Empowering Patients On Choices for Renal Replacement Therapy (Aim 3) (EPOCH-RRT): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Empowering Patients On Choices for Renal Replacement Therapy (Aim 3) (EPOCH-RRT) (NCT02488317), a randomized na study. The primary outcome is Preference for Shared Decision Making.

2. Clinical Question

The clinical question concerns the decision-making process rather than a clinical outcome such as survival or a change in kidney function. The primary measure is titled Preference for Shared Decision Making and uses a pick-one approach to classify a participant’s preferred role. The study description says decision-making outcomes, including self-efficacy, were to be compared between patients receiving and not receiving a decision aid. However, the registry record does not report a numerical result for the primary outcome or a between-arm estimate.

The registry record describes an aim to compare measures related to decision-making between patients receiving and not receiving the aid. The question is therefore about whether the decision aid is associated with a different pattern of stated decision-role preferences, not whether one dialysis treatment is superior to another. No conclusion about the direction or size of any difference can be drawn from the registered outcome description alone.

3. Trial Design

The registry describes EPOCH-RRT Aim 3 as a randomized, parallel-group study. The intervention is a decision aid, classified in the record as an “other” intervention. One arm is described as receiving the tool; the control arm is described as not receiving it and testing knowledge without it. The record gives an enrollment of 234 and lists chronic kidney disease as the condition. It lists the phase as NA and the study status as completed.

The stated study aim is to compare measures related to decision-making between patients receiving and not receiving an aid focused on Peritoneal Dialysis and Hemodialysis. The record notes that patients, caregivers, and patient advocacy organizations were included as research partners to help address questions relevant to patients facing dialysis. That description explains the study’s patient-centered motivation, but it does not provide an account of how the aid was delivered or how adherence to the assigned approach was assessed.

The record lists a study start of 2015-05 and a primary completion of 2015-12. Arbor Research Collaborative for Health is listed as the sponsor. Although the allocation is identified as randomized, details of the randomization procedure, concealment, blinding, and handling of departures from assigned treatment are not reported in the registry record. The record also does not report the setting or eligibility criteria in the information available for this analysis.

Random allocation can support a fair comparison when assignment is implemented as intended and outcomes are collected comparably. That general rationale does not establish that the groups were balanced in this trial, nor does it establish that every participant received or avoided the aid as planned. Without outcome estimates and fuller analysis details, the design description identifies the intended comparison but does not establish its observed findings.

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.

Population: Baseline information was collected for 140, but only 133 participants completed the study.

6. Primary Endpoint

The registry lists the primary outcome as Preference for Shared Decision Making, measured over 6 months.

The primary outcome is titled Preference for Shared Decision Making and is assessed at 6 months. The record describes use of the Control Preferences Scale, with a pick-one approach to identify whether a patient prefers an active, passive, or collaborative role in decision-making about dialysis treatment. The outcome is recorded as a number of participants rather than as a continuous score.

This type of measure is used to describe the role a patient prefers in a healthcare decision. The three categories distinguish taking an active role, leaving the decision more to others, or making the decision collaboratively. These categories represent different preferences; they are not inherently a ranking of better and worse choices. A count for each category can describe how preferences are distributed, while a comparison between arms would require the registry to report the corresponding results and an analysis method.

The registry record gives the measure’s rationale as a clinically relevant, easily administered, valid, and reliable assessment of preferred roles. Those statements describe the scale in the record; they are not trial-specific evidence that the measure was completed by every participant or that the study found a particular preference pattern. The primary-outcome result is not reported in the registry record. The record also does not provide category-specific counts, missing-outcome counts, or an estimate of the difference between arms.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does randomization contribute?

General: Randomization assigns participants to comparison groups using a chance-based process. When implemented appropriately, it can reduce systematic differences between groups at the start of a study, including differences that researchers did not measure. It does not guarantee identical groups, and it does not by itself prevent differences in follow-up, treatment delivery, or outcome recording after assignment. The registry identifies this study as randomized, but details of the sequence generation and concealment are not reported in the registry record.

The parallel design means the intended comparison is between groups assigned to different approaches, rather than a description of one participant switching between approaches. The registry record describes a decision-aid arm and a no-aid control arm. It does not report whether the primary analysis followed assignment regardless of what participants actually received, or whether another analysis population was used.

How should a categorical endpoint be understood?

General: A categorical outcome places each recorded response into one of a set of labels. Here, the record describes a pick-one preference among active, passive, and collaborative decision roles. Counts in those categories describe how many participants fall into each category; they do not behave like measurements on a continuous scale where a mean difference is automatically meaningful. A comparison should respect the categories and explain what contrast is being estimated.

The registry characterizes the primary outcome as a number of participants, while the listed outcome has several categories. The record explicitly notes that the primary outcome has several categories or time points and was not reduced to one comparison. No category-specific results or statistical model are reported. It is therefore not possible to identify a particular contrast, judge its uncertainty, or determine whether category ordering or grouping was used in an analysis.

What do estimates and confidence intervals convey?

General: An effect estimate describes the observed contrast between groups according to a specified analysis. A confidence interval describes the range of values reasonably compatible with the data and the assumptions of that analysis. A narrow interval indicates greater precision than a wide interval, but its interpretation depends on the endpoint and the scale of the estimate. A confidence interval is not a guarantee that the true value lies within its endpoints, and it does not replace consideration of the clinical meaning of the contrast.

No primary-outcome estimate or confidence interval is reported in the registry record. There is consequently no basis for describing the magnitude, direction, or precision of an observed decision-preference difference. A p-value is also not reported. In general, a p-value assesses how compatible data are with a specified null model under the assumptions of a test; it does not measure the size or importance of an effect. Without an identified analysis and its results, statistical significance cannot be assessed.

What do the planning assumptions mean?

General: A sample-size calculation asks how many participants would be needed under specified assumptions to detect a target effect with a chosen error rate and power. It is a planning exercise, not an observed trial outcome. The registry’s stated illustration assumptions are a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation. The target effect is described as a stated fraction of the observed effect. The record cautions that participant counts are per arm for non-time-to-event outcomes and that total event counts are used for time-to-event outcomes.

Those assumptions do not reveal the trial’s original design rationale or prove that the study was planned using them. The registry record does not report the sample-size justification, a target effect for the primary endpoint, or a primary-outcome estimate from which to interpret the effect-based illustration. Any planning value based on an observed effect should not be called observed power. Observed power is not a substitute for reporting the estimate and its uncertainty.

Why can event counts need more than a simple comparison?

General: For count outcomes, the choice of model depends on how counts arise and how variable they are. If the data vary more than a simple model expects, over-dispersion can make standard errors too small and confidence intervals too narrow unless the analysis accounts for that extra variation. For categorical preferences, the modeling question is different: the analysis needs to handle the category structure and define a meaningful comparison between groups. These are general considerations, not descriptions of methods used in this trial.

The registry lists 0 analyses, so the statistical methods for the primary endpoint are not reported in the registry record. It is not possible to determine whether the investigators used a model-based comparison, a test of category distributions, an adjusted analysis, or another method. The record also does not identify covariates, assumptions checks, or a strategy for handling incomplete or inconsistent responses.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists 5 outcome measures and 0 analyses. The primary outcome itself has several response categories, and the record notes that it was not reduced to one comparison. A set of outcomes or possible category comparisons can raise multiplicity concerns: when many hypotheses are tested, the chance of at least one chance finding can increase. That is a general statistical issue, not evidence that a particular multiplicity problem occurred in this study.

The registry record does not report a hierarchy of primary and secondary analyses beyond identifying the primary outcome, nor does it report a multiplicity adjustment or a prespecified testing strategy. Without analysis results, it is not possible to assess how the study handled multiple outcomes or comparisons. The record’s listing of outcome measures alone does not show that each was analyzed or tested.

Missing-data handling is also not reported in the registry record. The record does not give the number of participants with completed primary-outcome responses, the amount or pattern of missingness, or an imputation or sensitivity-analysis approach. Missing outcomes can affect both the precision and validity of a comparison, particularly if the likelihood of missingness differs by arm or relates to participants’ preferences. Those possibilities are general concerns; the record does not establish that missing data occurred or how they were handled.

The analysis population is not reported in the registry record. It is therefore unclear whether the intended comparison was based on randomized assignment, receipt of the decision aid, or another defined group. Details on exclusions after randomization and on adherence to assigned approaches are also not reported. These omissions limit interpretation of how any future or external estimate would relate to the randomized comparison.

11. 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 reports 0 serious adverse events and 0 other adverse events in the intervention arm, and the same counts in the control arm. For each event category, 117 participants at risk are listed in each arm. The record’s deaths field is blank; deaths are not reported in the registry record.

These entries are counts in the registry’s adverse-event fields, not a complete account of all possible harms or burdens. A zero count means that no event was entered for that category in the posted results; it does not establish that no event occurred outside the recorded categories, that every participant was assessed equally, or that the intervention is free of risk. The record does not provide event definitions, collection procedures, follow-up completeness, or a comparison estimate with uncertainty.

Serious-event counts and other-event counts should be read separately, and the listed at-risk group should not be mistaken for the number who experienced an event. These safety entries do not answer the primary question about preferences for shared decision-making. They also do not provide evidence about clinical outcomes of dialysis or the broader experience of using the decision aid.

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

The registry record documents a randomized, parallel comparison of a decision aid with no decision aid and defines a primary measure of preferred decision-making role at 6 months. That information clarifies what the study set out to compare. The record does not report the primary-outcome results, so it cannot show whether the intervention arm had a different distribution of active, passive, or collaborative preferences from the control arm.

Without category counts or an estimated contrast, the direction and magnitude of any difference are unknown. Without a confidence interval or other uncertainty measure, the precision of an effect cannot be judged. Without an analysis description, it is also unclear how the categories were compared, which participants contributed to the analysis, or how missing responses were handled. The completed study status does not fill those gaps.

The posted safety counts are limited to the serious and other adverse-event categories in the record. They show 0 events recorded in each category for each arm, with 117 at risk listed for each category in each arm. They do not establish absence of all harms or provide a full assessment of intervention burden. Deaths are not reported in the registry record.

Even if a difference in preferences were reported, it would describe the measured decision-role outcome. It would not by itself establish that participants made better-informed decisions, that their choices matched their values, or that any particular dialysis treatment was preferable. Those conclusions would require suitable measures and evidence beyond the endpoint definition. The registry’s study description mentions self-efficacy as an example of a decision-making outcome, but no result for that measure is reported in the information available.

The appropriate interpretation is therefore bounded: the record identifies a patient-centered question and a randomized design, but the posted information does not allow an assessment of the primary comparison’s result or uncertainty. No claim that the aid did or did not change preferences is justified by the outcome information reported in the registry record.

13. Why This Trial Matters Statistically

This trial addresses a statistical question that differs from the usual comparison of clinical events or treatment efficacy. Its primary endpoint concerns patients’ preferred role in a decision about dialysis treatment. Such a measure is inherently categorical and patient-centered: the outcome describes whether a person favors an active, passive, or collaborative role, rather than assigning a clinical score to the decision itself.

That distinction matters when interpreting results. Category counts and their distribution can be informative, but the statistical contrast must be stated clearly. A comparison of each category separately, a global comparison of the overall distribution, and a contrast that combines categories answer different questions. The registry record notes that the primary outcome has several categories and was not reduced to one comparison, while listing 0 analyses. This makes the absence of a stated analysis strategy particularly important for interpretation.

The randomized design indicates an intended comparison between receiving and not receiving the decision aid. Randomization is useful for reducing the influence of baseline confounding in expectation, but the strength of the resulting evidence also depends on outcome completion, adherence, analysis choices, and transparent reporting. Those details and the primary-outcome findings are not reported in the registry record.

The registry’s patient-partnership description also highlights a broader measurement issue: a statistically clear difference in a preference category would still need to be interpreted in terms of what the categories mean to patients. Statistical significance alone would not determine whether a difference is important or whether it reflects a better decision process. In the absence of outcome estimates, even that interpretation cannot begin.

The trial therefore matters statistically because it illustrates why endpoint structure, analysis specification, and complete reporting are essential when the target is a preference rather than a conventional clinical event. The record defines the question, but the reported information is insufficient to evaluate its answer.

14. 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 main limitation for interpretation is that the primary-outcome result is not reported in the registry record. No category counts, between-arm estimate, confidence interval, or p-value are available. This prevents assessment of direction, magnitude, and precision. The record also lists 0 analyses, so the method used to compare the preference categories is not reported.

The endpoint is described as a pick-one measure with active, passive, and collaborative roles, but the record does not report how responses were distributed, whether participants could change their preferences, or how incomplete responses were treated. Missing-data amounts and methods are not reported. The analysis population, post-randomization exclusions, and adherence to the assigned approach are also not reported, limiting assessment of how the planned randomized comparison was implemented.

Additional design details—including the randomization procedure, concealment, blinding, and setting—are not reported in the registry record. The available description does not establish whether outcome collection was complete or comparable between arms. The record gives enrollment and dates, but these details alone do not explain how many participants contributed usable primary-outcome data.

The safety entries are similarly limited. The record reports zero serious and other adverse events in each arm, but does not report event definitions, collection methods, or deaths. These fields cannot establish that all relevant harms or burdens were captured. The record also does not provide results for the other listed outcome measures or a multiplicity strategy. These limitations constrain interpretation of the registry information; they do not prove that a particular problem occurred in the conduct of the study.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Empowering Patients On Choices for Renal Replacement Therapy (Aim 3) (EPOCH-RRT) — NCT02488317.
• Registry record last updated 2017-05-10; status last verified 2017-04.
• Results first posted 2016-12-28.
• 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.

18. Record Summary

EPOCH-RRT Aim 3 is listed as a completed, randomized, parallel study sponsored by Arbor Research Collaborative for Health. The registry identifies chronic kidney disease as the condition, lists enrollment of 234, and describes a decision-aid arm and a control arm without the aid. The study aim is to compare decision-making measures related to choices about Peritoneal Dialysis and Hemodialysis.

The primary outcome is Preference for Shared Decision Making, assessed at 6 months using a pick-one classification of active, passive, or collaborative decision roles. The registry lists 5 outcome measures and 0 analyses. It does not report the primary-outcome result, an effect estimate, or inferential analysis details.

The adverse-event record lists 0 serious and 0 other events in each arm, with 117 participants at risk listed for each category in each arm. Deaths are not reported in the registry record. The available information describes the intended design and endpoint, but does not establish the observed effect of the decision aid on decision-role preferences.