Efficacy and Safety of Oral Salmon Calcitonin in Patients With Knee Osteoarthritis (OA 2 Study): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Efficacy and Safety of Oral Salmon Calcitonin in Patients With Knee Osteoarthritis (OA 2 Study) (NCT00704847), a randomized phase 3 study. The primary outcome is Joint Space Width (JSW) in the Medial Tibia-femoral Knee Joint in the Signal Knee Measured by X-ray Change From Baseline Over 24 Months.

2. Clinical Question

The posted primary comparison is a mean difference in change. A positive difference means the calcitonin group’s mean change was less negative than the placebo group’s mean change. That describes the direction of the recorded group averages; by itself, it does not establish that the difference is clinically important or attributable to treatment. The record does not report a threshold for a clinically meaningful difference.

3. Trial Design

The registry describes a phase III, randomized, double-blind, multicenter, parallel-group, placebo-controlled study sponsored by Nordic Bioscience A/S. Its title identifies the study as the OA 2 Study and describes oral salmon calcitonin in patients with knee osteoarthritis. The registered enrollment is 1030. The study start is recorded as 2008-06, and primary completion as 2011-05. The current registry status is TERMINATED; the reason for termination is not reported in the registry record.

The two registered arms were SMC021 Oral Calcitonin and SMC021 Placebo. For the primary endpoint, the posted arm summaries include 520 participants in the calcitonin group and 508 in the placebo group. These are the endpoint analysis counts, not an additional statement about the overall randomized population. The record defines the primary endpoint population as randomized participants who received at least one dose; it also says 2 randomized participants received no study drug and were excluded.

The primary endpoint was change in medial tibia-femoral joint-space width in the signal knee from baseline to 24 months. The signal knee was selected before randomization, based on eligibility criteria. The registry also lists outcomes relating to bone and cartilage metabolism markers, MRI-assessed knee disease progression, function and physical activity, stiffness, health-related quality of life, and pain. Numerical summaries are posted for the primary endpoint and for two secondary outcomes, stiffness and health-related quality of life. Numerical results for the other listed secondary outcomes are not reported in the registry record.

The registry gives no protocol details about the randomization procedure, masking implementation, adherence, or treatment exposure beyond the study descriptions and endpoint population definition. The statistical plan, original sample-size justification, and reason for termination are not reported in the registry record. Those omissions limit what can be inferred about the intended analysis and the circumstances surrounding the study’s completion.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Overall Study

Analysis population for the primary outcome, as defined in the registry: The number of participants analysed for this outcome is the intent-to-treat (ITT) population. ITT is the number of randomized subjects who received at least one dose of study medication. There were 2 patients in the randomized population who did not receive any study drug and where therefore excluded from the ITT 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 Joint Space Width (JSW) in the Medial Tibia-femoral Knee Joint in the Signal Knee Measured by X-ray Change From Baseline Over 24 Months, measured over Change from baseline to 24 months.

The primary endpoint is change from baseline to 24 months in joint-space width in the medial tibia-femoral joint of the signal knee. The signal knee was chosen before randomization based on the inclusion and exclusion criteria. Joint-space width is the distance between the two bones in the knee joint, measured in mm by X-ray. The registry states that this width decreases with disease progression. It also records a lower eligibility limit of 2 mm and says there was no upper limit as long as the other inclusion and exclusion criteria were met.

This endpoint was used to assess structural change in the knee over the trial’s stated follow-up period. A change score compares the measured width at the later assessment with the baseline measurement. A less negative mean change describes less average narrowing over that interval than a more negative mean change; it does not, on its own, identify why the groups differed. Because the endpoint is an X-ray measurement, it should be distinguished from the separately listed questionnaire outcomes for pain, stiffness, function, physical activity, and health-related quality of life.

The registry reports the endpoint as a mean with standard deviation. The posted arm means are changes, not final joint-space widths. The record does not report a clinically meaningful difference threshold, detailed X-ray reading procedures, or how missing endpoint measurements were handled.

7. Results Posted in the Registry

• SMC021 Oral Calcitonin: mean -0.32 versus -0.40 on control; difference 0.08 (95% CI -0.01 to 0.17), standardized effect d = 0.11; z-test p = 0.080.

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.

Questionnaire to Assess Stiffness in the Signal Knee.

• 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:
  • SMC021 Oral Calcitonin: mean -26.10 versus -20.30 on control; difference -5.80 (95% CI -23.23 to 11.63), standardized effect d = -0.04; z-test p = 0.514.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between SMC021 Oral Calcitonin and the control. The interval gives the range of effects that remain compatible with the data.

Questionnaire to Assess Health-related Quality of Life

• 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:
  • SMC021 Oral Calcitonin: mean 13.70 versus 11.82 on control; difference 1.88 (95% CI -3.00 to 6.76), standardized effect d = 0.05; z-test p = 0.450.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between SMC021 Oral Calcitonin and the control. The interval gives the range of effects that remain compatible with the data.

The registry lists secondary outcomes spanning biological markers, MRI-assessed structural progression, function and physical activity, stiffness, health-related quality of life, and pain. Numerical comparisons are posted for stiffness and health-related quality of life. The record does not report numerical results for the other listed secondary outcomes, so the overall pattern across all the planned secondary domains cannot be assessed from the registry results.

For the two outcomes with posted comparisons, the confidence intervals include zero and the p-values are unadjusted. Their point estimates do not provide a consistent, decisive statistical confirmation of the primary radiographic contrast. At the same time, the secondary measures address different constructs and use percentage-of-change scales; the record does not describe scale scoring sufficiently to interpret every direction of change as favorable or unfavorable. Their findings should not be collapsed into a single overall claim.

Secondary outcomes can add context by examining dimensions not captured by joint-space width. For example, a radiographic measure and a questionnaire measure may capture different aspects of disease experience. Agreement across outcomes could strengthen coherence, while disagreement could prompt further investigation, but neither pattern can be assessed fully when results are missing and analysis methods are not described.

Multiplicity matters because testing many outcomes creates more opportunities for apparently notable results to occur by chance. The registry lists multiple secondary outcomes, but does not report adjustment for multiple comparisons, a hierarchy of endpoints, or a prespecified strategy for interpreting them together. The posted unadjusted secondary results therefore carry less weight than a single prespecified, fully described analysis. They are best treated as descriptive evidence alongside the primary result, not as independent proof that confirms or overturns it.

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

• Bone & Cartilage Metabolism Biochemical Marker Change (Percentage).
• Knee Disease Progression Assessed by MRI
• Questionnaire to Assess Function and Physical Activity
• Questionnaire to Assess Stiffness in the Signal Knee.
• Questionnaire to Assess Health-related Quality of Life
• Questionnaire to Assess Pain

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is continuous: joint-space width is measured in mm, and the registry reports change from baseline as a mean and standard deviation. A mean describes the group’s average change. The standard deviation describes how spread out participant measurements are around that group mean. It is not the uncertainty around the group mean or the treatment contrast.

What does the mean difference represent?

The primary comparison is the calcitonin mean change minus the placebo mean change, as expressed by the posted estimate. Its value is 0.08 mm. Because both group means are negative, a positive difference corresponds to a less negative average change in the calcitonin arm. This is a difference between group averages, not a count of participants whose joint-space width improved or worsened. The record does not report individual participant outcomes, so the distribution of personal changes cannot be reconstructed from the means and standard deviations alone.

What does the confidence interval add?

A confidence interval describes the uncertainty around an estimated contrast under the statistical procedure used to construct it. It gives a range of values compatible with the observed data and the assumptions of that procedure. For the primary comparison, the posted interval runs from -0.009 to 0.170. Since that range includes zero, the summary does not rule out no average difference; it also includes values on either side of zero. The interval’s span conveys that the estimate is not exact. The registry does not report the interval method or confidence level, so those details should not be assumed.

How should the p-value be read?

The primary unadjusted p-value is 0.08. In general, a p-value describes how surprising data at least as inconsistent with a specified null model would be if that model and the analysis assumptions held. It is not the probability that the null hypothesis is true, the probability that the result is due to chance, or a measure of clinical importance. Here, the p-value is explicitly unadjusted. The registry does not report a prespecified analysis threshold or whether this calculation matches the trial’s planned analysis.

Why do variability and analysis details matter?

The pooled standard deviation for the primary comparison is 0.731 mm, and the standardized effect size is 0.109. Standardization expresses a mean difference relative to the spread of the measurements, which can help describe its scale apart from the original unit. It does not by itself establish clinical relevance. The reported values are summary-statistic calculations; the registry reports 0 analyses and does not describe a fitted model, covariate adjustment, or other analysis procedures. A more elaborate analysis should not be inferred from the group summaries.

General statistical explanation: over-dispersion means that observed variability is larger than a model expects. When a model does not account for that extra variability, standard errors and confidence intervals can be too optimistic. The registry does not report a count outcome for the primary endpoint or a model for over-dispersion; the primary endpoint is continuous, so over-dispersion is not a stated feature of this analysis.

What can the secondary comparisons establish?

The posted stiffness and health-related quality-of-life outcomes are also continuous mean changes, expressed as percentage of change. Their reported mean differences, confidence intervals, and unadjusted p-values are descriptive comparisons. The scales are not described in enough detail to infer a clinical interpretation of every direction of change. The registry does not report an analysis plan for these endpoints or adjustments for testing multiple outcomes. They therefore should be treated as exploratory or descriptive evidence rather than independent confirmation of the primary finding.

The meaning of the endpoint population also matters. The registry defines it as randomized participants who received at least one dose of study medication, and says 2 randomized participants who received no study drug were excluded. That definition is not identical to including every randomized participant regardless of treatment exposure. The record does not explain how missing measurements among included participants were handled, whether any additional exclusions occurred, or whether sensitivity analyses were performed.

General statistical explanation: randomized comparisons are intended to preserve the benefits of random assignment. Excluding participants after randomization can affect that balance, depending on the reasons for exclusion and how outcomes are missing. The registry’s brief population description is not enough to assess that possibility. No inference about the effects of missing data or exclusions can be made without the omitted analysis details.

11. Confidence Intervals and Precision

The primary mean difference is 0.08 mm, with a confidence interval from -0.009 to 0.170. The interval extends across zero, so the observed summaries are compatible with a small negative difference, no difference, or a positive difference. The point estimate is only one location within that range. Readers should consider the full interval rather than treating the estimate as an exact effect.

Precision concerns how tightly the estimate is bounded by uncertainty; it is different from whether the observed estimate is important. A narrow interval can still surround an unimportant difference, and a potentially important estimate can be imprecise. The registry does not specify a clinically meaningful threshold for joint-space width, so the interval cannot be classified against such a threshold from the record.

The standard deviations are 0.693 mm for calcitonin and 0.768 mm for placebo. These describe participant-level spread, not the precision of the contrast. The posted standard errors for the group means are 0.030 mm and 0.034 mm, respectively. The uncertainty interval for the contrast brings the comparison into focus, while the separate arm summaries describe each group. These quantities answer related but distinct questions.

The confidence level and construction method are not reported in the registry record. The posted comparison is unadjusted, and there is no reported model or analysis plan with which to verify whether the interval reflects the prespecified trial analysis. The record also does not give a missing-data method or sensitivity analyses. Those limitations qualify how precisely the interval can be interpreted as evidence about the intended trial estimand.

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 planning figures are participants per arm, not total event counts. They are illustrations based on the observed primary effect, rather than the trial’s original sample-size design. The listed assumptions are a two-sided significance level of 0.05, equal allocation between two arms, a normal approximation, and an effect to detect set to the stated fraction of the observed effect. The trial’s original design assumptions and sample-size justification are not reported in the registry record.

For planning at 100% of the observed effect, the illustration gives 1,311 participants per arm for 80% power and 1,755 participants per arm for 90% power. At 75% of the observed effect, the corresponding figures are 2,331 and 3,120 participants per arm. At 50% of the observed effect, they are 5,244 and 7,019 participants per arm. These are hypothetical planning quantities under the stated assumptions, not observed trial enrollment targets or participant counts.

In general, power is the probability that a specified statistical procedure will detect an effect of the size assumed for planning, if that effect is present and the assumptions hold. The 80% column represents planning for an 80% chance of detection under those conditions; the 90% column represents planning for a 90% chance. Seeking higher power or planning to detect a smaller effect requires more participants. The figures are not observed power: observed power is not a useful substitute for interpreting the estimated effect and its confidence interval.

The curve expresses the same planning relationship across assumed effect sizes. Its values depend on the planning assumptions and should not be read as evidence that the trial had the corresponding probability of success. They are illustrations, not a reconstruction of the design decisions made when the study was planned.

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 = 0.080; pooled standard deviation σ = 0.731.
2. n = 2 × (1.96 + 0.842)² × 0.731² / 0.080² = 1,310.8, rounded up to 1,311 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 <- 0.08                      # observed difference in means
sigma <- 0.731023741424528                     # 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(704847)
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 = 1,311 (80% power), 1,755 (90% power)
 75% of observed effect: n per arm = 2,331 (80% power), 3,120 (90% power)
 50% of observed effect: n per arm = 5,244 (80% power), 7,019 (90% power)
power.t.test: 1311.7 
Simulated power at n = 1,311 per arm: 0.804 (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. Multiplicity, Missing Data and Analysis Populations

The record lists a primary outcome and multiple secondary outcomes, but does not report a multiplicity strategy. It is not reported whether secondary endpoints were arranged in a testing hierarchy, whether family-wise error was controlled, or whether any formal adjustment was planned. The posted p-values for the primary and numerical secondary comparisons are unadjusted. Without a prespecified strategy, interpreting several endpoint-specific p-values as if they were separate confirmatory tests can overstate the strength of evidence.

Missing-data methods are not reported in the registry record. It does not state how missing month-24 measurements were handled, whether values were imputed, whether participants with incomplete follow-up were excluded from particular outcomes, or whether sensitivity analyses assessed departures from assumptions. The outcome-specific analysis counts are posted, but those counts alone do not explain why measurements are absent or how missingness might affect the comparisons.

The record defines the primary analysis population as randomized subjects who received at least one dose of study medication. It states that 2 randomized participants who received no study drug were excluded. The registry does not provide further detail about analysis populations, such as whether a per-protocol population was used, or whether the stated definition was applied identically to every outcome. The details needed to assess deviations from randomization or differences between analysis populations are not reported in the registry record.

The registry reports 0 analyses, despite posting outcome summaries and comparison calculations. It does not give the statistical analysis plan, model specification, covariate adjustments, handling of protocol deviations, or interim-analysis procedures. No interim analysis or stopping rule should be inferred. These gaps limit assessment of whether the posted results reflect the planned analysis and how robust they are to alternative reasonable choices.

15. 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 2% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: osteoarthritis (1.6%), arthralgia (1.2%), prostate cancer (1.0%), atrial fibrillation (0.6%), prostatic specific antigen increased (0.6%).

The posted serious-adverse-event counts are 77 among 520 participants in the SMC021 Oral Calcitonin group and 80 among 508 participants in the SMC021 Placebo group. The record also lists other adverse-event counts of 474 among 520 and 459 among 508, respectively. These are counts with events as recorded; the registry summary does not provide a detailed account of event duration, severity grading, relatedness, or how events were collected.

The serious adverse event terms appearing most often in the registry’s aggregate list include osteoarthritis at 1.6%, arthralgia at 1.2%, prostate cancer at 1.0%, atrial fibrillation at 0.6%, prostatic specific antigen increased at 0.6%, and angina pectoris at 0.6%. The list is not broken out by treatment arm in the posted summary, so these percentages should not be assigned to either group. The record does not report death counts. These summaries describe reported events; they do not establish that an event was caused by study treatment or that the treatment was safe or unsafe.

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

The primary estimate describes the difference in average radiographic change between the randomized groups among the endpoint population defined in the registry. The calcitonin group’s mean change was less negative than the placebo group’s mean change, and the posted mean difference was positive. The confidence interval includes zero, leaving uncertainty about the direction and size of the underlying average difference. The result should therefore be understood as an estimate with meaningful uncertainty, not as a categorical finding.

The endpoint measures X-ray joint-space width, not pain, daily functioning, or a participant’s overall experience. A structural difference cannot automatically be translated into a patient-perceived benefit. The registry does not report a clinical-importance threshold or enough scoring information to establish the significance of the secondary questionnaire changes. The record also lacks the planned analysis details and missing-data methods needed to fully evaluate robustness.

These results do not establish that oral salmon calcitonin works or does not work. Nor do they establish that a difference is clinically important, explain why the trial was terminated, or determine the cause of reported adverse events. A confidence interval and p-value address statistical uncertainty under analytical assumptions; they do not resolve those broader questions. The registry record should be read as a limited account of posted results, rather than as a complete clinical or statistical report.

Interpretation also depends on what was measured and who was included. The recorded endpoint population excludes 2 randomized participants who received no study drug, and the registry does not explain how missing outcome measurements among other participants were handled. The potential effect of those choices cannot be quantified from the posted summaries. Conclusions should remain tied to the defined analysis population and the specific radiographic endpoint.

17. Why This Trial Matters Statistically

This trial is statistically informative because it pairs a structural primary outcome with several secondary domains while leaving important analysis details unreported. It illustrates why a point estimate should be considered together with its confidence interval: the primary estimate is positive, but the interval spans zero and allows a range of plausible contrasts. The result’s direction alone is not a sufficient summary of its evidential strength.

It also demonstrates the distinction between measurement variability and uncertainty in a group comparison. The standard deviations describe spread among participant outcomes, while the confidence interval describes uncertainty around the estimated contrast. The difference between those ideas matters when interpreting continuous outcomes such as joint-space width and questionnaire change.

The study record also makes clear why statistical plans matter. Without a reported multiplicity approach, missing-data method, or model specification, readers cannot fully evaluate how the primary and secondary findings were selected and analyzed. Multiple endpoints can provide useful context, but unadjusted secondary results should not be treated as a collection of independent confirmations.

Finally, the planning illustrations show how sample-size needs depend on the effect a future study aims to detect. Smaller assumed effects require more participants under the stated assumptions. These planning calculations are distinct from the trial’s actual results and cannot substitute for the reported effect estimate, its uncertainty, or a transparent account of the original design.

18. 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.

Several limitations are specific to the available registry record. The study status is TERMINATED, but the reason for termination is not reported in the registry record. The trial’s original sample-size justification and statistical analysis plan are also not reported, so the planning illustrations should not be mistaken for the design used in the trial.

The primary outcome has posted arm means, standard deviations, and an unadjusted comparison, but the record does not describe the confidence-interval method, model, covariate adjustment, X-ray reading procedures, or a clinically meaningful threshold. It also does not report the handling of missing measurements, sensitivity analyses, interim analyses, or the role of protocol deviations. These omissions limit assessment of the precision and robustness of the result beyond the posted summary.

The registry lists six secondary outcomes, but numerical comparisons are posted for only stiffness and health-related quality of life. The record does not report results for the remaining listed secondary outcomes, nor does it provide a multiplicity strategy or sufficient scoring detail for interpreting the direction of every questionnaire change. A complete account of consistency across outcome domains is therefore unavailable.

The safety summary includes group counts and an aggregate list of serious adverse event terms, but the term percentages are not allocated by arm in the posted information. Death counts are not reported. The record also does not provide detailed event definitions, relatedness assessments, or follow-up context. These limitations prevent a full comparative safety assessment from the registry summary alone.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Efficacy and Safety of Oral Salmon Calcitonin in Patients With Knee Osteoarthritis (OA 2 Study) — NCT00704847.
• Registry record last updated 2019-04-24; status last verified 2019-04.
• Results first posted 2012-09-12.
• 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.

22. Record Summary

NCT00704847 is a phase III randomized, double-blind, multicenter, parallel, placebo-controlled study of oral salmon calcitonin for knee osteoarthritis. Registered enrollment is 1030; the study start is 2008-06, primary completion is 2011-05, and the registry status is TERMINATED.

The primary outcome was mean change in medial tibia-femoral joint-space width in the signal knee from baseline to 24 months. The posted mean difference was 0.08 mm, with a confidence interval from -0.009 to 0.170 and an unadjusted p-value of 0.08. Numerical secondary comparisons are posted for stiffness and health-related quality of life. The record does not report the reason for termination, original sample-size justification, statistical analysis plan, multiplicity strategy, or missing-data methods.