Fluoxetine in Pulmonary Arterial Hypertension (PAH) Trial: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Fluoxetine in Pulmonary Arterial Hypertension (PAH) Trial (NCT03638908), a na phase 2 study. The primary outcome is Pulmonary Vascular Resistance (PVR).

2. Clinical Question

The protocol summary describes a hypothesis that PVR would be lower after 24 weeks of fluoxetine. That statement is a planned hypothesis, not a reported finding. Without the numerical endpoint result, the record does not establish the direction or size of the observed change, or how precisely it was estimated. The absence of a concurrent control group also means that any before-and-after change would not, by itself, distinguish a treatment-associated change from other explanations such as time-related changes or measurement variability.

3. Trial Design

This was a phase 2, open-label, single-group clinical trial of fluoxetine in PAH. The registry lists no allocation and no masking. There is no separate control arm, so the study does not make a randomized, like-for-like comparison between fluoxetine and another treatment or control condition. Its primary comparison is within participants, using PVR measured at baseline and Week 24.

The trial record lists 8 participants enrolled and a completed status. The brief summary describes a planned study of 18 patients, while the enrollment record states 8. The record does not explain the difference between those figures. For the primary endpoint, it reports that 6 of the 8 enrolled participants had complete baseline and Week 24 data. Reasons the other participants did not contribute complete endpoint data are not reported in the registry record.

The listed intervention is fluoxetine, with a scheduled increase from 20 mg daily in Weeks 1-4 to 80 mg daily in Weeks 13-24. The primary endpoint timeframe is baseline and Week 24. The summary also describes assessments of other hemodynamic measures, quality of life, the QIDS-SR depression scale, functional class, and six-minute walk distance. A secondary outcome listed in the record is 5-HIAA level. Although the registry lists 10 outcome measures, it lists no analyses. Details of the protocol’s statistical analysis plan, including any adjustments or handling of missing measurements, 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.

Before assignment: This was an open label study and all subjects were assigned to receive Fluoxetine.

Overall Study

Analysis population for the primary outcome, as defined in the registry: 6 out of the 8 subjects enrolled had complete- baseline and week 24- data for analysis

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 Pulmonary Vascular Resistance (PVR), measured over Baseline and Week 24.

The primary endpoint is change in pulmonary vascular resistance (PVR), measured in woods unit, between baseline and Week 24. The record defines the change as Week 24 PVR minus baseline PVR, then describes the endpoint as the mean of those participant-level changes. PVR is calculated as pulmonary artery mean pressure minus wedge pressure, divided by Fick cardiac output. The record specifies that Fick cardiac output is used rather than thermodilution, stating that Fick appears to have greater precision but not accuracy.

PVR is a hemodynamic measure that combines pressure and cardiac output information. In general, measuring a continuous endpoint at baseline and follow-up allows the analysis to describe the average within-participant change over time. A mean can convey the overall direction and scale of changes, while the standard deviation describes how spread out participant-level changes are around that mean. Neither statistic alone establishes why a change occurred.

The registry identifies the endpoint and its measurement timeframe, but does not report the mean change, standard deviation, confidence interval, or a test result. It therefore does not provide a numerical estimate of the primary endpoint’s observed change or its precision.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary endpoint is continuous: each participant’s PVR is measured at baseline and Week 24, and the difference is defined as Week 24 minus baseline. The record says that the mean of these changes is the primary summary. The reported endpoint population comprises 6 participants with complete measurements. No numerical mean or standard deviation is reported, and the record lists no analyses. The explanations below describe statistical concepts in general and do not imply that the trial used a particular analysis beyond what its endpoint description states.

What does a mean change describe?

In general, a mean change is calculated by averaging participant-level differences between follow-up and baseline. With the direction defined as follow-up minus baseline, a negative mean would indicate lower follow-up measurements on average, and a positive mean would indicate higher follow-up measurements on average. The registry defines the direction of subtraction but does not report the mean, so the direction and size of the observed average change cannot be determined from the record.

Using participant-level changes focuses the summary on within-person differences rather than treating the two time points as unrelated groups. In general, this can account for the fact that repeated measurements from the same person are linked. However, the mean alone does not show whether individual changes were similar or whether a small number of larger changes influenced the average. The standard deviation can describe the spread of participant-level changes, but its value is not reported.

What would uncertainty measures add?

In general, a confidence interval gives a range of values that are compatible with the data under a specified statistical model and its assumptions. Its width conveys precision: a wider interval indicates less precise estimation than a narrower interval. A confidence interval is not a substitute for the raw data or a control group, and it does not resolve limitations in design or missingness. No confidence interval is reported for this endpoint.

In general, a hypothesis test compares data with a stated null hypothesis using a specified test procedure. A p-value describes how unusual the observed data, or data more extreme, would be under that null model. It is not the probability that the null hypothesis is true, nor does it measure the size or clinical importance of an effect. The registry lists no analyses and reports no p-value, test statistic, or inferential conclusion for the primary endpoint.

Why does the single-group design matter?

In general, a within-participant baseline-to-follow-up analysis can estimate change over time among those measured at both occasions. Without a concurrent control group, however, it cannot by itself separate a change associated with treatment from changes that might have occurred for other reasons. Random allocation to treatment groups, when used, can help balance measured and unmeasured factors on average; this trial is listed as single-group, with allocation and masking both listed as none. The registry record does not report another design feature that would supply a randomized control comparison.

Open-label means the trial is not masked. In general, lack of masking can affect some aspects of measurement or reporting, depending on the outcome and its assessment. PVR is described as derived from hemodynamic measurements, but the record does not report procedures for blinded measurement or other safeguards. Such procedures should not be assumed.

What can be inferred from the analysis population?

The registry reports complete baseline and Week 24 data for 6 of 8 enrolled participants. In general, an analysis restricted to participants with complete measurements estimates change among those participants. If missingness is related to participants’ characteristics or outcomes, that estimate may not represent the full enrolled group. The record does not report missing-data methods, reasons for incomplete data, or an analysis that includes participants without both measurements.

Sample-size calculations based on an observed effect, when available, are planning illustrations rather than observed trial findings or observed power. The registry record gives assumptions for such an illustration but reports no numerical sample-size estimates in the information available for interpretation. It also does not report the trial’s own sample-size justification. Those planning assumptions cannot replace the missing primary outcome estimate or an analysis of its uncertainty.

Finally, multiple outcomes can raise questions about how findings are interpreted when many hypotheses are examined. In general, multiplicity procedures may be used to control error rates across a defined family of tests, but their choice depends on the analysis plan. The record lists 10 outcome measures and no analyses; it does not report a multiplicity plan or indicate which outcomes, beyond the named primary endpoint, were tested. No claim about adjustment should therefore be inferred.

10. Confidence Intervals for Each Arm

No two-arm comparison is shown for this trial, but each arm's mean can be given with a t-based 95% confidence interval computed from the posted mean, spread and number of participants.

R code and output

arm <- c("Fluoxetine")
n   <- c(6)
m   <- c(6.7)
sp  <- c(2.7)
sd  <- sp      # posted standard deviations
for (i in seq_along(n)) {
  se <- sd[i] / sqrt(n[i]); h <- qt(0.975, n[i] - 1) * se
  cat(sprintf("CARM|%s|%d|%.3f|%.3f|%.3f|%.3f\n", arm[i], n[i], m[i], sd[i], m[i] - h, m[i] + h))
}

CARM  Fluoxetine  6  6.700  2.700  3.867  9.533

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 10 outcome measures and no analyses. It identifies change in PVR as the primary endpoint. The brief summary also names other hemodynamic endpoints, quality of life, QIDS-SR depression scale, functional class, and six-minute walk distance; 5-HIAA level appears as a listed secondary outcome. Numerical results for these outcomes are not reported in the registry record. A multiplicity strategy, such as a hierarchy of tests or adjustment across outcomes, is also not reported.

The primary endpoint description states that 6 of the 8 enrolled participants had complete baseline and Week 24 data. The record does not give reasons for incomplete measurements, describe their timing, or report how missing values were handled. It does not state whether the primary analysis included only complete cases or whether any other method was used. The stated endpoint population indicates the number with both measurements, but does not fully document the analysis procedure.

In general, missing outcome data can affect both the precision and interpretation of an estimate. A complete-case summary describes participants with the required measurements; if their outcomes differ systematically from those without complete measurements, the summary may not characterize all enrolled participants. The record does not provide enough information to assess that possibility.

The difference between 8 enrolled participants and the 18 patients described in the brief summary is not explained in the registry record. The protocol’s sample-size justification, analysis population definitions, missing-data plan, interim analyses, and detailed statistical analysis plan are not reported. No conclusion about whether those procedures were prespecified can be drawn from the posted record.

12. 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 non-serious adverse event terms

The registry reports 0 participants with serious adverse events among 8 at risk in the fluoxetine group. It reports 8 participants with other adverse events among 8 at risk, and 0 deaths. No serious adverse event terms are listed. These are participant counts in the registry’s safety categories; they are not measures of event frequency per unit of time.

The count of participants with other adverse events does not give the number of distinct events, their type, severity, timing, or duration. The record does not report whether events were considered related to fluoxetine, whether participants stopped treatment because of events, or whether additional events occurred outside the stated reporting period. Those details are not reported in the registry record.

In general, counts from a small single-group trial describe what was recorded among its participants, but cannot alone establish comparative safety. There is no control group in this trial against which to compare adverse-event counts. The zero count for serious adverse events and the zero count for deaths should be read as reported observations, not proof that such events cannot occur or that there is no risk. The registry’s safety figures do not establish a causal relationship between treatment and any event.

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

The record documents a completed, open-label, single-group phase 2 trial of fluoxetine in PAH and identifies mean change in PVR from baseline to Week 24 as its primary endpoint. It reports 8 enrolled participants and says 6 had complete data for that endpoint. It also records adverse-event counts. These details describe the trial’s design, endpoint, and available participant counts.

The posted record does not include the numerical mean PVR change, the standard deviation, a confidence interval, or a p-value. It therefore does not quantify the observed primary endpoint result or its precision. The planned hypothesis in the brief summary is not a result, and the trial’s completed status does not indicate whether that hypothesis was supported.

Even if a numerical before-and-after change were available, the single-group design would limit what could be attributed to treatment. A change over time is not the same as a randomized treatment effect because no concurrent control group is listed. The record also does not explain why 2 enrolled participants lack complete baseline and Week 24 data or describe methods for accounting for missing measurements.

The adverse-event counts are likewise descriptive. They report no participants with serious adverse events and no deaths, and 8 participants with other adverse events. They do not provide comparative estimates or sufficient detail to determine causality, event type, or timing.

Accordingly, the record supports a description of the planned study and the reported enrollment and endpoint completeness, but it does not support a numerical conclusion about the primary outcome or a causal conclusion about treatment. Other outcome results and detailed statistical procedures are not reported in the registry record.

14. Why This Trial Matters Statistically

This trial is statistically informative as an example of the distinction between a protocol hypothesis and a posted result. The registry specifies a continuous primary endpoint, the time points used, the direction of subtraction, the measurement formula, and the number of participants with complete endpoint data. Yet it lists no analyses and does not report the primary endpoint estimate. Design details alone cannot replace the result needed to assess magnitude or precision.

The record also illustrates why enrollment and analysis counts should be distinguished. It lists 8 enrolled participants and complete baseline and Week 24 data for 6 participants. The brief summary describes a plan involving 18 patients, but the record does not explain the difference. Reporting the analysis population alongside enrollment is useful, while reasons for incomplete data and their handling are also important for interpretation.

Its single-group structure clarifies the limits of a pre-post comparison. Within-participant measurements can describe change over time, but without a concurrent randomized comparator they do not isolate a treatment effect from other possible influences. This is a design limitation, not something a statistical test of the observed change could remove.

Finally, the record lists multiple outcome measures, but no analysis results or multiplicity plan. This underscores the value of accessible outcome estimates, uncertainty measures, analysis populations, and prespecified statistical methods. Without them, readers can understand what was intended and what was recorded about participation, but cannot assess the reported evidence for the primary question.

15. 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. First, it is a single-group, open-label study with no listed allocation or masking. The primary comparison is baseline versus Week 24 within participants, not a randomized comparison with a control group. Any observed change would therefore be difficult to distinguish from other changes over time or measurement-related variation.

Second, the registry reports 8 enrolled participants, while its brief summary describes a plan involving 18 patients. The reason for this difference is not reported in the registry record. The primary endpoint population is 6 participants with complete baseline and Week 24 data; the reasons for missing data and the methods used to handle them are not reported.

Third, the numerical primary endpoint result is absent. The record gives no mean change, standard deviation, confidence interval, or test result, and lists no analyses. Thus, neither the size nor precision of the observed change can be evaluated from the posted record. The record also does not report a detailed statistical analysis plan, sample-size justification, interim-analysis procedures, or multiplicity handling.

Finally, the safety information is limited to participant counts by broad category. Serious adverse event terms are not listed, and event details, timing, relatedness, and comparative safety data are not reported. These limits are especially relevant when interpreting counts from a single group.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Fluoxetine in Pulmonary Arterial Hypertension (PAH) Trial — NCT03638908.
• Registry record last updated 2020-06-26; status last verified 2020-06.
• Results first posted 2019-10-07.
• 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.

19. Record Summary

The registry record describes a completed phase 2, open-label, single-group trial of fluoxetine in PAH. It lists 8 enrolled participants and a 24-week dosing schedule that increases from 20 mg daily in Weeks 1-4 to 80 mg daily in Weeks 13-24. The primary endpoint is mean change in PVR from baseline to Week 24, measured in woods unit; 6 participants are reported to have complete data for this endpoint.

The brief summary describes a planned study of 18 patients, a figure not reconciled with the enrollment count. The registry lists 10 outcome measures but no analyses. It does not report a numerical primary endpoint result, uncertainty interval, or test result. Safety counts are 0 participants with serious adverse events, 8 with other adverse events, and 0 deaths among 8 at risk. Further statistical methods, reasons for incomplete endpoint data, and numerical results are not reported in the registry record.