitable analysis needs to reflect the design and the dependence structure of the observations. The registry record does not establish which approach, if any, was used.

The primary outcome is continuous: sleep efficiency is summarized as a mean. In general, a mean describes the average value in an analyzed group, while a confidence interval communicates uncertainty about an estimated mean or contrast. The record names a 95% confidence interval as the dispersion measure, but it reports no sleep-efficiency mean, confidence interval, or contrast. Thus, there is no numerical primary-outcome estimate to interpret.

A comparison of intervention effects ordinarily requires a clearly defined contrast, such as a difference between conditions or a change from one timepoint to another. Which contrast is appropriate depends on the design and the question. Here, the registry record notes that analysis was skipped because there were fewer than two arms in the primary outcome. It does not report an alternative comparison or a model-based estimate. The six listed sequences should not be treated as six independent treatment arms for the primary outcome without a reported analysis that defines them that way.

As a general concept, statistical significance and effect size answer different questions. A p-value, if reported, would summarize how compatible the observed data are with a specified null model; an effect estimate would describe the size and direction of the contrast. Neither can substitute for the other. The registry record does not report a primary-outcome p-value or effect estimate, and no conclusion about statistical significance can be drawn.

Sample-size planning is also distinct from analysis of observed outcomes. The record describes a planning illustration based on an observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; it states that the effect to detect is a stated fraction of the observed effect. Such illustrations are planning calculations, not observed results and not observed power. The record does not report the numerical planning outputs, a trial-specific sample-size justification, or whether those assumptions match the study’s sequential structure. The enrollment of 110 is an enrollment count, not evidence by itself that the study was adequately powered for the primary outcome.

Power is generally the probability that a prespecified analysis will detect an effect of a specified size under particular assumptions. It depends on the target effect, variability, allocation, analysis, and other design features. A power calculation cannot recover an unreported outcome estimate or resolve the absence of a comparison. Because the primary-outcome analysis is not reported, planning assumptions should not be used to imply that a result was observed or that the study could rule out an effect.

Confidence intervals are useful because they show a range of estimates compatible with the data and model, rather than only classifying a result as statistically significant or not. Their interpretation depends on the analysis being appropriate and the underlying data being sufficiently informative. No interval is posted for sleep efficiency, so its precision cannot be evaluated. The registry’s mention of a 95% confidence interval describes the intended statistical summary, not a finding.

Missingness can affect both the precision and interpretation of an outcome. The record specifically notes that some participants removed the actigraph and were omitted from the sleep-efficiency analysis. In general, excluding people with unavailable measurements can introduce bias if the likelihood of missingness is related to sleep, health, or other factors relevant to the outcome. The registry record does not report why watches were removed, how many measurements were unavailable, whether missingness was assessed, or whether any sensitivity analyses were conducted. Those omissions prevent assessment of the impact of missing actigraphy data.

The registry lists 12 outcome measures and 0 analyses. When many outcomes or comparisons are assessed, multiplicity can increase the chance of at least one apparently unusual result. General approaches include prespecifying a primary outcome, adjusting for multiple comparisons, or interpreting secondary outcomes as exploratory. The record identifies a primary endpoint but does not report a multiplicity strategy. In this case, the absence of a reported primary analysis is the central interpretive limitation; the record also does not show which secondary analyses, if any, were performed.

Finally, the record does not report an analysis population definition, such as whether results were intended to include all randomized participants or only those with usable measurements. It does not describe participant flow or provide outcome-specific denominators. Those details are important for judging whether an estimate reflects the randomized population or a subset. Without a posted estimate, denominators, and analysis description, the statistical methods cannot be evaluated beyond the design and endpoint information recorded.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists 12 outcome measures and 0 analyses. It identifies Sleep Efficiency (SE) as the primary outcome, but the primary analysis was skipped because the outcome had fewer than two arms. A strategy for controlling multiplicity across outcomes or comparisons is not reported in the registry record. In general, examining multiple outcomes can make isolated findings harder to interpret unless the analysis plan distinguishes confirmatory questions from exploratory ones. The posted record does not provide results for such an assessment.

Missing actigraphy data are specifically noted: some participants took off the wrist watch and were not included in the sleep-efficiency analysis. The record does not report how many participants were omitted, the amount of missing data, reasons for removal, or whether missingness differed by participant characteristics or sequence. It also does not report imputation, sensitivity analyses, or another strategy for handling missing measurements.

The analysis population is not described. The registry record does not state whether the intended outcome population was all randomized participants, participants with available actigraphy, or another group. No outcome-specific denominator is reported. Because both the analysis population and missing-data handling are unclear, the extent to which any potential estimate would represent the enrolled population cannot be assessed from the posted results.

11. 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 record reports one group, “Resident of the Skilled Nursing Facility,” with 110 at risk for serious adverse events and 0 affected. For other adverse events, it also reports 110 at risk and 0 affected. The record lists 9 deaths. These are event counts recorded for the group; they are not a comparison between intervention and control arms.

The registry does not report when the deaths occurred, their causes, or whether they were considered related to the intervention or study procedures. It also does not provide a timing window or a comparative analysis of event rates. The counts therefore cannot establish whether the intervention changed the risk of serious events, other adverse events, or death.

In general, safety counts need context, including how events were defined, the period of observation, completeness of reporting, and the groups being compared. A count of 0 recorded events does not by itself establish that no events occurred outside the reported collection or observation framework. The registry record does not provide enough detail to assess those issues. The listed counts should be read as descriptive registry entries, not as evidence of a causal safety effect.

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

The record describes a randomized, unmasked, sequential study of a behavioral program intended to teach nursing facility staff sleep-promoting strategies and environmental changes. It defines sleep efficiency as the primary outcome and describes measurement using wrist actigraphy. It also records that some participants removed the watch and were not included in the outcome analysis. These details establish the study’s topic and measurement approach, but they do not establish an intervention effect.

The primary endpoint is specified at Baseline, and the registry record does not report an analyzed comparison across conditions or timepoints. It lists 0 analyses and states that the primary-outcome analysis was skipped because there were fewer than two arms. Consequently, the posted results do not show whether sleep efficiency changed, the direction or size of any change, or how precisely a change was estimated. The absence of a reported result is not evidence that the intervention had no effect.

The record’s design description includes six sequences with different sustainability periods, but it does not report an analysis connecting those sequences to a primary-outcome contrast. It would be inappropriate to infer a between-sequence treatment effect or a durability pattern from the sequence descriptions alone. The registry also does not report the protocol details, analysis population, or statistical plan needed to determine how the planned comparisons were meant to be conducted.

The safety entries record 0 serious adverse events, 0 other adverse events, and 9 deaths in a group with 110 at risk. These figures are not accompanied by timing, causes, attribution, or a comparison group analysis. They therefore describe what the registry reports, but do not determine whether the intervention affected safety outcomes.

The central conclusion is limited: the registry record documents a completed study and its intended primary measure, but it does not provide a primary-outcome analysis from which an efficacy or precision conclusion can be drawn. Any claim that the intervention improved sleep, did not improve sleep, or produced a specific effect would go beyond the posted results.

13. Why This Trial Matters Statistically

This trial raises statistical questions that are important for studies conducted in real-world care settings. The intervention focused on staff practices and the environment in skilled nursing facilities, while the primary outcome was measured in residents using wrist actigraphy. The relationship between facility or unit assignment, resident-level measurements, and repeated measurement over time can shape the appropriate analysis. The registry record identifies a randomized sequential design but does not describe how those dependencies were handled.

The six sequences had different listed sustainability durations. Such a design makes it important to define which timepoint and comparison answer the primary question, and how sequence and time are represented in the analysis. The record identifies Baseline as the primary outcome timeframe, but does not report an analyzed comparison. This illustrates why an endpoint definition and a design description do not replace a posted estimand, estimate, and uncertainty interval.

Measurement availability also matters. Some participants removed their wrist actigraphs and were omitted from the sleep-efficiency analysis. When a measurement depends on device use, transparent reporting of missingness and analysis populations helps readers judge precision and possible selection effects. Those details are not reported in the registry record.

Finally, the record shows how the distinction between trial conduct and statistical evidence matters. Completion status and enrollment do not establish that a primary question was answered. A clear account of the analysis, the population included, the outcome estimate, and its uncertainty is needed to interpret the intervention effect. Because those primary results are not posted, this record is most informative about the planned context and measurement, rather than the magnitude or certainty of an outcome difference.

14. 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 trial-specific limitation is that no analyzed primary-outcome comparison is reported. The registry record lists 0 analyses and states that the primary-outcome analysis was skipped because there were fewer than two arms. It therefore supplies no sleep-efficiency estimate, confidence interval, or p-value from which to judge an intervention effect.

The primary outcome is described at Baseline, while the intervention and sustainability periods are also specified in the sequence descriptions. The record does not explain how baseline values were to be compared with later measurements or how the different sustainability durations would enter an analysis. The statistical plan and analysis population are not reported.

Actigraphy data were unavailable for some participants because they took off the watch, and those participants were not included in the outcome analysis. The number affected, reasons for watch removal, outcome-specific denominator, and missing-data methods are not reported in the registry record. The possible influence of these omissions on representativeness and precision cannot be assessed.

The record does not report details of random assignment, allocation concealment, or how the facility-and-unit sequences were related to the assignment process. It also does not report whether the statistical analysis accounted for grouping by facility or unit, repeated observations, or other features of the sequential design. These are uncertainties about the documented methods, not proof that a particular method was or was not used.

Safety reporting is limited to counts without timing, causes, attribution, or comparative analysis. The registry record also does not report a trial-specific sample-size justification or a multiplicity strategy. Together, these omissions limit evaluation of the study’s statistical evidence, even though the record identifies its intervention, endpoint, enrollment, and sequence structure.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Improving Sleep Using Mentored Behavioral and Environmental Restructuring — NCT03327324.
• Registry record last updated 2023-02-08; status last verified 2023-01.
• Results first posted 2023-02-08.
• 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.

18. Record Summary

NCT03327324, “Improving Sleep Using Mentored Behavioral and Environmental Restructuring,” is listed as a completed, randomized, unmasked, sequential behavioral study sponsored by NYU Langone Health. The record reports enrollment of 110 and describes six facility-and-unit sequences, each with Baseline, a 3 Month intervention, and a specified sustainability period.

The primary outcome was Sleep Efficiency (SE), expressed as a percentage of time in bed and measured using wrist actigraphy. Some participants removed the watch and were not included in the outcome analysis. The registry lists 12 outcome measures and 0 analyses, and states that the primary-outcome analysis was skipped because there were fewer than two arms. No primary-outcome estimate or confidence interval is reported.

Safety entries list 0 serious adverse events, 0 other adverse events, and 9 deaths among a group with 110 at risk. Timing, causes, attribution, and comparative safety results are not reported in the registry record.