A Study To Evaluate The Effect Of PF-04447943 On Qtc Interval In Healthy Adult Subjects: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study To Evaluate The Effect Of PF-04447943 On Qtc Interval In Healthy Adult Subjects (NCT02785770), a randomized phase 1 study. The primary outcome is Time-Matched Mean Difference in Corrected QT Interval Using Fridericia's Correction Method (QTcF) for PF-04447943 and Placebo at 0.5 Hour Post-Dose.

2. Clinical Question

3. Trial Design

This was a Phase 1 randomized, four-treatment, four-period crossover study in healthy adult subjects. The registry describes it as placebo- and positive-controlled, double-blind, open-label for the positive control, and sponsor-open. Its listed masking classification is triple. The study is marked completed, with a start date of 2016-06-29 and primary completion date of 2016-10-20. The registry records enrollment of 44.

The treatments were PF-04447943 25 mg, PF-04447943 100 mg, matching placebo, and moxifloxacin 400 mg. In a crossover design, participants receive treatments across periods, so comparisons can use measurements from the same participants under different treatment conditions. The registry’s primary analysis population is defined as participants who were randomized and treated and had at least 1 post-dose ECG measurement in at least 1 period. The available record does not describe treatment sequences, period spacing, or the details of how period and sequence effects were handled.

The primary endpoint was a time-matched mean difference in QTcF at 0.5 hour post-dose for each PF-04447943 dose versus placebo. The endpoint description identifies least-squares means for each reporting arm and least-squares mean differences for the dose-placebo comparisons. Secondary outcomes included other ECG intervals, heart rate, adverse events, physical examination findings, laboratory tests, and vital signs. The registry lists 20 outcome measures and 72 analyses, while the posted secondary outcome data include 12 measures. The original design assumptions, including any sample-size justification, 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.

Intervention Period 1 (1 Day)

Washout Period 1 (7 Days)

Intervention Period 2 (1 Day)

Reasons for not completing this period:

Washout Period 2 (7 Days)

Analysis population for the primary outcome, as defined in the registry: Electrocardiogram (ECG) analysis population included all participants randomized and treated who had at least 1 post-dose ECG measurement in at least 1 period.

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: Analysis set included all participants who were randomized in the study.

6. Primary Endpoint

The registry lists the primary outcome as Time-Matched Mean Difference in Corrected QT Interval Using Fridericia's Correction Method (QTcF) for PF-04447943 and Placebo at 0.5 Hour Post-Dose, measured over 0.5 hour post-dose.

The primary endpoint is the time-matched mean difference in corrected QT interval using Fridericia’s correction method, or QTcF, for PF-04447943 versus placebo at 0.5 hour post-dose. QT is an ECG interval; a correction method accounts for the relationship between the interval and heart rate. Fridericia’s method is the correction specified for this endpoint.

The registry reports least-squares mean QTcF values for each arm and least-squares mean differences for the two dose-placebo comparisons. A treatment contrast expresses the estimated difference in average QTcF between the specified dose and placebo at the stated time. The endpoint is relevant to drug development because QTc is an ECG measure used to assess potential effects on cardiac electrical timing. That general purpose does not, by itself, establish the clinical importance of any observed difference.

The population definition matters when interpreting the summaries: it includes randomized and treated participants with at least 1 post-dose ECG measurement in at least 1 period. The reported endpoint is specific to the 0.5-hour measurement; it should not be read as a summary of every post-dose time or as a description of longer-term outcomes. The registry record does not report a clinical threshold for interpreting the observed differences.

7. Results Posted in the Registry

• PF-04447943 25 mg: mean 404.09 versus 404.38 on control; difference -0.29 (95% CI -2.95 to 2.37), standardized effect d = -0.05; z-test p = 0.831.
• PF-04447943 100 mg: mean 408.92 versus 404.38 on control; difference 4.54 (95% CI 1.86 to 7.22), standardized effect d = 0.73; z-test p = <0.001.

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.

Number of Participants With Physical Examination Abnormalities

• 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:
  • PF-04447943 25 mg: 2.3% (1/43) versus 4.8% (2/42) on control; risk difference -2.4 percentage points (95% CI -10.3 to 5.4), risk ratio 0.49 (0.05–5.19); unadjusted two-proportion z-test p = 0.543.
  • PF-04447943 100 mg: 2.4% (1/42) versus 4.8% (2/42) on control; risk difference -2.4 percentage points (95% CI -10.3 to 5.5), risk ratio 0.50 (0.05–5.31); unadjusted two-proportion z-test p = 0.557.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between PF-04447943 25 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 PF-04447943 100 mg and the control. The interval gives the range of effects that remain compatible with the data.

Number of Participants With Laboratory Test Abnormalities

• 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:
  • PF-04447943 25 mg: 39.5% (17/43) versus 42.9% (18/42) on control; risk difference -3.3 percentage points (95% CI -24.2 to 17.6), risk ratio 0.92 (0.55–1.53); unadjusted two-proportion z-test p = 0.756.
  • PF-04447943 100 mg: 28.6% (12/42) versus 42.9% (18/42) on control; risk difference -14.3 percentage points (95% CI -34.6 to 6.0), risk ratio 0.67 (0.37–1.21); unadjusted two-proportion z-test p = 0.172.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between PF-04447943 25 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 PF-04447943 100 mg and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes broaden the record beyond QTcF, covering other ECG measures, heart rate, adverse events, physical examination findings, laboratory tests, and vital signs. The registry lists 20 outcome measures and 72 analyses, while 12 secondary measures are posted. This range helps describe what else was assessed, but it does not turn the primary result into a comprehensive account of every possible effect.

The posted secondary analyses include some baseline physical examination findings and laboratory abnormalities. Their estimates and uncertainty vary, and the available binary comparisons are based on relatively sparse counts. In general, a secondary finding can help contextualize the primary endpoint, but it is not an independent confirmation of the primary analysis unless the analysis and interpretation were planned and appropriately controlled. The registry record does not report a multiplicity adjustment for the secondary comparisons.

When many outcomes or comparisons are examined, the chance of at least one apparently notable result can increase. Unadjusted secondary p-values therefore carry less weight than a prespecified primary analysis, particularly when the record does not describe a multiplicity strategy. The secondary data should be treated as descriptive or exploratory unless the registry record establishes a confirmatory role. They do not replace the primary QTcF comparison, and differences across outcome types should not be combined into a single overall conclusion.

The registry’s stated primary population is ECG-based, whereas secondary measures can have their own relevant measurement windows and available participants. A finding on one measure should not be assumed to apply to another measure or time. Taken together, the secondary outcomes provide context for the scope of monitoring, but do not support a general claim about all safety or physiological effects.

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

• Time-Matched Mean Difference in Corrected QT Interval Using Fridericia's Correction Method (QTcF) for Moxifloxacin and Placebo
• Time-Matched Mean Difference in Heart Rate for PF-04447943 and Placebo
• Time-Matched Mean Difference in PR Interval for PF-04447943 and Placebo
• Time-Matched Mean Difference in QRS Interval for PF-04447943 and Placebo
• Number of Participants With Treatment-Emergent Adverse Events (AEs) and Serious Adverse Events (SAEs)
• Number of Participants With Physical Examination Abnormalities
• Number of Participants With Electrocardiogram (ECG) Abnormalities
• Number of Participants With Vital Sign Abnormalities

9. Statistical Interpretation

10. Statistical Methods Explained

The primary endpoint is continuous: QTcF is measured on a numerical scale, and the comparison is expressed as a mean difference. The registry identifies the reported arm summaries as least-squares means and the sponsor’s treatment contrasts as least-squares mean differences. Least-squares means are model-based averages; they can account for features of the analysis model rather than being simple arithmetic averages alone. The exact model specification beyond the registry’s description is not reported in the registry record.

What does the crossover design change?

In general, a crossover study assigns treatments across periods so that participants can contribute measurements under more than one treatment. This can make within-participant comparisons informative because each participant may serve as their own reference. In such a design, the analysis needs to address the structure of repeated observations, and may need to account for period and sequence features. The registry identifies a repeated-measures model for the primary comparisons, but details such as model terms, treatment sequence handling, carryover assumptions, or missing-period handling are not reported in the registry record. The simple recalculation is not a reconstruction of that model.

What does a mean difference tell the reader?

A mean difference is the estimated contrast between the average outcome under one treatment and the average under the comparator. A value below zero means the estimated mean in the named treatment arm is lower than the comparator; a value above zero means it is higher. For the primary outcome, the registry’s estimated contrast is -0.29 for 25 mg versus placebo and 4.54 for 100 mg versus placebo. Those are the reported model-based differences, not a percentage change or a measure of individual response.

Why are the recalculations labelled unadjusted?

The recalculated contrasts use the posted arm-level means and standard errors to produce simple comparisons. Their standard deviations are approximated from those standard errors, as indicated by the recalculation method. This approach does not use participant-level data and does not reproduce the sponsor’s repeated-measures analysis of a crossover trial. It is therefore a descriptive check rather than a substitute for the reported model. The two approaches produce the same stated point estimates for the primary contrasts but different uncertainty estimates and p-values. Since the record flags a discrepancy, the sponsor’s own analysis takes precedence.

How should the p-values be read?

A p-value describes how incompatible the observed data are with a specified null hypothesis under the assumptions of the analysis. It is not the probability that the null hypothesis is true, and it does not measure the size or importance of a difference. The sponsor reports p=0.820 for the 25 mg comparison and p< 0.001 for the 100 mg comparison. The recalculated unadjusted values are p=0.831 and p < 0.001, respectively. These inferential results are not identical, which is why the registry’s model-based results should guide interpretation.

Why does over-dispersion matter in general?

For outcomes that count events or classify participants, a simple model may assume that variation follows a particular distribution. Over-dispersion means the observed variability is greater than that assumption allows. When present, it can make standard errors too small and confidence intervals too narrow if not handled appropriately. The primary QTcF endpoint is continuous rather than a count endpoint, so this general issue is not a reason to reinterpret its reported result. The record does not give enough detail to assess dispersion assumptions for all secondary analyses.

The statistical methods and reporting details available for the primary endpoint are limited to the stated repeated-measures model, least-squares mean differences, confidence intervals, and p-values. The model’s full specification, analysis diagnostics, and missing-data procedure are not reported in the registry record. These omissions limit independent assessment of the model assumptions and of how all available repeated measurements contributed to the estimates.

11. Confidence Intervals and Precision

A confidence interval describes a range of values that are reasonably compatible with the data and the statistical model at its stated confidence level. It conveys uncertainty around an estimate; it does not guarantee that the true value lies within the particular interval. The sponsor reports 90% confidence intervals for both primary comparisons.

For 25 mg versus placebo, the reported difference is -0.29 milliseconds and the 90% interval runs from -2.38 to 1.80. The interval spans zero, so the data are compatible with a small difference in either direction under the sponsor’s analysis. The estimate is not precise enough to establish that the treatment and placebo means are exactly the same, and the result should not be described as proof of no effect.

For 100 mg versus placebo, the reported difference is 4.54 milliseconds and the 90% interval runs from 2.44 to 6.64. This interval is above zero, indicating a positive estimated difference in the sponsor’s model-based analysis. The interval also shows that the estimate is not a single known quantity: a range of effect sizes remains compatible with the data and model.

The unadjusted recalculated intervals are wider: -2.951 to 2.371 for 25 mg and 1.864 to 7.216 for 100 mg. They are not the sponsor’s model-based intervals and should not replace them, particularly because the registry flags a discrepancy. Confidence intervals address statistical uncertainty, not clinical importance. The registry record does not state a threshold that would distinguish a clinically important QTcF change from a smaller one.

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 illustrations translate an observed contrast into planning examples under specified assumptions. For the continuous primary endpoint and the binary secondary outcomes posted with planning illustrations, the figures are participants per arm. The record says time-to-event illustrations would give total events, but the outcomes are not presented as time-to-event planning rows.

The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each row uses an effect equal to the stated fraction of the observed effect. The 80% power column represents the illustrative number of participants per arm for an 80% chance of detecting that assumed effect under those planning assumptions, if the assumptions hold. The 90% power column has the same interpretation for a 90% chance. These are planning illustrations, not observed power.

For the 25 mg QTcF comparison, the row using 100% of the observed effect lists 7,308 participants per arm for 80% power and 9,784 for 90% power. The 75% row lists 12,992 and 17,393; the 50% row lists 29,232 and 39,134. For the 100 mg comparison, the corresponding rows list 30 and 40; 54 and 71; and 120 and 160. These examples illustrate how strongly the estimates depend on the observed effect size. Smaller effects need more participants to detect under otherwise identical assumptions.

The original sample-size justification and design assumptions for the trial are not reported in the registry record. These illustrations are based on the observed estimates and should not be read as the trial’s planned sample size or as a retrospective statement about its power. The planning rows are sensitive to the observed effect, which may itself be uncertain, especially when estimated from a small study.

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 = -0.290; pooled standard deviation σ = 6.257.
2. n = 2 × (1.96 + 0.842)² × 6.257² / -0.290² = 7,307.9, rounded up to 7,308 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 <- -0.29000000000002                      # observed difference in means
sigma <- 6.25715471650375                     # 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(2785770)
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 = 7,308 (80% power), 9,784 (90% power)
 75% of observed effect: n per arm = 12,992 (80% power), 17,393 (90% power)
 50% of observed effect: n per arm = 29,232 (80% power), 39,134 (90% power)
power.t.test: 7308.9 
Simulated power at n = 7,308 per arm: 0.801 (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. Analyses Reported by the Sponsor

The sponsor reports the primary contrasts using a repeated-measures model. For 25 mg versus placebo, the reported least-squares mean difference is -0.29, with a 90% confidence interval from -2.38 to 1.80 and p=0.820. For 100 mg versus placebo, the reported difference is 4.54, with a 90% confidence interval from 2.44 to 6.64 and p< 0.001.

The recalculation from posted arm summaries gives the same point estimates, but different intervals and p-values: -2.951 to 2.371 and p=0.831 for 25 mg, and 1.864 to 7.216 and p < 0.001 for 100 mg. These recalculations are simple unadjusted comparisons and do not reproduce the crossover repeated-measures model. They should therefore be read as a check on the posted summaries, not as an alternative primary analysis.

The registry flags a discrepancy between the sponsor analysis and the recalculation. The sponsor’s own analysis takes precedence. Its model accounts for the study’s repeated-measures structure in a way the simple recalculation does not. The available record does not include the full model specification or participant-level data, so an independent reproduction of that analysis is not possible from the posted summaries alone.

15. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when an analysis considers several outcomes, treatment contrasts, time points, or other comparisons. Even when each test uses a conventional significance threshold, examining many tests can increase the chance of at least one result appearing statistically notable by chance. Statistical methods for addressing this can include prespecified hierarchies or adjustments, but the appropriate approach depends on the study’s objectives and analysis plan.

The registry lists 20 outcome measures and 72 analyses, and 12 secondary measures are posted. The primary outcome itself includes two dose-placebo comparisons. The multiplicity strategy for these comparisons and the secondary analyses is not reported in the registry record. Accordingly, the reported p-values should be read as the results for their respective analyses, not as evidence that a family of comparisons has been controlled for multiplicity.

The primary ECG analysis population is defined as randomized and treated participants who had at least 1 post-dose ECG measurement in at least 1 period. The primary arm summaries report 43 participants for 25 mg and 42 for 100 mg and placebo. These figures describe the posted endpoint summaries; the record does not provide a full participant flow or explain every difference between enrollment and endpoint analysis counts.

Missing-data handling, including how incomplete periods or absent ECG measurements were treated in the repeated-measures model, is not reported in the registry record. The record also does not report sensitivity analyses for missing data or model assumptions. These details matter because exclusions and incomplete repeated measurements can affect which observations contribute to an estimate. The trial’s original sample-size justification and interim-analysis plan are also not reported in the registry record. No assumptions about those features should be made from the endpoint results alone.

16. 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 5% reporting threshold.

Most frequent non-serious adverse event terms

The posted serious adverse event summaries report zero serious adverse events in each listed treatment group and zero deaths. The numbers at risk for serious adverse events are 43 for PF-04447943 25 mg, 42 for PF-04447943 100 mg, 41 for moxifloxacin 400 mg, and 42 for placebo. These are the registry’s reported counts for the listed groups.

Zero reported serious events in these groups means that no serious adverse events were recorded in the posted summaries; it does not establish that the risk is absent. A study of this size and duration may provide limited information about uncommon events, and the registry record does not report a precision estimate for these safety counts. The summaries also list other adverse events affecting 7 participants in the 25 mg group, 3 in the 100 mg group, 3 in the moxifloxacin group, and 2 in the placebo group. Those counts are separate from the serious adverse event figures.

Safety findings should be interpreted in the context of this completed Phase 1 study in healthy adults, its listed treatment periods, and the reporting available in the registry. The posted summaries do not establish what might occur in other populations or with different treatment durations.

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

The primary result is a comparison of average QTcF at one specified post-dose time, not a statement about all ECG measurements or all possible effects of treatment. The sponsor’s repeated-measures analysis estimates a difference of -0.29 milliseconds for 25 mg versus placebo and 4.54 milliseconds for 100 mg versus placebo. Its reported intervals and p-values indicate different levels of statistical evidence for these two comparisons: the 25 mg interval includes zero, whereas the 100 mg interval is above zero.

Those results do not establish that the 25 mg dose and placebo are equivalent. An interval that includes zero can still include nonzero differences, and the reported interval does. Nor does a statistically clear difference for 100 mg, by itself, establish that the difference is clinically important. The registry record does not report a clinical interpretation threshold for QTcF.

The endpoint is an ECG measurement in healthy adult subjects, assessed at 0.5 hour post-dose. It does not directly measure symptoms, clinical events, or outcomes in patients with a condition. The study’s results should not be generalized automatically to other groups or treatment settings. The record identifies moxifloxacin as a positive control, but the primary comparisons discussed are PF-04447943 versus placebo.

Finally, the simple recalculations do not replace the sponsor’s repeated-measures results. The trial uses a crossover design, and the recalculations are unadjusted summaries rather than reproductions of that model. The discrepancy is flagged in the registry record, so the sponsor’s analysis takes precedence. Overall interpretation should consider the endpoint’s narrow scope, the uncertainty intervals, the analysis population, and the limitations in the available reporting.

18. Why This Trial Matters Statistically

This study illustrates how a clinical trial’s design and analysis affect interpretation even when the endpoint is a straightforward mean difference. QTcF was measured under multiple treatment conditions in a randomized crossover study, and the primary analysis used a repeated-measures model. Those features make the sponsor’s model-based estimate more relevant than a simple comparison of arm summaries.

It also demonstrates why a point estimate alone is not enough. The 25 mg contrast is close to zero, but its confidence interval covers values in both directions. The 100 mg contrast is positive and its sponsor-reported interval lies above zero. Precision and the range of compatible effects are essential to understanding what each estimate can support. A statistically notable result does not automatically indicate clinical importance, and an imprecise result does not demonstrate no effect.

The record also makes visible the distinction between trial results and planning illustrations. The planning examples use observed effects and stated assumptions; they do not recover the original design rationale or constitute observed power. This is especially important when observed effects are uncertain and smaller target effects would require larger planning samples.

Finally, the registry lists many outcome measures and analyses, while not reporting a multiplicity strategy or several analysis details. That context affects how much weight to give secondary results and how independently the primary model can be evaluated. The study is useful as an example of reading a time-specific ECG endpoint alongside its design, model, uncertainty, and reporting limits.

19. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The results concern healthy adult subjects in a Phase 1 study and a time-matched QTcF measurement at 0.5 hour post-dose. They do not describe effects in patients or outcomes beyond the measures and follow-up reported. The primary endpoint summaries include 43 participants for the 25 mg arm and 42 for each of the 100 mg and placebo arms, while enrollment is listed as 44. The registry defines the ECG analysis population but does not provide a complete participant flow explaining all available observations.

The reported analysis is identified as a repeated-measures model, but its full specification is not reported in the registry record. The treatment sequence details, period and carryover handling, model diagnostics, and missing-data procedures are also not reported. This limits independent assessment of the model and of how incomplete crossover measurements were handled.

The recalculated intervals and p-values differ from the sponsor’s reported results. The record flags this discrepancy, and the sponsor’s analysis takes precedence because the recalculations are simple unadjusted comparisons that do not reproduce the repeated-measures analysis. Participant-level data are not available in the posted summaries to independently reproduce the sponsor model.

The registry lists many outcome measures and analyses, but the multiplicity approach is not reported. Secondary results should therefore be interpreted cautiously. The sample-size planning illustrations are based on observed effects and stated assumptions, not the trial’s original design justification. The original assumptions and any interim-analysis plan are not reported in the registry record. Finally, the record does not provide a clinical threshold for interpreting the size of a QTcF difference, limiting conclusions about clinical importance.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Study To Evaluate The Effect Of PF-04447943 On Qtc Interval In Healthy Adult Subjects — NCT02785770.
• Registry record last updated 2018-12-28; status last verified 2018-06.
• Results first posted 2018-12-28.
• 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.

23. Record Summary

NCT02785770 is a completed Phase 1 randomized, four-treatment, four-period crossover study in healthy adult subjects. The listed treatments were PF-04447943 25 mg, PF-04447943 100 mg, matching placebo, and moxifloxacin 400 mg. Enrollment was 44.

The primary endpoint was the time-matched mean difference in QTcF at 0.5 hour post-dose. The sponsor’s repeated-measures analysis reported a difference of -0.29 for 25 mg versus placebo and 4.54 for 100 mg versus placebo. The corresponding 90% confidence intervals were -2.38 to 1.80 and 2.44 to 6.64. Because the registry flags a discrepancy with the simple recalculations, the sponsor’s analysis takes precedence.

The posted serious adverse event summaries report zero serious adverse events and zero deaths in each listed group. The registry does not report the full primary model specification, missing-data handling, multiplicity strategy, or original sample-size justification.