Extended Steroid in Use in Community Acquired Pneumonia (CAP)(e): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Extended Steroid in Use in Community Acquired Pneumonia (CAP)(e) (NCT01283009), a randomized phase 3 study. The primary outcome is 60-day Mortality.

2. Clinical Question

The study rationale in the record is that severe illness can involve an inflammatory response that contributes to organ damage. Glucocorticoids may reduce inflammation, so the trial was designed to evaluate whether adding methylprednisolone to usual pneumonia treatment would affect recovery and survival. The record refers to an earlier small preliminary trial as motivation, but no results from that earlier work are needed to interpret this trial's posted information. The clinical question is specific to the enrolled population and the treatment comparison; the record does not establish how findings might apply outside that setting.

3. Trial Design

NCT01283009, titled “Extended Steroid in Use in Community Acquired Pneumonia (CAP)(e),” is recorded as a completed phase 3 trial sponsored by the VA Office of Research and Development. The official title describes an evaluation of methylprednisolone in hospitalized veterans with severe community-acquired pneumonia. Enrollment was 584. The recorded study start is 2012-01-09, and primary completion is recorded as 2016-07-31.

The trial is described as randomized, double-masked, and parallel. Participants were assigned to one of two drug arms: an inactive substance or methylprednisolone. The summary says assignment would occur at each site, but the registry record does not report the detailed randomization procedure, allocation concealment method, or masking procedures. It also does not report treatment dose, timing beyond the summary's description of early treatment, duration, or adherence.

The trial's single listed primary outcome measure is 60-day mortality. It is defined as all-cause mortality by the end of study day 60 and is expressed as a count of participants. The record notes that the numbers are overall and by stratum, but the posted primary-outcome results are not available in the information recorded for this analysis. A reported primary-outcome comparison, its uncertainty, and its analysis population are therefore not described.

Study status is completed, but completion status alone does not supply outcome results or explain how the analysis was conducted. The record does not report the protocol's sample-size justification, prespecified statistical model, handling of protocol deviations, or details of stratification. These omissions matter when assessing how closely a reported analysis would reflect the planned comparison. The record also contains serious adverse event summaries for the two arms, which are distinct from a fully reported primary endpoint analysis.

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: Numbers provided are overall and by stratum.

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 60-day Mortality, measured over 60-day.

The primary endpoint is all-cause mortality at 60 days. It records whether each participant had died by the end of study day 60, and the outcome is expressed as a count of participants. A fixed-time mortality endpoint gives a clinically direct description of survival status at a specified point after study entry. Because it is all-cause mortality, it does not require assigning a cause of death to a particular illness or mechanism.

The endpoint was selected to address the study's stated question about survival in critically ill patients with severe community-acquired pneumonia. The rationale in the study summary is that uncontrolled inflammation may contribute to organ injury and death, and that glucocorticoid treatment might help regulate inflammation. The endpoint therefore relates to an outcome of central importance to that rationale. It does not, by itself, capture other dimensions of recovery, such as length of hospital stay or residual organ damage.

The registry record defines the endpoint, but the primary-outcome results needed to compare the arms are not reported in the record. There is no posted arm-specific primary-outcome count, treatment-effect estimate, confidence interval, or primary-outcome p-value in the available information. The adverse-event section lists death counts, but it does not identify those counts as the 60-day endpoint or give the endpoint's analysis details. Those figures cannot substitute for a clearly defined primary-outcome comparison.

Without a comparative estimate and its uncertainty, the endpoint definition explains what the investigators intended to measure, but not what the trial found about the treatment contrast.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Randomization is intended to make the treatment groups comparable, on average, with respect to both measured and unmeasured baseline characteristics. In a randomized comparison, differences in outcomes can therefore be interpreted as evidence about the assigned treatment contrast, subject to the conduct of the trial, the analysis population, missing data, and chance. Randomization does not guarantee identical groups, and it does not remove the need to report how the comparison was estimated.

What does the parallel, double-masked design mean?

A parallel design assigns each participant to one treatment arm and compares outcomes between arms. It differs from a design in which participants switch between treatments. Double masking indicates that the record describes the trial as masked for two parties, but it does not identify those parties or explain the masking procedures. The registry record does not report how the inactive substance was made to resemble the active treatment, how masking was maintained, or whether masking was assessed.

The record identifies random allocation but does not describe sequence generation, concealment, stratification factors, or any restrictions used in the randomization. These details can help readers judge the risk of selection bias and understand whether adjustments were part of the planned analysis. They are not reported in the registry record.

How should the endpoint be analyzed?

The stated primary outcome is binary for each participant: alive or dead by the end of study day 60. A trial report could present the count and proportion in each arm and an effect measure comparing the arms. Depending on the analysis plan, possible effect measures include a difference in risks, a risk ratio, or an odds ratio. These are general examples, not analyses reported for this trial. No primary-outcome estimate or confidence interval is reported in the registry information.

For a fixed-time binary endpoint, the analysis should make clear which participants are included, how outcome status is determined, and how participants with unknown status are handled. A participant's status at the endpoint cannot simply be assumed when it is missing. The registry record does not report the number with known status in either arm, the missing-data approach, or whether the primary analysis followed assigned treatment regardless of treatment received.

The endpoint is described as overall and by stratum. A stratified analysis can account for prespecified grouping factors in the design, but the registry record does not identify the strata, the statistical method used to combine them, or whether the primary result was adjusted. Without these details and the outcome results, the analysis cannot be independently reconstructed.

What does statistical uncertainty tell the reader?

A confidence interval is a range of values produced by a statistical procedure that reflects uncertainty in an estimate under its assumptions. A narrow interval indicates greater precision than a wide interval; an interval that includes the null value can be consistent with a range of effects that includes no difference. These are general interpretations, and no confidence interval for the primary endpoint is reported in the registry record. A p-value, if reported, would summarize how unusual data at least as incompatible with a null hypothesis would be under that hypothesis and the model. It would not measure the probability that the null hypothesis is true, nor would it indicate clinical importance. No primary-endpoint p-value is reported.

For count outcomes, the analysis model should reflect the outcome's structure. For example, count data may show over-dispersion when variability is greater than a simple model assumes; an analysis that ignores substantial over-dispersion can misstate uncertainty. This is a general statistical consideration, not a description of the method used in this trial. The primary outcome is recorded as a participant count with a binary event definition, and the record does not describe an over-dispersed count model or any other primary analysis model.

How should planning illustrations be read?

A planning sample-size calculation asks how much information might be needed to detect an effect under chosen assumptions. It is not a result observed in the trial and is not observed power. The record describes a planning illustration based on the observed effect, using a two-sided significance level of 0.05, equal allocation, and a normal approximation, with planning scenarios defined as fractions of the observed effect. Such an illustration depends on its assumptions and should not be mistaken for the trial's original sample-size justification. The trial's own sample-size justification is not reported in the registry record. The outcome comparison and corresponding values needed to assess the planning illustration are not reported in the available information.

More broadly, a statistical plan should specify the primary estimand, analysis population, model, covariate adjustment, handling of missing data, and any sensitivity analyses. The registry record does not report these details. Their absence limits assessment of whether a result, if available elsewhere, would follow the planned approach. It does not establish that the investigators did or did not use any particular method.

10. Multiplicity, Missing Data and Analysis Populations

The record lists one primary outcome measure, 60-day mortality, and one analysis. It does not report a primary-outcome result or a complete statistical analysis plan. The count of listed outcomes does not establish whether additional outcomes or subgroup analyses were prespecified elsewhere, nor does it describe any procedure for controlling the chance of false-positive findings across multiple comparisons. Multiplicity methods, including any hierarchy or adjustment, are not reported in the registry record.

The endpoint is defined as whether a participant had died by the end of study day 60. The record does not report how many participants had an ascertainable status, whether any outcome status was missing, or how such missingness was addressed. This matters because assumptions about participants with unknown status can influence both the estimate and its uncertainty. The registry record also does not describe sensitivity analyses for alternative assumptions about missing outcomes.

The analysis population is not reported in the available primary-outcome information. In randomized trials, an analysis based on assigned treatment generally retains participants in their randomized groups, while other approaches may use different inclusion rules. These are general distinctions; the record does not say which approach was used. It also does not report how crossovers, treatment discontinuation, or protocol deviations were handled.

Although the record notes that primary-outcome numbers are overall and by stratum, it does not identify the strata or explain the analysis. No details are reported about subgroup hypotheses, interaction testing, or whether any subgroup comparisons were exploratory. These omissions limit assessment of multiplicity and of the population represented by any estimate. They should not be taken as evidence that a particular analysis was or was not performed.

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.

Most frequent serious adverse event terms

Serious events with the highest arm-level rate: pneumonia (13.2%), cardiac failure congestive (8.4%), chronic obstructive pulmonary disease (5.6%), acute kidney injury (3.7%), multiple organ dysfunction syndrome (3.0%).

The serious adverse event summary reports participants at risk and participants affected in each arm. In the inactive-substance arm, 287 participants were at risk and 162 were affected; in the methylprednisolone arm, 297 were at risk and 167 were affected. These are counts as recorded, not a treatment-effect estimate. No statistical comparison, confidence interval, or p-value for serious adverse events is reported in the registry information.

The listed serious event terms include pneumonia at 13.2%, congestive cardiac failure at 8.4%, chronic obstructive pulmonary disease at 5.6%, acute kidney injury at 3.7%, multiple organ dysfunction syndrome at 3.0%, and respiratory failure at 3.0%. The record does not specify in this summary the denominators or time window for those percentages, or whether a participant could contribute to more than one term. The percentages should therefore be read as the reported term values, not as arm-specific comparative risks.

The adverse-event information also lists deaths of 84 in the inactive-substance arm and 79 in the methylprednisolone arm. It does not establish that these are the primary endpoint counts at study day 60. The primary endpoint is separately defined as all-cause mortality by that day, and its arm-specific results are not reported in the available primary-outcome information. Serious adverse event summaries and a defined mortality endpoint answer different reporting questions; one should not be used to reconstruct the other.

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

The record establishes that a randomized, double-masked, parallel phase 3 trial was conducted to compare methylprednisolone with an inactive substance in hospitalized veterans with severe community-acquired pneumonia. It identifies 584 participants enrolled and defines all-cause mortality by the end of study day 60 as the primary endpoint. It also reports serious adverse event counts by arm and lists several event terms.

Those details do not establish the trial's primary treatment effect. The registry record does not report the arm-specific 60-day mortality results, an estimated contrast, a confidence interval, or a primary-outcome p-value in the information available for this analysis. The death counts in the adverse-event summary cannot be substituted for the endpoint because the record does not confirm that they use the same time point, definition, or analysis population. No conclusion about whether mortality differed between the arms, or about the precision of such a difference, follows from the posted primary-outcome information.

Likewise, serious adverse event counts alone do not establish that one treatment is safer or less safe. A comparative interpretation would require consistent definitions, denominators, follow-up windows, and an appropriate analysis. The listed term percentages lack contextual details needed to interpret arm-specific risks or comparisons.

The study rationale describes a hypothesis about inflammation and glucocorticoid treatment. A biological rationale is not itself evidence that the treatment altered mortality or recovery in this trial. Results from an earlier preliminary study, as mentioned in the registry summary, also cannot fill the gap in the primary results for this trial.

The practical interpretation is bounded by what is reported: the design and endpoint are documented, but the primary clinical outcome comparison is not. The record's completed status should not be confused with a complete public account of the trial's statistical findings. A result would need an effect estimate, its uncertainty, and analysis details before readers could assess its magnitude, precision, and robustness.

13. Why This Trial Matters Statistically

This trial matters statistically because it poses a direct randomized comparison around a clearly defined, patient-important endpoint: all-cause mortality by study day 60. The design is identified as randomized, double-masked, and parallel, and the record describes a phase 3 study with enrollment of 584. Those features make transparent reporting of the primary estimate and its uncertainty especially important: readers need both the study design and the outcome evidence to evaluate the treatment contrast.

A mortality endpoint is relatively straightforward to define as a participant-level event at a fixed time. Yet even a clear endpoint requires adequate reporting of who was analyzed, how outcome status was established, and how missing status was handled. For a randomized comparison, the estimated effect and confidence interval communicate the direction, scale, and precision of the observed contrast. The registry record does not report those primary-outcome statistics in the information available for this analysis.

The record also illustrates why results from different reporting categories should not be conflated. Death counts occur in the adverse-event summary, while 60-day mortality is the named primary endpoint. Without confirmation that the counts share the endpoint definition and analysis population, they cannot recreate the primary result. Serious adverse event counts and individual event terms likewise require denominators and context for valid comparison.

Finally, a planning illustration based on an observed effect is conceptually separate from the original trial design and from observed power. Its assumptions can help illustrate the sample size associated with a hypothetical target effect, but it cannot replace the trial's sample-size justification or primary results. That justification is not reported in the registry record. The main statistical lesson is that design labels and event summaries are not enough: interpretation depends on a complete, prespecified analysis and an estimate accompanied by uncertainty.

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 largest limitation for interpreting the primary question is that the registry record does not report the primary outcome results in the available information. It defines 60-day mortality but supplies no arm-specific endpoint counts, comparative estimate, confidence interval, or primary-outcome test result. The adverse-event death counts are not explicitly linked to the primary endpoint definition, so they cannot resolve this omission.

Important design and analysis details are also not reported in the registry record. These include the full eligibility criteria, treatment dose and schedule, detailed randomization and masking procedures, the sample-size justification, analysis model, analysis population, missing-data approach, and handling of protocol deviations. Although the record says outcome numbers are overall and by stratum, it does not describe the strata or the method used to analyze them.

The safety summary has additional interpretive limits. The record gives at-risk and affected counts for serious adverse events, but does not provide a statistical comparison or confidence interval. It does not state the denominators and time windows for the listed serious event percentages, or whether participants could be represented under multiple terms. The death counts in that summary are not identified as day-60 endpoint results.

The registry record mentions a planning illustration based on the observed effect, but the trial's own sample-size rationale is not reported. A planning exercise cannot establish whether the original study was designed for a particular effect size, and it is not observed power. The record also does not report details of multiplicity control, subgroup analyses, or sensitivity analyses.

These reporting gaps limit independent assessment of effect size, precision, robustness, and safety comparisons. They do not demonstrate that a particular design or analysis decision was inappropriate; rather, the relevant details are not reported in the registry record.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Extended Steroid in Use in Community Acquired Pneumonia (CAP)(e) — NCT01283009.
• Registry record last updated 2020-10-08; status last verified 2020-09.
• Results first posted 2020-10-08.
• 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

NCT01283009 is a completed, randomized, double-masked, parallel phase 3 trial sponsored by the VA Office of Research and Development. The official title describes methylprednisolone in hospitalized veterans with severe community-acquired pneumonia. Enrollment was 584, with an inactive-substance arm and a methylprednisolone arm. The recorded study start is 2012-01-09, and primary completion is recorded as 2016-07-31.

The primary endpoint is all-cause mortality by the end of study day 60. The registry record defines the endpoint but does not report its arm-specific results or comparative statistical analysis in the available information. Serious adverse event counts and selected event terms are recorded, but they do not substitute for the primary endpoint comparison. The detailed statistical plan, analysis population, missing-data methods, and trial-specific sample-size justification are not reported in the registry record.