Use of Pulsed Electromagnetic Fields (PEMF) After Breast Reconstruction Surgery: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Use of Pulsed Electromagnetic Fields (PEMF) After Breast Reconstruction Surgery (NCT01262599), a randomized phase 4 study. The primary outcome is Pain Score Measured by Visual Analog Scale.

2. Clinical Question

The clinical question concerns postoperative recovery after pedicled TRAM flap breast reconstruction. The registry describes patients scheduled for reconstruction of a single breast immediately following mastectomy. The comparison is between an active Ivivi Torino II pulsed electromagnetic field device and an inactive sham device, with postoperative pain measured using a visual analog scale.

The posted results report a mean score of 3.48 in the sham group and 1.67 in the active-device group, with 16 participants in each group. The recalculated unadjusted difference is reported as -1.81 units, where a negative value reflects a lower mean in the active-device group than in the sham group. This describes the observed groups; it does not by itself establish how broadly the result applies to other patients or settings.

3. Trial Design

The registry identifies this as a completed, randomized, parallel-group, phase 4 trial sponsored by Columbia University. It records quadruple masking and an enrollment of 32. The posted group counts are 16 for the sham device and 16 for the active PEMF device. The record describes the study as double-blind in its summary, while its design fields classify masking as quadruple; it does not explain the roles included in that masking classification.

The intervention comparison was an inactive sham PEMF device versus the Ivivi Torino II PEMF device. The population described in the summary was patients scheduled for pedicled TRAM flap reconstruction of a single breast immediately following mastectomy. The registry's summary explains that this surgery uses tissue from the patient's abdomen to create a breast, and identifies donor-site pain and postoperative complications as relevant context. Those statements describe the motivation for studying recovery, not outcomes measured in every participant.

The primary outcome was mean pain on a visual analog scale at 12 hours. Two secondary outcomes are posted: number of narcotic pain medications at 24 hours and cytokine levels at 24 hours. The registry reports a superiority analysis using ANOVA for the primary comparison, with p < 0.01. No analysis results are listed in the registry record for the secondary outcomes, though group summaries and recalculated comparisons are available.

The original protocol's sample-size assumptions, rationale for enrollment, planned handling of missing observations, and intended analysis population are not reported in the registry record. The record also does not describe whether interim analyses were planned or conducted. Consequently, the enrollment count and observed group sizes alone do not establish what design assumptions informed the trial.

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 Pain Score Measured by Visual Analog Scale, measured over 12 hours.

The primary endpoint is the mean patient-reported pain score on a visual analog scale at 12 hours after surgery. The registry describes the scale as ranging from 0, meaning no pain, to 10, meaning worst possible pain. Higher values indicate more pain; lower values indicate less pain. The posted group summaries are means with standard deviations.

A patient-reported pain scale is used to quantify a subjective experience in a standardized form. In general, converting responses to a defined scale makes it possible to summarize responses across participants and compare group averages. The mean represents the center of the observed scores, while the standard deviation describes how dispersed individual scores are around that mean. Neither statistic alone describes every participant's experience.

The registry's endpoint description mentions recording pain at 12 hours and assessing it in comparison with previous timepoint pain levels, with 1 hour post-op given as an example. However, the posted primary result is the mean score at 12 hours for each group. A separate numerical result for an earlier timepoint is not reported in the registry record. The endpoint should therefore be read as a comparison of the reported 12-hour group means, rather than as a reported change from an earlier measurement.

Pain was a relevant outcome in the study rationale because the registry describes abdominal donor-site pain as a contributor to postoperative hospital stay. That background explains why pain was selected for study; it does not establish an effect on hospital stay, which is not among the posted outcomes.

7. Results Posted in the Registry

• PEMF Device: mean 1.67 versus 3.48 on control; difference -1.81 (95% CI -2.18 to -1.44), standardized effect d = -3.43; z-test p = <0.001.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Number of Narcotic Pain Medications

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • PEMF Device: mean 20.00 versus 34.00 on control; difference -14.00 (95% CI -15.63 to -12.37), standardized effect d = -5.95; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for PEMF Device than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Levels of Cytokines

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • PEMF Device: mean 168.00 versus 320.00 on control; difference -152.00 (95% CI -171.51 to -132.49), standardized effect d = -5.40; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for PEMF Device than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The secondary outcomes show a consistent descriptive direction: the active-device group has lower posted means than the sham group for both the medication-count and cytokine measures. The recalculated comparisons likewise point in the direction of lower values in the active-device group. This broad pattern is compatible with the primary result's direction, but it should be treated as contextual rather than as separate confirmation.

The registry does not list sponsor-reported statistical analyses for either secondary outcome. The recalculated comparisons are unadjusted, and the registry does not report a multiplicity procedure for considering the primary and secondary outcomes together. When several outcomes are examined, the chance of obtaining at least one apparently strong result can increase, depending on how the tests are related and what analysis plan was specified. Without a reported adjustment or prespecified hierarchy, individual secondary p-values carry less weight than a result arising from a clearly defined primary analysis.

In general, secondary outcomes can help describe whether observed patterns extend beyond the primary measure, but they do not automatically validate it. The number of narcotic pain medications and cytokine levels are different measures from patient-reported pain, recorded at 24 hours rather than the primary endpoint's 12-hour timepoint. The registry does not explain the relationship between these endpoints in a formal analysis plan or report whether the secondary outcomes were confirmatory or exploratory.

These outcomes should therefore be interpreted as additional posted summaries that motivate further scrutiny, not as evidence that every outcome was independently tested under a prespecified confirmatory framework. Details about the cytokine measure, its analysis, and the handling of the secondary outcomes are not reported in the registry record.

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

• Number of Narcotic Pain Medications
• Levels of Cytokines

9. Statistical Interpretation

10. Statistical Methods Explained

The registry reports a randomized, parallel-group comparison of an active device with a sham device. Random assignment is intended, in general, to make the groups comparable on average at the start of a study, including for characteristics that are not measured. It does not guarantee balance in every small trial, and the registry record does not provide baseline characteristics or details about the randomization procedure.

What does the mean difference represent?

The primary outcome is continuous: each participant has a score on a numeric pain scale, and the registry reports a mean for each group. The recalculated mean difference compares the active-device mean with the sham mean. Its estimate is -1.81 units. Because the active-device mean is subtracted from the sham mean in the reported direction? No: the reported estimate corresponds to the active-device mean minus the sham mean, so a negative estimate means the active-device average is lower. The confidence interval uses the same contrast and is entirely below zero.

For a continuous outcome, the mean summarizes the average score but can be affected by unusually high or low individual observations. The standard deviation describes the spread of scores within a group; it is not the uncertainty around the group mean. The posted standard deviations are 0.61 for the sham group and 0.43 for the active-device group. General interpretation of these summaries assumes that the mean is a useful representation of the group and that the analysis is suitable for the distribution of the data. Participant-level distributions are not reported in the registry record.

How should the p-values be understood?

The registry reports a superiority analysis using ANOVA, with p < 0.01. In general, ANOVA compares group means relative to the variability in the data under the analysis model. The registry does not provide the ANOVA statistic, degrees of freedom, model details, or a confidence interval for that sponsor analysis. The posted recalculation gives p < 0.001 using an unadjusted comparison based on the group summaries. A p-value is not the probability that the null hypothesis is true, and it does not measure the size or clinical importance of an effect.

The recalculation is a summary-statistic analysis rather than a reconstruction of the sponsor's full analysis. It cannot account for covariate adjustment, participant-level patterns, or other analysis decisions that are not contained in the posted summaries. No discrepancy is flagged in the registry record, but the sponsor's reported method remains distinct from the unadjusted recalculation.

Why does dispersion matter?

General statistical explanation: the standard deviation describes how much individual observations vary around their group mean. More variability generally makes a mean difference harder to estimate precisely for a fixed sample size. If data vary more than a simple model assumes, a model that ignores that extra variability can understate uncertainty. This issue is often described as over-dispersion in settings where a model expects a particular degree of variation. The posted endpoints are reported as continuous means and standard deviations; the record does not report an over-dispersion assessment or a diagnostic analysis.

For the primary comparison, the recalculation reports a pooled standard deviation of 0.528 and a standardized effect estimate of -3.43. These are derived from the posted group summaries. Standardization expresses the mean difference relative to a measure of within-group spread, but it does not replace the original scale: the pain score and its units remain more directly interpretable clinically. A large standardized estimate in a small study should not be assumed to predict the same magnitude in another sample.

What does randomization establish, and what does it not?

General explanation: randomization supports causal comparison when assignment is implemented as intended and participants are analyzed in a way that preserves the randomized comparison. But the registry record does not describe the allocation sequence, concealment procedures, adherence, exclusions, or analysis population. It identifies the study as randomized and quadruple-masked, and its summary calls it double-blind and placebo-controlled. The record does not explain whether masking remained intact or who was masked.

The registry's posted group counts match the reported enrollment, but that alone does not establish that all randomized participants contributed to every endpoint. Missing-data counts and the treatment of missing outcomes are not reported in the registry record. These details matter because a difference in who contributes data, or how incomplete observations are handled, can affect the estimated contrast.

In general, a confidence interval gives a range of values compatible with the data and analysis model at a stated confidence level. The recalculated interval for the primary mean difference is from -2.176 to -1.444. It describes uncertainty in that unadjusted estimate under the calculation's assumptions. It does not account for uncertainty from unreported design features, selective outcome reporting, or multiple comparisons.

11. Confidence Intervals and Precision

The recalculated primary mean difference is -1.81 units, with a 95% confidence interval from -2.176 to -1.444. The estimate compares the active-device group with the sham group. Because the interval lies below zero, the posted summary data are compatible with a lower mean score in the active-device group under this unadjusted comparison. The interval does not state that the true effect has a particular probability of being inside its limits.

In general, a confidence interval describes sampling uncertainty around an estimate under the model and calculation used. A narrower interval indicates greater precision for that estimate under those assumptions; it does not remove potential bias or establish that the study population represents other patients. Here, the interval is based on group summaries from 16 participants per group. The registry does not report participant-level data, baseline characteristics, or missing-data handling, so the interval cannot account for uncertainties those details might raise.

The sponsor's ANOVA result is reported as p < 0.01, but no confidence interval is reported for that analysis in the registry record. The recalculated interval is attached to the summary-statistic comparison, not the sponsor's ANOVA result. These should be kept distinct when reading the evidence.

The posted standard deviations describe variation among observed scores within each group. They do not measure the precision of the difference directly. Precision of a group contrast depends on the sample sizes, variability, and analysis assumptions. General caution is warranted because a small trial can give an estimate whose apparent precision under a simple calculation does not capture uncertainty about study conduct, representativeness, or analysis choices.

12. Planning a Similar Trial: Sample Size

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

The sample-size figures are planning illustrations expressed as participants per arm, not total events. They use the observed effect as a starting point and show scenarios for detecting the full observed effect, 75% of it, or 50% of it. The stated planning 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.

The 80% power column means the illustrated number of participants per arm for a design with an 80% probability of detecting the specified effect under those planning assumptions, if that effect is real and the assumptions hold. The 90% column uses a higher target probability of detection. These are design calculations, not results from the completed trial. The original trial's own sample-size assumptions and justification are not reported in the registry record.

The curve should be read as a planning aid for how the target effect and required sample size relate, not as a display of observed trial performance. Smaller effects generally require more participants to detect reliably than larger effects, all else equal. The listed illustrations are based on effects observed in this trial, and an observed effect can be unstable, particularly in a small study. The values are not observed power.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = -1.810; pooled standard deviation σ = 0.528.
2. n = 2 × (1.96 + 0.842)² × 0.528² / -1.810² = 1.3, rounded up to 2 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- -1.81                      # observed difference in means
sigma <- 0.527730992078351                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(1262599)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 2 (80% power), 2 (90% power)
 75% of observed effect: n per arm = 3 (80% power), 4 (90% power)
 50% of observed effect: n per arm = 6 (80% power), 8 (90% power)
power.t.test: 2.7 
Simulated power at n = 2 per arm: 0.462 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Analyses Reported by the Sponsor

The sponsor's posted primary analysis is an ANOVA superiority test comparing the two groups, reported as p < 0.01. The registry record does not provide an ANOVA estimate, confidence interval, test statistic, or further model details. This is the sponsor's own inferential result and should be distinguished from the summary-statistic recalculation.

The recalculation uses the posted means, standard deviations, and group sizes. It reports an unadjusted mean difference of -1.81 units, a 95% confidence interval from -2.176 to -1.444, and p < 0.001. The estimate has the same direction as the observed group means: the active-device group mean is lower than the sham group mean. The difference in reported p-value does not by itself imply a contradiction, because the reported sponsor test and the recalculated comparison are not necessarily identical procedures.

No discrepancy is flagged in the registry record. If a discrepancy were identified, the sponsor's own analysis would take precedence over a simple recalculation from summary statistics. In any case, the recalculation cannot reproduce model choices or participant-level analyses that the registry does not report. The most faithful reading is to cite the ANOVA result as the sponsor-reported analysis and use the recalculated estimate and interval as an additional, unadjusted summary-based description.

15. Multiplicity, Missing Data and Analysis Populations

The registry reports a primary endpoint and two secondary outcomes, but a multiplicity plan is not reported in the registry record. Multiplicity refers, in general, to the statistical issue that arises when several hypotheses or outcomes are tested. If many tests are considered, the likelihood of at least one small p-value can rise unless the analysis plan accounts for the number and relationship of the tests. The registry does not indicate whether the secondary outcomes were intended as confirmatory, exploratory, or part of a hierarchy, nor whether any adjustment was planned.

Missing-data handling is also not reported in the registry record. The posted group sizes are 16 for each arm, but the record does not say whether every randomized participant contributed data to each endpoint, whether any observations were missing, or how missing values were handled. Equal posted counts do not alone establish that there were no missing observations or exclusions.

The analysis population is not reported in the registry record. The registry identifies the study as randomized, but does not state whether the sponsor analyzed all randomized participants, only participants who received the assigned device, or another defined group. In general, these choices can affect how closely an analysis preserves the benefits of randomization and how the estimate applies to the participants assigned to treatment.

Details of the randomization sequence and allocation concealment are not reported in the registry record. The masking fields identify quadruple masking, while the narrative calls the study double-blind; the roles masked and the success of masking are not described. Interim analyses and stopping rules are not reported in the registry record. These omissions limit assessment of the safeguards against bias and the intended statistical framework; they do not establish that such procedures were absent.

16. 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.

The registry's adverse-event figures list 0 serious adverse events among 16 participants at risk in the sham group and 0 among 16 at risk in the active-device group. It also lists 0 other adverse events among 16 participants at risk in each group. These are the posted group counts and should be read as recorded in the registry.

Zero recorded events in these groups does not prove that an intervention is risk-free or that an event could not occur in a larger or different population. In general, safety estimates from a small group have limited ability to detect uncommon events. The registry record does not provide detailed event definitions, collection procedures, or a narrative describing how adverse events were monitored.

Deaths are not reported in the registry record. The event counts are not a substitute for the primary pain outcome or for the two secondary measures. They describe the adverse-event information posted for this study and should not be generalized beyond the trial participants and the record's reporting.

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

The results describe a randomized comparison in which the posted mean pain score at 12 hours was lower in the active-device group than in the sham group. The recalculated unadjusted mean difference is -1.81 units, with a 95% confidence interval from -2.176 to -1.444. The sponsor reports an ANOVA superiority analysis with p < 0.01. Together, these statistics describe the direction and precision of the posted primary comparison under the analyses reported.

They do not establish that every participant experienced less pain, because group means summarize rather than display individual responses. They also do not establish that the observed difference applies to people outside the trial population, or that it persists at timepoints beyond the reported 12-hour endpoint. The registry's discussion of donor-site pain and postoperative recovery provides context, but the posted primary result is not a result for hospital stay or recovery duration.

The registry describes two further outcomes at 24 hours, but the record does not list sponsor-reported statistical analyses for them. Their consistent direction is descriptive and should not be treated as independent confirmation, particularly without a reported multiplicity plan. The adverse-event figures record no serious or other events in either group, but the sample size and limited reporting do not establish absence of risk.

In general, statistical evidence is strongest when the design, analysis plan, participant flow, endpoint definitions, and missing-data procedures are transparent and when findings are replicated in relevant settings. Several of those details are not reported in the registry record. The findings should therefore be understood as the results posted for this completed trial, with uncertainty about aspects of conduct and analysis that cannot be assessed from the record.

No treatment recommendation follows from these statistics. They describe group-level outcome data and their calculated uncertainty; judgments about clinical importance require context beyond a p-value or mean difference, and that context is not fully reported in the registry record.

18. Why This Trial Matters Statistically

This trial is statistically informative because it makes a direct randomized comparison between an active device and an inactive sham in a defined postoperative setting. The sham comparison is relevant to a patient-reported pain outcome, where expectations and reporting context can matter. The record also identifies masking, although the difference between its quadruple-masking classification and double-blind narrative is not explained.

The posted data offer more than a significance result: group means, standard deviations, and sample sizes allow an unadjusted estimate of the mean difference and its confidence interval. That distinction matters because a p-value alone does not describe the size or precision of a contrast. The primary confidence interval provides a range of estimates compatible with the summary-data analysis, while remaining limited by what the registry reports.

The trial also illustrates why statistical interpretation depends on the full reporting context. The sponsor's ANOVA result, the recalculated comparison, and the secondary summaries are not interchangeable. The absence of a reported multiplicity plan, analysis population, missing-data handling, and original sample-size assumptions constrains how confidently the numerical results can be interpreted.

Finally, the secondary outcomes and adverse-event counts show that a trial record may include multiple types of evidence, each with different strengths. Descriptive summaries can help characterize patterns, but they do not substitute for prespecified inferential analysis. Reading these elements together helps distinguish the observed data from broader claims that the record cannot support.

19. Limitations

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

The posted enrollment is 32, with 16 participants in each group, so the evidence comes from a small trial. The registry record does not report baseline characteristics, detailed eligibility criteria, participant flow, or the analysis population. Without those details, it is difficult to assess the composition of the analyzed groups or how representative they are of the intended population.

The primary endpoint is reported at 12 hours, while the registry description mentions assessment in comparison with earlier pain timepoints. Numerical results for other primary timepoints are not reported in the registry record. The secondary outcomes are reported at 24 hours, but the record does not provide detailed definitions for the cytokine measure or sponsor analyses for either secondary outcome.

The sponsor reports an ANOVA p-value, but further model details and a sponsor-reported confidence interval are not reported in the registry record. The summary-based recalculations are unadjusted and cannot account for participant-level data or unreported model choices. The original sample-size rationale, missing-data procedures, multiplicity strategy, interim analyses, and stopping rules are also not reported in the registry record.

The record describes masking as quadruple in its design fields and double-blind in its narrative, without clarifying who was masked or whether masking was maintained. The safety figures are limited to the posted event counts and do not include detailed event descriptions or monitoring methods. These limitations concern what can be assessed from the registry record; they do not establish that omitted procedures were not conducted.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Use of Pulsed Electromagnetic Fields (PEMF) After Breast Reconstruction Surgery — NCT01262599.
• Registry record last updated 2018-09-25; status last verified 2018-08.
• Results first posted 2015-08-10.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

23. Record Summary

The registry record describes a completed, randomized, parallel-group, phase 4 trial sponsored by Columbia University, with quadruple masking and enrollment of 32. It compares an Ivivi Torino II PEMF device with an inactive sham device after pedicled TRAM flap breast reconstruction.

The primary outcome is mean visual analog scale pain score at 12 hours. The sponsor reports an ANOVA superiority result with p < 0.01. The unadjusted summary-statistic recalculation reports a mean difference of -1.81 units, with a 95% confidence interval from -2.176 to -1.444 and p < 0.001. Two secondary outcomes are posted at 24 hours; sponsor analyses for them are not reported in the registry record.

The registry lists 0 serious adverse events and 0 other adverse events in each group. The original sample-size justification, analysis population, missing-data approach, and multiplicity plan are not reported in the registry record.