Effect of CPAP on Postoperative Delirium: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Effect of CPAP on Postoperative Delirium (NCT01816685), a randomized na study. The primary outcome is Presence of Postoperative Delirium.

2. Clinical Question

3. Trial Design

The registry record describes a randomized, single-masked, parallel-group trial sponsored by Duke University. The phase is listed as not applicable. The listed enrollment is 135. Random allocation is reported, but the method used to generate the allocation sequence, how assignments were concealed, and which people were masked are not reported in the registry record. The specific sites and recruitment procedures are also not reported in the registry record.

The intervention was a device-based regimen. Participants assigned to CPAP were instructed to use an autotitrating positive airway pressure device whenever they slept before surgery and on postoperative days 0, 1, and 2. The brief summary describes machines being loaned to randomly selected patients at high risk for OSA. The registry record does not report adherence to the instructions, duration of use, device settings, or whether use was monitored.

The comparison group is called Routine Care, but its care description is blank. Consequently, the registry record does not establish what usual care involved or whether participants in that group used positive airway pressure outside the trial intervention. This limits interpretation of the contrast as a comparison of specific care regimens.

The trial is listed as completed. Its start is recorded as 2013-06 and its primary completion as 2014-06. These dates describe the registry record; they do not provide information about follow-up completeness or the timing of individual assessments. The record lists two outcome measures and no analysis entries. It does not report how many participants were assessed for the primary outcome, how many were included in an analysis, or whether any participants discontinued or deviated from their assigned group.

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 Presence of Postoperative Delirium, measured over Postoperative day 2.

The primary outcome is titled Presence of Postoperative Delirium and is assessed on postoperative day 2. The registry description says that assessments used the Delirium Rating Scale-Revised-98 (DRS-R-98), a clinician-rated tool with 16 items. It includes a severity score ranging from 0 to 39 for repeated serial measurements and a total scale score ranging from 0 to 46, which adds diagnostic items for initial ratings. Individual items are rated from 0 to 3 points and have text descriptions for each rating.

Higher scores represent a larger number of delirium symptoms or increased symptom severity. The scale therefore provides a structured way to describe symptoms, rather than relying only on an unstructured impression. A score can capture variation in the number or severity of symptoms, which is relevant when the study question concerns postoperative delirium.

The outcome is described as continuous, with a mean and standard deviation as the stated summary format. However, the registry record does not report the group means, standard deviations, or a between-group estimate. Its outcome entry was not reduced to one comparison because it includes several categories or time points. The record therefore does not provide a single numeric primary-outcome result to interpret. It also does not clarify which DRS-R-98 score or scores were used in each analysis or how any categorical delirium classification was defined.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Randomization is intended to make assigned groups comparable, on average, with respect to both measured and unmeasured characteristics at the point of assignment. This supports a causal comparison when allocation is implemented appropriately and participants are analyzed in a way that respects assignment. The registry record identifies the allocation as randomized, but does not report sequence generation, concealment, or the handling of departures from assignment. Those omissions prevent a full assessment of the safeguards around randomization.

What does single masking mean?

Masking means keeping knowledge of assigned treatment from specified people involved in a study. A single-masked designation indicates that masking was intended for one party, but the registry record does not identify that party or explain how masking was maintained. Because CPAP use can be apparent to participants and care teams, the roles of assessors and analysts would be relevant to interpreting ratings. The record does not say whether delirium assessors were masked or whether the success of masking was evaluated. These details should not be inferred from the label alone.

Why does the outcome definition matter?

A continuous score summarizes a measured quantity along a scale. A group mean describes the average score, while a standard deviation describes variation among scores within a group. A comparison of means can be useful when the scale is treated as continuous and the score and assessment time are prespecified. Here, the listed outcome uses the DRS-R-98 and specifies postoperative day 2, but the record indicates that several categories or time points were not reduced to one comparison. It does not show which score or score summary was selected for a single analysis. Different score definitions or time points can answer different questions, so an interpretable estimate needs the relevant definition alongside the numerical result.

How should an effect estimate be read?

An effect estimate describes the size and direction of a contrast between groups. For a mean-based analysis, the estimate would express how the average measured score differed between assigned groups under the chosen analysis. A confidence interval describes the range of values reasonably compatible with the data and the analysis assumptions. A narrower interval indicates greater precision than a wider interval, while the interval's location helps show which directions and magnitudes remain plausible. The registry record does not report a primary-outcome estimate or confidence interval, so precision cannot be assessed from this entry.

A p-value, when reported, describes how incompatible the observed data are with a specified null model under the analysis assumptions. It does not measure the probability that the null hypothesis is true, nor does it show the clinical importance of an estimate. A p-value without an effect estimate and interval gives limited information about the magnitude and precision of a finding. No primary-outcome p-value is reported in this registry record.

Why can variation affect a mean comparison?

Standard deviations describe within-group spread, and greater variation can make a mean contrast less precise for a given amount of data. The distribution of scores and the analysis model also matter. A normal approximation is one general method for planning or inference when its assumptions are suitable. If the data are more variable than the model allows, uncertainty can be understated; this is one reason to assess model fit and consider robust or alternative methods. The registry record does not report diagnostics, distributional summaries, or a detailed analysis model for the primary outcome.

The planning assumptions recorded for illustrative calculations specify a two-sided significance level of 0.05, equal allocation, and a normal approximation. They describe hypothetical planning calculations tied to a stated fraction of an observed effect. They are not a report of observed power. Sample-size illustrations indicate how a design might be planned under chosen assumptions; they do not establish that a completed study had a particular probability of detecting an effect. The trial's own sample-size justification and the numeric inputs and outputs of any planning calculation are not reported in the registry record.

What does the analysis population determine?

An analysis population defines which participants contribute to an estimate. An intention-to-treat approach generally compares participants according to randomized assignment, while other approaches may restrict or reclassify participants according to treatment received or data availability. These choices can change the question answered and can introduce bias if they are not justified. The registry record does not report the analysis population, the number included in the primary analysis, or how adherence and post-randomization events were handled. It also does not report a missing-data method, so the impact of incomplete assessments cannot be evaluated.

These principles explain what statistical information is needed to interpret a trial result. In this record, the design and measurement are described to some extent, but the central outcome contrast and its uncertainty are absent. Statistical interpretation must therefore remain limited to the registered design and the reported event counts.

10. Multiplicity, Missing Data and Analysis Populations

The registry record lists two outcome measures, while the primary outcome description indicates that the measure involved several categories or time points and was not reduced to one comparison. When a study examines multiple outcomes, time points, or score definitions, there are multiple opportunities for a result to appear notable by chance. This general issue is called multiplicity. A prespecified hierarchy, adjustment procedure, or clearly stated primary comparison can help readers understand which result carries the main inferential claim. Such procedures are not reported in the registry record.

The record contains no analysis entries. It does not report which outcomes were analyzed, whether additional analyses were planned, or whether any adjustment for multiple comparisons was used. It also does not state whether the outcome definitions or analysis choices changed during the trial. The absence of these details does not show that a particular procedure was or was not used; those procedures are not reported in the registry record.

Missing data are also important for a postoperative assessment. Some participants might not have a score recorded for reasons related to surgery, illness, discharge, or assessment logistics, but the registry record does not report whether any scores were missing or why. It does not describe imputation, complete-case analysis, sensitivity analyses, or other methods for handling missing scores. The number assessed and the analysis population are not reported.

Likewise, the registry record does not describe an intention-to-treat analysis, a per-protocol analysis, or another analysis population. Without those details, the relationship between randomized assignment, treatment actually received, and the analyzed data cannot be assessed. These omissions limit evaluation of both multiplicity and potential bias from missing outcome information.

11. 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 lists serious adverse events for 0 of 58 participants at risk in the CPAP group and 0 of 56 participants at risk in the Routine Care group. It also lists other adverse events for 0 of 58 participants at risk in the CPAP group and 0 of 56 participants at risk in the Routine Care group. These are the counts recorded for the listed risk denominators; they do not establish that no adverse events occurred outside the registry's reporting framework or observation period.

No serious adverse-event terms are listed. Death counts are not reported in the registry record. The record does not describe event definitions, the period over which events were collected, how events were adjudicated, or whether all participants were monitored in the same way. The at-risk counts should therefore be read as the denominators accompanying the reported entries, not as evidence about outcome completion or the primary delirium analysis.

Zero recorded events in these entries do not establish that the groups had identical safety profiles or rule out uncommon events. With no event counts above zero, the registry entries provide little basis for comparing event patterns, and the missing details limit interpretation. These safety figures should be kept separate from the unreported primary-outcome comparison.

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

The registry record documents a randomized comparison of peri-operative CPAP with routine care among patients described as being at high risk for OSA. It identifies the DRS-R-98 assessment on postoperative day 2 as the primary outcome and describes the scale's symptom ratings. The record is listed as completed, and it includes adverse-event counts. These details establish the broad question and some features of how delirium was measured.

They do not establish the observed difference in delirium ratings. No group means, standard deviations, effect estimate, confidence interval, or p-value for the primary outcome are reported in the registry record. Consequently, the direction and magnitude of any difference cannot be determined from this record, and its precision cannot be judged. The trial's completion status is not itself evidence of a particular result.

The record also does not describe the precise routine-care regimen, the masked party, the analysis population, the handling of missing scores, or the statistical analysis plan. Those gaps matter for assessing what contrast was estimated and how reliable an estimate might be. The listed enrollment of 135 cannot be treated as the number with complete outcome data or as a sample size for a particular analysis.

The adverse-event entries record zero serious events and zero other events in each group for the stated risk denominators, while deaths are not reported. These entries do not answer the primary delirium question and should not be treated as a complete account of all possible safety outcomes.

The appropriate interpretation is therefore limited: the registry describes a completed randomized trial and its intended measurement, but does not report a single result for the central outcome. No claim about benefit, lack of benefit, or the size of an effect follows from the information reported in this record.

13. Why This Trial Matters Statistically

This trial addresses a statistical challenge common in peri-operative research: a potential risk factor, a device-based intervention, and a symptom outcome measured after surgery. Random allocation can support a fair comparison, while a structured clinician-rated scale can give the outcome a consistent definition. The registry describes both elements, making the question and measurement approach identifiable even though the results are not reported as a single comparison.

The DRS-R-98 description also illustrates why endpoint specification is important. The scale includes severity and total scores, and the registry notes several categories or time points rather than one reduced comparison. A clear choice of score and assessment time is needed to connect the clinical question to a statistical estimate. Without that choice and a reported estimate, readers cannot distinguish between the broad measurement plan and the specific result that would answer the primary question.

The record further illustrates that randomization alone is not enough for a complete statistical assessment. Information about masking, adherence, outcome availability, analysis populations, and missing-data handling helps determine how the assigned-group comparison should be interpreted. Those details are not reported in the registry record, as are the primary-outcome estimates and their uncertainty.

Finally, trial status and safety counts cannot fill the gap left by missing efficacy results. A completed record with reported event counts may still lack the estimate needed to assess the primary question. This distinction is important for researchers reading registries: design information, outcome definitions, and outcome results are separate parts of the evidence, and each contributes a different kind of information.

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 main limitation for interpreting this trial record is the absence of a single reported primary-outcome comparison. Although postoperative day 2 and the DRS-R-98 are identified, the record does not provide group-level scores, an effect estimate, or an uncertainty interval. It indicates that the outcome included several categories or time points and was not reduced to one comparison. The central result therefore cannot be evaluated from the registry entry.

Several design and conduct details are also not reported in the registry record. These include the definition of high risk for OSA, eligibility criteria, allocation-sequence and concealment methods, the identity of the masked party, and the contents of routine care. The record describes intended CPAP use but not adherence, device settings, or monitoring. These omissions constrain assessment of how the intervention and comparison were implemented.

The listed enrollment of 135 does not identify the number with a postoperative assessment or the number analyzed. Missing outcome data, reasons for missingness, analysis populations, and statistical methods are not reported. As a result, the possibility and direction of bias from incomplete assessments cannot be evaluated. The record also lists two outcome measures but does not describe multiplicity procedures or explain how the primary comparison was selected among the recorded categories or time points.

Safety interpretation is similarly limited by absent detail on event definitions, collection periods, and ascertainment. Death counts are not reported in the registry record. Taken together, these omissions mean the registry entry can describe the trial's broad design and measurement plan, but it cannot support a detailed assessment of the primary effect or the robustness of an analysis.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Effect of CPAP on Postoperative Delirium — NCT01816685.
• Registry record last updated 2015-08-31; status last verified 2015-08.
• Results first posted 2015-08-31.
• 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

The registry record describes a completed randomized, single-masked, parallel-group trial sponsored by Duke University. It lists enrollment of 135 and compares an autotitrating positive airway pressure device with routine care among patients at high risk for OSA. Participants assigned to CPAP were instructed to use the device whenever they slept before surgery and on postoperative days 0, 1, and 2.

The primary outcome is postoperative delirium assessed on postoperative day 2 with the DRS-R-98. The registry record does not report a single primary-outcome comparison or its uncertainty. It lists zero serious and zero other adverse events in each group for the stated at-risk counts; death counts are not reported in the registry record. The analysis population, missing-data methods, and detailed statistical plan are also not reported.