This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for Peripheral Nerve Stimulation for Shoulder Pain: Dose Response (NCT02928055), a randomized na study. The primary outcome is Brief Pain Inventory (BPI)- Short Form (SF) Question 3 (BPI-SF3).
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT02928055 |
| Phase | NA |
| Condition | Stroke, Shoulder Pain |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Single |
| Enrollment | 23 |
| Arms in the primary outcome | n/a |
| Lead sponsor | MetroHealth Medical Center |
| Results posted | Yes |
2. Clinical Question
Population
People with stroke and shoulder pain were the population of interest. The registry describes the pain condition as hemiplegic shoulder pain and reports an enrollment of 23. It does not give further eligibility details or a fuller description of the participants in the record.
Intervention
The intervention was implanted peripheral nerve stimulation delivered with an intramuscular electrical stimulator. The three treatment groups were assigned daily stimulation durations of 3 hr, 6 hr, or 9 hr. Each group description specifies three weeks of treatment following a one week electrode stabilization period.
Comparator
The described comparison is between the three stimulation-duration groups. The registry does not describe a no-stimulation, sham, or usual-care group. It also does not identify a single prespecified pair of groups as the sole comparison for the primary outcome.
Primary question
The stated objective was to explore the dose-response association between muscle contraction from peripheral nerve stimulation and pain reduction, alongside activities of daily living and quality of life. The listed primary outcome is worst pain on the Brief Pain Inventory Short Form Question 3, measured at the end of treatment and at end of treatment plus 3 mo. A numerical result for a specific comparison is not reported in the registry record.
3. Trial Design
The registry identifies this as a randomized, single-masked, parallel-group trial of peripheral nerve stimulation for shoulder pain after stroke. It records an enrollment of 23 and a completed status. The listed study start is 2016-10, and primary completion is listed as 2019-08. The phase is recorded as not applicable. The sponsor is MetroHealth Medical Center.
The parallel groups differ in the daily duration of stimulation: PNS (3 hr/day), PNS (6 hr/day), and PNS (9 hr/day). Each arm description states that stimulation was delivered for three weeks with an intramuscular electrical stimulator, after a one week electrode stabilization period. The record does not specify who was masked, how masking was maintained, or whether participants and outcome assessors were masked.
The study description frames the work as an exploration of mechanisms and associations. In addition to pain relief, it mentions subject-specific clinical and demographic information, sensory and pain-perception measures relevant to central sensitization, activities of daily living, and quality of life. These stated aims provide context for the dose-duration comparison, but the registry does not report numerical findings for those aims in the record described.
Random assignment can support comparisons among assigned groups, but the details needed to assess how assignment was generated or concealed are not reported in the registry record. The record also does not report a statistical analysis plan, a sample-size justification for the trial, or the specific analysis population. Those omissions limit what can be concluded about the planned conduct and analysis beyond the design features explicitly recorded.
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
| Stage | PNS (3 hr/Day) | PNS (6 hr/Day) | PNS (9 hr/Day) |
|---|---|---|---|
| STARTED | 8 | 8 | 7 |
| End of Treatment | 8 | 8 | 6 |
| 1 Month Follow up | 7 | 7 | 6 |
| 2 Month Follow up | 6 | 7 | 5 |
| 3 Month Follow up | 5 | 6 | 5 |
| COMPLETED | 5 | 6 | 5 |
| NOT COMPLETED | 3 | 2 | 2 |
Analysis population for the primary outcome, as defined in the registry: numbers analyzed differ at 3 month follow up due to study drop out of some subjects at 3 month follow up
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.
| Characteristic | Unit | PNS (3 hr/Day) | PNS (6 hr/Day) | PNS (9 hr/Day) | Total |
|---|---|---|---|---|---|
| Participants analysed | 8 | 8 | 7 | 23 | |
| Age, Categorical: <=18 years | Participants | 0 | 0 | 0 | 0 |
| Age, Categorical: Between 18 and 65 years | Participants | 8 | 7 | 4 | 19 |
| Age, Categorical: >=65 years | Participants | 0 | 1 | 3 | 4 |
| Age, Continuous | years | 55.75 (SD 6.25) | 57.5 (SD 8.64) | 62.86 (SD 5.64) | 58.52 (SD 7.35) |
| Sex: Female, Male: Female | Participants | 5 | 4 | 3 | 12 |
| Sex: Female, Male: Male | Participants | 3 | 4 | 4 | 11 |
| Ethnicity (NIH/OMB): Hispanic or Latino | Participants | 0 | 0 | 1 | 1 |
| Ethnicity (NIH/OMB): Not Hispanic or Latino | Participants | 8 | 8 | 6 | 22 |
| Ethnicity (NIH/OMB): Unknown or Not Reported | Participants | 0 | 0 | 0 | 0 |
| Race (NIH/OMB): American Indian or Alaska Native | Participants | 0 | 0 | 0 | 0 |
| Race (NIH/OMB): Asian | Participants | 0 | 0 | 0 | 0 |
| Race (NIH/OMB): Native Hawaiian or Other Pacific Islander | Participants | 1 | 0 | 0 | 1 |
| Race (NIH/OMB): Black or African American | Participants | 4 | 4 | 2 | 10 |
| Race (NIH/OMB): White | Participants | 3 | 4 | 5 | 12 |
| Race (NIH/OMB): More than one race | Participants | 0 | 0 | 0 | 0 |
| Race (NIH/OMB): Unknown or Not Reported | Participants | 0 | 0 | 0 | 0 |
| Region of Enrollment: United States | participants | 8 | 8 | 7 | 23 |
6. Primary Endpoint
The registry lists the primary outcome as Brief Pain Inventory (BPI)- Short Form (SF) Question 3 (BPI-SF3), measured over End of Treatment (EOT), EOT + 3 mo.
| Element | Registry entry |
|---|---|
| Result type | Mean |
| Measure of spread | Full Range |
| Unit | score on a scale |
| Endpoint classification | not reduced to a single comparison |
The primary endpoint is the Brief Pain Inventory Short Form Question 3, identified in the registry as BPI-SF3. It asks participants to rate their worst pain during the prior 7-d on a 0 to 10 numeric rating scale. A score of 0 means “No pain,” and a score of 10 means “Pain as bad as you can imagine.” The registry states that a lower score is better.
The record explains that the Brief Pain Inventory has strong measurement properties and is recommended for pain assessment in clinical trials. It also notes that the developers recommend the BPI-SF3 “pain worst” rating as the primary response metric. This makes the endpoint a patient-reported measure of pain intensity over a specified recall period, rather than a direct measure of muscle contraction or daily activity.
The listed assessment times are end of treatment (EOT) and EOT + 3 mo. These time points address pain at the end of the treatment period and at follow-up. However, the registry notes that the numbers analyzed differ at the follow-up assessment because some participants dropped out. Numerical outcome values and a defined comparison for these time points are not reported in the registry record. The endpoint description alone therefore identifies what was intended to be measured, but does not establish the size or precision of any treatment-group difference.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The record describes a randomized comparison of three peripheral nerve stimulation durations, but it does not reduce the primary outcome to one comparison. The registry explicitly notes that the primary outcome has several categories or time points and is not reduced to one comparison. It reports no analysis entries. As a result, there is no single posted estimate, confidence interval, or test result to interpret for the primary endpoint. Numerical BPI-SF3 results for the groups are not reported in the registry record.
This matters because an endpoint name and measurement schedule are not outcome findings. Knowing that the outcome is worst pain on a 0 to 10 scale, and that lower scores are better, does not indicate how participants’ scores changed or whether groups differed. Without group-level results and a defined contrast, the direction, magnitude, and precision of a treatment-duration association cannot be assessed from the registry record.
The stated scientific objective includes exploring dose response, but that phrase does not itself define a statistical test or establish a trend. A dose-response analysis would need to specify how the treatment durations are compared, what outcome summary is used, and how uncertainty is quantified. The record does not identify a single such primary analysis. It also does not report the statistical plan, so the intended handling of repeated measurements, baseline values, or other covariates is not available.
The follow-up outcome has a further interpretive qualification: the registry says that the numbers analyzed differ at EOT + 3 mo because of dropout. A follow-up estimate based only on observed participants may not represent everyone originally assigned, depending on the missing-data process and analysis method. The record does not report the number analyzed at each time point, reasons for missing outcome data, or methods used to address missingness.
The safety figures are descriptive counts by treatment group, not a substitute for the missing pain analysis. The registry lists serious and other adverse events, as well as zero deaths in each group. These counts can describe events recorded in the study, but they do not by themselves estimate comparative risk or establish a causal safety difference. Group sizes and event counts are limited, and no inferential safety analysis is reported in the registry record.
Overall, the record establishes the trial’s design features, the intended pain measure, follow-up timing, and certain safety counts. It does not provide enough primary-outcome statistics to judge the size or precision of differences among the stimulation-duration groups. The study’s stated aims should therefore be read as research questions, not as evidence that any particular duration produced a specific pain outcome.
9. Statistical Methods Explained
Statistical methods determine how observations are converted into estimates and how uncertainty is represented. The registry record identifies a randomized, single-masked, parallel design, but does not report a statistical analysis plan. It gives no model specification, endpoint summary method, approach to repeated measurements, adjustment strategy, or planned hypothesis test. These details are not reported in the registry record, so the analysis cannot be reconstructed from the design description alone.
What does randomization contribute?
General statistical explanation: random assignment is intended to make treatment groups comparable, on average, with respect to measured and unmeasured characteristics before treatment. It does not guarantee that groups will be balanced in a particular trial, especially when group sizes are limited. Interpretation also depends on following participants as assigned and reporting how outcomes were analyzed. The record identifies randomization, but its sequence-generation and allocation-concealment procedures are not reported in the registry record.
The three parallel groups differ by scheduled daily stimulation duration. A comparison could be framed as a set of pairwise differences, an overall comparison among groups, or a trend across the ordered durations. Those approaches answer different questions and may require different assumptions. The registry does not name one as the primary contrast; it specifically notes that the primary outcome is not reduced to one comparison. It is therefore not appropriate to present an unreported test or treat one possible contrast as the trial’s definitive primary analysis.
How should a treatment effect be read?
General statistical explanation: an effect estimate describes the size and direction of a contrast, while a confidence interval describes the range of values compatible with the data and the model under its assumptions. A narrow interval indicates greater precision than a wide interval, but neither the estimate nor interval alone establishes clinical importance. For the BPI-SF3, a lower score is defined as better, but the registry record does not report group scores, a treatment contrast, or a confidence interval. No effect magnitude or precision can therefore be interpreted.
A p-value, when reported, addresses compatibility between data and a specified null hypothesis under the analysis assumptions; it is not a measure of the probability that a treatment works. The registry record does not report a primary-outcome p-value or identify a single hypothesis test. Without a specified contrast and its estimate, a p-value would not resolve the central question of the size and precision of any duration-related difference.
What does multiplicity mean for this design?
General statistical explanation: multiplicity arises when several outcomes, time points, treatment comparisons, or analyses are examined. Testing many questions can increase the chance of a misleading positive result unless the analysis plan accounts for the number and structure of tests. Here, the registry identifies EOT and EOT + 3 mo for the primary endpoint, describes three treatment durations, and states that the primary outcome has several categories or time points rather than one comparison. The registry does not report a multiplicity strategy or identify which contrast and time point should govern a primary conclusion.
The study description also names exploratory aims involving clinical and demographic information, sensory and pain-perception measures, activities of daily living, and quality of life. The record does not give a complete outcome-specific analysis plan for these aims. Their mention indicates areas of interest, not evidence that all were measured, analyzed, or reported in a particular way.
How do the planning illustrations differ from trial power?
General statistical explanation: a sample-size calculation estimates the number of participants or events that a future study would need under stated assumptions. It is a planning exercise, not an observed result and not a measure of the probability that a completed study has detected a true effect. The planning assumptions described for the illustrations use a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; the effect to detect is expressed as a stated fraction of the observed effect. Time-to-event calculations use a total event count, while other calculations use participants per arm.
Those assumptions are not a report of the trial’s own design justification. The trial itself has three described stimulation-duration groups, while the illustration framework assumes a comparison between two arms. The registry does not identify which pairwise comparison, if any, should be used to interpret such planning quantities. Numerical planning outputs and a trial-specific sample-size justification are not reported in the registry record. Any planning quantity should be described as a hypothetical design illustration, not as observed power or evidence about the trial’s outcome.
How should missing follow-up data be handled?
General statistical explanation: missing outcome data can affect interpretation when the availability of a measurement is related to participants’ outcomes or characteristics. Complete-case analysis uses only observations with recorded outcomes; other approaches can make additional assumptions to use incomplete records. No method is universally appropriate without considering why data are missing and what assumptions are plausible. The registry says the number analyzed differs at EOT + 3 mo because of dropout, but does not report the missing-data method or the information needed to assess those assumptions.
10. Multiplicity, Missing Data and Analysis Populations
The primary outcome is specified at EOT and EOT + 3 mo, and the trial has three stimulation-duration groups. The registry states that the primary outcome has several categories or time points and is not reduced to one comparison. It does not report which treatment contrast or time point was designated as the primary inferential test, or whether any adjustment for multiple comparisons was planned. Without those details, a reader cannot identify a single confirmatory result or determine how multiple opportunities for comparison were handled.
The record notes that the number analyzed differs at EOT + 3 mo because some participants dropped out. It does not report time-specific analysis counts, reasons for dropout, whether participants who discontinued treatment continued outcome assessments, or a method for handling missing values. These details matter because a follow-up mean among observed participants may be affected if missingness is related to pain or other participant characteristics. The missing-data assumptions and any sensitivity analyses are not reported in the registry record.
The analysis population is also not specified. The record does not say whether the primary analysis followed randomized assignment, included only participants who completed treatment, or used another definition. The randomization status of the trial is reported, but that alone does not establish which population was used for the outcome analysis.
Finally, the description lists exploratory objectives involving clinical and demographic information, sensory and pain perception, activities of daily living, and quality of life. The analysis plan for those aims, including any prespecified hierarchy or multiplicity control, is not reported in the registry record. These omissions prevent a complete assessment of how outcome selection, missing data, and multiple testing may affect interpretation.
11. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| PNS (3 hr/Day) | 3 | 8 | 37.5% |
| PNS (6 hr/Day) | 2 | 8 | 25.0% |
| PNS (9 hr/Day) | 0 | 7 | 0.0% |
Deaths
| Arm | Deaths (participants affected of at risk) |
|---|---|
| PNS (3 hr/Day) | 0 of 8 |
| PNS (6 hr/Day) | 0 of 8 |
| PNS (9 hr/Day) | 0 of 7 |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 0% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| PNS (3 hr/Day) | 6 of 8 |
| PNS (6 hr/Day) | 6 of 8 |
| PNS (9 hr/Day) | 7 of 7 |
Most frequent serious adverse event terms
| Event term | Organ system | PNS (3 hr/Day) | PNS (6 hr/Day) | PNS (9 hr/Day) |
|---|---|---|---|---|
| other medical condition: Cancer | Neoplasms benign, malignant and unspecified (incl cysts and polyps) | 2 of 8 | 0 of 8 | 0 of 7 |
| Fall with injury | Musculoskeletal and connective tissue disorders | 0 of 8 | 1 of 8 | 0 of 7 |
| Orthopedic surgery | Musculoskeletal and connective tissue disorders | 0 of 8 | 1 of 8 | 0 of 7 |
| Anxiety(psychosocial) | Psychiatric disorders | 1 of 8 | 0 of 8 | 0 of 7 |
Most frequent non-serious adverse event terms
| Event term | Organ system | PNS (3 hr/Day) | PNS (6 hr/Day) | PNS (9 hr/Day) |
|---|---|---|---|---|
| Skin Irritation at Bandage site | Skin and subcutaneous tissue disorders | 5 of 8 | 2 of 8 | 4 of 7 |
| Accidental dislodgment of Lead | Skin and subcutaneous tissue disorders | 2 of 8 | 2 of 8 | 1 of 7 |
| Skin Irritation at electrode site | Skin and subcutaneous tissue disorders | 1 of 8 | 2 of 8 | 2 of 7 |
| Fall with injury | Musculoskeletal and connective tissue disorders | 1 of 8 | 2 of 8 | 1 of 7 |
| Skin Irritation under Sprint Pad | Skin and subcutaneous tissue disorders | 1 of 8 | 1 of 8 | 0 of 7 |
| Retained Lead Fragment | Skin and subcutaneous tissue disorders | 1 of 8 | 1 of 8 | 0 of 7 |
| Arm fatigue/weakness/pain | Musculoskeletal and connective tissue disorders | 0 of 8 | 0 of 8 | 2 of 7 |
Serious events with the highest arm-level rate: other medical condition: cancer (25.0%), fall with injury (12.5%), orthopedic surgery (12.5%), anxiety(psychosocial) (12.5%).
The registry reports serious adverse events for each stimulation-duration group: 3 among 8 at risk in PNS (3 hr/Day), 2 among 8 at risk in PNS (6 hr/Day), and 0 among 7 at risk in PNS (9 hr/Day). It reports other adverse events for 6 among 8, 6 among 8, and 7 among 7 at risk in those groups, respectively. The record reports 0 deaths in each group.
These are descriptive counts. They should be read with the listed group-specific numbers at risk, rather than as proof that event likelihood differs between groups. The registry does not report an inferential comparison of adverse-event rates or a method for attributing events to treatment. The counts alone do not establish causality, and their interpretation is limited by the small group sizes.
The registry also lists serious-event terms including “other medical condition: Cancer” at 25.0%, “Fall with injury” at 12.5%, “Orthopedic surgery” at 12.5%, and “Anxiety(psychosocial)” at 12.5%. The record does not explain the denominator for these percentages in the listed terms, so they should not be converted into counts or interpreted as group-specific rates. The serious-event counts and the listed terms describe what the registry records; they do not constitute a comparative safety analysis.
12. What the Result Does, and Does Not, Mean
The study addresses a clinically relevant statistical question: whether pain outcomes are associated with different daily durations of peripheral nerve stimulation among people with stroke and shoulder pain. The registry identifies randomized assignment to three stimulation-duration groups and defines a patient-reported pain endpoint. It also describes a follow-up assessment and exploratory interests beyond pain. These features establish the questions the study set out to investigate.
They do not, by themselves, establish the result. Numerical BPI-SF3 outcomes and a primary comparison are not reported in the registry record. The record therefore does not show whether pain scores differed among groups, how large any difference was, or how precisely an estimate was measured. The endpoint’s scoring direction—lower scores are better—cannot substitute for observed group results.
The stated interest in dose response is also not equivalent to a demonstrated dose-response pattern. Such a claim would require reported outcome values and an analysis that defines how the ordered treatment durations are evaluated. The record does not report that analysis or a result for it. Likewise, the descriptions of activities of daily living, quality of life, and central sensitization are study aims, not outcome findings in the information recorded.
The safety counts offer descriptive information about serious events, other events, and deaths. They do not establish that events were caused by stimulation, nor do they provide a formal comparison of safety across durations. At follow-up, dropout means the number analyzed differs; without reported missing-data methods and analysis populations, any available-case result would require cautious interpretation.
Accordingly, the registry record supports a description of the trial’s question, design, endpoint, and recorded safety counts, but not a conclusion about the efficacy or comparative safety of any stimulation duration. The absence of a posted primary comparison should not be mistaken for evidence of no difference; it means that the numerical outcome evidence needed to assess a difference is not reported in the registry record.
13. Why This Trial Matters Statistically
This trial is statistically interesting because its groups vary by treatment duration, making dose-response reasoning a central part of the stated objective. A duration comparison can be framed in more than one way: investigators may compare selected groups, evaluate an overall group effect, or assess an ordered trend. The choice changes the question being answered. The registry’s note that the primary outcome is not reduced to one comparison makes that distinction especially important.
The endpoint also combines a clear patient-reported construct with repeated assessment. BPI-SF3 asks about worst pain during the prior 7-d, and the record identifies both the end of treatment and a later follow-up. Those assessments can speak to different periods in the treatment course. Interpreting them requires knowing which time point is primary, how repeated observations are handled, and how dropout is addressed. Those analysis details are not reported in the registry record.
For biostatisticians and clinical researchers, the record is a reminder that trial design information and outcome evidence are different. Randomization and an endpoint definition describe the intended study structure; they do not provide an effect estimate or its uncertainty. A reader needs the actual group summaries, a clearly defined contrast, and a statistical analysis to evaluate the size, precision, and relevance of an outcome difference.
The record also illustrates why exploratory aims need careful separation from confirmatory claims. It mentions pain, sensory and pain-perception measures, activities of daily living, quality of life, and participant characteristics. When several questions are explored, a prespecified analysis strategy helps readers distinguish primary evidence from exploratory patterns. The registry does not report such a strategy. This limits interpretation but also clarifies exactly what additional reporting would be needed to assess the study statistically.
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 most consequential limitation for interpreting the primary endpoint is the absence of numerical outcome results and a defined primary comparison in the registry record. The endpoint and time points are described, but group-specific scores, estimates of differences, uncertainty intervals, and a test result are not reported. Consequently, the magnitude and precision of any pain difference cannot be assessed.
The trial enrollment is reported as 23, with the groups listed as PNS (3 hr/Day), PNS (6 hr/Day), and PNS (9 hr/Day). The registry does not provide further details in this record about participant characteristics, eligibility criteria, baseline comparability, or the allocation procedures. It also does not identify who was masked or report how masking was maintained. These omissions limit assessment of the trial’s implementation and the context for interpreting self-reported pain scores.
The record states that fewer participants were analyzed at EOT + 3 mo because of dropout, but does not report how many were missing, the reasons for missingness, or the analysis method. It does not specify the primary analysis population or whether missing-data sensitivity analyses were undertaken. As a result, potential effects of incomplete follow-up cannot be evaluated.
Safety information is recorded as group counts, but no inferential comparison or attribution method is reported. The record also does not give a trial-specific sample-size justification, statistical analysis plan, or multiplicity strategy. It describes several aims and assessment times without identifying a single primary comparison. These limitations are specific to interpreting this registry record and do not establish that the trial lacked procedures or analyses; those details are simply not reported in the registry record.
15. Related Tutorials
16. Related Statistical Calculators
17. Sources
- ClinicalTrials.gov: Peripheral Nerve Stimulation for Shoulder Pain: Dose Response — NCT02928055.
- Registry record last updated 2021-06-18; status last verified 2021-06.
- Results first posted 2021-06-18.
- 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.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
18. Record Summary
The completed randomized, single-masked, parallel-group study evaluated peripheral nerve stimulation for shoulder pain after stroke. The registry lists an enrollment of 23 and three daily stimulation durations: 3 hr, 6 hr, and 9 hr. Each group description specifies three weeks of treatment after a one week electrode stabilization period.
The primary endpoint is worst pain on BPI-SF3, rated on a 0 to 10 scale, with lower scores defined as better. Assessment times are EOT and EOT + 3 mo; the record notes dropout-related differences in the numbers analyzed at follow-up. A primary comparison and numerical primary-outcome results are not reported in the registry record.
Adverse-event counts are reported by group, including serious events, other events, and 0 deaths in each group. The registry does not report a statistical analysis plan, a trial-specific sample-size justification, or a multiplicity strategy.