Scleroderma Lung Disease: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Scleroderma Lung Disease (NCT00004563), a randomized phase 3 study. The primary outcome is Forced Vital Capacity.

2. Clinical Question

3. Trial Design

The Scleroderma Lung Study was registered as a phase 3, randomized, double-masked, parallel-group trial comparing cyclophosphamide with placebo. Its listed enrollment was 158. The registry records a start date of 1999-08, primary completion in 2013-05, and status as completed. It identifies The University of Texas Health Science Center, Houston, as the sponsor.

Participants were assigned to cyclophosphamide or matching placebo. The cyclophosphamide description gives an initial daily dose of 1 mg per kilogram of body weight, to the nearest 25 mg, with monthly dose increases of one capsule up to 2 mg per kilogram. The placebo is described as matching gel caps at a dose of 25 mg. These registry descriptions identify the assigned interventions; additional details about treatment delivery and adherence are not reported in the registry record.

The primary endpoint was FVC at 12 months, expressed as a percentage of predicted and adjusted for baseline FVC. Total lung capacity and DLCO at 12 months were also posted as secondary outcomes. The group summaries list 73 participants for cyclophosphamide and 72 for placebo for the primary and secondary measures. Those outcome counts differ from the enrollment figure, but a full accounting of the analysis populations is not reported in the registry record.

The registry notes that some people who withdrew, had treatment failure, or died had completed at least the six-month visit, and that a generalized estimating-equation regression model was fitted for these participants. The description ends before fully specifying the model or the overall handling of missing measurements. The original sample-size justification, planned interim analyses, and complete statistical analysis plan are not reported in the registry record.

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:

Analysis population for the primary outcome, as defined in the registry: The Number of Participants Analyzed is not consistent with the Participant Flow for the following reason: a number of the patients who withdrew, had treatment failure, or died had completed at least the six-month visit--for these patients, a generalized estimating-equation regression model was fitted, and data missing at 12 months were imputed.

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 Forced Vital Capacity, measured over 12 months.

Forced vital capacity measures the volume of air that can be forcibly exhaled after a full inhalation. In this trial, FVC was expressed as a percentage of the predicted value, so the endpoint represents lung-function capacity relative to a predicted reference. The primary measurement was taken at 12 months.

The registry specifies that the primary endpoint was FVC at 12 months after adjustment for baseline FVC. Baseline adjustment is used to account statistically for participants’ starting measurements when comparing follow-up results. In general, this can make the treatment-group comparison more informative than comparing follow-up values without regard to where participants began. The registry record does not provide the full model specification or the adjusted treatment estimate in the material posted.

The posted group means are 66.6 for cyclophosphamide and 65.6 for placebo, each expressed as a percentage of predicted. The accompanying unadjusted mean-difference estimate is 1.0. These summaries are useful for describing the observed groups, but the comparison is not itself a substitute for the baseline-adjusted primary analysis described in the registry. FVC is used as a lung-function measure relevant to the stated study aim of evaluating symptomatic pulmonary disease in systemic sclerosis; the record does not establish that one endpoint alone captures every aspect of disease status.

7. Results Posted in the Registry

• Cylophosphamide: mean 66.60 versus 65.60 on control; difference 1.00 (95% CI -3.58 to 5.58), standardized effect d = 0.07; z-test p = 0.668.

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.

Total Lung Capacity

• 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:
  • Cylophosphamide: mean 70.50 versus 64.70 on control; difference 5.80 (95% CI 0.67 to 10.93), standardized effect d = 0.37; z-test p = 0.027.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Cylophosphamide than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

DLCO

• 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:
  • Cylophosphamide: mean 42.80 versus 44.30 on control; difference -1.50 (95% CI -6.80 to 3.80), standardized effect d = -0.09; z-test p = 0.579.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Cylophosphamide and the control. The interval gives the range of effects that remain compatible with the data.

The two posted secondary endpoints broaden the description of lung function beyond the primary FVC measure. Their estimates do not point in the same direction: the total lung capacity comparison is positive, while the DLCO comparison is negative. Their intervals and unadjusted p-values also differ in how much uncertainty they convey. Together, these results form a mixed pattern rather than a consistent shift across every reported lung-function measure.

The secondary results should be considered in relation to, not instead of, the primary endpoint. The FVC comparison is the designated primary outcome, and its posted unadjusted estimate is uncertain. The secondary outcomes may add context about other aspects of lung function, but they do not supply the missing baseline-adjusted primary estimate or resolve uncertainty about the primary analysis.

Unadjusted secondary results carry less weight because examining multiple outcomes increases the possibility of a chance finding. In general, an adjustment or a prespecified hierarchy can help control how multiple comparisons are interpreted. The registry record does not report a multiplicity procedure, a hierarchy for these endpoints, or reported analyses for them. Their results are therefore appropriately read as exploratory or descriptive evidence unless additional analysis-plan information is available.

The registry lists total lung capacity and DLCO as secondary outcomes at 12 months, with group means and standard errors. Such summaries give a snapshot at the stated time point, but do not describe trajectories over time or participant-level changes. They also do not show whether the secondary outcomes were measured or analyzed in the same population and under the same model as the primary endpoint. Those details are not reported in the registry record.

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

• Total Lung Capacity
• DLCO

9. Statistical Interpretation

10. Statistical Methods Explained

The registry describes a randomized, double-masked, parallel-group comparison. Random assignment is intended to reduce systematic differences between groups at the start of a study, while masking is intended to limit the influence of expectations on care and assessment. These are general features of trial design; the registry record does not describe the randomization procedure or the operational details of masking.

What does a mean summarize?

FVC, total lung capacity, and DLCO are continuous measurements in the posted results. A group mean gives a central summary of the observed measurements in that group. It does not show how individual participants’ measurements were distributed, whether there were unusual values, or how outcomes changed for each participant. The registry lists the number analyzed alongside each mean, but individual-level data are not reported in the registry record.

What does the standard error describe?

The posted spread measure is the standard error. In general, a standard error describes the estimated uncertainty in a group mean, rather than the spread of individual participants’ measurements. A standard error can help readers assess how precisely a mean is estimated, but it should not be read as a range containing most individual observations. The registry does not post standard deviations for these measures; the displayed comparison calculations include pooled standard deviations, but those do not replace participant-level distribution information.

How should the mean difference be read?

The mean difference compares the cyclophosphamide group with the placebo group. A positive estimate means the reported mean was higher in the cyclophosphamide group; a negative estimate means it was lower. For FVC, the unadjusted estimate is 1.0, calculated from the posted group means. The associated confidence interval expresses a range of values compatible with the estimate and its statistical uncertainty under the calculation’s assumptions. The p-value describes how compatible the observed result is with a null model of no difference, under the assumptions of the test. Neither statistic alone establishes clinical importance or a treatment effect.

The comparison calculations are unadjusted summaries based on the posted group statistics. The primary endpoint description, in contrast, specifies adjustment for baseline FVC. A baseline-adjusted model uses both follow-up and baseline information to estimate the group comparison. Because the adjusted estimate and full model output are not reported in the registry record, the unadjusted calculation should not be presented as if it were the adjusted primary result.

Why might a generalized estimating-equation model be used?

In general, generalized estimating equations can analyze repeated measurements while accounting for the fact that measurements from the same participant are related. The registry says such a model was fitted for participants who withdrew, had treatment failure, or died after completing at least the six-month visit. The available wording does not fully describe the model, the measurements included, its assumptions, or how all missing outcomes were handled. Those details are not reported in the registry record.

Missing observations matter because the participants with observed follow-up measurements may differ from those without them. Statistical methods can use available repeated measurements, but the validity of the resulting estimates depends on assumptions about the missingness process and the model. The available registry description is not sufficient to determine those assumptions or to assess whether sensitivity analyses were conducted.

What is the role of the p-value?

A p-value is a measure of how unusual data at least as extreme as those observed would be under a specified null model and its assumptions. It is not the probability that the null hypothesis is true, and it does not measure the size or importance of a difference. The unadjusted FVC p-value is 0.668. It should be interpreted alongside the estimated difference and its confidence interval, while remembering that the registry describes a baseline-adjusted primary endpoint but does not post its complete analysis output.

For secondary outcomes, multiple comparisons create a general risk that some apparently notable results may occur by chance when many endpoints are examined. The registry posts two secondary measures and does not report an adjustment for multiplicity or a complete analysis plan. Their estimates are therefore best treated as supportive descriptive information rather than independent confirmation of the primary finding.

Finally, the planning illustrations use a normal approximation, a two-sided significance level of 0.05, and equal allocation between arms. They are based on specified fractions of the observed effect. These are planning assumptions for the illustrations, not evidence about the original design decisions; the trial’s own sample-size justification is not reported in the registry record.

11. Confidence Intervals and Precision

A confidence interval communicates the uncertainty around an estimated difference. In general, intervals that include values on both sides of no difference indicate that the data are compatible with more than one direction of effect under the statistical model. The interval does not guarantee that the true value lies within it, and its interpretation depends on the method and assumptions used to calculate it.

For the unadjusted FVC comparison, the mean difference is 1.0 and the confidence interval extends from -3.576 to 5.576. The range includes zero and spans negative as well as positive differences. This means the point estimate alone should not be treated as a precise description of the underlying group contrast. The unadjusted p-value of 0.668 is consistent with substantial uncertainty about whether the observed difference reflects a nonzero group difference under the comparison’s assumptions.

The interval pertains to the unadjusted summary-statistic comparison. The registry defines the primary endpoint as baseline-adjusted FVC at 12 months, but does not report the adjusted estimate or its interval in the available record. Consequently, precision for the specified adjusted primary analysis cannot be assessed from the posted summary comparison alone.

Confidence intervals address sampling uncertainty, not every source of uncertainty. They do not resolve limitations from incomplete outcome data, model assumptions, measurement, or selection of outcomes. The incomplete registry description of the analysis population and repeated-measures model leaves those issues difficult to evaluate. The interval should therefore be read as a measure of precision for the posted unadjusted calculation, rather than as a complete measure of certainty about the trial’s primary analysis.

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 sample-size table gives planning illustrations in participants per arm for these continuous-outcome comparisons; it does not report total events. The accompanying curve is a visual representation of how the planning requirement varies with the effect size used. The illustrations assume a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each row uses the stated fraction of the observed effect as the effect to detect. The original trial’s design assumptions and sample-size justification are not reported in the registry record.

The 80% power column gives the estimated participants per arm for a design with an 80% chance of detecting the specified effect under the listed assumptions, if that effect is the true difference and the assumptions hold. The 90% column gives the corresponding planning requirement for a higher chance of detecting that effect. These are prospective design calculations, not descriptions of what happened in the completed trial.

Read across each effect-size scenario rather than treating the entries as observed outcomes. Planning for 75% of the observed effect requires more participants per arm than planning for 100% of the observed effect; planning for 50% requires more again. This is the general relationship represented by the table and curve: smaller effects are harder to distinguish from random variation, so detecting them at a chosen power generally requires more participants.

These figures are not observed power. They are illustrations built from the observed effect and the stated assumptions, and should not be used to infer the probability that the completed study detected a true effect.

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 = 1.000; pooled standard deviation σ = 14.062.
2. n = 2 × (1.96 + 0.842)² × 14.062² / 1.000² = 3,104.0, rounded up to 3,105 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 <- 1                      # observed difference in means
sigma <- 14.0619408979764                     # 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(4563)
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 = 3,105 (80% power), 4,156 (90% power)
 75% of observed effect: n per arm = 5,519 (80% power), 7,388 (90% power)
 50% of observed effect: n per arm = 12,417 (80% power), 16,622 (90% power)
power.t.test: 3105 
Simulated power at n = 3,105 per arm: 0.799 (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 a study examines more than one outcome, comparison, time point, or analysis. Even if each individual test uses a conventional threshold, conducting multiple tests can increase the chance that at least one result appears notable by chance. Statistical approaches such as adjusting the threshold, controlling a family-wise error rate, or specifying a hierarchy can address this issue. These are general options, not methods confirmed for this trial.

The registry lists a primary endpoint and two secondary outcomes. It does not report a multiplicity adjustment, a testing hierarchy, or a complete statistical analysis plan. The unadjusted secondary p-values should therefore not be treated as if they had been protected against the effects of multiple testing. The registry lists no reported analyses for the secondary outcomes, although it does post group summaries and unadjusted comparisons.

Missing outcome data are also relevant. The registry’s primary-outcome description says that the number analyzed is not consistent with participant flow because some participants who withdrew, had treatment failure, or died had completed at least the six-month visit. It further says a generalized estimating-equation regression model was fitted for these participants, but the description is incomplete. The full analysis population, model specification, assumptions about missingness, and any sensitivity analyses are not reported in the registry record.

These omissions do not demonstrate that the analyses were inappropriate; they limit what can be independently assessed from the registry record. In particular, the reader cannot determine from the available description how the model incorporated follow-up measurements, how missing data after the six-month visit were handled, or whether the primary result was robust to alternative assumptions. The original sample-size justification and interim-analysis plan are also not reported in the registry record.

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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: death (3.8%), pneumonia and laryngospasm (3.8%), pneumonia (2.5%), leukopenia with pneumonia (1.3%), leukopenia with gastroenteritis (1.3%).

The registry lists 79 participants at risk for serious adverse events in each group. Serious adverse events affected 5 participants in the cyclophosphamide group and 5 in the placebo group. For other adverse events, 79 were listed at risk in each group; 20 participants were affected in the cyclophosphamide group and 4 in the placebo group. These are counts of affected participants, not estimates adjusted for exposure time or other factors.

The registry also lists serious-event terms and percentages, including Death at 3.8%, Pneumonia and Laryngospasm at 3.8%, Pneumonia at 2.5%, Leukopenia with pneumonia at 1.3%, and Leukopenia with gastroenteritis at 1.3%. These categories should not be assumed to be mutually exclusive. The death-count field is not reported in the registry record; the listed term and percentage do not provide a complete account of deaths by treatment group.

Adverse-event counts describe what was posted, but do not by themselves establish causation or comparative risk. The registry record does not provide further detail about event timing, duration, severity breakdown, or exposure-adjusted rates.

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

The posted results describe outcomes measured in a randomized comparison of cyclophosphamide and placebo among people with systemic sclerosis and symptomatic pulmonary disease. For the primary lung-function measure, the group means at 12 months were close in the unadjusted summaries, and the estimated difference had a confidence interval spanning negative and positive values. This indicates uncertainty in the posted unadjusted comparison; it does not establish that the groups were identical.

The registry’s stated primary endpoint is baseline-adjusted FVC, but its available numerical comparison is unadjusted. Without the adjusted estimate and complete model output, the posted summary cannot fully answer the primary statistical question as defined in the registry. The secondary measures add information about other lung-function outcomes, but they show a mixed pattern and are not reported with a multiplicity adjustment.

Statistical uncertainty is not the same as clinical importance. A mean difference and its interval describe the scale and precision of a group comparison, but the registry record does not state a threshold for a clinically important difference. It is therefore not possible to determine from these summaries alone how the estimated differences relate to a clinical decision threshold.

The results also do not establish what would happen to every individual participant. Group means can conceal variation in responses, and the posted summaries do not provide individual trajectories. Nor do the adverse-event counts by themselves show that an event was caused by treatment. Those counts should be interpreted as reported events in the study groups.

In short, the registry supports a careful description of the posted measurements, estimates, and safety counts. It does not provide enough detail to reconstruct the full primary analysis, judge all missing-data assumptions, or resolve every aspect of the benefit-risk question. The findings should be understood within those limits rather than converted into a categorical claim about treatment.

17. Why This Trial Matters Statistically

This trial illustrates why an endpoint’s definition and its reported numerical comparison need to be read together. The registry defines the primary endpoint as FVC at 12 months after adjustment for baseline FVC, while the posted group means and mean difference describe an unadjusted comparison. Distinguishing these estimands prevents a descriptive calculation from being mistaken for the full primary analysis.

It also demonstrates the importance of precision. The FVC point estimate is only one part of the result; its confidence interval shows that the unadjusted data remain compatible with differences in either direction. Reporting the interval alongside the estimate makes uncertainty visible and discourages overinterpretation of a single number or p-value.

The secondary outcomes illustrate a further statistical challenge: multiple outcomes can produce a varied pattern of estimates. Without a reported multiplicity plan, secondary p-values warrant particular caution. Their role is to add context, not automatically to confirm or overturn the primary result.

Finally, the registry’s incomplete description of the analysis population and repeated-measures model shows why transparent reporting matters. Knowing who contributed to each analysis, how incomplete follow-up was handled, and what assumptions were made is essential for judging how well the statistical results answer the clinical question. These reporting details matter alongside randomization, masking, and the posted outcome values.

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 reported for this trial. First, the primary endpoint is described as baseline-adjusted, but the available numerical comparison is unadjusted and the adjusted estimate is not reported in the registry record. The reported comparison therefore cannot stand in for the complete primary analysis.

Second, the registry notes a mismatch between the number analyzed and participant flow. It says that some participants who withdrew, had treatment failure, or died had completed at least the six-month visit and that a generalized estimating-equation model was fitted for them. The available wording is incomplete, and the record does not explain the full analysis population, the model specification, or sensitivity analyses for missing data.

Third, the secondary comparisons are unadjusted, and a multiplicity procedure is not reported in the registry record. Their estimates can describe the posted results, but the chance of findings arising among multiple comparisons is not addressed in the available record.

Fourth, the registry posts group means and standard errors rather than participant-level results. These summaries do not reveal the distribution of measurements, individual changes, or how outcomes varied over time. The original sample-size justification, interim-analysis plan, and complete statistical analysis plan are also not reported in the registry record. These gaps limit independent assessment of the analysis and its assumptions.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Scleroderma Lung Disease — NCT00004563.
• Registry record last updated 2015-03-27; status last verified 2015-03.
• Results first posted 2015-03-06.
• 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 Scleroderma Lung Study is registered as a completed phase 3, randomized, double-masked, parallel-group trial of cyclophosphamide versus placebo. Enrollment was 158. The registry lists FVC at 12 months, adjusted for baseline FVC, as the primary endpoint, with total lung capacity and DLCO as secondary outcomes.

The posted FVC summaries give means of 66.6 for cyclophosphamide and 65.6 for placebo, with an unadjusted mean difference of 1.0 and a confidence interval from -3.576 to 5.576. The registry also posts secondary lung-function summaries and adverse-event counts. The adjusted primary estimate, complete analysis-population details, multiplicity procedures, and full missing-data methods are not reported in the registry record.