Obstructive Sleep Apnea in World Trade Center Responders: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Obstructive Sleep Apnea in World Trade Center Responders (NCT01753999), a randomized na study. The primary outcome is Adherence to CPAP Pre-crossover.

2. Clinical Question

The study also set out to examine possible causes of sleep apnea in World Trade Center responders, including associations with upper airway disease, nasal inflammation, and nasal resistance. Those broader aims are distinct from the posted primary treatment comparison. The primary endpoint concerns use of the devices, not the nasal measures or a direct measure of sleep apnea severity.

The crossover structure means participants were intended to receive both treatments in sequence. However, the primary endpoint description specifies that its results are based on the first treatment period only, before crossover. It therefore provides a comparison between the groups’ initially allocated treatments rather than a reported within-person comparison across both periods. The registry record does not report a primary inferential test or a confidence interval for the adherence comparison.

3. Trial Design

The trial is identified as NCT01753999, titled Obstructive Sleep Apnea in World Trade Center Responders. Rutgers, The State University of New Jersey, is listed as the sponsor. The record describes a randomized, triple-masked, crossover study, with an enrollment of 317 and a status of completed. Its phase is listed as not applicable. The recorded start is 2012-12 and primary completion is 2017-03.

The two treatment methods were standard CPAP and CPAP-Flex. Standard CPAP uses constant pressure; CPAP-Flex decreases pressure during expiration. The registry describes assignment to one method for one month followed by crossover to the other method for the next month. The arm descriptions specify 4 weeks of use for each treatment. The primary endpoint timeframe is recorded as 5 weeks after initiation of treatment. The record does not explain the relationship between these different descriptions of treatment duration and endpoint timing.

The primary analysis is explicitly limited to the first treatment period, before crossover. It includes subjects who received the allocated treatment, defined in the record as having turned on the PAP device. The record says that some subjects took a device home but did not turn it on: n=11 in CPAP and n=16 in C-Flex. These subjects were assumed to have no use, while the primary endpoint population is described as those who did turn on the device. The record does not give further detail on the treatment of these non-users in the reported endpoint summaries.

The posted adherence groups have n=144 for standard CPAP and n=146 for CPAP-Flex. The broader trial enrollment is not the same as these endpoint group counts. The registry record does not report further detail on sequence generation, concealment, the masking procedures, a sample-size justification, or a prespecified statistical analysis plan. It also does not report how period or sequence effects were handled in analyses that might use both crossover periods.

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: Subjects without OSA were excluded from the trial.

Period 1 - Initial Assignment

Reasons for not completing this period:

Period 2 -Cross to Alternate Arm

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Subjects who received the allocated treatment (i.e., turned on the PAP device). The difference between the number allocated and the number of analyzed is the result of some subjects (n=11 in CPAP and n=16 in C-Flex) who did not turn on the device despite taking it home. These were assumed to have NOT received the allocated treatment.

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

Population: Participants assigned to the treatment arm.

6. Primary Endpoint

The registry lists the primary outcome as Adherence to CPAP Pre-crossover, measured over 5 weeks after initiation of treatment.

The primary endpoint is adherence to CPAP before crossover, measured in hours/night. It describes how much participants used the device during the first treatment period. Adherence is relevant to a device comparison because the treatment can only be used as intended when participants operate it; the endpoint therefore captures use rather than a direct physiological or clinical response.

The registry reports a median of 1.0 hours/night for standard CPAP, with an inter-quartile range from 0.0 to 3.32, and a median of 0.89 hours/night for CPAP-Flex, with an inter-quartile range from 0.0 to 3.17. The median gives the midpoint of the observed distribution, while the inter-quartile range describes the central spread. These summaries do not show every participant’s use or the shape of the full distribution.

Restricting the primary endpoint to the first treatment period avoids combining this result with use after participants switch devices. It also means the posted comparison does not present the full information that a paired analysis of both treatment periods might use. The registry record does not report a primary test of the difference, a confidence interval for the primary comparison, or details of an analysis that accounts for the crossover design.

Use of the device is not itself a measure of treatment efficacy. The registry lists CPAP efficacy separately as a secondary outcome, measured in events/hour. Keeping adherence and efficacy distinct is important: the adherence summaries describe device use, whereas the secondary endpoint describes an efficacy measure. Neither endpoint alone answers every question about the study’s broader aims concerning nasal pathology and sleep apnea.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

In general, an endpoint summary and an inferential comparison answer different questions. A median describes the center of observed values, and an inter-quartile range describes the middle spread. Neither alone estimates how uncertain the difference between treatment groups is. For the primary endpoint, the record reports these summaries but does not report a test or confidence interval for the treatment comparison. It is therefore appropriate to describe the observed values without assigning a statistical significance result to them.

In general, a confidence interval gives a range of values compatible with the data and the method used to construct it. For the secondary efficacy comparison, the recalculated interval extends to both negative and positive mean differences, so it includes the direction of a lower value for CPAP-Flex as well as the direction of a higher value. This breadth is a reminder that an estimate should be interpreted together with its uncertainty. The interval is not a guarantee that the true value lies within it, and it does not answer whether any difference would matter clinically.

The secondary endpoint is reported as a mean and standard deviation for each treatment group. A mean is sensitive to the magnitude of all observed values, while the standard deviation describes how dispersed those values are around the mean. The recalculated comparison uses the posted group means and standard deviations as an unadjusted summary-statistic comparison. It is not described as an adjusted analysis from the trial investigators. Its estimate is 0.23 events/hour for CPAP-Flex relative to standard CPAP, and its unadjusted p-value is 0.599.

In general, a p-value measures how unusual data at least as inconsistent with a specified null model would be if that model were true. It is not the probability that the null hypothesis is true, and it does not measure the size or importance of an effect. The secondary p-value should be interpreted as one component of an unadjusted comparison, alongside the estimate and its interval. The registry record does not identify an investigator-reported p-value or analysis for this outcome.

The reported planning illustration uses the observed secondary effect as its basis, with effect targets set at 100%, 75%, and 50% of that observed effect. Its stated assumptions are a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation. It gives participants per arm for 80% power and participants per arm for 90% power. For 100% of the observed effect, the corresponding figures are 2,683 and 3,591; for 75% of the observed effect, they are 4,769 and 6,384; and for 50% of the observed effect, they are 10,729 and 14,363.

These values are planning illustrations, not observed trial results and not observed power. They are based on the observed effect in the secondary comparison and should not be mistaken for the trial’s original sample-size calculation. The registry record does not report the trial’s sample-size justification. In general, planning estimates that use an observed effect can be unstable when the estimate is uncertain. Choosing a smaller target effect increases the planning requirement under the stated assumptions, as reflected in the listed illustrations, but these numbers do not establish what the study was designed to detect.

The trial is recorded as a crossover study. In general, crossover designs can allow comparisons within participants because each participant receives both treatments. Such designs also make period, sequence, and carryover considerations relevant to analysis. The posted primary endpoint deliberately uses only the first treatment period, before crossover, so it is a comparison of the initially assigned groups rather than a paired analysis across both periods. The registry record does not report an analysis using the later period or how crossover-related effects were evaluated.

Adherence is a special interpretive concern because the primary endpoint population is defined by turning on the device. The record says that subjects who took the device home but did not turn it on were assumed to have no use, while also defining the analyzed population as those who did turn it on. These statements indicate that non-use was recognized, but the posted summaries are described for the participants who activated the device. In general, excluding people who do not start treatment can change the population represented by an adherence summary. The registry record does not provide additional analyses that include all assigned participants or describe how those non-users were incorporated into the reported medians.

For safety outcomes, the record reports counts and numbers at risk by listed group. In general, a zero count means no event was recorded in that group’s reported data; it does not prove that the underlying risk is zero. Interpretation also depends on how groups, follow-up, and event collection were defined. Those details are not reported in the registry record, so the safety counts should be described as recorded counts rather than as a complete comparative safety analysis.

10. Multiplicity, Missing Data and Analysis Populations

The registry record lists CPAP efficacy as a secondary outcome and reports a recalculated comparison, but it does not report an analysis plan specifying how primary and secondary outcomes were prioritized or how multiple testing was handled. In general, multiplicity matters when several outcomes or comparisons are tested: without a prespecified strategy, the chance of at least one apparently notable result can increase. The posted secondary p-value is unadjusted, so it should not be interpreted as if an adjustment for multiple comparisons had been applied.

No primary inferential test or confidence interval for adherence is reported. The registry record also lists no reported analyses for the secondary outcome. The efficacy mean difference and p-value are therefore recalculations from the posted summary statistics, not an investigator-reported model. The record does not describe a statistical analysis plan, interim analyses, or methods for adjusting for baseline variables.

The primary endpoint population is defined as subjects who received their allocated treatment by turning on the device. The record notes that subjects who took devices home but did not turn them on were assumed to have no use, and gives n=11 in CPAP and n=16 in C-Flex for these subjects. It does not explain in detail how that assumption was used in the reported summaries or provide an analysis including all assigned participants. The record also does not describe reasons for missing efficacy observations or methods for handling missing data.

These gaps affect interpretation of both populations and uncertainty. In general, a complete-case result can differ from an analysis that includes all assigned participants, especially if the availability of measurements is related to outcomes or treatment. The registry record does not provide enough information to determine whether missingness was related to adherence or efficacy, or whether any sensitivity analyses were performed. The difference in endpoint group counts should therefore be noted without attributing it to a cause.

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 listed safety figures show zero serious adverse events, zero other adverse events, and zero deaths in the Standard CPAP group, with 267 at risk; the CPAP - Flex group also shows zero serious adverse events, zero other adverse events, and zero deaths, with 278 at risk. These are the recorded counts and group denominators in the registry record.

A zero count means that no event of the specified type was recorded in the listed group. It does not establish that the true risk is zero, nor does it show that risks are identical between treatments. In general, the strength of a safety comparison depends on the amount of observation, how events were collected and defined, and how follow-up was handled. The registry record does not provide details that would support a more complete comparative safety analysis.

The at-risk counts are not the same as the primary adherence endpoint group counts, and the record does not explain the relationship between these safety populations and the treatment periods in the crossover design. The figures should therefore be read as reported group-level counts rather than combined with the adherence or efficacy summaries. No serious adverse event terms are listed.

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

The primary result describes device use during the first treatment period among participants who turned on their allocated device. The posted medians are 1.0 hours/night for standard CPAP and 0.89 hours/night for CPAP-Flex, with inter-quartile ranges of 0.0 to 3.32 and 0.0 to 3.17, respectively. These values describe the observed distributions; the registry record does not report a statistical test or interval for the primary comparison.

The secondary efficacy result gives a recalculated unadjusted mean difference of 0.23 events/hour for CPAP-Flex relative to standard CPAP, with p = 0.599. The comparison interval spans both directions of difference, so the result is not a precise estimate of which treatment has a higher or lower mean. This secondary result should not be converted into a claim that the treatments are equivalent or different.

Neither endpoint alone answers the study’s broader question about possible underlying causes of sleep apnea in World Trade Center responders. The registry describes examination of upper airway disease, nasal inflammation, and nasal resistance, but the posted numerical results concern adherence and CPAP efficacy. They do not establish relationships between nasal pathology and sleep apnea.

In general, absence of a statistically notable result is not proof of no effect, and a small observed estimate is not proof that treatments are equivalent. Equivalence or non-inferiority conclusions require an appropriate design, prespecified margins, and analysis; those details are not reported in the registry record. Similarly, adherence summaries are not a substitute for efficacy outcomes, and efficacy summaries do not reveal whether participants used the devices as intended.

The record’s limitations include the absence of a reported primary inferential comparison, missing detail about the analysis plan and missing data, and limited posted analyses of the crossover structure. The findings should therefore be understood as descriptive registry results with an unadjusted secondary comparison, not as a complete account of treatment effects or of the study’s broader mechanistic aims.

13. Why This Trial Matters Statistically

This trial raises several useful statistical issues. Its crossover structure makes the distinction between first-period and within-participant comparisons important. The primary endpoint is explicitly pre-crossover, so the posted adherence summaries compare initially assigned groups rather than showing a paired analysis across both treatment periods. That distinction affects what the estimates represent and what information is used.

Adherence also makes the analysis population central to interpretation. The endpoint is described for subjects who turned on their allocated PAP device, while the record notes that some subjects took the device home but did not turn it on and were assumed to have no use. A treatment-use endpoint can be sensitive to whether non-users are included, excluded, or represented by an assumed value. The registry does not provide enough detail to reconcile that assumption with the analyzed population or to assess its effect on the summaries.

The secondary efficacy comparison illustrates why an estimate should be read alongside its uncertainty and analysis status. The unadjusted recalculation gives a mean difference of 0.23 events/hour and p = 0.599, while the interval spans both negative and positive differences. The registry lists no reported analysis for this outcome. This makes it important to distinguish an arithmetic comparison from a reported trial analysis, and to avoid treating the unadjusted p-value as a complete answer to the study question.

The planning illustrations offer another teaching point: sample-size requirements based on an observed effect are not observed power and are not evidence of the trial’s original design rationale. The record does not report its own sample-size justification. Interpreting the planned or observed results requires keeping those distinct quantities separate, along with recognizing that multiplicity, missing data, and model specification are not fully described.

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 registry record does not report the statistical analysis plan, the primary inferential comparison, or a confidence interval for adherence. As a result, the primary endpoint can be interpreted descriptively, but the precision of a treatment contrast cannot be assessed from a reported primary test. The record also does not report the original sample-size justification, so the planning illustrations should not be mistaken for the trial’s design basis.

The primary adherence summaries are restricted to participants who turned on the device. The record notes subjects who took a device home but did not turn it on and says they were assumed to have no use, but it does not fully explain how those subjects were handled in the reported summaries. It also does not report sensitivity analyses or an analysis covering all allocated participants. This leaves uncertainty about how the reported adherence distribution represents non-use.

The secondary efficacy comparison is based on fewer participants than the adherence summaries, and the reason for the difference is not reported. The registry does not describe missingness by treatment, reasons for unavailable measurements, or methods for handling missing data. The recalculated comparison is unadjusted and the record lists no reported analyses for that outcome. No multiplicity strategy is reported.

Although the study is described as a crossover, the primary endpoint is pre-crossover and the registry record does not report how data from the later treatment period were analyzed or whether sequence, period, or carryover effects were evaluated. The record also does not explain the difference between its treatment-period descriptions of one month and 4 weeks, or how that relates to the primary endpoint timeframe of 5 weeks after initiation. These gaps limit reconstruction of the analysis and the interpretation of the complete crossover study.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Obstructive Sleep Apnea in World Trade Center Responders — NCT01753999.
• Registry record last updated 2025-04-30; status last verified 2025-04.
• Results first posted 2018-07-20.
• 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

NCT01753999 is a completed randomized, triple-masked crossover study sponsored by Rutgers, The State University of New Jersey, with enrollment of 317. It compares standard CPAP with CPAP-Flex in World Trade Center responders with sleep apnea. The primary endpoint is adherence in hours/night during the first treatment period, before crossover. The posted medians are 1.0 for standard CPAP and 0.89 for CPAP-Flex; no primary inferential comparison is reported.

The secondary CPAP efficacy result is reported in events/hour and has a recalculated unadjusted mean difference of 0.23 for CPAP-Flex relative to standard CPAP, with p = 0.599. The registry record does not report an investigator analysis, a multiplicity plan, or detailed methods for missing data and crossover analyses. Safety listings record zero serious adverse events, zero other adverse events, and zero deaths in both listed groups.