Discontinuation of Disease Modifying Therapies (DMTs) in Multiple Sclerosis (MS): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Discontinuation of Disease Modifying Therapies (DMTs) in Multiple Sclerosis (MS) (NCT03073603), a randomized phase 4 study. The primary outcome is Number of Participants Developing a New MS Relapse and/or MRI Brain Lesion Over the Course of the Study Duration.

2. Clinical Question

3. Trial Design

The registry describes a Phase 4, randomized, single-masked, parallel-group study sponsored by the University of Colorado, Denver. The record lists enrollment of 259 and a completed status. Participants were assigned to a Drug Continuation Arm or a Drug Discontinuation Arm. In the continuation strategy, participants remained on their current disease-modifying therapies without changes. In the discontinuation strategy, participants stopped those therapies, with no other treatment changes described in the arm summary.

The study’s background focuses on MS and the possibility that relapses and new MRI changes become less common as people age, especially among people aged 55 or older who have had no relapse or MRI change for a prolonged period. This is the rationale stated in the registry summary, not a substitute for the full eligibility criteria. The exact criteria for age, prior stability, and medication use are not reported in the registry record. The identity of the person or group masked is also not reported in the registry record.

The primary outcome was assessed over 18-24 months, based on time of enrollment. The registry describes the primary measure as the count of participants who developed a new relapse and/or a new brain MRI lesion. It also lists disability progression confirmed at 6 months and several Neuro-QoL domains among secondary outcomes. The record identifies 21 outcome measures and has 19 secondary outcomes posted.

The enrollment figure is a study-level registry figure; the primary outcome has arm-specific denominators of 128 and 131. Other outcomes have their own denominators, which differ. The registry record does not report the original design assumptions, the planned analysis population, or the detailed schedule and conduct of all assessments beyond the outcome descriptions.

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

Reasons for not completing this period:

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 Number of Participants Developing a New MS Relapse and/or MRI Brain Lesion Over the Course of the Study Duration, measured over 18-24 months, based on time of enrollment.

The primary endpoint counts participants who experienced at least one of two kinds of disease activity: a new MS relapse or a new brain MRI lesion. A participant meeting either condition is counted in the outcome; the endpoint is not a count of the total number of relapses or lesions. The registry describes the result as the proportion in each group with a new relapse and/or new lesion over the study duration.

The stated timeframe is 18-24 months, based on time of enrollment. The combined endpoint allows the study to capture clinical disease activity and MRI evidence within one participant-level measure. In general, combining outcome components can make an endpoint more inclusive of different manifestations of the condition, but it also means the overall result does not by itself reveal how much each component contributed. Separate component-specific primary results are not reported in the registry record.

The posted counts are 6 of 128 in the Drug Continuation Arm and 16 of 131 in the Drug Discontinuation Arm. The comparison reports a risk difference of 0.075, with the discontinuation group as the comparison group against continuation. These counts describe the observed outcomes in the registry record. They do not, on their own, establish why events occurred or whether the pattern would be the same in other populations, follow-up periods, or treatment settings.

7. Results Posted in the Registry

• Drug Discontinuation Arm: 12.2% (16/131) versus 4.7% (6/128) on control; risk difference 7.5 percentage points (95% CI 0.8 to 14.2), risk ratio 2.61 (1.05–6.45); unadjusted two-proportion z-test p = 0.030.

8. Secondary Efficacy Endpoints

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

Number With Disability Progression Confirmed at 6 Months Using the Expanded Disability Status Scale (EDSS)

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Drug Discontinuation Arm: 12.3% (16/130) versus 11.1% (14/126) on control; risk difference 1.2 percentage points (95% CI -6.7 to 9.1), risk ratio 1.11 (0.56–2.17); unadjusted two-proportion z-test p = 0.766.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Drug Discontinuation Arm and the control. The interval gives the range of effects that remain compatible with the data.

Change in Neuro-QoL (Quality of Life) Short Form Scores -- Upper Extremity Function

• 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:
  • Drug Discontinuation Arm: mean -0.60 versus -1.10 on control; difference 0.50 (95% CI -1.08 to 2.08), standardized effect d = 0.08; z-test p = 0.535.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Drug Discontinuation Arm and the control. The interval gives the range of effects that remain compatible with the data.

Change in Neuro-QoL (Quality of Life) Short Form Scores -- Lower Extremity Function

• 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:
  • Drug Discontinuation Arm: mean -1.10 versus -0.20 on control; difference -0.90 (95% CI -2.29 to 0.49), standardized effect d = -0.17; z-test p = 0.204.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Drug Discontinuation Arm and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes cover disability progression and several quality-of-life domains. The posted estimates do not form a clear, consistent pattern of differences across those varied measures. The disability progression recalculation has a small positive risk difference with an interval spanning zero. The two posted Neuro-QoL mean differences point in different directions, and their intervals also span zero. These outcomes add context about other aspects of participants’ experience but do not resolve the primary endpoint’s analysis discrepancy.

Secondary measures address distinct questions from the primary combined relapse/MRI outcome. Disability progression and quality-of-life scores are not interchangeable with new relapse or MRI lesion occurrence. Similarity or difference in one measure cannot be assumed to establish a corresponding result in another. The available secondary summaries therefore should be read as complementary outcomes, not as substitutes for the prespecified primary question.

Unadjusted secondary comparisons carry less inferential weight when many outcomes are examined. In general, considering many tests increases the opportunity for a seemingly notable result to arise by chance, even if each test uses a conventional threshold. This does not make secondary outcomes uninformative; it means that isolated results need cautious interpretation and should be considered alongside the full pattern, uncertainty, and analysis plan. The registry identifies 21 outcome measures and 19 posted secondary outcomes, but a multiplicity strategy is not reported in the registry record.

The record’s posted comparisons are aggregate summaries, and the participant denominators vary across outcomes. It does not report a single secondary analysis population or detailed missing-data procedures. The observed pattern is consequently best treated as descriptive evidence across several domains, with uncertainty and analysis multiplicity limiting strong claims.

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

• Number With Disability Progression Confirmed at 6 Months Using the Expanded Disability Status Scale (EDSS)
• Change in Neuro-QoL (Quality of Life) Short Form Scores -- Upper Extremity Function
• Change in Neuro-QoL (Quality of Life) Short Form Scores -- Lower Extremity Function
• Change in Neuro-QoL (Quality of Life) Short Form Scores -- Fatigue
• Change in Neuro-QoL (Quality of Life) Short Form Scores -- Sleep Disturbance
• Change in Neuro-QoL (Quality of Life) Short Form Scores -- General Concerns
• Change in Neuro-QoL (Quality of Life) Short Form Scores -- Executive Function
• Change in Neuro-QoL (Quality of Life) Short Form Scores -- Communication

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is binary at the participant level: each participant either met the definition of a new relapse and/or MRI lesion during follow-up or did not. For a binary endpoint, a proportion describes the share of the analyzed group with the event. The posted primary proportions are 0.047 for continuation and 0.122 for discontinuation. The associated counts and denominators are 6 of 128 and 16 of 131.

What does the risk difference express?

A risk difference subtracts one group’s event proportion from the other group’s proportion. Here the comparison is discontinuation against continuation, and the recalculated estimate is 0.075. Its positive direction corresponds to a higher observed event proportion in the discontinuation arm. This measure is expressed on the proportion scale, so it describes an absolute contrast rather than a relative multiplier. The confidence interval gives a range of values compatible with the recalculation’s assumptions and data; it does not remove uncertainty or establish that the observed difference will recur in another study.

Why are there two sets of inferential results?

The registry posts a sponsor analysis using an exact binomial test for a difference in proportions. It identifies that analysis as non-inferiority and reports an estimate of 0.075, an interval from 0.006 to 0.150, and p = 0.521. A separate simple, unadjusted recalculation from the posted group summaries yields an interval from 0.008 to 0.142 and p = 0.03. The estimates are numerically similar, but the inferential results differ. The record flags a discrepancy; therefore the sponsor’s analysis takes precedence. The recalculation is not a correction to the sponsor’s result and should not be used to overrule it.

In general, an exact binomial procedure is based on the probability model for binary event counts. Different procedures can produce different intervals and p-values, particularly when counts are limited or when the test is designed for a different objective. Here the sponsor labels the test non-inferiority, whereas the recalculation is a simple unadjusted comparison. The non-inferiority margin, detailed hypotheses, and analysis implementation are not reported in the registry record. Consequently, the reported p-value cannot be interpreted as a stand-alone proof of non-inferiority, and the recalculation cannot be assumed to reproduce the sponsor’s full analysis.

What do superiority and non-inferiority mean?

In general, a superiority analysis asks whether the data support a difference in a specified direction or any difference, depending on the test. A non-inferiority analysis asks whether a strategy is not worse than a comparator by more than a prespecified margin. These answer distinct questions. A non-inferiority finding depends on the margin, the planned analysis, and how uncertainty relates to that margin. The registry record reports a sponsor analysis classified as non-inferiority, but does not report the margin or enough of the plan to assess the conclusion independently. The recalculated p-value is not evidence that the sponsor’s non-inferiority analysis should be replaced by a superiority test.

How should the secondary comparisons be read?

The disability progression outcome is also binary, while the Neuro-QoL outcomes are continuous scale changes. For a continuous measure, a group mean summarizes the average change, and the standard deviation describes variation among observed values. A mean difference contrasts the group averages. The posted Neuro-QoL comparisons include intervals that extend across zero, indicating that the recalculated estimates are compatible with values on either side of no mean difference under the calculation’s assumptions. The registry also posts sponsor p-values from two-sided t-tests for the Neuro-QoL outcomes. As with the primary endpoint, an unadjusted recalculation and a sponsor analysis should not be treated as interchangeable when their methods or results differ.

For binary outcomes, the unadjusted p-value tests compatibility with a specified null contrast under its assumptions. For continuous outcomes, the t-test evaluates a difference in means under its modeling assumptions. A p-value is not the probability that a hypothesis is true, nor does it measure the size or practical importance of an effect. Estimates and intervals are needed to understand direction and precision. Multiple secondary outcomes also raise the chance of apparently notable findings when many comparisons are considered; the record does not report a multiplicity adjustment.

All recalculated comparisons are simple and unadjusted, based on posted summary statistics. They do not account for covariates, time to event, or other participant-level features. Whether the sponsor used additional details not captured in the registry summaries is not reported in the registry record. The randomized allocation is part of the study design, but the posted aggregate results alone do not show adherence, missingness patterns, or the exact population included in each sponsor analysis.

11. Confidence Intervals and Precision

Confidence intervals show the range of effect estimates reasonably compatible with the observed data under a particular statistical method and its assumptions. They are not guarantees that the true effect lies inside the interval, nor do they express the probability that a specific hypothesis is true. A wider interval indicates less precision than a narrower interval on the same scale. Precision depends on the amount of information in the data and on the analysis method.

For the primary comparison, the unadjusted recalculation reports a risk difference of 0.075 and an interval from 0.008 to 0.142. The sponsor’s reported non-inferiority analysis gives the same stated estimate, 0.075, but an interval from 0.006 to 0.150. The difference between these intervals is relevant: they arise from analyses that do not produce the same inferential result. The registry flags the discrepancy, so the sponsor’s own analysis takes precedence. The non-inferiority margin is not reported in the registry record, preventing assessment of the sponsor interval against that decision threshold.

The intervals around the secondary comparisons also indicate uncertainty. For the disability progression risk difference, the recalculated interval extends from -0.067 to 0.091. For the Neuro-QoL mean differences, the posted recalculated intervals extend from -1.081 to 2.081 for upper extremity function and from -2.288 to 0.488 for lower extremity function. These ranges include values on both sides of zero. They therefore do not pin down a single direction of effect at the level of precision expressed by those intervals.

Interval width should not be interpreted without considering the measure and the analysis. A risk-difference interval is on the proportion-difference scale; a Neuro-QoL interval is on its scale units. The registry provides group-specific denominators for each measure, and these differ across outcomes. Missing-data handling and the precise analysis populations are not reported in the registry record, so the posted intervals cannot answer how alternative assumptions about missing observations might change precision.

12. Planning a Similar Trial: Sample Size

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

The sample-size figures are planning illustrations based on the observed effect, not the original trial’s design calculation. For the primary binary outcome, the figures are participants per arm, not a combined total and not total events. The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each scenario uses the stated fraction of the observed effect: 100%, 75%, or 50%.

The 80% power column gives the estimated number of participants per arm under those assumptions for a future study to have an 80% chance of detecting the effect specified in that scenario, if the assumed effect and planning model are appropriate. The 90% power column gives the corresponding number per arm for a 90% chance. These are prospective planning quantities; this is not observed power.

For the primary outcome, planning for 100% of the observed effect gives 214 participants per arm for 80% power and 285 for 90% power. Planning for 75% gives 342 per arm for 80% power and 457 for 90% power. Planning for 50% gives 680 per arm for 80% power and 909 for 90% power. The pattern illustrates a general feature of power calculations: smaller effects are harder to distinguish from random variation, so they require more participants under otherwise unchanged assumptions.

The curve represents the same kind of planning relationship visually; it should be read as an illustration under the stated assumptions, not as an observed trial outcome. These figures do not show that the completed trial had a particular probability of finding a result. The original trial’s sample-size justification and assumptions are not reported in the registry record. The observed effect used for these planning scenarios is itself uncertain, so the illustrations should not be treated as a guaranteed design requirement for another study.

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, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.047; treatment proportion p₁ = 0.122; average p̄ = 0.085.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.3934 = 0.7710.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.3897 = 0.3280.
4. n = (0.7710 + 0.3280)² / (0.0753)² = 213.2, rounded up to 214 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
p0 <- 0.046875                          # control proportion
p1 <- 0.122137404580153                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^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 (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(3073603)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 214 (80% power), 285 (90% power)
 75% of observed effect: n per arm = 342 (80% power), 457 (90% power)
 50% of observed effect: n per arm = 680 (80% power), 909 (90% power)
power.prop.test: 213.2 
Simulated power at n = 214 per arm: 0.811 (20000 replicates)

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

14. Analyses Reported by the Sponsor

The registry’s own analysis reports an exact binomial test for the difference in proportions and classifies it as non-inferiority. It gives an estimate of 0.075, an interval from 0.006 to 0.150, and p = 0.521. A separate simple unadjusted recalculation from the posted counts reports the same estimate, 0.075, but an interval from 0.008 to 0.142 and p = 0.03. The registry flags a discrepancy between the sponsor’s reported analysis and the recalculation.

For interpretation, the sponsor’s own analysis takes precedence. The recalculation is useful as a transparent comparison based on aggregate counts, but it is not a reconstruction of the full analysis plan and does not replace the registry’s reported test. A different p-value does not by itself reveal which method was planned or whether the sponsor result is correct; the discrepancy means the results should be distinguished rather than blended.

The sponsor labels its analysis non-inferiority, but the non-inferiority margin, its justification, and the decision rule are not reported in the registry record. The reported estimate and p-value alone do not establish the clinical or statistical conclusion of that test. Readers should therefore treat the sponsor result as the controlling posted analysis while recognizing that the information needed to independently evaluate its non-inferiority interpretation is absent.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists multiple outcomes, including the primary combined relapse/MRI measure, disability progression, and quality-of-life domains. It identifies 21 outcome measures and 19 posted secondary outcomes. When several outcomes are tested, the chance of at least one apparently unusual result can increase. In general, multiplicity procedures can control a chosen error rate or organize which analyses are confirmatory and which are exploratory. The registry record does not report whether an adjustment was planned or applied, nor does it report a hierarchy or other strategy for interpreting the outcome set.

The secondary recalculations are unadjusted. Their p-values should not be treated as though each were the sole planned test. The same caution applies when comparing several domains or deciding whether a pattern is consistent. No adjusted secondary estimates or multiplicity-adjusted p-values are reported in the registry record. The trial’s full statistical analysis plan is not reported in the registry record.

The reported denominators differ between outcomes. For the primary measure, the arm denominators are 128 and 131; for disability progression they are 126 and 130; and the Neuro-QoL measures have still different denominators. Those differences may reflect different available data or analysis sets, but the reasons are not reported in the registry record. The missing-data amounts, reasons for missing observations, methods for handling missing data, and sensitivity analyses are also not reported in the registry record.

The record does not report the exact analysis populations used for the sponsor’s primary non-inferiority test, including whether all randomized participants or a defined subset were analyzed. The original design assumptions, interim analysis procedures, and detailed statistical plan are not reported in the registry record. These omissions limit evaluation of how outcomes were selected, how missingness affected the comparisons, and whether the posted analyses followed the intended plan. They do not establish that a particular procedure was or was not used; they mean the registry record does not document it.

16. 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.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: ms relapse (0.8%).

The posted safety figures list serious adverse events in 1 participant in the Drug Continuation Arm and 1 participant in the Drug Discontinuation Arm, with denominators of 128 and 131 at risk, respectively. The registry lists deaths as 1 in the continuation arm and 2 in the discontinuation arm. Other adverse events are listed for 66 participants in continuation and 70 in discontinuation, with the same arm-specific denominators. The most frequent serious adverse event term listed is MS relapse, at 0.8%.

These figures are descriptive registry summaries. They do not, by themselves, establish that assignment to either strategy caused an adverse event or death. The record does not provide detailed event timing, severity beyond the serious-event categorization, relatedness assessments, or participant-level information. It also does not describe how recurrent events were handled in the summary. The counts should be interpreted within the posted study groups and safety reporting, not as a complete comparison of all possible harms or as evidence about an individual’s risk.

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

The posted results describe a randomized comparison of continuing current DMT and discontinuing current DMT in people with MS. For the primary combined outcome, the registry reports 6 participants with a new relapse and/or MRI brain lesion in continuation and 16 in discontinuation. The unadjusted recalculation indicates a higher observed proportion in discontinuation, but it conflicts in inferential result with the sponsor’s reported non-inferiority analysis. The sponsor analysis takes precedence, and its margin and decision rule are not reported in the registry record.

This result does not establish that stopping DMT is safe or unsafe for every person with MS, nor does it establish that either strategy is appropriate for a particular person. The registry’s rationale emphasizes older people and prolonged absence of disease activity, but detailed eligibility criteria are not reported. The posted aggregate outcome does not identify which primary component—relapse or MRI lesion—accounted for each event, and it does not provide enough information to assess all differences among therapy formulations or participant subgroups.

The sponsor’s analysis is labeled non-inferiority, but the information needed to judge that claim fully is not posted. A non-inferiority interpretation depends on a prespecified margin, the analysis population, and the relationship between the interval and that margin. The registry record does not report these details. The unadjusted recalculation is not a replacement for the sponsor’s analysis, especially given the flagged discrepancy.

Secondary outcomes provide additional, uncertain information about disability progression and quality of life. They do not overturn or settle the primary result, and multiple unadjusted comparisons warrant caution. The safety counts are also descriptive and do not establish causality. Overall, the posted findings are evidence from this study’s reported groups and follow-up, with important uncertainty about analysis details and generalizability.

18. Why This Trial Matters Statistically

This trial poses a statistical question that is especially sensitive to how the analysis objective is defined. A comparison intended to test non-inferiority is not interpreted in the same way as a routine test for any difference. The sponsor’s classification of the primary analysis as non-inferiority makes the absent margin and decision rule important: without them, readers cannot independently assess the meaning of the reported interval and p-value.

The difference between the sponsor’s result and a simple unadjusted recalculation is also instructive. The recalculated estimate matches the sponsor’s stated estimate, yet the intervals and p-values differ. This demonstrates why an effect estimate alone does not fully describe an analysis. The chosen method, test, assumptions, and inferential objective all shape uncertainty and conclusions. The discrepancy should be made visible rather than resolved by selecting whichever p-value appears more persuasive.

The primary endpoint combines clinical relapse and MRI lesion occurrence. This creates a participant-level measure of disease activity, but component-specific results would be needed to understand which events drive the combined outcome. The study also reports outcomes with varying denominators, illustrating why analysis populations and missing-data procedures matter when interpreting more than one measure.

Finally, this record provides a useful setting for distinguishing observed results from planning illustrations. The sample-size figures are future-study scenarios based on the observed effect and stated assumptions, not evidence about the probability that this completed study detected an effect. Clear separation of descriptive counts, sponsor analyses, recalculations, uncertainty, and planning calculations helps prevent statistical summaries from being mistaken for broader clinical conclusions.

19. Limitations

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

The registry summary describes the study rationale in relation to older people and prolonged absence of relapses or MRI changes, but the detailed eligibility criteria are not reported in the registry record. It is therefore not possible to confirm from the posted summary exactly how closely the enrolled population matches that rationale or to determine how results might vary across age or disease-activity groups. Subgroup-specific primary results are not reported in the registry record.

The interventions are strategies involving current DMTs rather than one specified medication. The record describes ~14 formulations/doses. It does not provide posted outcome comparisons by specific therapy, dose, or other treatment features. This limits interpretation of whether the pooled strategy comparison represents similar effects across the different treatments included.

The primary outcome combines relapse and MRI lesion occurrence, and separate primary component results are not reported in the registry record. The outcome’s stated follow-up spans 18-24 months based on enrollment, so the record does not present a single identical follow-up duration for every participant. Detailed event timing and participant-level data are not reported.

The flagged discrepancy between the sponsor’s analysis and the unadjusted recalculation is a central limitation for interpreting the primary inference. The non-inferiority margin and detailed statistical plan are not reported, so the sponsor’s conclusion cannot be independently checked from the registry information. The record also does not report missing-data methods, exact analysis populations, adherence summaries, or sensitivity analyses. These limitations constrain assessment of the primary and secondary comparisons; they do not establish that any specific unreported procedure was omitted.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Discontinuation of Disease Modifying Therapies (DMTs) in Multiple Sclerosis (MS) — NCT03073603.
• Registry record last updated 2023-08-16; status last verified 2023-08.
• Results first posted 2022-10-06.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

23. Record Summary

The registry describes a completed, randomized, single-masked, parallel-group Phase 4 study sponsored by the University of Colorado, Denver, comparing continuation with discontinuation of current disease-modifying therapy for multiple sclerosis. Enrollment is listed as 259. The primary outcome was participants developing a new MS relapse and/or MRI brain lesion over 18-24 months based on enrollment.

The posted primary counts are 6 of 128 in continuation and 16 of 131 in discontinuation. The sponsor reports an exact binomial non-inferiority analysis, while a simple unadjusted recalculation differs in its interval and p-value; the registry flags this discrepancy, and the sponsor’s analysis takes precedence. The non-inferiority margin and detailed analysis plan are not reported in the registry record. Secondary results and safety counts provide further descriptive information, with uncertainty, varying denominators, and unreported multiplicity and missing-data procedures limiting interpretation.