Tiotropium Respimat Versus HandiHaler on SaO2 and Sleep in COPD Patients: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Tiotropium Respimat Versus HandiHaler on SaO2 and Sleep in COPD Patients (NCT02331940), a randomized na study. The primary outcome is Sleeping Oxygen Saturation.

2. Clinical Question

The study aim was to compare the devices in relation to sleeping oxygen saturation and sleep quality in people with COPD. The primary outcome was mean sleeping oxygen saturation, expressed as a percentage of oxygen saturation. The registry describes the endpoint as a mean, with standard deviation as its measure of spread. The reported group means were both 94.4, and the recalculated mean difference for Respimat versus HandiHaler was 0.0.

The clinical question is specifically about a comparison between two ways of delivering tiotropium, not a comparison of tiotropium with no treatment or with a different drug. The registry lists sleepiness and hospitalization rate among the secondary outcomes, but numerical results for hospitalization rate are not reported in the registry record. The posted sleepiness results are a separate secondary measure and should not be treated as a substitute for the primary oxygen-saturation outcome.

3. Trial Design

The registry describes a randomized, parallel-group trial with no masking. The comparison was between tiotropium delivered by the HandiHaler device and tiotropium delivered via the Respimat Soft Mist Inhaler. The HandiHaler arm description identifies it as a single-dose dry powder inhaler. The Respimat arm description identifies the Soft Mist Inhaler. The study sponsor was the University of Crete.

Enrollment was 200. The registry lists the study as completed, with a start date of 2010-03 and a primary completion date of 2013-12. The phase is recorded as NA. The primary endpoint was assessed 6 months after treatment initiation. The endpoint results list 93 participants for HandiHaler and 95 for Respimat. The registry record does not explain how those endpoint counts relate to the enrollment figure or report the reasons for any differences.

Random allocation means that assignment to the two groups was intended to occur by chance. In general, randomization can help balance characteristics between groups, including characteristics that are not measured, but the registry record does not report how the allocation sequence was generated or concealed. The trial was unmasked, so participants and study personnel were not blinded to the assigned device, according to the masking entry. Whether outcome assessors or analysts were masked is not reported in the registry record.

The registry record does not report treatment duration beyond the assessment timing, detailed eligibility criteria, the planned analysis population, or the planned sample-size justification. It also does not describe an interim-analysis plan or the details of the statistical analysis plan. Those omissions limit how fully the design and the handling of the results can be assessed from the registry information alone.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Overall Study

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 Sleeping Oxygen Saturation, measured over 6 months after treatment initiation.

The primary endpoint is mean sleeping oxygen saturation, reported as a percentage of oxygen saturation and assessed 6 months after treatment initiation. Oxygen saturation describes the proportion of hemoglobin carrying oxygen, and measurements during sleep can characterize oxygen levels during that period. The registry identifies this endpoint as a continuous measure summarized by its mean and standard deviation.

Using a mean provides a concise summary of the values observed in each group. The standard deviation describes how dispersed individual values are around that group mean; it is not a measure of uncertainty in the mean itself. The endpoint was chosen to address the study’s stated question about sleeping oxygen saturation in people with COPD. General statistical context: a continuous endpoint allows the group comparison to be expressed as a difference in average values, while retaining information about the scale of the measurement.

The registry reports a mean of 94.4 for HandiHaler and 94.4 for Respimat, with standard deviations of 2.4 and 2.2, respectively. These are group summaries at the stated follow-up time. They do not describe every participant’s measurement, the change from each participant’s starting value, or the pattern of measurements at other times. Baseline oxygen-saturation results and change-from-baseline results are not reported in the registry record.

The endpoint is a measurement, not by itself a complete account of sleep quality or other aspects of COPD. The registry lists sleepiness and hospitalization rate as secondary outcomes, but these address different questions. The primary endpoint result should therefore be interpreted as a comparison of mean sleeping oxygen saturation at the specified assessment time, rather than as a summary of every outcome relevant to the study aim.

7. Results Posted in the Registry

• Respimat: mean 94.40 versus 94.40 on control; difference 0.00 (95% CI -0.66 to 0.66), standardized effect d = 0.00; z-test p = 1.000.

8. Secondary Efficacy Endpoints

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

Sleepiness

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Respimat: mean 6.10 versus 6.50 on control; difference -0.40 (95% CI -2.07 to 1.27), standardized effect d = -0.07; z-test p = 0.639.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Respimat and the control. The interval gives the range of effects that remain compatible with the data.

The registry lists sleepiness and hospitalization rate among the secondary outcomes. Numerical results are posted for sleepiness, while numerical results for hospitalization rate are not reported in the registry record. The secondary material therefore offers an additional continuous outcome but does not provide a complete numerical account of every outcome named in the record.

Across the posted primary and secondary continuous outcomes, the estimates should be considered separately because they measure different aspects of the study aim. The primary endpoint concerns mean sleeping oxygen saturation; sleepiness concerns a scale measured in units on a scale. The registry’s outcome summaries and recalculations do not establish that a result on one measure explains or confirms a result on another. The overall pattern is best described as separate estimates with their own uncertainty, rather than as a single combined treatment effect.

Secondary outcomes generally carry less confirmatory weight when their analysis plan, role in the testing sequence, and adjustment for multiple comparisons are not reported. Testing several outcomes can increase the chance that at least one result appears notable by chance. General statistical context: multiplicity procedures can control that risk, but the registry record does not report whether any such procedure was planned or used. The posted secondary results are unadjusted, so their p-values, where calculated, should be interpreted as exploratory evidence rather than as independent confirmation of the primary finding.

The registry does not report a prespecified hierarchy among the secondary outcomes, a multiplicity adjustment, or a complete analysis plan for them. The planning illustrations attached to the sleepiness comparison are not trial-design evidence and do not increase the confirmatory status of that outcome. The secondary results can add context to the primary comparison, but they do not resolve the limitations of the primary estimate or substitute for unreported results on hospitalization rate.

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

• Sleepiness
• Hospitalization Rate

9. Statistical Interpretation

10. Statistical Methods Explained

What does the mean difference describe?

The primary endpoint is continuous, so the group summaries are means and standard deviations. The mean is the average value in a group. The standard deviation describes the spread of individual values around that mean. These statistics answer different questions: a group can have a particular average while individuals vary around it, and the standard deviation does not tell how precisely the average has been estimated.

The comparison is expressed as Respimat minus HandiHaler. The posted means are both 94.4, so the recalculated unadjusted mean difference is 0.0. A positive difference would mean a higher average in the Respimat group under that direction of subtraction; a negative difference would mean a lower average. This direction is important when reading a contrast, especially when results are reported for multiple outcomes or comparisons.

What is the recalculated test?

The recalculated comparison is an unadjusted two-group comparison based on the group summary statistics. It is not an adjusted model and does not account for covariates or baseline measurements. The recalculated unadjusted p-value is 1.0. In general, a p-value describes how unusual data at least as extreme as those observed would be under a specified null hypothesis and statistical model. It is not the probability that the null hypothesis is true, and it does not measure the size or importance of a difference.

The registry’s own reported analysis is a one-sided t-test with p = 0.88. A one-sided test evaluates evidence in a specified direction rather than in either direction, but the registry record does not identify the direction or provide the test statistic and other analysis details. The recalculated unadjusted result and the sponsor-reported result should therefore be understood as different entries with different stated methods, not as two versions of precisely the same calculation.

How should the confidence interval be read?

The recalculated 95% confidence interval for the primary mean difference is -0.658 to 0.658. In general, a confidence interval combines an estimate with its sampling uncertainty under the assumptions of the method. A narrower interval indicates greater precision for the estimated contrast; a wider interval indicates less precision. The interval is expressed on the same scale as the mean difference and includes zero. Its inclusion of zero means the recalculation does not distinguish the observed contrast from no difference at the chosen interval level; it does not establish that all meaningful differences have been excluded.

Confidence intervals depend on assumptions about the data and the calculation. General assumptions for a two-group t-based comparison include independent observations and a suitable model for the distribution of the mean difference. The posted summary statistics alone do not allow a full assessment of data shape, outliers, dependence, or protocol deviations. The registry record does not report diagnostic checks or a detailed analysis plan.

What role does randomization play?

In general, random allocation is used to reduce systematic differences between groups at the point of assignment. It does not guarantee that every measured or unmeasured characteristic will be balanced, particularly in a finite sample. The registry identifies the study as randomized, but does not report sequence generation, allocation concealment, or baseline characteristics. The primary result is therefore read in the context of the stated randomized design, while details needed to evaluate how assignment was implemented are not reported in the registry record.

The study is listed as unmasked. General statistical context: knowledge of assigned treatment can affect some aspects of trial conduct or assessment, depending on the outcome and procedures. Sleeping oxygen saturation is a measured endpoint, but the registry record does not describe measurement procedures, assessor masking, or whether outcome assessment could be influenced by awareness of assignment. No more specific conclusion about bias can be drawn from the masking entry alone.

What do planning illustrations mean?

The sleepiness results include sample-size planning illustrations based on fractions of the observed effect. These are planning scenarios, not observed trial results and not observed power. They assume a two-sided significance level of 0.05, equal allocation, and a normal approximation. The displayed counts are participants per arm for 80% power or participants per arm for 90% power for the stated fraction of the observed effect. Such calculations show how the required sample size changes when a planning effect is smaller or larger; they do not establish what the trial was designed to detect.

The registry explicitly describes these as illustrations based on the observed effect, not the trial’s own design. The sample-size justification used when the trial was planned is not reported in the registry record. The planning values should not be used to infer the study’s original target, its achieved power, or whether the study was adequately powered for the primary endpoint.

11. Confidence Intervals and Precision

The recalculated primary mean difference is 0.0, with a 95% confidence interval from -0.658 to 0.658. The interval gives a sense of precision by displaying a range of contrasts compatible with the data under the calculation’s assumptions. It is centered on the reported mean difference and extends in both directions, so the posted summaries are compatible with a modest difference favoring either group as well as with no difference.

The range is stated in the endpoint’s units, percentage of oxygen saturation. Its interpretation should remain on that scale: it concerns the estimated difference between group means, not the spread of individual values. The group standard deviations are 2.4 for HandiHaler and 2.2 for Respimat. Those values describe within-group variation, while the confidence interval describes uncertainty in the estimated between-group mean difference.

Precision is not the same as clinical importance. Whether values near either confidence-interval limit would matter depends on a meaningful threshold or prespecified margin. Such a threshold is not reported in the registry record. Nor is an equivalence margin reported, so the interval cannot be used as a formal demonstration that the devices are equivalent.

The interval is a recalculated, unadjusted estimate based on posted summary statistics. The registry’s sponsor-reported one-sided t-test gives p = 0.88 but does not report a confidence interval. The registry does not describe enough detail to reconstruct that analysis. Baseline values, missing-data handling, and the analysis population are also not reported in the registry record. These omissions limit how completely precision can be evaluated beyond the stated recalculation.

12. Analyses Reported by the Sponsor

The registry reports a sponsor analysis using a one-sided t-test, with p = 0.88 for the comparison between the two groups. It does not report a test statistic, confidence interval, or details of the one-sided hypothesis direction. The sponsor result is therefore best described as the reported result of that stated analysis, without inferring additional methodological details that are absent from the registry record.

The recalculation uses the posted group means, standard deviations, and sample sizes to make an unadjusted comparison. It gives a mean difference of 0.0, a 95% confidence interval from -0.658 to 0.658, and an unadjusted p-value of 1.0. These are not the same output as the registry’s one-sided test. In particular, the recalculation does not attempt to reproduce an analysis whose full specification is not reported.

The registry does not flag a discrepancy. The different p-values should not be treated as evidence that one calculation is necessarily erroneous: the stated test direction and other analysis choices can affect inferential results, and the record does not give enough detail to determine the precise reason for the difference. The sponsor’s reported analysis and the unadjusted recalculation should remain distinct when interpreting the result. Neither entry, on its own, establishes equivalence between the devices.

13. Multiplicity, Missing Data and Analysis Populations

The registry record lists one primary outcome and names sleepiness and hospitalization rate among the secondary outcomes. It also reports four outcome measures and one analysis. Numerical endpoint details are posted for the primary sleeping oxygen-saturation measure and for sleepiness. Numerical results for hospitalization rate are not reported in the registry record. The record does not explain how the listed outcome measures relate to the posted analyses or identify a complete hierarchy for testing them.

Multiplicity arises when several outcomes, time points, subgroups, or statistical tests are examined. In general, the more tests conducted, the greater the chance that at least one will appear statistically unusual even if there is no corresponding underlying effect. Statistical approaches can address multiplicity, for example by adjusting significance thresholds or by specifying an ordered testing strategy. Whether any such approach was prespecified or used is not reported in the registry record. Consequently, secondary findings should not be interpreted as though the registry had documented control of the overall false-positive risk.

The record reports endpoint sample sizes, but does not state the analysis population, such as whether results used all randomized participants, participants who received treatment, or participants with an available outcome. It does not report missing-data counts by outcome, reasons for missing measurements, or any imputation or sensitivity-analysis method. The endpoint group counts alone do not establish how missing observations were handled or whether analysis followed the assigned groups.

Baseline characteristics, baseline endpoint values, protocol deviations, and details of participant flow are not reported in the registry record. The record also does not report a prespecified statistical analysis plan, interim analyses, or a sample-size justification for the trial. These are distinct reporting gaps: they limit assessment of the planned testing strategy, the handling of incomplete data, and the relation between the enrolled and analyzed populations. They should not be filled in by assumptions about common trial practice.

14. 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 zero serious adverse events among 93 participants at risk in the HandiHaler group and zero among 95 participants at risk in the Respimat group. It also lists zero other adverse events among 93 and 95 participants at risk, respectively. The registry does not provide serious adverse-event terms in the posted list. Death counts are not reported in the registry record.

These figures describe the counts recorded in the adverse-event summary. They do not show that no adverse events could occur, or establish that the groups have the same safety profile. A zero count in a study does not rule out events that are uncommon, occur outside the observation period, or were not captured or reported. The registry record does not provide event definitions, collection procedures, follow-up duration for safety, or details about the severity and relationship assessments for other events.

The study’s primary endpoint was sleeping oxygen saturation, not a safety comparison. The adverse-event figures can be read as a limited description of what was posted for this trial, but they are not sufficient on their own to characterize all possible harms or compare safety with high precision.

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

The primary result describes a comparison of average sleeping oxygen saturation at 6 months after treatment initiation. Both posted group means were 94.4, and the recalculated unadjusted mean difference was 0.0. Its 95% confidence interval, -0.658 to 0.658, indicates uncertainty around that estimate under the calculation’s assumptions. The registry separately reports a one-sided t-test with p = 0.88.

These findings do not prove that the devices have identical effects. A zero difference between the posted means is an observed summary, not proof that the underlying group averages are exactly the same. Likewise, an interval that includes zero does not demonstrate equivalence. Formal equivalence or non-inferiority conclusions require a relevant margin and a suitable prespecified analysis, neither of which is reported in the registry record.

The result also does not describe changes from baseline, individual response patterns, or outcomes at time points other than the reported assessment. Baseline values and repeated measurements are not reported in the registry record. The endpoint is one measure related to the study aim; it does not by itself capture sleep quality as a whole, sleepiness, hospitalization, or all other outcomes relevant to COPD.

The trial record identifies randomized allocation, parallel groups, and no masking. Those design entries provide context, but details about allocation methods, analysis populations, missing data, and outcome assessment procedures are not reported. The findings should therefore be read as the posted group summaries and their associated analyses, rather than as a complete account of every aspect of trial conduct or every possible effect of the devices.

In general, statistical significance, precision, and clinical importance are separate considerations. The reported p-values address statistical evidence under particular testing approaches; the confidence interval conveys uncertainty around the recalculated mean difference; and clinical importance depends on thresholds not reported in the registry record. No medical advice or treatment recommendation follows from these statistics alone.

16. Why This Trial Matters Statistically

This trial illustrates why an estimated effect, its uncertainty, and the analysis method need to be read together. Identical posted group means give a recalculated mean difference of 0.0, but the confidence interval remains important because it shows that the estimate is not perfectly precise. A point estimate alone cannot establish that the groups are equivalent or exclude every potentially meaningful difference.

It also shows why a registry’s reported analysis should be distinguished from a recalculation based on summary data. The registry reports a one-sided t-test with p = 0.88; the unadjusted recalculation gives p = 1.0 and a 95% confidence interval. These entries have different stated forms and should not be conflated. Without the full analysis specification, the recalculation cannot be assumed to reproduce the sponsor’s method.

The design information matters, too. Randomization can support a fair group comparison in general, while the unmasked status and missing details about allocation, outcome assessment, analysis populations, and missing data leave important questions unaddressed. Those limitations are about the strength and interpretability of the evidence, not proof that bias occurred.

Finally, the secondary results and planning illustrations demonstrate the distinction between exploratory findings and trial-design evidence. Planning calculations based on an observed effect are illustrations, not observed power and not a substitute for the original sample-size justification. For biostatistical interpretation, the central task is to keep the target outcome, estimate, uncertainty, analysis specification, and reporting gaps visible at the same time.

17. 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.

Several limitations specific to this trial constrain interpretation. The primary endpoint is reported at 6 months after treatment initiation, but baseline sleeping oxygen-saturation results and changes from baseline are not reported in the registry record. The posted comparison therefore concerns the group means at the stated follow-up, not a comparison of within-participant change.

The registry reports a randomized design but does not describe sequence generation or allocation concealment. It lists the trial as unmasked, but does not state whether endpoint assessors or analysts were masked. Measurement procedures for sleeping oxygen saturation are also not reported. These omissions prevent a detailed assessment of how assignment and outcome collection were implemented.

Endpoint counts are posted for the two arms, but the registry does not identify the analysis population or explain how those counts relate to the enrollment figure. Missing-data counts, reasons for missing outcomes, and methods for addressing missingness are not reported. Participant flow and protocol deviations are likewise not described in the registry record.

The sponsor-reported analysis is described as a one-sided t-test, but the hypothesis direction and further analytical details are not reported. The registry also does not report an adjustment for multiple outcomes, a hierarchy for secondary testing, or a complete statistical analysis plan. Numerical results for hospitalization rate are not reported in the registry record, despite the outcome being listed. Death counts are not reported in the registry record, and the adverse-event summary does not provide further details about event definitions or ascertainment.

Finally, the original sample-size justification is not reported. Planning illustrations attached to sleepiness are based on the observed effect and are not evidence of the trial’s intended detectable effect. These reporting limitations make it inappropriate to infer unreported design choices or data-handling decisions.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Tiotropium Respimat Versus HandiHaler on SaO2 and Sleep in COPD Patients — NCT02331940.
• Registry record last updated 2017-03-30; status last verified 2017-02.
• Results first posted 2017-03-30.
• 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.

21. Record Summary

NCT02331940 is a completed randomized, parallel-group, unmasked trial sponsored by the University of Crete. It compared tiotropium delivered by HandiHaler with tiotropium delivered via Respimat in people with COPD. Enrollment was 200, and the primary endpoint was mean sleeping oxygen saturation at 6 months after treatment initiation.

The posted primary means were 94.4 in both groups. The recalculated unadjusted mean difference was 0.0, with a 95% confidence interval from -0.658 to 0.658 and an unadjusted p-value of 1.0. The registry reports a one-sided t-test with p = 0.88. These are distinct analyses, and the registry does not report enough detail to reconstruct the sponsor’s test.

Sleepiness results are posted as a secondary outcome; numerical hospitalization-rate results are not reported in the registry record. The record does not report the original sample-size justification, analysis population, missing-data methods, or multiplicity procedure. The estimates therefore describe the posted outcomes and analyses, with these reporting limitations in view.