Epidemiology and Prevention of Methicillin Resistant Staphylococcus Aureus (MRSA) Transmission in the Community: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Epidemiology and Prevention of Methicillin Resistant Staphylococcus Aureus (MRSA) Transmission in the Community (NCT00966446), a randomized na study. The primary outcome is Recurrent Infection.

2. Clinical Question

3. Trial Design

This was a randomized, parallel-group study of MRSA transmission and prevention in the community. The registry lists an enrollment of 223 and three arms: unsupervised decolonization, supervised decolonization, and no intervention. The trial is marked completed. Its start is listed as 2009-09 and primary completion as 2014-09. The listed phase is NA. The sponsor was the University of Pennsylvania.

The two decolonization arms involved households. In the unsupervised arm, the regimen was Mupirocin ointment twice daily during the first 7 days of study enrollment, plus Hibliclens body wash on day 1 and day 7, according to the instructions. The supervised-arm description reports the same regimen and begins to describe study-staff involvement, but does not give the rest of that description. The no-intervention arm did not use an active MRSA decolonization protocol. The record does not report further details about supervision, adherence monitoring, or other aspects of implementation.

Masking is listed as none, so the trial was not masked according to the registry entry. Allocation is listed as randomized, and the model is parallel. The posted primary outcome is recurrent infection: confirmed new MRSA infections within 6 months, in study participants who self-reported re-infections. Outcome counts were posted for 48 participants in the unsupervised arm, 49 in the supervised arm, and 52 in the no-intervention arm. The record does not explain how these outcome denominators relate to the overall enrollment or whether all enrolled participants contributed to the primary outcome.

The registry also lists time to clearance of MRSA colonization as a secondary outcome. Further information about the methods used to measure or analyze that outcome is 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

Analysis population for the primary outcome, as defined in the registry: These study participants who self-reported re-infections of MRSA

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 Recurrent Infection, measured over Within 6 months.

The primary endpoint is recurrent infection, recorded as a count of participants with confirmed new MRSA infections within 6 months. The stated population is study participants who self-reported re-infections of MRSA. The posted counts are 1 for unsupervised decolonization, 3 for supervised decolonization, and 1 for no intervention. This endpoint focuses on new infection rather than colonization alone; the registry separately lists time to clearance of MRSA colonization as a secondary outcome.

Counting participants with an event gives a direct summary of how many people in each arm experienced the specified outcome during the stated timeframe. The risk-difference comparisons express the difference between the proportions with an event in each compared arm. These summaries are useful for describing the observed outcome in the trial population, but their interpretation depends on the outcome definition, the participants included, and the completeness of follow-up.

The registry describes infections as confirmed and the population as people who self-reported re-infections, but does not report the confirmation procedure or how self-reports were collected. It also does not report a detailed endpoint-adjudication process, the handling of incomplete follow-up, or an analysis population definition. Those details are important context when interpreting a sparse event count, because a small number of recorded events can make estimates imprecise and sensitive to how outcomes were identified.

7. Results Posted in the Registry

• Supervised Decolonization: 6.1% (3/49) versus 2.1% (1/48) on control; risk difference 4.0 percentage points (95% CI -3.8 to 11.9), risk ratio 2.94 (0.32–27.27); unadjusted two-proportion z-test p = 0.317.
• No Intervention: 1.9% (1/52) versus 2.1% (1/48) on control; risk difference -0.2 percentage points (95% CI -5.7 to 5.3), risk ratio 0.92 (0.06–14.35); unadjusted two-proportion z-test p = 0.954.

8. Statistical Interpretation

9. Statistical Methods Explained

The posted comparisons use a risk difference, which is the difference between the event risk in one arm and the event risk in the reference arm. A risk difference of zero represents equal observed risks. A positive estimate means the event risk was higher in the arm named in the comparison than in the unsupervised decolonization reference arm; a negative estimate means it was lower. These are unadjusted arm-level comparisons based on the posted event counts and denominators.

What does the confidence interval describe?

A confidence interval describes the uncertainty around an estimated effect under the statistical method used. It gives a range of values compatible with the data and model at the stated confidence level. It is not a range containing a fixed proportion of individual outcomes, and it is not the probability that the true effect lies within this particular interval. For the supervised-versus-unsupervised comparison, the risk-difference interval is -0.038 to 0.119. For the no-intervention-versus-unsupervised comparison, it is -0.057 to 0.053. Both intervals include zero, so neither rules out no difference between the compared arms.

How should the p-values be read?

The posted p-values are unadjusted: 0.317 for supervised decolonization compared with unsupervised decolonization, and 0.954 for no intervention compared with unsupervised decolonization. A p-value measures the compatibility of data with a specified null model, assuming that model and the analysis assumptions hold. It does not give the probability that the null hypothesis is true, and a large p-value does not demonstrate equivalence. The registry record does not report enough detail to identify a fuller testing plan or any adjustment for multiple comparisons.

Why do sparse events matter?

The outcome counts are 1, 3, and 1 across the three arms. With few events, each observed event has substantial influence on the estimated arm risks and their difference. This can produce considerable uncertainty, as reflected in the intervals. General statistical principle: sparse binary outcomes may also challenge approximations used by some methods, particularly when estimating relative effects. The record does not identify the method used to construct the posted intervals, so no claim can be made about whether a particular sparse-data method was used.

What is the distinction between risk and risk difference?

The arm-level risks are posted as 0.021 for unsupervised decolonization in both comparisons, 0.061 for supervised decolonization, and 0.019 for no intervention. The risk difference summarizes the absolute contrast between the relevant pair of risks. The registry also posts relative-effect estimates: for supervised versus unsupervised decolonization, the risk ratio is 2.939 with an interval from 0.317 to 27.27; for no intervention versus unsupervised decolonization, it is 0.923 with an interval from 0.059 to 14.35. These intervals are very broad and include values on either side of the no-relative-difference value of one. Such relative estimates should be interpreted cautiously when events are sparse.

The registry does not report an adjusted model, covariates, the exact interval-construction procedure, or a detailed analysis plan. It also does not report whether the posted comparisons were prespecified or how participants with missing outcome information were handled. Those omissions restrict assessment of the analytic choices and their implications. The posted estimates remain useful descriptions, but they should not be mistaken for a complete account of the trial's statistical methods.

10. Confidence Intervals and Precision

The confidence intervals indicate substantial uncertainty around both risk-difference estimates. For supervised decolonization compared with unsupervised decolonization, the estimated risk difference is 0.040 and its interval ranges from -0.038 to 0.119. For no intervention compared with unsupervised decolonization, the estimate is -0.002 and the interval ranges from -0.057 to 0.053. In each case, the interval includes zero and permits differences in either direction.

Precision depends in part on how much outcome information the trial contains. The posted event counts are 1, 3, and 1, so the primary endpoint is based on few events. In general, fewer events tend to leave greater uncertainty about risks and contrasts, even when the participant count is larger than the event count. The wide relative-risk intervals reinforce that the relative magnitude is not well pinned down: 0.317 to 27.27 for supervised versus unsupervised decolonization, and 0.059 to 14.35 for no intervention versus unsupervised decolonization.

An interval crossing zero is not evidence that the true difference must be zero. It means the posted estimate is compatible with a range that includes no difference as well as differences in either direction. The registry does not report a target precision, a prespecified clinically important difference, or a formal interpretation threshold. Therefore, the intervals can be described as broad, but the record does not establish what degree of uncertainty would have been acceptable for the trial's objectives.

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 illustrations report participants per arm, not total events. They use the observed risk difference as a starting point and show participant requirements when planning for 100%, 75%, or 50% of that observed effect. The listed assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is the stated fraction of the observed effect. These are planning calculations, not the trial's original design assumptions; the original sample-size justification is not reported in the registry record.

The 80% power column gives the number of participants per arm that the planning calculation estimates would be needed for an 80% chance of detecting the specified effect under the stated assumptions, if that effect and model were correct. The 90% power column gives the corresponding per-arm requirement for a 90% chance. Power is a design property under assumptions, not a guarantee that a future study will detect an effect or that an effect is real.

For the supervised-versus-unsupervised contrast, planning for 100% of the observed effect gives 378 participants per arm for 80% power and 505 for 90% power. Planning for 75% gives 593 and 793 participants per arm, while planning for 50% gives 1,153 and 1,543 participants per arm. For the no-intervention contrast, the corresponding requirements are much larger: 119,989 and 160,631 per arm for 100% of the observed effect; 215,404 and 288,364 for 75%; and 489,357 and 655,110 for 50%. The smaller the effect to be detected, the more participants the calculation requires. These figures are not observed power; they are planning illustrations based on the observed effects, whose estimates are uncertain.

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 = 0.021; treatment proportion p₁ = 0.061; average p̄ = 0.041.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.2805 = 0.5498.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.2791 = 0.2349.
4. n = (0.5498 + 0.2349)² / (0.0404)² = 377.4, rounded up to 378 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 <- 0.0208333333333333                          # control proportion
p1 <- 0.0612244897959184                          # 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(966446)
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 = 378 (80% power), 505 (90% power)
 75% of observed effect: n per arm = 593 (80% power), 793 (90% power)
 50% of observed effect: n per arm = 1,153 (80% power), 1,543 (90% power)
power.prop.test: 377.4 
Simulated power at n = 378 per arm: 0.819 (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 one secondary outcome, time to clearance of MRSA colonization. It reports three outcome measures and zero analyses in the record summarized by the posted results. The recurrent-infection outcome has two posted comparisons against the unsupervised decolonization arm. The registry record does not report whether those comparisons were designated in advance, whether additional analyses were performed, or whether any adjustment for multiple comparisons was planned. The unadjusted p-values should therefore be read as individual comparison results, not as values adjusted to account for a family of tests.

Multiplicity matters when many outcomes, comparisons, or analytic choices are examined: the chance of finding at least one apparently unusual result can rise. General statistical principle: prespecifying a primary comparison and an adjustment approach can help readers understand how a collection of tests should be interpreted. The original analysis plan and any multiplicity strategy are not reported in the registry record, so it is not possible to determine whether the posted comparisons were the full planned set.

The registry record does not report the amount or pattern of missing outcome data, the follow-up completeness for the primary endpoint, or a method for handling missing values. It also does not define an intention-to-treat, per-protocol, or other analysis population for these results. The primary outcome population is described as study participants who self-reported re-infections, but further detail about who was eligible for the posted counts is not given. These gaps prevent assessment of whether participants were analyzed according to randomized assignment, whether all eligible participants contributed data, and how missing or unconfirmed reports affected the estimates.

There is no reported imputation procedure, sensitivity analysis for missing data, or description of exclusions. The registry record also does not report interim analyses, stopping rules, or a statistical analysis plan. Those details should not be inferred from the randomized design or from the results themselves.

14. 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

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: heart disease (4.1%), abscess (4.1%), anorexia nervosa (2.6%), hypertension (2.6%), hemmorage (1.4%).

For serious adverse events, the registry lists 5 affected participants among 76 at risk in the unsupervised decolonization group, 7 among 74 in the supervised decolonization group, and 4 among 73 in the no-intervention group. These are counts as posted, and they should be read alongside the reported group-specific numbers at risk rather than as comparisons adjusted for follow-up or other differences. The record does not provide a statistical comparison of these serious adverse event counts.

The registry also lists other adverse events affecting 17 among 76 participants in the unsupervised group, 29 among 74 in the supervised group, and 0 among 73 in the no-intervention group. The record does not explain the collection or coding procedures, the observation period for adverse events, or whether these counts represent participants with at least one event or event totals. Death counts are not reported in the registry record.

Listed serious-event terms include heart disease and abscess, each at 4.1%; anorexia nervosa and hypertension, each at 2.6%; and “hemmorage” and “homocide,” each at 1.4%, using the registry's spellings. These terms are not assigned to specific arms in the information reported. The counts alone do not establish cause or comparative safety.

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

The posted results describe a small number of recurrent infections in the outcome population: 1 among 48 participants assigned to unsupervised decolonization, 3 among 49 assigned to supervised decolonization, and 1 among 52 assigned to no intervention. The estimated risk difference for supervised versus unsupervised decolonization is 0.040, while the estimate for no intervention versus unsupervised decolonization is -0.002. Both confidence intervals include zero and span values in both directions.

These results do not establish that the compared approaches have the same risk, nor do they establish that one approach causes more or fewer recurrent infections. The counts are sparse, and the uncertainty intervals are broad relative to the point estimates. The unadjusted p-values do not resolve that uncertainty, and they do not give the probability that a treatment effect exists. Interpretation also depends on how the registry's self-reported re-infection population and confirmed new infections were defined and followed; further details are not reported in the registry record.

The randomized, parallel design is relevant to the intended arm comparisons, but the posted outcome counts alone do not describe all features of trial conduct, adherence, follow-up, or analysis populations. No outcome result for time to clearance of colonization is reported in the information summarized. The findings therefore speak to the posted recurrent-infection endpoint and should not be extended to other outcomes or populations.

The planning illustrations show how sample-size requirements change under specified assumptions when targeting fractions of the observed effects. They do not establish the sample size needed for every future study, because the observed effects are uncertain and the assumptions may not match a later study. In particular, very large projected requirements for the no-intervention comparison arise from planning around an observed risk difference close to zero; that illustration should not be interpreted as proof that the true difference is exactly zero.

16. Why This Trial Matters Statistically

This trial illustrates a recurring challenge in clinical and community research: a randomized comparison can still produce limited precision when the primary outcome occurs infrequently. The three posted arm counts are 1, 3, and 1. With such sparse event data, the point estimates may appear simple to interpret, but their intervals show that a broad range of underlying differences remains compatible with the observations.

The comparison between absolute and relative effects is also important. A risk difference describes the absolute contrast between arms, while a risk ratio describes their relative contrast. When event counts are very low, relative estimates can be especially unstable, as reflected by the broad risk-ratio intervals posted for both comparisons. Readers benefit from seeing the event counts, absolute effects, relative effects, and uncertainty together rather than relying on a single point estimate or p-value.

The registry's population description matters statistically as well as clinically. The endpoint is defined among participants who self-reported re-infections, and the outcome is described as confirmed new infections. Without further detail on ascertainment, follow-up, missing data, and analysis populations, it is difficult to assess how completely the posted counts represent the randomized groups.

Finally, sample-size planning based on an observed effect is sensitive to that effect's uncertainty. The planning illustrations make the relationship between a smaller target effect and a larger participant requirement explicit, but they are not a substitute for the trial's original design justification, which is not reported in the registry record. This example reinforces why effect estimates should be read together with their precision and the methods used to obtain them.

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 results are based on few events, limiting precision. The confidence intervals include zero for both posted risk differences, and the relative-effect intervals are broad. The registry does not report the method used to calculate the intervals or the test underlying the p-values, so the analytical details cannot be independently assessed from the record.

The endpoint population is described as study participants who self-reported re-infections, while the outcome is described as confirmed new MRSA infections. The procedures for self-report, confirmation, and outcome adjudication are not reported in the registry record. The outcome denominators differ from the overall enrollment, but the record does not explain this relationship or state how many participants lacked outcome information.

The registry does not report an analysis plan, adjusted analyses, multiplicity procedures, missing-data methods, analysis populations, interim analyses, or stopping rules. It also does not report adherence results or detailed follow-up information for the primary endpoint. These omissions limit assessment of how trial conduct and analytic decisions may have influenced the posted comparisons.

The description of supervised decolonization is incomplete in the registry record, ending during its account of study-staff involvement. The record therefore does not provide a complete explanation of how supervision differed in practice from unsupervised decolonization. Safety reporting also lacks detail on adverse-event collection, attribution, observation time, and the meaning of the “other” adverse-event counts; death counts are not reported.

Finally, the posted results address recurrent infection and list time to clearance of colonization as a secondary outcome, but no secondary-outcome result is reported in the information analyzed. Conclusions should remain limited to the outcome counts and comparisons that the registry record reports.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Epidemiology and Prevention of Methicillin Resistant Staphylococcus Aureus (MRSA) Transmission in the Community — NCT00966446.
• Registry record last updated 2019-07-05; status last verified 2019-07.
• Results first posted 2016-06-29.
• 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

NCT00966446 is a completed randomized, unmasked, parallel-group study sponsored by the University of Pennsylvania. The registry lists enrollment of 223 and three arms: unsupervised decolonization, supervised decolonization, and no intervention. The primary outcome is recurrent infection, defined as confirmed new MRSA infections within 6 months among participants who self-reported re-infections.

Posted event counts were 1 among 48 participants for unsupervised decolonization, 3 among 49 for supervised decolonization, and 1 among 52 for no intervention. The risk-difference comparisons use unsupervised decolonization as the reference, and both confidence intervals include zero. The record lists time to clearance of MRSA colonization as a secondary outcome, but does not report its result in the information analyzed.

Important details—including the original sample-size justification, full statistical analysis plan, missing-data handling, analysis populations, and complete supervised-arm description—are not reported in the registry record. The sparse recurrent-infection counts and broad uncertainty intervals are central to interpreting the posted comparisons.