Antibiotics Usage in Pediatric Orthopaedic Percutaneous Surgery (APOPS): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Antibiotics Usage in Pediatric Orthopaedic Percutaneous Surgery (APOPS) (NCT03261830), a randomized phase 4 study. The primary outcome is Number of Participants With the Presence of Post-Operative Infection.

2. Clinical Question

3. Trial Design

APOPS (NCT03261830) is recorded as a phase 4, randomized, triple-masked, parallel-group trial. It enrolled 160 participants. The registry describes two groups: pre-operative antibiotics and saline placebo. The antibiotic group could receive cefazolin, or clindamycin in cases of documented cefazolin allergy; the comparator group received saline placebo. The record labels the allocation as randomized, but the method used to generate or conceal the allocation sequence is not reported in the registry record.

The listed study dates run from 2017-08-18 to primary completion on 2023-03-10, and the trial status is completed. Participants were to be followed for 3-6 weeks, depending on age. The primary outcome timeframe is six weeks. The record says the infection assessment was made by a blinded attending surgeon at cast removal, but does not specify which other participants or staff were masked or how masking was maintained.

Alongside infection, the listed secondary outcomes are Visual Analog Pain Scale, fractures fully healed, need for repeat casting, and loss of fixation. The registry lists five outcome measures in total. Although these outcomes are identified, their results are not reported in the registry record. The sponsor is Sumit Gupta. Details of the eligibility criteria, sample-size justification, original analysis plan, and the practical conduct of follow-up 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.

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 the Presence of Post-Operative Infection, measured over Six weeks.

The primary endpoint counts participants with post-operative infection, rather than counting separate infection episodes. The registry defines a superficial infection as a clinical diagnosis involving increased erythema, granulation tissue, or purulence at pin sites that resolves with a short course of oral antibiotics. A deep infection is defined by septic arthritis or osteomyelitis requiring return to the operating room for debridement or a prolonged course of intravenous antibiotics. If the attending physician judged that either kind of infection occurred at any time during treatment, the participant was categorized as infected.

This binary endpoint makes the comparison straightforward to describe: each participant is classified as having or not having the defined event during the specified follow-up. The endpoint combines superficial and deep infections, which differ in clinical features and treatment as described in the registry. The posted result therefore does not separately estimate the frequency of each category. Infection was assessed by the blinded attending surgeon at cast removal, with the primary timeframe listed as six weeks. The registry does not report additional details about outcome adjudication or how participants without an assessment were handled.

7. Results Posted in the Registry

• Pre-operative Antibiotics: 2.4% (2/82) versus 2.6% (2/78) on control; risk difference -0.1 percentage points (95% CI -5.0 to 4.7), risk ratio 0.95 (0.14–6.59); unadjusted two-proportion z-test p = 0.960.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is binary: a participant either met the registry’s definition of post-operative infection during the treatment period or did not. For a binary outcome, the basic descriptive information is the number of participants with the event and the number of participants assessed in each group. The record reports 2 infections among 82 participants in the antibiotic group and 2 among 78 participants in the saline-placebo group. These are the posted group counts and denominators; the registry does not give a separate account of missing outcome assessments.

What does the risk difference describe?

The risk difference expresses the absolute contrast between the groups’ observed event risks. The posted estimate is -0.001 for antibiotics relative to saline placebo. A negative estimate means the observed risk was lower in the antibiotic group according to the direction used for this comparison. The estimate is very close to zero, but it is not an exact statement that the underlying risks are equal. The interval from -0.050 to 0.047 describes a range of values compatible with the data under the interval method’s assumptions. The method and confidence level used for the interval are not reported in the registry record.

How do relative measures differ?

The risk ratio compares the event risks as a ratio, while the odds ratio compares the event odds. Both are relative measures, unlike the risk difference, which is absolute. The registry reports a risk ratio of 0.951, with an interval from 0.137 to 6.588, and an odds ratio of 0.95, with an interval from 0.131 to 6.915. The estimates are near one, the value representing no relative difference, but the intervals are wide. In particular, these estimates should not be read as precise measures of the size of an effect. With few observed events, changes in the event count can have a substantial influence on relative estimates.

What does the p-value contribute?

The unadjusted p-value is 0.96. In general, a p-value measures how unusual data at least as incompatible with a specified null hypothesis would be if that null hypothesis and the test assumptions held. It does not give the probability that the null is true, and it does not measure the clinical importance of an effect. The specific statistical test, its assumptions, and any adjustment strategy are not reported in the registry record. The p-value should therefore be read alongside the event counts and effect estimates, rather than as a stand-alone conclusion.

Why do sparse events matter?

In general, when events are uncommon, the estimated risks and especially relative measures can be sensitive to a small number of observations. Confidence intervals communicate this uncertainty; broad intervals signal that the data do not pin down the effect very closely. A non-significant result does not establish that an intervention and comparator are equivalent. Equivalence requires a prespecified margin and an analysis designed to assess whether the full uncertainty interval falls within that margin. Neither an equivalence margin nor a non-inferiority framework is reported in the registry record.

What can be inferred about the analysis?

The registry records a randomized parallel-group design and reports an unadjusted comparison, but the analysis method is not specified. The record does not report whether the primary analysis followed intention-to-treat, per-protocol, or another population definition. It also does not report how missing outcome data, protocol deviations, or participants who did not complete follow-up were handled. Those details matter because analysis choices can affect the comparison, particularly when event counts are sparse. No additional statistical procedure should be assumed from the posted summary alone.

10. Confidence Intervals and Precision

The risk-difference estimate is -0.001, and its reported interval extends from -0.050 to 0.047. In general, a confidence interval gives a range of effect values compatible with the data under a specified statistical method. The registry does not state the confidence level or interval method, so those details cannot be characterized further from the record. The important feature of the posted interval is that it crosses zero and extends in both directions. Thus, the point estimate is close to no difference, but the data remain compatible with a range of absolute differences.

The relative measures are also imprecise. The risk ratio is 0.951, with an interval from 0.137 to 6.588; the odds ratio is 0.95, with an interval from 0.131 to 6.915. For ratios, the value representing no relative difference is one. The reported intervals extend on either side of that reference and are broad, so the relative effect is not estimated precisely. The sparse infection counts are relevant to this uncertainty: there were 2 recorded infections in each group.

Precision is distinct from whether a result crosses a conventional significance threshold. The unadjusted p-value is 0.96, but neither that value nor a point estimate near zero proves that the groups have the same underlying risk. To support a claim of equivalence, the study would need an appropriate prespecified margin and an analysis addressing it; neither is reported in the registry record. The posted estimates are best understood as inconclusive about the size of a possible difference, rather than as precise evidence of no difference.

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 sample-size figures are planning illustrations based on the observed effect, not the trial’s original design assumptions. For this binary outcome, the figures are 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. Each planning row uses the stated fraction of the observed effect as the effect to detect. The registry record does not report the trial’s own sample-size justification or original design assumptions.

The 80% power column describes the planned number of participants per arm under those assumptions for a design with 80% power to detect the specified effect. The 90% power column gives the corresponding planning figure for 90% power. In general, greater power requires more participants when other assumptions are held fixed. The rows use 100%, 75%, and 50% of the observed effect; the smaller-effect rows require more participants per arm. For example, planning for 100% of the observed effect lists 244,727 participants per arm for 80% power and 327,620 for 90% power. Planning for 75% lists 437,720 and 585,983, respectively; planning for 50% lists 990,830 and 1,326,440.

These figures are not observed power and do not describe the number enrolled in APOPS. They are conditional illustrations: changing the assumed effect or planning assumptions would change the figures. The record describes this calculation as using equal allocation, while the observed randomized groups had different sizes. No additional assumptions about the trial’s original planning should be inferred from this illustration.

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 = 0.026; treatment proportion p₁ = 0.024; average p̄ = 0.025.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.2209 = 0.4329.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.2209 = 0.1859.
4. n = (0.4329 + 0.1859)² / (-0.0013)² = 244,726.9, rounded up to 244,727 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 <- 0.0256410256410256                          # control proportion
p1 <- 0.024390243902439                          # 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(3261830)
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 = 244,727 (80% power), 327,620 (90% power)
 75% of observed effect: n per arm = 437,720 (80% power), 585,983 (90% power)
 50% of observed effect: n per arm = 990,830 (80% power), 1,326,440 (90% power)
power.prop.test: 244726.9 
Simulated power at n = 244,727 per arm: 0.798 (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 five outcome measures: the primary infection outcome and secondary outcomes covering pain, fracture healing, repeat casting, and loss of fixation. Four secondary outcomes are listed, but their results are not reported in the registry record. The record does not report a hierarchy for testing outcomes, a multiplicity adjustment, or a prespecified strategy for interpreting several comparisons. When multiple outcomes are assessed, the chance of at least one apparently notable result can increase; in general, multiplicity procedures address that issue by adjusting interpretation or analysis. Whether any such procedure was used in this trial is not reported.

The registry reports zero analyses in its summary, while also posting an unadjusted comparison for the primary outcome. It does not provide a full statistical analysis plan or describe the analysis method for that comparison. The analysis population—such as intention-to-treat or per-protocol—is not reported in the registry record. No conclusion about adherence to a specific analysis population can be drawn from the posted summary.

Missing-data handling is also not reported. The registry does not state whether every randomized participant had an endpoint assessment, how incomplete follow-up was treated, or whether any imputation or sensitivity analysis was used. Follow-up was planned for 3-6 weeks depending on age, and the primary outcome timeframe is six weeks, but that description does not establish how actual follow-up completeness was handled. The record also does not report interim analyses, stopping rules, or the approach to protocol deviations. These absences limit the ability to reconstruct the full analysis from the registry summary.

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

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: osteomyelitis (1.2%).

The adverse-event record lists one serious adverse event among 82 participants in the pre-operative antibiotics group and none among 78 in the saline-placebo group. It lists no other adverse events in the antibiotic group and one in the saline-placebo group. Deaths are recorded as zero in both groups. Osteomyelitis is listed as the top serious adverse-event term at 1.2%; the record does not provide a group-specific breakdown for that term in the posted summary.

These entries are counts recorded in the registry, not estimates that establish whether an event was caused by an intervention. The serious and other adverse-event categories are reported separately, and the record does not supply further detail about event definitions, severity grading, timing, or attribution. With the small number of recorded events, the counts alone cannot characterize the range of possible safety outcomes. They should be read as the registry’s reported event information rather than as a comprehensive estimate of comparative safety.

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

The posted primary result is that 2 participants in each randomized group were recorded as having post-operative infection. The risk-difference estimate is -0.001, and the interval from -0.050 to 0.047 includes values in both directions. The relative estimates also have wide intervals. Together, these results describe little observed separation in the group counts but substantial uncertainty about the size of any underlying difference.

This is not evidence that the groups are exactly equivalent. A study can fail to show a clear difference while still being too imprecise to rule out meaningful differences. The registry does not report a prespecified equivalence margin or non-inferiority analysis. The unadjusted p-value of 0.96 should not be treated as the probability that antibiotics and placebo have the same effect, nor as a measure of the clinical importance of the result.

The result applies to the recorded endpoint and the trial population described in the registry: pediatric patients undergoing percutaneous fixation for supracondylar humerus fractures. It does not by itself settle questions about other patient groups, procedures, infection definitions, or follow-up periods. The registry lists additional outcomes, including pain, healing, repeat casting, and loss of fixation, but their results are not reported in the registry record. It also does not provide enough statistical detail to determine how missing data, analysis populations, or multiplicity were handled.

The appropriate interpretation is therefore limited: the posted data show equal observed infection counts in the two groups, while the effect estimates remain imprecise. They do not establish that a true difference is absent, and they should not be extended beyond the trial’s recorded setting and endpoint.

16. Why This Trial Matters Statistically

This trial illustrates a common statistical challenge in clinical research: an important outcome can occur infrequently, leaving a randomized comparison with few events. When that happens, even a small difference in event counts can influence estimates, and confidence intervals can remain wide. Reporting the absolute and relative effect estimates alongside the counts makes that uncertainty visible rather than reducing the interpretation to a p-value.

The registry’s primary endpoint also combines superficial and deep infection under a participant-level definition. That approach yields a clear overall event classification, but the total count does not reveal how many events fell into each category. The clinical descriptions indicate that these categories differ, so the overall comparison answers a narrower question than a breakdown by infection type would. Such a breakdown is not reported in the registry record.

The result is also a useful reminder that a large p-value is not a test of equivalence. The recorded p-value is 0.96, while the risk-difference interval spans values on both sides of zero. Without a prespecified equivalence margin and an analysis designed to assess it, the result cannot establish that the groups’ risks are sufficiently similar. For biostatisticians and clinical researchers, this distinction is central when communicating a null or non-significant finding.

Finally, a registry summary may identify outcomes without presenting their results or the analysis details needed for a complete statistical appraisal. Here, the secondary outcomes are named, but their results, missing-data procedures, and analysis populations are not reported in the registry record. Recognizing those limits helps keep conclusions tied to what the posted infection data can support.

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 primary outcome is based on a small number of recorded infections, and the posted effect intervals are broad. This limits precision and makes the size of any difference uncertain. The endpoint combines superficial and deep infection, so the overall comparison does not show how the two infection categories contributed to the event counts. The registry does not report category-specific results.

The registry record does not report the allocation-generation or concealment method, the detailed roles covered by triple masking, or the eligibility criteria. It describes follow-up as 3-6 weeks depending on age and gives a six-week primary timeframe, but actual follow-up completeness and missing-data handling are not reported. The analysis population, statistical test, confidence level, and full analysis plan are also not reported. This prevents a detailed assessment of how the comparison was conducted.

Results for the four listed secondary outcomes are not reported in the registry record, and the approach to multiple outcomes is not described. The record does not report an equivalence margin, so the similar observed infection counts cannot establish equivalence. The planning illustrations are based on the observed effect and are not a record of the trial’s original sample-size justification. These specific gaps limit what can be concluded from the registry summary, independently of the outcome estimates themselves.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Antibiotics Usage in Pediatric Orthopaedic Percutaneous Surgery (APOPS) — NCT03261830.
• Registry record last updated 2024-08-09; status last verified 2024-07.
• Results first posted 2024-08-09.
• 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

APOPS (NCT03261830) is a completed, phase 4 randomized parallel-group trial with triple masking, concerning antibiotics before percutaneous surgery for pediatric supracondylar humerus fractures. The registry reports enrollment of 160 and compares pre-operative antibiotics with saline placebo. Follow-up was planned for 3-6 weeks depending on age; the primary infection outcome timeframe was six weeks.

The primary outcome counted participants with post-operative infection. The registry reports 2 infections in each group and a risk-difference estimate of -0.001, with an interval from -0.050 to 0.047 and an unadjusted p-value of 0.96. The estimate is imprecise, and the record does not establish equivalence. Four secondary outcomes are listed, but their results are not reported in the registry record. Details of the original sample-size justification, analysis population, missing-data handling, and statistical plan are also not reported.