Placebo-Controlled Trial of Antibiotic Therapy in Adults With Suspect Lower Respiratory Tract Infection (LRTI) and a Procalcitonin Level: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Placebo-Controlled Trial of Antibiotic Therapy in Adults With Suspect Lower Respiratory Tract Infection (LRTI) and a Procalcitonin Level (NCT03341273), a randomized na study. The primary outcome is Clinical Improvement at Day 5 Visit (D5V).

2. Clinical Question

The question is framed as non-inferiority rather than as a test asking only whether one group has a higher improvement percentage. A non-inferiority analysis asks whether the observed result is sufficiently compatible with a prespecified acceptable loss in outcome. The acceptable margin is essential to interpreting that question, but it is not reported in the registry record. The sponsor reports a non-inferiority analysis, while the posted arm-level results also allow a simple unadjusted comparison of the percentages meeting the endpoint. These are related but not interchangeable summaries: the sponsor’s analysis addresses the stated trial objective, and the unadjusted calculation describes the difference in the posted proportions without establishing the trial’s non-inferiority criterion.

The eligibility description focuses on outpatients with suspected LRTI and a low PCT measurement at enrollment. It does not establish that every participant had a bacterial infection, nor does the registry record report a diagnostic confirmation requirement. The clinical question therefore applies to the defined trial population and endpoint, not to all people with respiratory symptoms or to other care settings.

3. Trial Design

This was a randomized, double-blinded, placebo-controlled, parallel-group, multicenter non-inferiority study sponsored by the National Institute of Allergy and Infectious Diseases. The registry lists phase as NA. The study compared oral azithromycin with matching placebo among adults presenting as outpatients with suspected LRTI and PCT <= 0.25 ng/mL. Both groups were assigned a five-day treatment schedule. The registry lists randomization and double masking, but does not report the specific sequence-generation or masking procedures.

The listed start date is 2017-12-08, and the primary completion date is 2020-03-30. The current status is terminated. The registry lists enrollment as 514. Arm descriptions state N=337 for each group, while the primary endpoint results include 249 participants in the azithromycin group and 250 in the placebo group. How these figures relate, including the reason for the difference between the enrollment listing and arm-description counts, is not reported in the registry record.

The primary outcome was clinical improvement at the Day 5 visit. The study description says efficacy was assessed from the first dose through approximately Day 28. Clinical improvement at Day 11 and Day 28 was also listed as a secondary outcome. Other secondary outcomes included symptom improvement or worsening, a desirability-of-outcome ranking using a RADAR approach, and health-care visits for persistent or worsening LRTI. The registry lists 19 outcome measures, with 18 secondary outcomes posted.

The registry record does not report the original sample-size justification, the planned non-inferiority margin, or the protocol’s detailed statistical analysis plan. It also does not explain why the trial was terminated. Those omissions limit assessment of how closely the reported analysis and observed enrollment correspond to the original design.

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 ITT population includes all participants with PCT \<= 0.25 ng/mL who were randomized to receive study product.

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.

Population: Intent-to-Treat (ITT) population includes all participants with procalcitonin (PCT) = 0.25 ng/mL who were randomized to receive study product.

6. Primary Endpoint

The registry lists the primary outcome as Clinical Improvement at Day 5 Visit (D5V), measured over Day 5 Visit.

The primary endpoint is the percentage of participants meeting the registry’s definition of clinical improvement at the Day 5 visit. Improvement requires meeting all of the listed conditions: improvement in at least two symptoms present at enrollment, or in one symptom and at least one enrollment vital-sign abnormality; no deterioration in a qualifying symptom and no new vital-sign abnormality; no fever on the preceding day or at the visit; and no medically attended ambulatory visit or hospitalization for persistent or worsening LRTI after randomization.

This composite definition combines symptom and vital-sign change with the absence of fever and subsequent medical attendance for worsening or persistent illness. It is intended to capture more than a change in a single symptom. Because each component must be considered together, the endpoint represents a defined clinical-improvement state rather than an isolated symptom response. The posted unit is the percentage of participants.

A Day 5 assessment is suited to the trial’s stated question about early response after the treatment period began. Using a prespecified composite can make the outcome more consistent across participants, but its interpretation depends on the component rules and on how missing component information is handled. The registry record gives the endpoint definition and identifies the intention-to-treat population, but does not report a detailed missing-data rule or the original rationale for selecting this endpoint and visit.

The endpoint should be read as a measured outcome in this trial population, not as a direct measure of every aspect of illness or later health. The Day 11 and Day 28 clinical-improvement outcomes provide later assessments, but they do not replace the primary Day 5 outcome.

7. Results Posted in the Registry

• Azithromycin: 69.1% (172/249) versus 63.2% (158/250) on control; risk difference 5.9 percentage points (95% CI -2.4 to 14.2), risk ratio 1.09 (0.96–1.24); unadjusted two-proportion z-test p = 0.165.

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.

Clinical Improvement at Day 11 Visit (D11V)

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Azithromycin: 81.1% (202/249) versus 76.0% (190/250) on control; risk difference 5.1 percentage points (95% CI -2.1 to 12.3), risk ratio 1.07 (0.97–1.17); unadjusted two-proportion z-test p = 0.163.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Azithromycin and the control. The interval gives the range of effects that remain compatible with the data.

Clinical Improvement at Day 28 Visit (D28V)

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Azithromycin: 88.0% (219/249) versus 82.0% (205/250) on control; risk difference 6.0 percentage points (95% CI -0.3 to 12.2), risk ratio 1.07 (1.00–1.16); unadjusted two-proportion z-test p = 0.063.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Azithromycin and the control. The interval gives the range of effects that remain compatible with the data.

Number of Participants Exhibiting Improvement in One or More LRTI Symptoms or Fever at Day 11 Visit

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Azithromycin: 97.6% (206/211) versus 98.2% (221/225) on control; risk difference -0.6 percentage points (95% CI -3.3 to 2.1), risk ratio 0.99 (0.97–1.02); unadjusted two-proportion z-test p = 0.664.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Azithromycin and the control. The interval gives the range of effects that remain compatible with the data.

The registry lists 18 secondary outcomes, including clinical improvement at later visits, symptom changes, worsening, health-care use, and a desirability-of-outcome ranking. Among the later clinical-improvement results with posted arm-level estimates, the azithromycin percentages are higher than the placebo percentages. Their unadjusted comparison intervals include zero, so the estimates do not provide a precise separation between groups. This broad pattern is consistent with the primary comparison’s direction in the posted proportions, but it does not independently resolve the primary uncertainty.

The secondary outcomes address different aspects and time points of illness. A later clinical-improvement measure is not identical to the Day 5 composite, while symptom-specific counts and health-care visits reflect other outcomes. Differences in outcome definition, visit timing, and available denominators make it inappropriate to treat all secondary measures as repeated versions of one test. Several listed secondary outcomes are not accompanied by results in the record used for these comparisons, so no overall effect pattern can be established across every listed outcome.

Unadjusted secondary results carry less weight because many outcomes are considered and the registry record does not report a multiplicity adjustment or a prespecified hierarchy for interpreting them. When multiple tests are examined, an apparently notable result can occur by chance even if no underlying effect exists. Secondary outcomes can help describe consistency and generate questions, but they should not be used as standalone confirmation of efficacy or non-inferiority. The primary endpoint remains the prespecified focus identified by the registry, subject to the limits of the missing design and analysis details.

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

• Clinical Improvement at Day 11 Visit (D11V)
• Clinical Improvement at Day 28 Visit (D28V)
• Composite Overall Desirability of Outcome Ranking (DOOR) Assessed Employing a Superiority Analysis Using the "Response Adjusted for Days of Antibiotic Risk (RADAR)" Approach at Day 5 Visit
• Number of Participants Exhibiting Improvement in One or More LRTI Symptoms or Fever at Day 11 Visit
• Number of Participants Exhibiting Improvement in One or More LRTI Symptoms or Fever at Day 28 Visit
• Number of Participants Exhibiting Improvement in at Least Two Presenting Signs or Symptoms at Day 5 Visit
• Number of Participants Exhibiting Worsening or Deterioration in One or More Symptoms at Day 5 Visit
• Number of Participants Reporting One or More Hospitalization or Visits to an Emergency Department (ED), Outpatient Clinic, or Urgent Care Center (After Randomization) for Worsening or Persistent Lower Respiratory Tract Infection

9. Statistical Interpretation

10. Statistical Methods Explained

What does a risk difference describe?

A risk difference compares the proportion of participants with an outcome in one group with the proportion in another. A value of zero corresponds to equal proportions under the chosen contrast direction. A positive or negative value indicates which group has the higher proportion only when the direction of the subtraction is known. The unadjusted recalculation for the primary outcome uses azithromycin relative to placebo and reports 0.059, with a 95% confidence interval from -0.024 to 0.142. The sponsor’s non-inferiority result is reported as -6, with a confidence interval from -15 to 2, but the registry record does not state its contrast direction. The two numbers should not be treated as directly comparable without that information and the sponsor’s analysis details.

The recalculation is unadjusted: it compares the posted summary proportions without adjustment for other characteristics or a more complex model. It is useful as a transparent description of the observed binary outcome, but it is not a substitute for the trial’s prespecified analysis. The registry record does not identify the method used for the sponsor’s reported analysis.

How does a confidence interval help?

A confidence interval expresses uncertainty around an estimate. It gives a range of values compatible with the data and the method used, under the procedure’s assumptions. It is not a range containing a known true value with certainty, and it does not mean that every value in the interval is equally likely. The primary recalculated interval extends from a small negative risk difference to a positive one. Since it crosses zero, it does not rule out equal proportions under this unadjusted comparison. Its width also shows that the estimate is not highly precise.

The intervals for each arm’s percentage describe uncertainty in each group proportion. They are not the same as the interval for the difference between groups. A comparison should use the interval for the contrast, because uncertainty in a difference depends on both groups and on the method used to calculate that contrast.

What is the distinction between superiority and non-inferiority?

A conventional superiority question asks whether the data support a difference in a specified direction or any difference, depending on the test. A non-inferiority question instead asks whether the outcome with the intervention is not worse than the comparator by more than a prespecified acceptable amount. That acceptable amount is the non-inferiority margin. The analysis needs a defined contrast direction, margin, and statistical procedure. The registry labels the sponsor analysis as non-inferiority and posts its estimate and confidence interval, but does not report the margin or method. Without them, readers cannot determine from the summary alone how the formal criterion was set or applied.

A non-significant superiority comparison would not, by itself, demonstrate non-inferiority. These are different statistical questions. Similarly, a non-inferiority label does not establish that the criterion was satisfied; interpretation requires the margin and the relevant confidence-bound rule.

Why do binary outcomes and denominators matter?

The primary outcome is binary for each participant: the participant either meets the defined clinical-improvement criteria or does not. The registry reports percentages and group denominators. The primary endpoint population is described as randomized participants with PCT <= 0.25 ng/mL, but the record does not fully explain why the primary outcome denominators are smaller than the counts stated in the arm descriptions. Differences between randomized, enrolled, and analyzed populations can affect interpretation, especially when outcome status is unavailable for some participants. A detailed accounting of these populations and missing outcomes is not reported in the registry record.

Why can over-dispersion matter?

As a general statistical point, a simple binary-outcome calculation often relies on a model for the variability of observed proportions. Over-dispersion means that outcomes vary more than the model assumes, potentially because observations are more alike within clusters or because other sources of variation are not represented. If that extra variability is present but ignored, standard errors may be too small and confidence intervals too narrow. The registry identifies the study as multicenter, but does not report whether clustering by center was accounted for in the sponsor’s analysis or whether over-dispersion was assessed. This is a general consideration, not a claim that over-dispersion occurred in this trial.

What do the posted p-values and secondary comparisons mean?

The unadjusted p-value for the primary recalculation is 0.165. A p-value describes how unusual results at least as incompatible with a specified null model would be if that model and the analysis assumptions held. It is not the probability that the null hypothesis is true, and it does not quantify the size or clinical importance of an effect. It also does not evaluate whether a non-inferiority margin has been met.

Multiple outcome comparisons create a multiplicity issue: as the number of tests increases, the chance of at least one apparently unusual result can increase if each result is considered in isolation. The registry lists many secondary outcomes, but adjustment procedures and a hierarchy for interpreting them are not reported. Accordingly, the unadjusted secondary comparisons are best treated as descriptive or exploratory rather than as independent confirmation of the primary finding.

These explanations distinguish the simple recalculation from the formal trial analysis. The posted results support a careful description of the observed proportions and their uncertainty. They do not supply the missing design and analysis details needed to reconstruct the sponsor’s complete non-inferiority procedure.

11. Confidence Intervals and Precision

The primary arm estimates are 69% for azithromycin, with a 95% confidence interval from 61% to 77%, and 63% for placebo, with a 95% confidence interval from 54% to 71%. The group intervals are fairly broad, so the percentages should not be interpreted as exact population values. The interval for the unadjusted risk difference, 0.059 with a 95% confidence interval from -0.024 to 0.142, is the more direct summary of uncertainty in the comparison. It includes zero and permits values on either side of zero under the recalculation’s assumptions.

These intervals reflect sampling uncertainty as calculated from the posted summary statistics. They do not capture every possible source of uncertainty, such as bias from missing outcomes, issues with the endpoint definition, or unreported design and analysis choices. The registry record does not report the detailed approach to missing data or the sponsor’s statistical method. A narrow confidence interval would not resolve such concerns, and a broad interval does not imply that the true effect is exactly zero.

The sponsor reports a non-inferiority risk difference of -6 with a 95% confidence interval from -15 to 2. The confidence limits alone do not establish whether the result satisfies the non-inferiority criterion, because the registry record does not report the margin or contrast direction. The sponsor interval and the recalculated interval arise from reported analyses with different available details; they should be interpreted within their own definitions rather than combined into a single estimate.

Confidence intervals communicate precision, not clinical importance. The registry record does not report a threshold for a clinically meaningful difference, so the observed uncertainty cannot be classified against such a threshold. The same caution applies to later outcome intervals: their estimates are descriptive and do not eliminate uncertainty about the primary comparison.

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 figures are planning illustrations based on the observed effect, not the trial’s original design calculation. For these binary outcomes, the figures are participants per arm; the general convention for time-to-event outcomes would instead be total events, but the displayed planning context for this comparison is participants per arm. The listed assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is the stated fraction of the observed effect.

The 80% power column describes the planned number of participants per arm that would give an 80% chance of detecting the specified effect under those assumptions, if that effect and the statistical model were correct. The 90% power column gives the corresponding planning number for a higher chance of detection. These are prospective design illustrations; they are not evidence about the chance that the completed trial detected an effect.

Rows that plan for 100%, 75%, and 50% of the observed effect illustrate how the assumed target affects the required sample size. Smaller effects are harder to distinguish from random variation, so planning for a smaller effect requires more participants. The calculations use the observed estimate as their starting point, which is uncertain; they should not be treated as a recommendation or as a reconstruction of the trial’s original sample-size justification.

The registry record does not report the trial’s original design assumptions, including its planned target effect or non-inferiority margin. The planning illustrations therefore answer a different question from whether the completed study had adequate power for its original objective. They are not observed power, and they do not replace interpretation of the estimates and confidence intervals.

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, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.632; treatment proportion p₁ = 0.691; average p̄ = 0.661.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6693 = 1.3117.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6680 = 0.5622.
4. n = (1.3117 + 0.5622)² / (0.0588)² = 1,016.9, rounded up to 1,017 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
p0 <- 0.632                          # control proportion
p1 <- 0.690763052208835                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^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 (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(3341273)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 1,017 (80% power), 1,361 (90% power)
 75% of observed effect: n per arm = 1,828 (80% power), 2,447 (90% power)
 50% of observed effect: n per arm = 4,154 (80% power), 5,560 (90% power)
power.prop.test: 1016.9 
Simulated power at n = 1,017 per arm: 0.802 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-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. Analyses Reported by the Sponsor

The registry posts a sponsor analysis labeled non-inferiority. It reports a risk difference of -6, with a 95% confidence interval from -15 to 2, and does not report a p-value or name the statistical method. The record also does not state the contrast direction or the non-inferiority margin. Those details are necessary to judge the formal non-inferiority criterion, so the posted result should be reported without asserting that the criterion was or was not met.

The unadjusted recalculation from the posted arm summaries estimates a risk difference of 0.059, with a 95% confidence interval from -0.024 to 0.142 and an unadjusted p-value of 0.165. It describes the observed group proportions using a simple comparison. It does not reproduce the sponsor’s analysis, and it does not use information about the sponsor’s analysis method that is not reported in the registry.

The sign of a risk difference depends on the order of subtraction. Because the sponsor’s contrast direction is not stated, the sponsor estimate cannot be mapped directly to the recalculated direction from the posted record alone. In addition, the sponsor’s result is explicitly labeled non-inferiority, whereas the recalculation is an unadjusted comparison. The two summaries therefore have different interpretive roles. The sponsor result is the posted formal analysis; the recalculation is a separate descriptive check based on the arm-level summaries. Neither should be used to fill in the unreported margin or analysis plan.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists 19 outcome measures, including 18 secondary outcomes, and reports four analyses. This number of outcomes raises a multiplicity consideration: interpreting many unadjusted comparisons as if each were the only test can overstate the strength of evidence. The registry record does not report whether the sponsor adjusted for multiplicity, designated a hierarchy among secondary outcomes, or treated particular secondary findings as confirmatory. Without that information, secondary comparisons should be read cautiously and not as a collection of independent confirmations.

The record identifies the primary outcome population as the intention-to-treat population, defined as randomized participants with PCT <= 0.25 ng/mL who received study product. It does not report a detailed account of participant flow into each analysis, the handling of participants with missing endpoint data, or whether all randomized participants were included in the primary analysis as defined. The primary outcome denominators are listed as 249 and 250, while the arm descriptions state N=337 in each group. How these figures relate is not reported in the registry record.

Missing outcome data can affect both the estimated percentages and their uncertainty, especially if the likelihood of missingness differs by group or is related to clinical status. This is a general statistical concern; the registry does not establish that missing data caused bias in this study. The record also does not describe imputation, sensitivity analyses, or rules for classifying participants with incomplete endpoint components.

The non-inferiority setting makes these details particularly relevant. Analysis population choices and missing-data assumptions can influence a non-inferiority assessment, and the registry does not provide enough information to evaluate their impact. The sponsor’s method, planned margin, original analysis plan, and any interim analyses are not reported in the registry record. No conclusions about those procedures should be inferred from the posted summary alone.

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

The posted results describe the percentage of participants meeting a specific Day 5 clinical-improvement definition in two randomized groups of outpatients with suspected LRTI and PCT <= 0.25 ng/mL. The observed percentage was higher in the azithromycin group than in the placebo group, but the unadjusted risk-difference interval includes zero. This means the recalculation does not precisely distinguish a difference from no difference; it does not establish that the groups are identical.

The sponsor’s result is labeled non-inferiority, but the non-inferiority margin and contrast direction are not reported. Consequently, the posted record does not allow a reader to verify the formal criterion from the estimate and interval alone. A non-significant unadjusted comparison is not proof of equivalence, and it is not a substitute for the sponsor’s non-inferiority analysis.

The findings apply to the trial population and outcome as defined. They do not establish outcomes for people outside the outpatient setting, participants with PCT above the eligibility threshold, or other treatment schedules. The endpoint combines symptom and vital-sign improvement with absence of deterioration, fever, and medically attended care for worsening or persistent LRTI. It is not a direct measurement of every possible patient outcome.

The registry lists the study as terminated but does not report why. It also does not provide the original sample-size justification, detailed statistical plan, missing-data procedures, or multiplicity approach. Those gaps limit assessment of the full design and of how the reported findings relate to the original objectives. The results should therefore be read as estimates with stated uncertainty, rather than as a definitive answer about treatment effectiveness or non-inferiority.

17. Why This Trial Matters Statistically

This trial illustrates why the statistical question matters as much as the observed percentages. The study was framed around non-inferiority, a design that requires a prespecified margin and a clearly defined contrast. The registry reports the sponsor’s non-inferiority estimate but not those critical details. A reader can describe the posted result, but cannot fully audit the formal criterion from the record alone.

It also illustrates the difference between a formal trial analysis and an unadjusted recalculation. The latter provides a transparent comparison of the posted arm summaries and a confidence interval for that comparison. It is useful for understanding the estimate’s uncertainty, but it cannot recreate an analysis plan that is not reported. Keeping those roles distinct prevents a simple recalculation from being mistaken for the sponsor’s primary inferential method.

The record also highlights the importance of clear population accounting. Enrollment, arm-description counts, and primary outcome denominators are all reported, but how they relate is not explained. Clear accounting helps readers understand which participants contribute to each estimate and how incomplete outcomes are handled.

Finally, multiple secondary outcomes can broaden the clinical description, but they complicate inference when adjustment and outcome hierarchy are not reported. The general lesson is to consider the primary question, effect estimate, confidence interval, analysis population, and design details together. Statistical interpretation is strongest when those elements are reported transparently.

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.
• Participant counts were reconstructed from posted percentages and arm sizes, so small rounding differences are possible.

Several limitations specific to this record constrain interpretation. The study is listed as terminated, but the reason for termination is not reported in the registry record. The record gives enrollment as 514 while each arm description states N=337, and the primary endpoint results use denominators of 249 and 250. It does not explain how these counts relate, whether they refer to different stages or populations, or why the primary endpoint denominators differ from the arm-description counts.

The registry does not report the planned non-inferiority margin, the original sample-size justification, or the statistical method used for the sponsor’s non-inferiority analysis. It also does not state the contrast direction for the sponsor’s reported risk difference. Those omissions prevent a complete independent assessment of the formal non-inferiority conclusion. The record does not report detailed missing-data handling, sensitivity analyses, multiplicity procedures, or the analysis hierarchy for secondary outcomes.

Although the study is identified as multicenter, the record does not say whether the sponsor’s analysis accounted for center. The posted summary also does not give enough detail to assess the completeness of outcome ascertainment or the influence of missing endpoint components. These are limitations of the available reporting; they do not establish that any particular bias occurred.

The results are specific to the stated outpatient population, PCT threshold, treatment schedule, and composite Day 5 endpoint. The registry record does not establish how well they generalize beyond those conditions. The posted adverse-event summary reports no serious events or deaths in either group, but it does not supply detailed serious-event terms. This summary should not be treated as a complete account of safety outcomes.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Placebo-Controlled Trial of Antibiotic Therapy in Adults With Suspect Lower Respiratory Tract Infection (LRTI) and a Procalcitonin Level — NCT03341273.
• Registry record last updated 2023-06-28; status last verified 2020-03-12.
• Results first posted 2021-08-24.
• 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 NCT03341273, a randomized, double-blinded, placebo-controlled, multicenter parallel-group non-inferiority study of azithromycin versus placebo in adults with suspected outpatient LRTI and PCT <= 0.25 ng/mL. The primary endpoint was clinical improvement at the Day 5 visit. The listed status is terminated; the reason is not reported in the registry record.

Posted primary outcome estimates were 69% for azithromycin and 63% for placebo. The unadjusted recalculation gives a risk difference of 0.059, with a 95% confidence interval from -0.024 to 0.142. The sponsor’s reported non-inferiority risk difference is -6, with a 95% confidence interval from -15 to 2. The registry does not report the non-inferiority margin or contrast direction, so the formal criterion cannot be independently assessed from the posted summary alone.

Later clinical-improvement outcomes have estimates in the same direction, with unadjusted intervals including zero. Interpretation is limited by unreported details on the original design assumptions, missing data, multiplicity, and the relationship between the listed enrollment, arm-description counts, and endpoint denominators.