Acetaminophen for the Reduction of Oxidative Injury in Severe Sepsis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Acetaminophen for the Reduction of Oxidative Injury in Severe Sepsis (NCT01739361), a randomized phase 2 study. The primary outcome is F2-isoprostanes After 72 Hours of Acetaminophen or Placebo.

2. Clinical Question

3. Trial Design

The registry identifies this as a phase IIa randomized controlled trial with a parallel-group design. Participants were assigned to acetaminophen or placebo. Masking is recorded as quadruple, although the record does not describe which roles were masked or how masking was implemented. The study was sponsored by Vanderbilt University and is listed as completed. The registry reports enrollment of 44. The trial start is recorded as 2013-04, with primary completion in 2013-12.

The treatment regimen was acetaminophen at 1 gram by mouth or by enteral feeding tube every six hours for 72 hours. The placebo group received placebo by either route on the same schedule and for the same duration. The primary measurement was scheduled at 72 hours after randomization. The record describes follow-up for in-hospital mortality through the end of the hospital stay, with an average stay of 5 weeks; that duration is the registry's description, not a fixed follow-up time for every participant.

The posted primary group summaries include 18 participants in the acetaminophen group and 22 in the placebo group. Those are the denominators attached to the primary endpoint summaries; the record does not explain the difference between these analysis denominators and the overall enrollment figure. It also does not describe the allocation procedure, concealment, or specific masking procedures in further detail. Details about the trial's sample-size justification and whether any interim analyses were planned are not reported in the registry record. The trial design therefore establishes a randomized comparison, but the registry does not provide all operational details needed to reconstruct how assignment and follow-up were conducted.

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

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 F2-isoprostanes After 72 Hours of Acetaminophen or Placebo, measured over 72 hours after randomization.

The primary endpoint was F2-isoprostanes measured 72 hours after randomization, with results expressed in pg/mL. The registry identifies F2-isoprostanes as a marker of oxidative stress, specifically lipid peroxidation. The trial's stated primary question was whether this marker would be lower among participants assigned to acetaminophen than among those assigned to placebo.

The endpoint is a laboratory measure linked to the biological process under study. It is not itself a clinical outcome such as acute kidney injury, length of stay, or mortality. A change in a marker can help assess whether a treatment is associated with a change in a targeted process, while not establishing by itself whether patients experience a corresponding clinical benefit. The registry describes acetaminophen's proposed role in relation to hemoprotein-mediated lipid peroxidation, but a proposed mechanism is a rationale for testing, not evidence that the mechanism changed in this trial.

The endpoint is summarized by medians and interquartile ranges. A median describes the central position of observed values, while the interquartile range describes the spread of the middle portion of those values. The record reports the group medians and interquartile-range bounds, but not an estimated between-group difference with a confidence interval. The sponsor's reported comparison uses the Wilcoxon (Mann-Whitney) method.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary comparison used the Wilcoxon (Mann-Whitney) method, a rank-based approach for comparing observations from two groups. Rather than requiring the measurements to follow a particular bell-shaped distribution, this method uses the relative ordering of observations. It is often useful for continuous outcomes when data may be skewed or when the analysis is framed around distributions or ranks. The registry reports this method for F2-isoprostanes, whose group summaries are medians and interquartile ranges.

What do the medians and interquartile ranges describe?

A median is the middle value when observations are ordered. It gives a summary of the center that is less affected by unusually high or low observations than a mean can be. The interquartile range describes the spread between the lower and upper quartiles, covering the middle portion of the distribution. These summaries communicate central tendency and variability, but they do not by themselves quantify the uncertainty around a between-group effect. In particular, overlap between interquartile ranges is not a formal test of whether groups differ.

The Wilcoxon (Mann-Whitney) test assesses whether the observed ranks in one group tend to be higher or lower than those in the other, under its null model. Its p-value is not a direct estimate of how far apart the group medians are. Interpreting the test specifically as a test of medians requires assumptions about the shapes of the distributions; when shapes differ, the test can reflect broader distributional differences. The registry does not report a rank-based effect estimate or a confidence interval for the primary comparison.

How should the p-value be read?

A p-value describes how unusual data at least as incompatible with the null model as the observed data would be, assuming that model and the analysis assumptions hold. It does not give the probability that treatment has no effect, and it does not measure the probability that the finding is due to chance. Nor does it quantify whether a difference is important to patients. The reported p = 0.353 should therefore be read alongside the observed group summaries and the absence of a reported primary-effect confidence interval, not as proof of equality or proof of a treatment effect.

What does a confidence interval add?

In general, a confidence interval gives a range of effect estimates compatible with the data and a specified procedure. Its width conveys precision: a wider interval generally reflects greater uncertainty. The registry does not report a confidence interval for the primary F2-isoprostane comparison. For the secondary mortality result, the posted unadjusted recalculation includes an interval for the risk difference. Such an interval is a measure of uncertainty around that recalculated estimate, not a guarantee that the true value lies within the particular interval or a replacement for a trial's prespecified analysis.

How do the binary outcome calculations differ?

In-hospital mortality is a binary outcome: each participant either had the recorded event or did not. The registry lists group counts, and a simple unadjusted recalculation expresses the contrast as a risk difference. A risk difference compares the event proportions between groups; it is an absolute contrast, unlike a relative measure such as a risk ratio. The posted recalculation is not an adjusted estimate and does not account for additional participant characteristics. The registry does not report a sponsor analysis for this secondary endpoint.

Small numbers of events can make binary-outcome estimates imprecise. A confidence interval may include values consistent with benefit, little difference, or harm, and a single point estimate cannot resolve that uncertainty. The recalculated result should thus be treated as descriptive and exploratory, particularly because the registry record does not report an analysis plan for secondary outcomes or a multiplicity adjustment.

What do the planning illustrations mean?

The planning illustrations use the observed mortality effect as their starting point and show hypothetical participants per arm for 80% power and 90% power under stated assumptions. They consider planning for 100% of the observed effect, 75% of the observed effect, and 50% of the observed effect. The assumptions are a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation. These figures are planning illustrations, not observed power and not the trial's own sample-size justification. Smaller effects require larger planned group sizes under the stated framework. The illustrations do not establish that the observed effect is a reliable estimate of the effect a future study should expect.

More generally, power is a design property conditional on assumptions about the effect to detect, variability or event rates, significance threshold, and analysis. It is not a post-study measure of whether a result is true. A planning calculation based on the observed effect may be especially unstable when the trial has relatively few outcome events, because the observed estimate can vary substantially. The registry's stated planning assumptions should be kept distinct from unreported design decisions.

10. Analyses Reported by the Sponsor

For the primary endpoint, the registry reports the sponsor's Wilcoxon (Mann-Whitney) comparison and its p-value of 0.353. The primary group summaries are medians with interquartile-range bounds; no separate recalculated primary effect estimate or interval is reported. The reported test is therefore the main inferential result for F2-isoprostanes in the registry.

For in-hospital mortality, the registry posts event counts but no sponsor-reported comparison. The risk difference and associated interval displayed with that outcome are an unadjusted recalculation from the posted summary statistics. They should not be mistaken for a prespecified or adjusted sponsor analysis. A simple recalculation can provide a useful check or descriptive contrast, but it cannot recover details of a full analysis plan that the record does not describe.

The registry does not identify a discrepancy between the posted information and its recalculation. Even so, the distinction in analysis source matters: the primary inference comes from the reported rank-based test, while the mortality contrast is an unadjusted recalculation. The two results address different outcomes and should not be treated as interchangeable evidence.

11. Multiplicity, Missing Data and Analysis Populations

The registry identifies one primary endpoint and lists in-hospital mortality and serum creatinine after treatment as secondary outcomes. It does not describe how the primary and secondary hypotheses were prioritized statistically beyond the stated primary and secondary labels. A plan for controlling multiplicity, such as adjusting significance thresholds across multiple comparisons or using a prespecified testing sequence, is not reported in the registry record.

Multiplicity matters because examining multiple outcomes can increase the chance of finding a result that appears noteworthy even when no true effect is present. This is a general statistical concern, not evidence that any particular result in this trial is false. When adjustment procedures and outcome hierarchies are not reported, secondary comparisons are best treated as exploratory and interpreted with attention to their uncertainty.

Handling of missing data and the definition of analysis populations are not reported in the registry record. The primary summaries include 18 participants assigned to acetaminophen and 22 assigned to placebo, while the record reports enrollment of 44. The registry does not explain how those endpoint denominators were determined, whether measurements were unavailable, or how any such cases were handled. It would therefore be inappropriate to assume a particular missing-data method or to infer why the denominators differ from enrollment.

The registry does not report whether the primary analysis followed intention-to-treat principles, used a per-protocol population, or applied another population definition. It also does not report whether analyses were adjusted for baseline characteristics, how protocol deviations were handled, or whether the primary analysis was prespecified in a detailed statistical plan. Interim analyses and stopping rules are not reported in the registry record. The trial's sample-size justification is also not reported; the separate planning illustrations based on the observed mortality effect are not evidence of the original design rationale.

These omissions limit the ability to reconstruct the complete inferential framework. They do not demonstrate that the analyses were conducted improperly; they mean that the registry record does not provide those details. The primary test and posted outcome summaries can be described as recorded, while broader claims about analysis conduct or confirmatory status remain unsupported.

12. 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 0% reporting threshold.

Most frequent serious adverse event terms

Serious events with the highest arm-level rate: ast or alt >400 u/l (9.5%).

The registry reports serious adverse events among participants at risk in each group. In the acetaminophen group, 2 participants were recorded with serious adverse events among 21 at risk; in the placebo group, 1 was recorded among 23 at risk. Other adverse events are recorded as 0 in each group, with 21 at risk in the acetaminophen group and 23 at risk in the placebo group. These are event summaries, not estimates of comparative safety adjusted for differences between participants or exposure.

The listed serious adverse event term is “AST or ALT >400 U/L,” with a reported value of 9.5%. The registry entry does not specify an arm breakdown for that term in the posted information, so it should not be assigned to either group. Death counts in the adverse-event record are not reported in the registry record; in-hospital mortality is listed separately as a secondary outcome.

These figures are descriptive. The small event counts and lack of detail about event timing, severity grading, attribution, and the term's group distribution limit interpretation. They do not by themselves establish a safety difference between acetaminophen and placebo.

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

The primary result describes F2-isoprostane values 72 hours after randomization in the two randomized groups. The median was lower in the acetaminophen group than in the placebo group, while the reported Wilcoxon comparison did not provide statistically clear evidence of a difference. This is a measured result for the laboratory marker at the stated time point, not a direct demonstration of a clinical outcome.

The result does not establish that the treatments have identical effects. A lack of statistically clear evidence can reflect a small or uncertain difference, and the registry does not report a primary-effect confidence interval to quantify the range of values compatible with the analysis. Conversely, the lower observed median does not by itself establish that acetaminophen caused a reduction in oxidative injury. Group summaries, a p-value, and an endpoint's biological rationale do not prove a treatment mechanism or patient benefit.

The clinical outcomes require separate consideration. The registry lists in-hospital mortality and serum creatinine after treatment as secondary outcomes. Mortality has posted counts and an unadjusted recalculation, but this is exploratory and uncertain; a creatinine result is not reported in the registry record. The primary laboratory finding cannot be used as a substitute for results on these clinical outcomes.

The trial's findings apply to the population, treatment schedule, comparator, endpoint, and follow-up described in the registry. The record does not supply enough detail about eligibility, missing-data handling, analysis populations, or the full statistical plan to extend interpretation beyond the posted summaries. The findings are best understood as evidence about the observed comparison in this trial, rather than a general conclusion about all patients with severe sepsis or all uses of acetaminophen.

14. Why This Trial Matters Statistically

This trial raises a useful statistical distinction between a mechanistic laboratory endpoint and outcomes that matter directly to patients. F2-isoprostanes were chosen as a marker of oxidative stress and lipid peroxidation, allowing the primary comparison to focus on the biological process described in the trial rationale. That choice makes the endpoint relevant to the stated hypothesis, but it also means that a result on the marker cannot automatically be interpreted as a result on mortality or other clinical outcomes.

The study also illustrates why both effect summaries and uncertainty matter. The group medians describe the observed central values, while the Wilcoxon test evaluates a rank-based comparison. Neither element alone gives a complete account of the magnitude and precision of a treatment effect. The record's absence of a primary-effect confidence interval limits assessment of that precision, even though the reported group summaries and p-value remain informative.

Secondary outcomes demonstrate a further challenge: clinical outcomes may be sparse and uncertain, and multiple outcomes raise questions about how confirmatory claims should be controlled. The posted mortality recalculation is useful as an exploratory description, but it does not substitute for a prespecified sponsor analysis. The unreported creatinine result and unreported multiplicity plan also constrain the overall picture.

Finally, the planning illustrations show how future sample-size requirements depend on the effect a study aims to detect. Because these illustrations use the observed mortality effect and stated assumptions, they are not observed power or proof that the original trial was adequately sized for its primary endpoint. The design rationale is not reported in the registry record. Separating planning assumptions, observed summaries, and inferential tests is essential to avoid overstating what this trial can establish.

15. Limitations

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

Several limitations specific to this registry record affect interpretation. The primary endpoint is a single laboratory measure at 72 hours after randomization. The record does not report serial primary measurements or a primary between-group effect estimate with a confidence interval, so the time course and precision of the primary contrast cannot be assessed from the posted analysis.

The record reports the primary medians and interquartile-range bounds but does not provide additional distributional detail. The group summaries do not show every participant's value, and the Wilcoxon test result does not quantify the size of a difference in the marker. The analysis population and handling of unavailable measurements are not reported in the registry record. Although enrollment is listed as 44, the primary summaries use group denominators of 18 and 22; the record does not explain that relationship.

The secondary outcome reporting is incomplete. In-hospital mortality is posted, but the record does not report a sponsor comparison for it; the available contrast is an unadjusted recalculation. Serum creatinine after treatment is listed, but its result is not reported in the registry record. The relationship between the posted adverse-event summaries and the separate mortality outcome is not explained in detail.

Other details relevant to interpretation are also unavailable in the registry record, including eligibility criteria beyond the stated population, allocation and masking procedures, the statistical analysis plan, multiplicity handling, interim analyses, missing-data methods, and the sample-size justification. These absences do not establish that such procedures were not used; they limit what can be assessed from the record. The trial's phase IIa designation and randomized design provide context, but neither resolves these reporting limitations.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Acetaminophen for the Reduction of Oxidative Injury in Severe Sepsis — NCT01739361.
• Registry record last updated 2017-12-26; status last verified 2017-11.
• Results first posted 2015-03-23.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

The registry describes a completed, randomized, quadruple-masked, parallel phase IIa trial sponsored by Vanderbilt University. It compared acetaminophen with placebo in adults with severe sepsis and detectable cell-free hemoglobin. Acetaminophen was given at 1 gram every six hours for 72 hours, by mouth or enteral feeding tube; the placebo group followed the same schedule.

The primary endpoint was F2-isoprostanes at 72 hours after randomization. The acetaminophen group had a median of 33.4 pg/mL and the placebo group 40.15 pg/mL. The reported Wilcoxon (Mann-Whitney) comparison had p = 0.353. The record also lists in-hospital mortality and serum creatinine after treatment as secondary outcomes. Mortality is posted with counts and an unadjusted recalculation; a creatinine result is not reported in the registry record. Details of the full statistical plan, missing-data handling, and multiplicity procedures are not reported in the registry record.