Effectiveness of a Modern Educational Intervention in Breast Cancer Patients: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Effectiveness of a Modern Educational Intervention in Breast Cancer Patients (NCT03351075), a randomized na study. The primary outcome is Self-reported Change in Pain-related Disability.

2. Clinical Question

3. Trial Design

This was a randomized, triple-masked, parallel-group trial. The registry lists an enrollment of 184 and marks the study as completed. Its phase is listed as NA. The sponsor was Universitaire Ziekenhuizen KU Leuven. The registry records a start date of 2017-11-27 and a primary completion date of 2021-03-05.

Both groups received standard physical therapy. The intervention group also received a modern educational program, while the control group received a traditional biomedical educational program. This design compares the two educational approaches in the context of the shared physical therapy program. The registry classifies the study allocation as randomized and the model as parallel. Details about the randomization procedure, how masking was maintained, and which roles were masked are not reported in the registry record.

The primary outcome was change in self-reported pain-related disability between baseline and one year. The record also lists secondary outcomes concerning pain intensity, central sensitization symptoms, somatosensory functions, endogenous pain facilitation and inhibition, and upper limb function. There are 19 outcome measures and 52 analyses recorded, with 18 secondary outcomes posted.

The primary analysis included subjects with missing values at one or more timepoints using likelihood-based estimation. The number of participants assigned to each group, details of treatment exposure, and the analysis population definitions beyond that missing-data statement are not reported in the registry record. The record also does not report a protocol-based sample-size justification, an interim analysis plan, or the detailed statistical analysis plan.

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.

Before assignment: In total, 493 women were eligible, of which 184 were included in the study.

Intensive Treatment Phase (4 mo FU)

Reasons for not completing this period:

Maintenance Treatment Phase (12 mo FU)

Reasons for not completing this period:

End of Trial (18 mo FU)

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Using likelihood-based estimation, subjects with a missing at one or more timepoints were still included in the analysis.

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 Self-reported Change in Pain-related Disability, measured over Change between baseline and one year follow-up assessment.

The primary endpoint was self-reported change in pain-related disability between baseline and the one-year follow-up assessment. It was measured using the Pain Disability Index, which assesses the impact of pain on a person’s ability to participate in essential life activities. The scale ranges from 0 to 70, and higher scores indicate greater disability due to pain.

A change measure asks how the outcome differs between the baseline assessment and a later assessment, rather than describing only the score at one timepoint. Comparing changes between groups is intended to assess whether the groups’ courses over that period differed. The registry reports each group’s mean change with a 95% confidence interval and gives a net mean difference from a multivariate linear model.

Pain-related disability is relevant to the trial’s stated focus because it concerns the effects of pain on daily activities, rather than pain intensity alone. The registry separately lists pain intensity and upper limb function among the secondary outcomes. The interpretation of the primary change depends on the scale’s direction and the group contrast: higher Pain Disability Index scores represent greater disability, but the posted summary does not separately describe the sign convention used for the net difference.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is continuous: the Pain Disability Index produces a numerical score, and the registry reports a mean change. A mean describes the average value in a group. Here, the group means of 4.22 and 5.53 represent reported changes from baseline to the one-year follow-up, not average endpoint scores at one visit. Higher scores on the underlying disability scale indicate greater disability, but the registry summary does not state how the change values were signed or how the net difference was ordered. The reported contrast is therefore best quoted as a net mean difference rather than assigned an unreported direction of benefit.

What does the confidence interval convey?

A confidence interval expresses statistical uncertainty around an estimated quantity. In general, a 95% confidence interval is generated by a method that, under its assumptions, would cover the quantity of interest in 95% of repeated samples. It is not a range containing 95% of individual participants’ scores, and it does not assign a probability to the true effect after the data have been observed. The intervals around the group means describe uncertainty about those means. The interval around the net mean difference describes uncertainty about the between-group estimate.

For the net difference, the interval runs from -5.28 to 2.65 and includes zero. Zero represents no difference in the contrast as parameterized. The interval’s span indicates that the estimate is not highly precise: the data and model are compatible with effects on either side of zero. Interpreting the interval also depends on whether the model assumptions and missing-data assumptions are reasonable; those details are not fully reported in the registry record.

What does the model contribute?

The registry names a multivariate linear model for the primary comparison. In general, a linear model estimates relationships among numerical outcomes and explanatory variables, while a multivariate model can account for more than one outcome or response dimension in a model. The registry does not describe the exact model specification, which variables were included, how repeated measurements were represented, or whether any covariates were adjusted for. Those details should not be inferred from the model label alone.

The posted estimate from that model is a net mean difference of -1.31, with a 95% confidence interval from -5.28 to 2.65 and p = 0.516. The p-value measures how unusual a result at least as incompatible with the null hypothesis would be under a specified null model and its assumptions. It is not the probability that the null hypothesis is true, nor does it measure clinical importance. A p-value also does not replace the estimate and confidence interval, which communicate the scale and uncertainty of the result.

How were missing observations handled?

The registry says that likelihood-based estimation was used and that subjects with a missing value at one or more timepoints were still included in the analysis. In general, likelihood-based methods estimate model parameters using the observed parts of participants’ records, rather than automatically excluding everyone with any missing measurement. Such methods rely on assumptions about the relationship between missingness and observed or unobserved data. The registry does not describe those assumptions, the missingness pattern, or any sensitivity analysis that tests alternative assumptions. It also does not report the number of missing values for the primary outcome.

Including incomplete records is not the same as observing every scheduled measurement, and it does not remove the need to consider whether missingness could affect the result. The interpretation of the reported estimate should therefore retain uncertainty about the missing-data process. The exact analysis population and any additional exclusions are not reported in the registry record.

Why can variability affect a comparison?

In general, greater variation among participants makes group means harder to estimate precisely, all else equal. The posted results give confidence intervals rather than a standard deviation for each group. A confidence interval reflects uncertainty in the estimated mean and depends on factors such as sample size and variability, as well as on the statistical model. The interval widths should not be interpreted as the spread of individual scores.

Over-dispersion is a general concern when observed variability is greater than a model assumes. If a model understates variability, uncertainty estimates can be too narrow and tests can appear more conclusive than warranted. The registry does not report an over-dispersion assessment or indicate that over-dispersion was relevant to this continuous endpoint. It would not be appropriate to assume that such a problem occurred, or that it was assessed, from the information posted.

How should planning illustrations be read?

In general, sample-size calculations based on an observed effect are planning illustrations, not evidence that a trial had a particular probability of detecting an effect. Their results depend on assumptions, including the target effect, significance level, allocation, and statistical approximation. The registry describes its illustration as using a two-sided significance level of 0.05, equal allocation, a normal approximation, and a stated fraction of the observed effect. The illustration is not the trial’s own design justification, and it is not observed power. A protocol-based sample-size justification is not reported in the registry record.

The study has multiple recorded outcomes and analyses. In general, making many statistical comparisons can increase the chance of at least one apparently notable result when each is assessed independently. Methods for controlling that issue depend on the prespecified testing strategy. A multiplicity plan is not reported in the registry record, so the number of analyses alone does not reveal how the investigators handled this concern.

10. Analyses Reported by the Sponsor

The registry reports the primary comparison as the result of a multivariate linear model: a net mean difference of -1.31, with a 95% confidence interval from -5.28 to 2.65 and p = 0.516. This is the sponsor’s reported analysis in the registry record and is the principal between-group estimate to use when describing the posted primary result.

The record does not flag a discrepancy between the reported analysis and a recalculation. It does not provide a separate recalculated estimate or its numerical details in the information available for this account. The group-specific means and confidence intervals are also posted, but they are not a substitute for the model-based net contrast: comparing two separate confidence intervals is not the same as estimating the uncertainty of a between-group difference.

A recalculation from summary statistics, when available, is a simple unadjusted check and may not reproduce a model-based analysis that accounts for features of the study data. The registry’s reported model result should therefore remain central to interpretation. Any planning illustration based on an observed effect is separate from the result itself and should not be read as observed power.

11. Multiplicity, Missing Data and Analysis Populations

The record lists 19 outcome measures, 18 posted secondary outcomes, and 52 analyses. The secondary outcomes address pain-related disability, pain intensity, central sensitization symptoms, touch and temperature-related somatosensory functions, endogenous pain facilitation and inhibition, and upper limb function. The registry does not report a prespecified hierarchy for testing these outcomes, a family of primary and secondary hypotheses, or a multiplicity adjustment plan. Without that information, the posted counts alone cannot show whether or how the investigators controlled the chance of false-positive findings across analyses.

The primary endpoint analysis used likelihood-based estimation. The registry states that subjects with a missing value at one or more timepoints were still included. This is relevant because the analysis was not limited to participants with complete observations at every timepoint. The record does not report the quantity or pattern of missing primary outcome data, the assumptions about missingness, sensitivity analyses, or how missingness was handled for each secondary outcome.

The posted summary also does not specify all analysis populations, exclusions after randomization, or whether the primary analysis followed an intention-to-treat principle. It reports the primary comparison and gives group-specific analysis counts, but those details are not enough to reconstruct every step from enrollment to analysis. Information about allocation concealment, protocol deviations, and adherence to the planned analysis is not reported in the registry record.

These limitations do not change the reported estimate, but they constrain how fully it can be evaluated. The primary finding should be read as the posted model-based result, with the missing-data and multiple-analysis context in mind. The registry does not provide enough detail to determine whether alternative missing-data assumptions or another multiplicity strategy would change the interpretation.

12. 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.

The registry reports serious adverse events affecting 0 participants in the intervention group and 0 participants in the control group. The number at risk for the serious adverse event summary was 92 in each group. No serious adverse event terms are listed among the top serious terms. For other adverse events, 0 participants were affected in each group, with 92 at risk in each group.

The death count was 0 in the intervention group and 1 in the control group. These are descriptive counts in the registry record. The record does not state whether the death was considered related to the study intervention, nor does it provide further detail about its circumstances. A death count alone does not establish causation.

The safety groups each have 92 participants at risk, while the primary outcome analysis lists 82 participants in the intervention group and 84 in the control group. These counts refer to different posted summaries and should not be treated as interchangeable. The registry does not explain the difference in analysis populations or provide additional exposure details in this record. The adverse-event figures should therefore be read as the counts reported, not as a complete assessment of comparative safety or risk.

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

The primary result describes a comparison of mean change in pain-related disability between two randomized groups that both received standard physical therapy and differed in their educational program. The reported net mean difference was -1.31. Its 95% confidence interval, from -5.28 to 2.65, includes zero, and the reported p-value is 0.516. The estimate is uncertain and does not establish a clear between-group difference on this endpoint.

This result does not demonstrate that the groups are equivalent. A nonsignificant result is not proof that any difference is absent. Establishing equivalence or non-inferiority requires a design and decision criteria suited to those questions; such criteria are not reported in the registry record. The estimate and interval also do not identify a single precise effect size.

The result is specific to the primary endpoint, the posted analysis, the participants included in that analysis, and the comparison of the two educational programs alongside standard physical therapy. It does not compare physical therapy with no treatment, and it does not by itself answer whether either program changes every aspect of pain or functioning. Other outcomes are listed, but their results are not included in the primary comparison.

The trial record also does not provide enough information to determine how sensitive the estimate is to missing-data assumptions, how multiple analyses were handled, or how participants’ characteristics and treatment exposure affect generalizability. The result is therefore most accurately described as an imprecise model-based estimate with a confidence interval spanning zero, rather than as a definitive finding about the overall value of either educational approach.

These statistical findings are not clinical advice. They describe what the registry reports for one trial’s primary outcome and do not determine what an individual patient should do.

14. Why This Trial Matters Statistically

This trial addresses a statistical question that is easy to misstate: the comparison is between two educational programs, each delivered alongside standard physical therapy. Recognizing the shared component matters when interpreting the treatment contrast. The randomized parallel-group design is intended to compare outcomes between those groups, rather than to isolate the effect of physical therapy itself.

The primary endpoint is a patient-reported measure of disability associated with pain, with assessment of change from baseline to one year. That focus goes beyond pain intensity alone and connects the measured outcome to the impact of pain on essential life activities. The record also includes outcomes on pain intensity, sensory function, pain facilitation and inhibition, and upper limb function, reflecting several distinct dimensions rather than one interchangeable measure.

The statistical result illustrates why an estimate should be read together with its uncertainty. The model-based difference is -1.31, but its confidence interval extends from -5.28 to 2.65 and includes zero. Reporting the point estimate alone could imply more certainty than the data support; reporting only that the p-value is 0.516 would omit the range of estimates compatible with the analysis.

The registry also makes clear why analysis details matter. The primary analysis included participants with missing observations through likelihood-based estimation, but the missing-data assumptions and sensitivity analyses are not reported. With 19 outcome measures and 52 analyses recorded, the absence of a posted multiplicity plan also limits assessment of how results across outcomes should be interpreted together. These are not reasons to replace the posted result; they are reasons to describe its scope and uncertainty precisely.

15. Limitations

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

The registry summary does not report detailed eligibility criteria or participant characteristics. The extent to which the enrolled sample represents the broader population of people undergoing breast cancer surgery therefore cannot be assessed from this record alone. Treatment details such as the educational program’s duration, delivery, and adherence are also not reported in the registry record.

The primary endpoint is described as change between baseline and one year, but the timing of any other assessments and the handling of repeated measurements are not fully specified. The record names a multivariate linear model without giving its full specification, covariates, or assumptions. The sign convention for the net mean difference is not stated in the posted summary, so the negative estimate should be quoted without assigning it an unreported direction of benefit.

Missing values were addressed using likelihood-based estimation, with participants who had missing data at one or more timepoints retained in the analysis. However, the amount and pattern of missingness, assumptions about the missing-data process, and sensitivity analyses are not reported. These details are needed to assess how robust the estimate may be if missingness is related to observed or unobserved outcomes.

The record lists 19 outcome measures and 52 analyses but does not report a multiplicity strategy or a complete statistical analysis plan. It also does not report a protocol-based sample-size justification or interim analysis plan. Finally, the posted safety counts do not include further details about the single death in the control group or its relationship, if any, to the study. These limitations restrict interpretation of the primary result and the wider set of trial findings.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Effectiveness of a Modern Educational Intervention in Breast Cancer Patients — NCT03351075.
• Registry record last updated 2024-11-21; status last verified 2022-05.
• Results first posted 2024-11-21.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

NCT03351075 is a completed randomized, triple-masked, parallel-group trial sponsored by Universitaire Ziekenhuizen KU Leuven, with enrollment of 184. It compared standard physical therapy plus a modern educational program with standard physical therapy plus a traditional biomedical educational program in a study focused on pain after breast cancer surgery.

The primary endpoint was mean change in self-reported pain-related disability from baseline to one year, measured with the Pain Disability Index. The reported net mean difference was -1.31, with a 95% confidence interval from -5.28 to 2.65 and p = 0.516. The interval includes zero and the estimate is uncertain.

The primary analysis used likelihood-based estimation and retained participants with missing data at one or more timepoints. The registry lists 19 outcome measures and 52 analyses. Details about multiplicity, missing-data assumptions, and several aspects of the statistical plan are not reported in the registry record.