A Placebo-controlled Trial of Daliresp on Chronic Obstructive Pulmonary Disease (COPD): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Placebo-controlled Trial of Daliresp on Chronic Obstructive Pulmonary Disease (COPD) (NCT01572948), a randomized na study. The primary outcome is Induced Sputum Proline-glycine-proline (PGP) Levels at Baseline.

2. Clinical Question

The study summary describes an aim to examine whether roflumilast is associated with changes in inflammatory biomarkers and sputum neutrophilia, and to relate those findings to lung function, sputum scores, and quality of life. It also says the investigators sought to assess findings regardless of concomitant medication use, including inhaled corticosteroids. Those aims provide context for the trial, but the posted primary outcome is narrower: a baseline biomarker measure. The distinction matters because an objective described in the study summary is not itself evidence that the corresponding outcome was measured, analyzed, or reported. Further details on those objectives and analyses are not reported in the registry record.

3. Trial Design

The registry record describes a completed, randomized, triple-masked, parallel-group, placebo-controlled trial sponsored by the University of Alabama at Birmingham. The record lists a start date of 2012-06 and a primary completion date of 2014-06. It records enrollment of 27 and lists the phase as NA. The active group is called Roflumilast in the primary outcome record and Daliresp in the intervention description; the comparator group is Placebo.

The posted primary outcome has a baseline mean of 0.74 ng/ml in the placebo group, with a standard deviation of 0.25, and a mean of 0.64 ng/ml in the roflumilast group, with a standard deviation of 0.21. The posted group sizes are 16.0 and 11.0, respectively. The registry comparison is an unadjusted mean difference of -0.1 ng/ml, with p = 0.261. These are baseline values, not a record of change from baseline.

The design labels describe important structural features, but not every practical detail. “Randomized” identifies the allocation approach in the registry; the random sequence method and allocation concealment are not reported in the registry record. “Triple” is the recorded masking designation, but which groups of people were masked and how masking was maintained are not reported. The parallel model indicates separate groups rather than a crossover comparison. The registry does not report treatment duration, adherence, or the timing of the primary outcome beyond identifying it as baseline. It also does not report a phase beyond the listed NA designation.

Although the study summary refers to several clinical and mechanistic objectives, the posted record does not provide enough detail to reconstruct the full protocol or analysis plan. In particular, eligibility criteria, participant flow, baseline clinical characteristics, the intended treatment period, and the original design assumptions are not reported in the registry record. Those absences limit how fully the randomized design and the relationship between the stated aims and posted results can be assessed.

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 Induced Sputum Proline-glycine-proline (PGP) Levels at Baseline, measured over baseline.

The primary measure is induced-sputum proline-glycine-proline (PGP) level at baseline, expressed as a mean in ng/ml with a standard deviation. PGP is the biomarker named in the endpoint; the registry record does not give a fuller description of its measurement method, laboratory procedures, or interpretation. The study summary places this measure within a broader interest in pulmonary inflammation and sputum neutrophilia.

In general, a continuous measure retains information about the amount measured rather than reducing it to categories such as present or absent. Reporting a group mean describes the central value, while the standard deviation describes variation among observations around that mean. Those summaries can help readers compare the observed group distributions, although they do not show the full distributions or establish why the values differ.

The timing is central to interpretation: the registry explicitly labels the primary outcome as baseline. Its posted contrast therefore describes a difference between group means at baseline, not a treatment-associated change after exposure. A baseline measure may describe the groups entering a trial or provide a reference for later comparisons, but no later primary PGP result is reported in the registry record. The record also does not describe why this particular PGP measure was selected as the primary endpoint or how it was intended to connect to the broader clinical aims. The posted numbers can be interpreted as a baseline comparison only.

7. Results Posted in the Registry

• Roflumilast: mean 0.64 versus 0.74 on control; difference -0.10 (95% CI -0.27 to 0.07), standardized effect d = -0.43; z-test p = 0.261.

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.

Induced Sputum Neutrophil Count

• 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 errors.
• Result:
  • Daliresp: mean 80.00 versus 83.00 on control; difference -3.00 (95% CI -3.67 to -2.33), standardized effect d = -3.34; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Daliresp than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The registry lists induced sputum neutrophil count among its secondary outcomes three times, while the detailed secondary result available for interpretation is a one-month measure reported as a mean percentage of sputum neutrophils. No other secondary outcome results are included in the detailed posted results. The trial summary mentions pulmonary function, sputum scores, quality of life, and mechanistic aims, but those mentions do not establish that results for those measures were posted. Their detailed results are not reported in the registry record.

The available secondary result is a distinct time point and outcome from the primary baseline PGP measure. It therefore adds information about a different inflammatory measure, but it does not answer the primary baseline comparison or supply a later PGP result. The secondary result can be considered alongside the primary finding as part of the broader set of posted evidence, while keeping the two outcomes and their timing separate. The overall pattern cannot be characterized across a broad collection of outcomes when detailed results are available for only one secondary measure.

In general, unadjusted secondary results carry less weight than a prespecified primary analysis because they may be among several outcomes, time points, or comparisons examined. When multiple tests are conducted, the chance of seeing an apparently unusual result by chance can rise unless multiplicity is addressed. The registry record does not report a multiplicity plan, a hierarchy of outcomes, or sponsor-reported analyses. It also records no reported analyses. These omissions make it difficult to know how the secondary result was selected or whether other analyses were conducted but not posted.

Accordingly, the secondary finding should be treated as a reported signal requiring context rather than as confirmation of the broader study aims. No inference about unreported outcomes, or about a consistent pattern across all intended measures, is supported by the registry record.

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

• Induced Sputum Neutrophil Count
• Induced Sputum Neutrophil Count
• Induced Sputum Neutrophil Count

9. Statistical Interpretation

10. Statistical Methods Explained

What does a mean difference describe?

For a continuous outcome, a mean difference compares the average measurement in one group with the average in another. The posted primary contrast is roflumilast minus placebo, so a negative value indicates a lower average in the roflumilast group for this measure. Here the estimate is -0.1 ng/ml. A mean difference keeps the outcome in its original units, which helps readers connect the contrast to the measured biomarker. It does not describe every participant’s value, and it does not indicate whether an observed difference is important in clinical terms.

In general, the standard deviation describes dispersion of individual measurements around a group mean. The primary outcome record reports standard deviations of 0.25 for placebo and 0.21 for roflumilast. Standard deviations should not be confused with standard errors: a standard deviation characterizes variation in observations, while a standard error characterizes uncertainty in an estimate such as a mean. The registry identifies the primary dispersion measure as a standard deviation. The one-month secondary outcome instead labels its dispersion as a standard error.

What does the p-value say?

The primary comparison has an unadjusted p-value of 0.261. In general, a p-value describes how compatible the observed data are with a specified null model, under the assumptions used for the test. It is not the probability that the null hypothesis is true, nor does it measure the size or clinical importance of a difference. The registry does not report a sponsor analysis or enough detail to identify a complete inferential procedure. The comparison displayed is a recalculation from posted summary statistics.

A p-value above a conventional threshold does not prove that groups are identical. It can reflect a small underlying difference, noisy measurements, limited information, or other features of the data and analysis. Conversely, a small p-value does not establish that an effect is important or unbiased. The result should therefore be read together with the estimated difference, its uncertainty, timing, and the outcome’s role in the study.

Why does the confidence interval matter?

In general, a confidence interval gives a range of effect estimates compatible with the data and the interval procedure. It communicates precision in the units of the outcome, not just whether a threshold was crossed. For the primary comparison, the recalculated interval includes zero. That means the posted data are compatible with no mean difference as well as values on either side of zero under the interval method. The interval does not establish that all such values are equally plausible, and it does not address bias from missing data or unreported design choices.

The interval is based on a recalculation from summary statistics. The registry record does not provide a sponsor-reported interval or analysis details. Because the record does not flag a discrepancy, there is no stated disagreement to resolve between the recalculation and a sponsor analysis. Still, the recalculation is an unadjusted summary comparison, not evidence that all trial-level features were incorporated.

Why can over-dispersion matter?

In general, over-dispersion occurs when observed variability is greater than a model assumes. When a model assumes less variability than the data contain, standard errors may be too small and intervals too narrow, making results appear more precise than they are. If variability is greater than assumed, the reverse can also occur. This is a general statistical concern, not a claim that over-dispersion was present in this trial. The registry identifies the primary measure as continuous and reports group standard deviations, but it does not report a distributional assessment or a model for the outcome.

For continuous outcomes, analysts may consider whether a mean adequately summarizes the data and whether assumptions behind a comparison are reasonable. The registry does not report raw observations, diagnostics, transformations, robust methods, or adjusted models. Those methods cannot be inferred from the posted group summaries.

What can recalculated comparisons establish?

The posted comparison is labeled as a mean difference and is recalculated without adjustment from the summarized group data. Such a calculation offers a transparent way to describe the observed contrast, but it may not match an analysis that accounts for covariates, repeated observations, or other protocol features. No reported sponsor analysis is listed, and no analysis population is described. The record does not state whether the primary result used all randomized participants, only participants with a measurement, or another population.

General statistical interpretation also depends on the question’s timing. A baseline comparison describes group values at baseline. A treatment effect would ordinarily require a comparison tied to post-baseline measurements, change over time, or another specified estimand. The posted primary endpoint does not supply that later comparison. Thus, statistical language about the baseline contrast should not be extended into a conclusion about treatment-related change.

11. Confidence Intervals and Precision

The primary mean difference is -0.1 ng/ml, with a recalculated confidence interval that includes zero. The interval therefore allows for a range of underlying group differences that includes no difference as well as differences in either direction. The p-value of 0.261 is consistent with the fact that the posted estimate is not sharply distinguished from zero under the unadjusted comparison. Neither the p-value nor the interval proves that the group means are the same.

In general, precision is the degree to which an estimate is constrained to a relatively narrow range. A confidence interval is useful because it puts that uncertainty on the outcome’s scale. A narrow interval can indicate a more precise estimate than a wide one, while still depending on the validity of the analysis assumptions. The posted primary interval is a recalculation from summary statistics; sponsor-reported analysis details are not reported in the registry record.

The group sizes recorded for the primary outcome are 16.0 for placebo and 11.0 for roflumilast. With limited observations, estimates can be sensitive to the measured values, and uncertainty may remain substantial. The registry does not provide individual-level data, so the influence of unusually high or low observations cannot be assessed. It also does not report missingness, the number with outcome data relative to the number randomized, or whether the analysis adjusted for baseline features.

Precision should also be separated from relevance. An interval describes statistical uncertainty in a mean difference; it does not identify a threshold for a meaningful change in PGP. No such threshold is reported in the registry record. Most importantly, the primary measure is at baseline. Even a more precise baseline difference would not, by itself, estimate a treatment effect over time.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning illustrations express the required sample as participants per arm, not total events. The displayed planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each planning row uses a stated fraction of the observed effect as the effect to detect. The registry record does not report the original trial’s design assumptions or sample-size justification, so these illustrations should not be mistaken for the rationale behind the trial’s enrollment.

The 80% power column describes the number of participants per arm that the planning calculation estimates would give an 80% chance of detecting the specified effect under those assumptions. The 90% power column gives the corresponding planning estimate for a 90% chance. These are hypothetical planning targets, not statements about the probability that the posted trial found a true effect. This is not observed power.

Rows that plan for a smaller fraction of the observed effect require more participants per arm. That is a general feature of power calculations: when the difference one wants to detect is smaller, more information is usually needed to distinguish it from random variation under a chosen design. The illustrations are anchored to the observed estimate, so they inherit its uncertainty and may not represent a clinically meaningful target. They also use simplified assumptions and do not establish what a future study should do.

The calculations are not a retrospective verdict on whether the trial was adequately designed. The original protocol’s assumptions, anticipated effect, allowance for missing outcomes, and any planned adjustments are not reported in the registry record. The planning figures should be read as conditional illustrations based on the stated calculation settings, rather than as evidence about the trial’s own statistical planning.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = -0.100; pooled standard deviation σ = 0.235.
2. n = 2 × (1.96 + 0.842)² × 0.235² / -0.100² = 86.6, rounded up to 87 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- -0.1                      # observed difference in means
sigma <- 0.234819079292974                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(1572948)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 87 (80% power), 116 (90% power)
 75% of observed effect: n per arm = 154 (80% power), 207 (90% power)
 50% of observed effect: n per arm = 347 (80% power), 464 (90% power)
power.t.test: 87.5 
Simulated power at n = 87 per arm: 0.795 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when several outcomes, time points, or statistical comparisons are examined. In general, testing many hypotheses without a plan to account for that process can make at least one apparently strong result more likely to occur by chance. This does not mean that a secondary finding is false; it means its evidential weight depends partly on what else was measured and how the analyses were planned. The registry lists the induced sputum neutrophil count three times among secondary outcomes and records six outcome measures, but does not provide a multiplicity adjustment or a hierarchy explaining how results were to be interpreted.

The detailed secondary outcomes include one reported one-month neutrophil result, while the primary PGP measure is at baseline. The registry reports no sponsor analyses. It does not state whether other outcomes were measured but not reported, whether there were additional planned time points, or whether the posted analysis was prespecified. The analysis plan, including any interim analyses, is not reported in the registry record.

Missing-data methods and analysis populations are also not reported. The record does not describe whether participants who lacked an outcome measurement were excluded, whether any imputation was used, or whether the primary analysis followed an intention-to-treat principle. Participant flow, withdrawals, reasons for missing measurements, and outcome-specific completeness are not reported. These details matter because an estimate based only on observed measurements can be affected if the availability of data is related to participant characteristics or outcomes.

The registry does not report the randomization sequence method, allocation concealment, or which people were masked under the triple-masking label. These are not evidence of a flaw; they are details unavailable in the record. Without a protocol or fuller statistical analysis plan, it is not possible to determine the planned treatment contrasts, handling of covariates, treatment of multiple outcomes, or whether reported analyses follow the original plan. Interpret the posted comparisons as limited summaries rather than a complete account of trial conduct and analysis.

15. Safety Results

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

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent non-serious adverse event terms

The posted serious adverse-event counts are zero in the placebo group among 16 participants at risk and zero in the roflumilast group among 11 participants at risk. No serious adverse-event terms are listed. These counts describe what the registry reports; they do not establish that no serious events occurred outside the recorded observation or that the groups have identical safety profiles. The record does not report follow-up duration for safety or further detail about event collection.

Other adverse events are recorded for 4 participants in the placebo group among 16 at risk and 5 participants in the roflumilast group among 11 at risk. These are counts, not a comparison adjusted for exposure or follow-up. No causal relationship between an event and treatment can be inferred from the counts alone. Death figures are not reported in the registry record. The posted safety information is therefore limited and should be read as a record of reported counts, not as a complete safety assessment.

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

The primary posted result describes baseline PGP levels, not the effect of treatment. The mean was 0.74 ng/ml in the placebo group and 0.64 ng/ml in the roflumilast group, with an unadjusted mean difference of -0.1 ng/ml and p = 0.261. The recalculated confidence interval includes zero. This pattern indicates uncertainty around the baseline contrast; it does not show that the groups were identical, and it does not estimate how either group changed after treatment.

The study summary describes broader objectives involving inflammatory biomarkers, sputum neutrophilia, pulmonary function, sputum scores, quality of life, concomitant medication use, and a proposed mechanistic pathway. Those objectives are not all represented by detailed outcome results in the registry record. The available one-month secondary neutrophil measure is separate from the baseline primary PGP endpoint. No conclusion about unreported outcomes or the full set of stated aims can be drawn from the available results.

In general, a statistically uncertain estimate is not proof that an intervention has no effect, just as a small p-value is not proof of a clinically important effect. Statistical results describe data under particular measurement and analysis choices. Clinical meaning requires context, including what size of change would matter, how outcomes were collected, and whether the result is consistent with other evidence. A threshold for meaningful PGP change is not reported in the registry record.

The limited enrollment, absent sponsor analyses, and missing information about analysis populations and missing data further constrain interpretation. The trial record is useful for identifying what was measured and what summary results were posted, but it does not support a broad conclusion about treatment benefit, lack of benefit, or clinical use. No medical advice follows from these statistical summaries.

17. Why This Trial Matters Statistically

This trial illustrates why endpoint timing can be as important as the name of an outcome. The primary endpoint is a biomarker that fits the study’s stated interest in pulmonary inflammation, but its recorded timeframe is baseline. A baseline group contrast may describe the groups at entry; it cannot stand in for a post-treatment effect or a change-from-baseline comparison. Distinguishing those questions prevents a descriptive result from being interpreted as an intervention effect.

It also demonstrates the value of separating an estimate from its uncertainty. The mean difference and p-value give different but related information, while the confidence interval indicates that the primary contrast is not precisely separated from zero. None of these quantities alone settles whether the difference matters clinically. A registry record is most informative when it includes outcome timing, complete analysis details, and context for interpreting the scale of an effect.

The secondary finding highlights another statistical issue: a striking unadjusted result can coexist with uncertainty about multiplicity and reporting. Without a reported hierarchy or analysis plan, readers cannot know how the result relates to other planned outcomes or comparisons. That uncertainty is especially relevant when the study summary names several clinical and mechanistic aims but detailed posted results are limited.

Finally, the record shows the limits of registry summaries for reconstructing an analysis. Important details about missing data, participant flow, treatment duration, randomization procedures, and sponsor analyses are not reported. Recognizing what is and is not documented is part of sound statistical interpretation: it preserves the value of the posted numbers without asking them to answer questions they do not address.

18. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

Several limitations are specific to the information recorded for this trial. The primary endpoint is a baseline PGP measurement, so the posted primary comparison does not document a treatment-related change. The registry summary describes a broader set of scientific aims, but detailed results for those aims are not reported in the registry record. The only detailed secondary result available is not a later PGP measurement.

The registry records enrollment of 27 and outcome group sizes of 16.0 for placebo and 11.0 for roflumilast. It does not report participant flow or explain whether the outcome group sizes reflect all randomized participants, missing measurements, or another analysis population. Eligibility criteria and baseline clinical characteristics are also not reported, limiting assessment of the study population and how it relates to the intended stable moderate to severe COPD population.

The primary comparison is recalculated from summary statistics, and sponsor-reported analyses are absent. The record does not describe the original sample-size rationale, a statistical analysis plan, adjustment for covariates, multiplicity handling, missing-data methods, or any interim analyses. It also lacks detail about measurement procedures for PGP, treatment duration, and the methods used to preserve masking. These gaps constrain assessment of both the estimate and the trial conduct.

The adverse-event information is similarly limited: serious-event counts and other-event counts are posted, but event terms, safety follow-up details, and deaths are not fully reported. These are limitations of the available registry record, not proof that the underlying study did not collect further information. The posted results should be interpreted within the scope of what is documented.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: A Placebo-controlled Trial of Daliresp on Chronic Obstructive Pulmonary Disease (COPD) — NCT01572948.
• Registry record last updated 2015-10-19; status last verified 2015-09.
• Results first posted 2015-07-31.
• 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.

22. Record Summary

The registry record describes a completed randomized, triple-masked, parallel-group, placebo-controlled trial of roflumilast in COPD, sponsored by the University of Alabama at Birmingham. It records enrollment of 27 and lists the primary outcome as mean induced-sputum PGP level at baseline. The posted baseline means are 0.74 ng/ml for placebo and 0.64 ng/ml for roflumilast; the unadjusted mean difference is -0.1 ng/ml, with p = 0.261.

The primary comparison is a baseline contrast, not a treatment-effect estimate over time. A detailed one-month secondary neutrophil result is also posted, but the registry lists no sponsor analyses and does not report a multiplicity plan, missing-data methods, or analysis populations. Serious adverse-event counts are zero in both groups; other adverse-event counts are posted, while deaths are not reported in the registry record. Interpretation is consequently limited to the outcomes and summaries recorded.