The Prevention of Erythropoietin on Cardiac Surgery-associated Acute Kidney Injury: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for The Prevention of Erythropoietin on Cardiac Surgery-associated Acute Kidney Injury (NCT03007537), a randomized na study. The primary outcome is Number of Participants With Acute Kidney Injury.

2. Clinical Question

The registry identifies acute kidney injury as the primary outcome and describes it as occurrence according to the KDIGO guideline. The posted results count 15 participants with the event in the control group and 9 in the erythropoietin group; each group has a reported denominator of 46. These counts describe the observed outcome in the posted groups. The comparison does not by itself establish that the intervention caused any difference.

The registry also lists renal replacement therapy as a secondary outcome, but it does not post results for that outcome. The record does not give details of eligibility criteria, the clinical characteristics of participants, or the specific population included in the primary-outcome analysis. Those omissions limit how precisely the findings can be applied to a particular group of patients or clinical setting.

3. Trial Design

The registry describes a randomized, double-masked, parallel-group trial. Participants were assigned to an erythropoietin group or a control group. The intervention descriptions specify one subcutaneous injection, given 1day before cardiac surgery: 10000 IU erythropoietin or 0.9% sodium chloride 1ml. The registry does not identify which people were masked, how the random sequence was generated, or how allocation was concealed.

The record lists enrollment as 101 and gives reported primary-outcome denominators of 46 for each group. These are different registry fields: the enrollment figure should not be treated as the denominator for the posted comparison. The record does not explain the relationship between the enrollment figure and the outcome denominators, or describe exclusions, withdrawals, or the analysis population.

The study is listed as terminated. Its start date is 2015-10-01 and its primary completion date is 2018-12-28. The reason for termination is not reported in the registry record. The phase is listed as NA. The sponsor is Nanjing First Hospital, Nanjing Medical University.

Acute kidney injury was assessed over a 7 days timeframe, according to the KDIGO guideline. The registry lists renal replacement therapy as a secondary outcome, but does not post its results. The record includes no detailed account of the trial protocol, prespecified statistical analysis, or planned sample-size justification. Consequently, the design features that are explicitly recorded can be described, but other design and conduct details cannot be inferred.

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 Number of Participants With Acute Kidney Injury, measured over 7 days.

The primary endpoint is the number of participants with acute kidney injury during the 7 days after cardiac surgery. The registry defines the event as occurrence according to the KDIGO guideline. It is a binary outcome: each participant is classified according to whether the event occurred during the stated timeframe. The posted group results give event counts and denominators rather than a measure of the severity or duration of kidney injury.

General: A binary clinical endpoint can make an outcome comparison straightforward to interpret because it focuses on whether an event occurred within a defined period. The event definition and observation window are important: changing either can change what the endpoint captures. Here, the registry names the guideline and timeframe, but does not provide additional operational details about how events were identified or adjudicated.

Acute kidney injury is relevant to the trial’s stated focus on cardiac surgery-associated kidney injury. The record’s stated objective is to examine prevention of this outcome and provide evidence about renal function and prognosis after surgery. That stated objective does not establish that a difference in the endpoint occurred or that any observed difference was caused by the assigned intervention. The registry lists renal replacement therapy as a secondary outcome, but the result for that measure is not reported in the registry record.

7. Results Posted in the Registry

• Erythropoietin Group: 19.6% (9/46) versus 32.6% (15/46) on control; risk difference -13.0 percentage points (95% CI -30.8 to 4.7), risk ratio 0.60 (0.29–1.23); unadjusted two-proportion z-test p = 0.154.

8. Statistical Interpretation

9. Statistical Methods Explained

General: What does a risk difference express? A risk difference compares the proportion of participants with an event in one group with the proportion in another. It is expressed on an absolute scale, which can help readers understand the size and direction of a contrast. In this trial, the comparison is erythropoietin versus control. The posted estimate points toward fewer events in the erythropoietin group. Its confidence interval crosses the value corresponding to no difference, so the estimate is uncertain. No additional risk-difference value is needed to interpret that direction and uncertainty.

General: What does an unadjusted comparison mean? An unadjusted comparison uses the reported group outcome counts without accounting for other variables. It is useful as a direct description of the observed contrast, but it does not correct for possible group imbalances or other features of trial conduct. Randomization is intended to make groups comparable, but the registry record does not post baseline characteristics or enough information about allocation and analysis to evaluate those issues. The unadjusted result should therefore be read as a summary comparison, not as a complete account of the trial’s analysis.

General: How should the p-value be read? The posted unadjusted p-value is 0.154. A p-value describes the compatibility of data with a specified statistical model under its assumptions; it is not the probability that a treatment effect is true, nor is it a measure of the size or clinical importance of an effect. The registry record does not state the trial’s original testing plan, including whether this was its prespecified primary analysis. The p-value is consequently informative as part of the posted recalculation, but its role in the sponsor’s planned analysis is not reported in the registry record.

General: What does the confidence interval add? A confidence interval describes uncertainty around an estimate under a statistical procedure. In repeated use under the procedure’s assumptions, intervals at a stated confidence level have a defined coverage property. A confidence interval is not a range containing a known fixed effect with a specified probability after the data have been observed. Here, the reported risk-difference interval includes no difference. That means the posted data do not sharply distinguish the observed direction from no difference, and values on the other side of no difference also remain compatible with the interval.

General: Why does the outcome type matter? The primary outcome is a count of participants with an event, not a continuous measurement or a time-to-event outcome. Methods for binary outcomes compare event occurrence between groups. The registry record does not state the method used in its own analysis, and no sponsor analysis is posted. The displayed unadjusted comparison is a recalculation from summary counts, so it should not be described as an adjusted or time-to-event analysis.

General: What is over-dispersion? Over-dispersion occurs when variability in data exceeds what a statistical model assumes. It can make standard errors too small if the model does not account for the extra variability. This concept is most relevant to some count-data models; the primary endpoint is reported as participants with an event, and the registry does not report using an over-dispersed count model. It would therefore be inappropriate to claim that over-dispersion was present or that the trial addressed it.

General: What can be inferred about the statistical plan? The registry record lists no posted analyses and does not report the original analysis plan, interim analyses, covariate adjustment, missing-data procedures, or analysis-population definitions. These details should not be reconstructed from the posted event counts. The recalculated comparison is a transparent, unadjusted way to describe the available summary results, but it cannot establish which analyses were planned or performed by the sponsor.

For the planning illustrations, the stated assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The target effect is set to a stated fraction of the observed effect. Those are assumptions for an illustrative future-study calculation, not evidence about the original trial’s design or analysis. The original design assumptions are not reported in the registry record.

10. Confidence Intervals and Precision

The event counts are 15 of 46 in the control group and 9 of 46 in the erythropoietin group. The unadjusted risk-difference confidence interval extends from a negative value to a positive value, crossing no difference. This breadth and crossing indicate that the observed direction is not estimated precisely enough to rule out no difference or an effect in the opposite direction under the interval procedure. The p-value for the unadjusted comparison is 0.154.

A confidence interval conveys more than whether a result passes a threshold. It indicates the range of values that remain compatible with the data under the statistical method and its assumptions. When an interval includes no difference, that does not prove the groups are identical. It means that the available data do not tightly isolate a nonzero difference. Conversely, the observed direction of the point estimate does not establish that the same direction would appear in another study.

Precision depends in part on the amount of outcome information and on how often the outcome occurs. The registry reports 46 participants in each outcome group, with 15 and 9 events, respectively. The record does not report the original sample-size rationale or give enough detail to assess whether the trial reached its intended information target. The overall enrollment field is 101, but it does not replace the denominators used for the posted outcome comparison, and the registry does not explain their relationship.

Interpretation should account for the unreported analysis population, handling of missing outcomes, and trial conduct details. These omissions do not change the arithmetic of the posted summary comparison, but they limit a fuller assessment of its precision and generalizability. The interval is best read as evidence of substantial uncertainty around the estimated contrast, rather than as a definitive estimate of benefit or harm.

11. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning figures are participants per arm, not total event counts. They are illustrative calculations based on the observed effect, rather than the original trial’s documented design. The stated assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, a normal approximation, and an effect to detect set at the stated fraction of the observed effect. The trial’s own original sample-size assumptions are not reported in the registry record.

For planning at 100% of the observed effect, the illustration gives 177 participants per arm for 80% power and 237 participants per arm for 90% power. At 75% of the observed effect, it gives 328 participants per arm for 80% power and 438 participants per arm for 90% power. At 50% of the observed effect, it gives 765 participants per arm for 80% power and 1,023 participants per arm for 90% power.

In plain terms, the 80% power column represents the number of participants per arm estimated to give an 80% chance of detecting the assumed effect under the listed calculation assumptions. The 90% power column uses a higher target chance of detection, which generally calls for more participants. Smaller assumed effects also require more participants: the illustrations increase as the target moves from the full observed effect to a smaller fraction.

These figures are planning illustrations, not observed power. They depend on treating the observed effect as the reference for future planning, and the observed effect itself is uncertain. They should not be read as the number enrolled in this trial or as evidence about the trial’s original sample-size target.

12. 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 = 0.326; treatment proportion p₁ = 0.196; average p̄ = 0.261.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6210 = 1.2171.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6141 = 0.5168.
4. n = (1.2171 + 0.5168)² / (-0.1304)² = 176.7, rounded up to 177 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 <- 0.326086956521739                          # control proportion
p1 <- 0.195652173913043                          # 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(3007537)
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 = 177 (80% power), 237 (90% power)
 75% of observed effect: n per arm = 328 (80% power), 438 (90% power)
 50% of observed effect: n per arm = 765 (80% power), 1,023 (90% power)
power.prop.test: 176.7 
Simulated power at n = 177 per arm: 0.807 (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.

13. Multiplicity, Missing Data and Analysis Populations

The registry lists acute kidney injury as the primary outcome and renal replacement therapy as a secondary outcome. It does not post results for the secondary outcome. The record does not report a multiplicity strategy describing how it would handle testing more than one outcome or more than one analysis. Without that information, it is not possible to assess whether the posted p-value was adjusted for multiple comparisons or how the trial’s original analysis treated multiple endpoints.

The registry lists zero analyses and does not provide a detailed statistical analysis plan. It does not report interim analyses, stopping rules, prespecified subgroup analyses, covariate adjustment, or procedures for selecting among analyses. These details should be regarded as not reported in the registry record, rather than assumed to have been absent from the trial itself.

Missing-data handling is also not reported. The record does not state whether any participants lacked primary-outcome information, how such information was handled, or whether the analysis followed an intention-to-treat approach. The reported outcome denominators are 46 in each group, while enrollment is listed as 101; the registry does not explain the relationship. It is not possible to infer from these fields alone whether participants were excluded, withdrew, or had missing data.

The record does not define analysis populations or describe participant flow after randomization. Nor does it provide baseline group characteristics or reasons for discontinuation. These omissions constrain assessment of comparability, missingness, and the impact of analysis choices. The unadjusted recalculation from the posted counts describes the available summary contrast, but it cannot answer questions about unreported analysis populations or missing outcomes.

General: Multiplicity matters because testing several outcomes or analyses can increase the chance that at least one result appears statistically notable even when there is no underlying effect. Prespecified handling of multiple tests helps clarify how results should be interpreted. Because the registry does not report the trial’s multiplicity plan, no conclusion can be made about whether such a plan existed or how it was applied.

14. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

Most frequent serious adverse event terms

Serious events with the highest arm-level rate: pulmonary infection (21.7%), thrombotic events (0.0%).

The registry reports serious adverse events affecting 10 participants in the control group and 9 in the erythropoietin group. It reports no other adverse events and no deaths in either group. These are descriptive counts; the record does not provide a comparative safety analysis or enough detail to assess event timing, severity categories, or causal attribution to treatment.

The registry also lists pulmonary infection at 21.7% and thrombotic events at 0.0% among its top serious-event terms. The record does not specify a group assignment for these terms in the posted list, so they should not be attributed to one treatment group. A listed percentage is not, by itself, a comparison between groups.

Absence of reported deaths or other adverse events in these data does not establish that no such risks exist. The safety information should be interpreted in light of the reported group counts and the lack of detail about ascertainment and analysis. It does not support medical advice or a definitive conclusion about comparative safety.

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

The posted primary-outcome results describe fewer recorded acute kidney injury events in the erythropoietin group than in the control group: 9 participants versus 15, with a reported denominator of 46 in each group. The unadjusted comparison has p = 0.154, and its confidence interval spans no difference. Taken together, these results show a direction in the observed counts but leave considerable uncertainty about the underlying group difference.

They do not establish that erythropoietin prevents acute kidney injury, nor do they demonstrate that the groups have identical outcomes. A nonsignificant p-value is not proof of no effect, and an observed difference in event counts is not proof that the intervention caused a difference. The interval’s coverage of no difference is a reason to avoid a definitive conclusion from this comparison.

The interpretation is also limited by missing registry information. The reason for the terminated status is not reported. The record does not explain why the enrollment figure is 101 while the primary-outcome groups each have a denominator of 46. It does not report the original statistical plan, analysis population, missing-data handling, or participant flow. The secondary outcome of renal replacement therapy is named but its result is not posted.

The safety counts are descriptive, with serious adverse events reported for 10 control participants and 9 erythropoietin-group participants, and no other adverse events or deaths reported in either group. The record does not provide a formal comparative analysis. These data should not be used to make individual treatment decisions or to claim a comprehensive assessment of benefit and risk.

The appropriate conclusion is narrow: the registry records an unadjusted primary-outcome comparison with fewer events in the erythropoietin group, but the estimate is imprecise and compatible with no difference. The available record does not resolve the uncertainty or supply the methodological detail needed for a stronger causal interpretation.

16. Why This Trial Matters Statistically

This trial is statistically instructive because its observed event counts point in one direction while the uncertainty around the comparison remains substantial. Readers can distinguish the direction of the point estimate from the strength and precision of evidence: fewer events were recorded in the erythropoietin group, but the confidence interval crosses no difference and the unadjusted p-value is 0.154. That distinction prevents an observed contrast from being mistaken for a conclusive treatment effect.

The record also illustrates why denominators and analysis populations matter. Enrollment is listed as 101, whereas the primary-outcome comparison reports 46 participants in each group. Without a registry explanation, the enrollment field cannot be substituted for the outcome denominators, and the flow from enrollment to analysis cannot be reconstructed. Clear reporting of participant flow and missing data would help readers understand which observations support the result.

Planning calculations further show how conclusions about future study size depend on assumptions about the effect worth detecting. The illustrations use fractions of the observed effect, with smaller assumed effects requiring more participants per arm. Because the observed effect is uncertain, these calculations are planning aids rather than a guarantee of a future result or a description of the original design.

Finally, the registry’s lack of a posted analysis plan and secondary-outcome result demonstrates the value of complete reporting. Details about prespecification, multiplicity, missing-data handling, and analysis populations help determine how much confidence to place in an estimate. Their absence does not invalidate the posted counts, but it limits the scope of statistical conclusions that can be drawn from this record.

17. 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 study is listed as terminated, but the reason for termination is not reported in the registry record. The enrollment field is 101, while the primary-outcome results give a denominator of 46 for each group; the record does not explain this relationship. Participant flow, withdrawals, exclusions, and the analysis population are not described, so the posted denominators cannot be reconciled with enrollment.

The registry does not report eligibility criteria, baseline characteristics, random-sequence generation, allocation concealment, or which roles were masked. It names the primary outcome definition and timeframe but does not provide further information about event ascertainment or adjudication. The original statistical analysis plan, sample-size justification, missing-data procedures, and any interim analyses are also not reported.

Renal replacement therapy is listed as a secondary outcome, but its result is absent. No sponsor analysis is posted, and the registry lists no analyses. The unadjusted comparison is therefore a recalculation from the available summary counts, not a substitute for a fully reported trial analysis. The safety record gives counts and selected event terms but does not report a formal comparative analysis or detailed event-level information.

These limitations restrict assessment of precision, risk of bias, and generalizability. They do not justify assuming that particular procedures were absent; they mean those details are not reported in the registry record. The interpretation should remain limited to the posted outcome counts and the uncertainty in their unadjusted comparison.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: The Prevention of Erythropoietin on Cardiac Surgery-associated Acute Kidney Injury — NCT03007537.
• Registry record last updated 2023-01-26; status last verified 2023-01.
• Results first posted 2023-01-26.
• 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.

21. Record Summary

NCT03007537 is a randomized, double-masked, parallel-group trial sponsored by Nanjing First Hospital, Nanjing Medical University. It compared one subcutaneous injection of 10000 IU erythropoietin with 0.9% sodium chloride 1ml, applied 1day before cardiac surgery. The record lists enrollment as 101, a start date of 2015-10-01, a primary completion date of 2018-12-28, and status as terminated.

The primary outcome was the number of participants with acute kidney injury according to the KDIGO guideline during 7 days. The posted counts were 15 among 46 control participants and 9 among 46 erythropoietin-group participants. The unadjusted comparison has p = 0.154, and its confidence interval includes no difference. Renal replacement therapy is listed as a secondary outcome, but its result is not reported in the registry record. The reason for termination and key analysis details are also not reported.