PA-824-CL-007: Phase IIa Evaluation of Early Bactericidal Activity in Pulmonary Tuberculosis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for PA-824-CL-007: Phase IIa Evaluation of Early Bactericidal Activity in Pulmonary Tuberculosis (NCT00567840), a randomized phase 2 study. The primary outcome is Early Bactericidal Activity (EBA) Measured as the Daily Rate of Change in log10 CFUs (Colony Forming Units) of M. Tuberculosis in Sputum on Solid Media (Days 0-14)..

2. Clinical Question

3. Trial Design

This was a randomized, quadruple-masked, parallel-group Phase IIa trial. The registry lists enrollment of 69 participants and a completed status. The record gives a start date of 2007-08 and a primary completion date of 2007-12. It identifies Global Alliance for TB Drug Development as the sponsor.

Five arms are described: PA-824 at 200 mg, 600 mg, 1000 mg, or 1200 mg once daily, and Rifafour. The trial description specifies 14 consecutive days of oral PA-824 treatment and a standard tuberculosis treatment control group. The registry does not explain the allocation procedure, the masking arrangements, or whether masking applied identically to every intervention.

The primary outcome was a continuous measure of early bactericidal activity, expressed as the daily rate of change in log10 CFUs of M. tuberculosis in sputum on solid media between Day 0 and Day 14. The registry also lists outcomes covering shorter solid-media intervals and changes in time to positive signal in liquid culture. These additional measures provide related but distinct ways of describing early changes in sputum culture results.

Although the trial enrolled 69 participants, the number contributing to the primary endpoint varied by arm. The registry explains that some EBA values could not be calculated because results were missing; the endpoint counts represent participants with available results for the relevant time period. The original protocol, detailed allocation ratios, prespecified analysis populations, and trial-specific statistical analysis plan are not reported in the registry record. The posted summaries therefore do not establish how the trial was planned to handle missing outcomes, multiple comparisons, or departures from the intended treatment regimen.

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

Analysis population for the primary outcome, as defined in the registry: Some of the EBA values could not be calculated due to missing results. The number of participants given are the number of participants for whom the results were available for this time period and therefore for whom the relevant EBA could be calculated.

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 Early Bactericidal Activity (EBA) Measured as the Daily Rate of Change in log10 CFUs (Colony Forming Units) of M. Tuberculosis in Sputum on Solid Media (Days 0-14)., measured over Day 0 and Day 14.

The primary endpoint measures the daily rate of change in sputum M. tuberculosis colony-forming units on solid media from Day 0 to Day 14. It is expressed in log10 CFU/ml. A colony-forming unit is a culture-based measure of organisms capable of forming colonies under the assay conditions. The endpoint condenses measurements across the treatment period into a rate, making it possible to describe early change in bacterial counts rather than only an endpoint value.

Early bactericidal activity is used in early-stage tuberculosis research to characterize short-term changes in sputum culture measurements during treatment. It is a pharmacodynamic outcome: it describes a biological response over a limited observation period. It is not, by itself, a measure of long-term clinical outcomes or a substitute for outcomes assessed over longer follow-up. The registry record does not explain the expected direction of the endpoint or give a protocol-defined threshold for interpreting a difference.

The registry reports an arm mean and standard deviation for each group, with the number of participants whose results could be used. Missing results prevented calculation of some EBA values. Consequently, the endpoint summaries apply to participants with available results for this period, and the record does not establish whether those participants are representative of everyone randomized. The reported numbers describe the observed summaries; they do not alone establish a causal or clinically meaningful difference between dose groups.

7. Results Posted in the Registry

• PA-824 600 mg: mean 0.11 versus 0.11 on control; difference 0.00 (95% CI -0.04 to 0.04), standardized effect d = 0.02; z-test p = 0.960.
• PA-824 1000 mg: mean 0.09 versus 0.11 on control; difference -0.01 (95% CI -0.07 to 0.04), standardized effect d = -0.21; z-test p = 0.559.
• PA-824 1200 mg: mean 0.09 versus 0.11 on control; difference -0.02 (95% CI -0.07 to 0.04), standardized effect d = -0.26; z-test p = 0.535.

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.

Early Bactericidal Activity (EBA) Measured as the Daily Rate of Change in log10 CFUs (Colony Forming Units) of M. Tuberculosis in Sputum on Solid Media (Days 0-2).

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • PA-824 600 mg: mean 0.10 versus 0.11 on control; difference -0.01 (95% CI -0.29 to 0.26), standardized effect d = -0.03; z-test p = 0.926.
  • PA-824 1000 mg: mean 0.03 versus 0.11 on control; difference -0.08 (95% CI -0.38 to 0.22), standardized effect d = -0.20; z-test p = 0.584.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between PA-824 600 mg and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between PA-824 1000 mg and the control. The interval gives the range of effects that remain compatible with the data.

Early Bactericidal Activity (EBA) Measured as the Daily Rate of Change in log10 CFUs (Colony Forming Units) of M. Tuberculosis in Sputum on Solid Media (Days 2-14).

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • PA-824 600 mg: mean 0.11 versus 0.11 on control; difference 0.01 (95% CI -0.05 to 0.06), standardized effect d = 0.10; z-test p = 0.806.
  • PA-824 1000 mg: mean 0.10 versus 0.11 on control; difference -0.01 (95% CI -0.06 to 0.04), standardized effect d = -0.18; z-test p = 0.655.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between PA-824 600 mg and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between PA-824 1000 mg and the control. The interval gives the range of effects that remain compatible with the data.

EBA Expressed as the Change in Time to Positive (TTP) Signal in Liquid Culture for M. Tuberculosis (Days 0-14)

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • PA-824 600 mg: mean 4.78 versus 3.82 on control; difference 0.96 (95% CI -1.04 to 2.96), standardized effect d = 0.36; z-test p = 0.348.
  • PA-824 1000 mg: mean 4.87 versus 3.82 on control; difference 1.05 (95% CI -1.25 to 3.34), standardized effect d = 0.35; z-test p = 0.372.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between PA-824 600 mg and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between PA-824 1000 mg and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes examine early bactericidal activity through other time intervals and through time to positive signal in liquid culture. Taken together, they provide related descriptions of short-term sputum culture changes, but they are not interchangeable with the primary solid-media measure. Differences in assay, time window, and measurement scale mean that consistency across outcomes can be informative without making one endpoint a direct substitute for another.

Across the posted secondary summaries, the arm means vary by outcome and dose. Recalculated comparisons generally have confidence intervals that include zero, and the registry does not list reported analyses for these outcomes. The secondary findings therefore do not provide precise inferential confirmation of a dose-group difference. They should be considered alongside, rather than as a replacement for, the primary endpoint, whose own comparisons are also uncertain.

These outcomes may help describe whether the observed pattern is similar across measurement approaches and periods. However, multiple endpoints and multiple dose comparisons create many opportunities to observe differences by chance. When the results are unadjusted and no prespecified hierarchy is reported, an isolated estimate or p-value deserves less weight than a result supported by a clear analysis plan and corroborating evidence.

The registry record does not report a multiplicity correction, a hierarchy among endpoints, or a prespecified strategy for interpreting concordance across solid and liquid culture results. Nor does it provide a formal sponsor analysis for the posted secondary measures. Consequently, the secondary summaries are best treated as exploratory descriptions of the available data. They can inform questions for further research, but do not on their own establish a reliable ranking of doses or a definitive treatment effect.

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

• Early Bactericidal Activity (EBA) Measured as the Daily Rate of Change in log10 CFUs (Colony Forming Units) of M. Tuberculosis in Sputum on Solid Media (Days 0-2).
• Early Bactericidal Activity (EBA) Measured as the Daily Rate of Change in log10 CFUs (Colony Forming Units) of M. Tuberculosis in Sputum on Solid Media (Days 2-14).
• EBA Expressed as the Change in Time to Positive (TTP) Signal in Liquid Culture for M. Tuberculosis (Days 0-14)
• EBA Expressed as the Change in Time to Positive (TTP) Signal in Liquid Culture for M. Tuberculosis (Days 0-2)
• EBA Expressed as the Change in Time to Positive (TTP) Signal in Liquid Culture for M. Tuberculosis (Days 2-14) Show Description: [Not Specified]

9. Statistical Interpretation

10. Statistical Methods Explained

The primary endpoint is continuous, and the registry reports a mean and standard deviation for each arm. A mean gives the arithmetic average of the endpoint values among participants included in that arm's summary. The standard deviation describes the spread of individual values around that average. It is not the standard error of the mean and is not a confidence interval. The sample size alongside each summary indicates how many participants had results available for the endpoint.

What does a mean difference represent?

The recalculated comparisons subtract the mean in the PA-824 200 mg reference group from the mean in another PA-824 group. A positive estimate means the second arm's reported mean is higher on the recorded scale; a negative estimate means it is lower. The numerical sign does not by itself imply a beneficial or harmful direction, because the registry record does not define a threshold or explain the clinical interpretation of a difference of a given size.

These are simple, unadjusted re-calculations from the posted arm summaries. They do not incorporate participant-level data or adjustment for covariates. The trial is randomized, but randomization alone does not transform a summary-statistic calculation into the sponsor's prespecified analysis. The registry reports no outcome analyses, and a detailed statistical analysis plan is not reported in the registry record. The recalculations therefore help describe the posted data, but do not establish what analysis the trial investigators intended to use.

How should a confidence interval be understood?

A confidence interval gives a range of estimates compatible with the data under the statistical method used. In repeated use of the same method under its assumptions, intervals constructed this way have a stated coverage property. For a mean difference, an interval spanning zero includes no difference as one value compatible with the data. It does not show that the true difference is exactly zero. The width of an interval reflects precision: wider intervals indicate less precise estimates and leave a broader range of plausible differences.

The calculations use the available arm sample sizes and pooled variability for the comparisons. A pooled standard deviation combines variability from the two arms for the purpose of a two-group comparison. The standard errors quantify the estimated sampling variability of the arm means. These methods rely on assumptions about the data and the comparison; the registry does not report diagnostics or a protocol analysis that would allow those assumptions to be checked against participant-level results.

Why do small endpoint counts matter?

The endpoint counts are small relative to the full enrollment, and counts differ across arms. Smaller samples generally make estimates less precise, particularly when observations vary. This is visible in the confidence intervals, which cover a range of both positive and negative mean differences. The primary outcome description says values could not be calculated for some participants because results were missing. The registry does not report why results were missing by participant or arm, whether missingness was related to observed outcomes, or whether any imputation or sensitivity analysis was used.

Missing observations can affect both the number of participants contributing to a mean and the interpretation of that mean. If participants with available measurements differ systematically from those without them, the observed summaries may not represent all randomized participants. This is a general concern about incomplete outcome data, not a claim that such a pattern occurred in this trial. The posted record does not provide enough information to assess it.

Why can variability affect planning?

In general, greater variability makes it harder to estimate a group difference precisely, while smaller effects require more information to distinguish them from random variation. The observed variability differs across endpoints and arms. The primary endpoint standard deviations are modest in absolute terms, but the observed differences between means are also small. Planning calculations based on small observed differences can therefore imply large participant requirements. Such illustrations depend on assumptions and are not a substitute for a prospectively justified sample-size calculation.

The posted planning illustrations use a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. They use fractions of the observed effect as target effects. These are planning assumptions for the illustrations only; the original trial design assumptions are not reported in the registry record. A nominal significance level describes a decision threshold in a test, while power describes the chance of detecting a specified effect if that effect exists under the planning model. Neither quantity shows that the trial's own design had those properties.

The record also contains multiple outcomes and multiple dose comparisons. When several hypotheses are examined, the chance of at least one apparently unusual result can increase if no adjustment or prespecified testing sequence is used. No multiplicity strategy is reported in the registry record. The unadjusted p-values should therefore be interpreted as individual descriptive tests, rather than as a controlled family of confirmatory tests.

11. Confidence Intervals and Precision

Precision is limited by the endpoint counts and the uncertainty around the recalculated differences. The primary summaries include 12 participants in the 200 mg group, 14 in the 600 mg group, 15 in the 1000 mg group, 11 in the 1200 mg group, and 7 in the Rifafour group. The registry notes that missing results prevented some EBA calculations. These counts are the participants with available results for the relevant period, not necessarily the number randomized to each arm.

For the 600 mg comparison with 200 mg, the estimated mean difference is 0.001, with a 95% confidence interval from -0.038 to 0.040. For 1000 mg, the estimated difference is -0.015, with a 95% confidence interval from -0.065 to 0.035. For 1200 mg, it is -0.018, with a 95% confidence interval from -0.075 to 0.039. Each interval covers zero and includes values on either side of it. Thus, the point estimates should not be read as precise evidence of a difference in either direction.

A confidence interval is more informative than a point estimate alone because it communicates uncertainty about the size and direction of a difference. Here, the intervals do not rule out either a lower or a higher mean relative to the 200 mg group. They also do not establish equivalence: that would require a defined range of differences considered acceptably small, together with an appropriate design and analysis. Such a margin is not reported in the registry record.

The recalculated intervals use posted summary statistics and do not account for a broader analysis plan, multiplicity, or missing-data adjustments. They are consequently one way to describe the precision of the available summaries, not a replacement for a full analysis using the trial data. The registry does not report sponsor confidence intervals for these comparisons.

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 figures are participants per arm for 80% power and participants per arm for 90% power for the comparisons that use two treatment arms. They are not total events. The calculations assume a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The target effect is set to the stated fraction of the observed effect. The original trial's own sample-size justification and design assumptions are not reported in the registry record.

The 80% power column gives an illustrative number of participants per arm under which the planning model would detect the specified target effect with 80% power, if its assumptions hold. The 90% power column gives the corresponding illustration for 90% power. These are planning quantities, not descriptions of the chance that the completed trial found an effect. The posted sample-size calculations are not observed power.

Rows that plan for 75% or 50% of the observed effect ask what would be needed to detect a smaller effect than the full observed estimate. Smaller effects require more participants in these illustrations, because they are harder to distinguish from random variation. Planning figures can become especially large when the observed difference is close to zero, as in the primary 600 mg comparison. That does not mean those exact numbers would be appropriate for a future study: planning should be based on a scientifically justified target effect and suitable assumptions, not only on a noisy estimate from a small trial.

The calculations are separate two-arm illustrations and do not represent the original multi-arm trial design. They do not include any reported adjustment for multiple comparisons or missing outcomes. The registry record does not report the trial's original sample-size rationale, so these figures should be read only as conditional planning examples based on the stated assumptions.

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 = 0.001; pooled standard deviation σ = 0.051.
2. n = 2 × (1.96 + 0.842)² × 0.051² / 0.001² = 41,159.5, rounded up to 41,160 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 <- 0.001                      # observed difference in means
sigma <- 0.0512054684579684                     # 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(567840)
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 = 41,160 (80% power), 55,101 (90% power)
 75% of observed effect: n per arm = 73,173 (80% power), 97,958 (90% power)
 50% of observed effect: n per arm = 164,639 (80% power), 220,404 (90% power)
power.t.test: 41160.5 
Simulated power at n = 41,160 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

The registry lists a primary outcome, several secondary outcomes, and more than one PA-824 dose comparison. Testing multiple outcomes or contrasts can increase the chance of at least one small p-value arising by chance if every result is considered separately. The reported recalculated p-values are unadjusted. A multiplicity plan, such as an adjustment or a prespecified testing order, is not reported in the registry record. This limits the strength of confirmatory interpretations based on individual comparisons.

Missing outcome data are directly relevant to the primary endpoint. The registry explains that some EBA values could not be calculated because results were missing, and that the reported counts are those with available results for the time period. The primary endpoint counts differ among arms. The registry does not explain the reasons for missing results, give participant-level missingness patterns, describe how missing data were handled, or report sensitivity analyses. It is therefore not possible to determine from the posted record whether the available-case summaries could be affected by informative missingness.

The analysis populations are not reported in the registry record. In particular, the record does not specify whether the posted arm summaries correspond to an intention-to-treat population, a per-protocol population, or another defined set of participants. The total enrollment should not be assumed to equal the number analyzed for every outcome.

There are no analyses listed as reported for the primary outcome, and the secondary outcomes likewise have no reported analyses. The registry record also does not include the trial-specific statistical analysis plan, details of any interim analyses, or a complete account of the model used by the sponsor. These absences prevent assessment of whether the recalculated comparisons align with planned analyses. The available statistics support description of the posted summaries, but do not resolve questions about prespecification, multiplicity, or the impact of missing data.

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 serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: hemoptysis (12.5%).

The registry reports serious adverse events for each arm as counts among participants at risk. The counts were 1 of 15 for PA-824 200 mg, 0 of 15 for PA-824 600 mg, 0 of 16 for PA-824 1000 mg, 0 of 15 for PA-824 1200 mg, and 1 of 8 for Rifafour. These are serious-event figures, not counts of all adverse events. The record lists Hemoptysis at 12.5% among the top serious terms, but does not give an arm-specific attribution for that term in the posted information.

Other adverse events were also reported, with counts among those at risk of 6 of 15, 4 of 15, 5 of 16, 7 of 15, and 2 of 8 in the respective arms. These counts should not be conflated with serious adverse events. The registry does not report death counts in the posted record. It also does not provide enough detail about event timing, severity grading beyond the serious-event classification, relatedness to treatment, or statistical comparisons between arms.

The arm counts are small, and descriptive differences in counts should not be interpreted as evidence that one regimen is safer or less safe. These figures characterize events recorded in this trial; they do not establish comparative safety.

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

The posted primary endpoint summaries show mean daily rates of change in log10 CFUs that were close across the PA-824 dose groups. The recalculated unadjusted differences versus PA-824 200 mg are small in their point estimates, and all reported confidence intervals include zero. The results therefore do not precisely establish a difference between these PA-824 groups on this endpoint. They also do not establish that the groups are equivalent, because the registry record does not report an equivalence margin or a corresponding analysis.

The Rifafour arm has a reported primary mean, but no recalculated primary comparison against Rifafour is listed. It would be inappropriate to infer a comparative effect from the arm means alone, especially with a small available endpoint count and no reported analysis. The trial's other early bactericidal activity outcomes offer additional descriptions of short-term culture measures, but the posted secondary analyses are also not reported and their unadjusted results warrant caution.

Early bactericidal activity is a short-term biological measure. These data do not, by themselves, establish longer-term clinical outcomes or answer questions that require follow-up beyond the measured period. The record does not provide a basis for conclusions about outcomes not included in the posted results.

More generally, lack of a statistically clear difference is not proof that no difference exists. The uncertainty intervals cover multiple possible values, and limited endpoint counts constrain precision. Conversely, an observed numerical difference is not necessarily meaningful or reproducible. Its interpretation depends on the measurement scale, the prespecified clinical context, the quality of the analysis, and other evidence, much of which is not reported in the registry record.

The safety counts describe events recorded in the trial and should not be used alone to compare risk between groups. Small group sizes and incomplete reporting of event details limit such interpretation. Overall, the record supports a cautious description of the observed short-term endpoint summaries, not a definitive claim about efficacy, equivalence, or comparative safety.

17. Why This Trial Matters Statistically

This trial is statistically instructive because it combines a multi-arm randomized design with a continuous early bactericidal activity endpoint and small numbers of participants contributing to each endpoint summary. Randomization provides a framework for comparing assigned groups, while the observed arm counts and missing endpoint values show why the precision and analysis population matter when interpreting those comparisons.

The primary mean differences are close to zero for the reported PA-824 dose comparisons, but their confidence intervals are wide enough to include values on both sides of zero. This illustrates why a point estimate or p-value alone cannot answer how precisely a study estimates a difference. The endpoint counts, standard deviations, and intervals all contribute to that assessment.

The record also illustrates the distinction between a trial's design and a later planning illustration. The posted sample-size figures are based on fractions of observed effects under stated assumptions, whereas the trial's own sample-size justification is not reported. Using observed effects from a small study can produce planning numbers that are unstable, especially when the estimated effect is near zero. Future planning requires a defensible target effect and assumptions suited to the intended study.

Finally, the registry record makes visible several common reporting challenges: multiple outcomes, unadjusted comparisons, incomplete endpoint data, and no posted formal analyses. These do not erase the descriptive information in the results, but they limit what can be concluded from it. Distinguishing descriptive patterns from confirmatory evidence is an important part of sound interpretation in early clinical research.

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 available record. The primary outcome is based on participants with calculable EBA values, and the registry reports that some values were unavailable because results were missing. The arm-level reasons and patterns of missingness are not reported, so the effect of missing outcomes cannot be assessed.

The posted primary comparisons are recalculated unadjusted comparisons of PA-824 dose groups against the 200 mg group. A corresponding primary comparison with Rifafour is not reported. Nor does the record include sponsor-reported inferential analyses for the primary or secondary outcomes. As a result, the posted contrasts cannot be checked against a complete trial analysis or a prespecified statistical plan.

There are multiple outcomes and dose contrasts, but the registry record does not report a strategy for controlling multiplicity or a hierarchy for interpreting endpoints. The analysis populations and handling of missing data are also not reported. These omissions limit confirmatory interpretation of unadjusted p-values and confidence intervals.

Enrollment was 69, while endpoint counts were lower and varied across arms. The available summaries do not include participant-level data, baseline characteristics, covariate-adjusted estimates, or details that would permit evaluation of model assumptions. The record does not report a clinical threshold for interpreting the size of the primary measure or a prespecified equivalence margin.

Finally, the primary endpoint describes short-term bactericidal activity, not longer-term outcomes. Safety reporting includes arm-level serious and other adverse-event counts, but death counts are not reported in the registry record and event details are limited. These restrictions mean that the results should be interpreted as a partial statistical account of a short-duration Phase IIa trial.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: PA-824-CL-007: Phase IIa Evaluation of Early Bactericidal Activity in Pulmonary Tuberculosis — NCT00567840.
• Registry record last updated 2017-05-15; status last verified 2017-01.
• Results first posted 2017-04-04.
• 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, quadruple-masked, parallel Phase IIa trial with 69 participants. It compared four once-daily PA-824 dose groups with Rifafour in adults with newly diagnosed, uncomplicated, smear-positive pulmonary tuberculosis.

The primary outcome was the daily rate of change in sputum log10 CFUs on solid media from Day 0 to Day 14. Available-case arm summaries and recalculated unadjusted comparisons are posted for PA-824 doses versus the 200 mg group. The confidence intervals include zero, and no sponsor analysis is listed in the record. Secondary outcomes cover other culture periods and liquid-culture time to positive measures.

Missing primary outcome results, lack of a reported multiplicity strategy, and absence of a trial-specific statistical analysis plan limit interpretation. Safety information is reported as arm-level event counts, with death counts not reported in the registry record. The results characterize short-term biological measurements and recorded events; they do not establish longer-term outcomes or definitive comparative effects.