A Randomized Trial of Two Formal Group Programs for Multiple Sclerosis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Randomized Trial of Two Formal Group Programs for Multiple Sclerosis (NCT01918800), a randomized na study. The primary outcome is Modified Fatigue Impact Scale.

2. Clinical Question

3. Trial Design

The registry identifies this as a completed, randomized, single-masked, parallel-group trial of two behavioral education programs for people with multiple sclerosis. The interventions were Fatigue: Take Control and MS: Take Control. The registry reports total enrollment of 282. Its brief summary describes a planned 1:1 chance of assignment to either program and says the study involved four VA sites.

The registry summary describes 6 weekly 2-hour group sessions for each program and 10 study visits within 11 months, including baseline, classes, and follow-up data collection. It also notes an additional final 12-month follow-up visit for participants in Portland. The recorded primary endpoint timeframe is 4 months. The actual allocation sequence, concealment procedures, and masking details beyond the single-masked designation are not reported in the registry record.

For the primary outcome, the posted results include 109 participants in each arm. Those endpoint counts are smaller than the total enrollment, but the reason for the difference and the flow of participants from enrollment to analysis are not reported in the registry record. The original sample-size justification, intended analysis population, and protocol-level statistical analysis plan are also not reported in the registry record. The registry's brief summary mentions an intention to enroll up to 600 participants, while the recorded enrollment is 282; the reason for the difference is not reported.

The trial compares two education programs rather than an education program with no program. Therefore, the observed contrast addresses the relative outcomes of these two active approaches under the trial conditions, not the effect of either program against no intervention.

4. Participant Flow and Accounting

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

Overall Study

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

6. Primary Endpoint

The registry lists the primary outcome as Modified Fatigue Impact Scale, measured over 4 months.

The primary endpoint was the Modified Fatigue Impact Scale (MFIS) at 4 months. It is a self-report, retrospective measure of fatigue symptoms. The registry describes the full-length instrument as 21 items scored from 0 to 4, with a total score from 0 to 84. It also reports a coefficient alpha of .81. The MFIS produces a total score and physical, cognitive, and psychosocial subscale scores; lower scores indicate less fatigue.

Fatigue is a relevant outcome for this trial because the programs were designed for people with multiple sclerosis and one program specifically focused on fatigue. A standardized self-report scale allows outcomes to be represented numerically and compared between randomized arms. The posted primary results report group means and standard deviations, along with a mean difference and confidence interval.

The MFIS describes participants' reported fatigue impact; it is not a direct measure of every symptom or aspect of quality of life. The registry record does not report a threshold for a clinically important difference, a responder definition, or the smallest difference that participants would consider meaningful. Statistical uncertainty around the group contrast can therefore be described from the posted estimates, but the record does not establish how a difference of a particular size should be interpreted by participants.

7. Results Posted in the Registry

• MS: Take Control: mean 46.70 versus 46.10 on control; difference 0.60 (95% CI -2.60 to 3.80), standardized effect d = 0.05; z-test p = 0.713.

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.

Multiple Sclerosis Self Efficacy Scale

• 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:
  • MS: Take Control: mean 1,352 versus 1,434 on control; difference -82.00 (95% CI -150.58 to -13.42), standardized effect d = -0.32; z-test p = 0.019.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for MS: Take Control than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Beck Depression Inventory II (BDI-II)

• 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:
  • MS: Take Control: mean 10.70 versus 9.50 on control; difference 1.20 (95% CI -0.84 to 3.24), standardized effect d = 0.16; z-test p = 0.250.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between MS: Take Control and the control. The interval gives the range of effects that remain compatible with the data.

Pittsburgh Sleep Quality Index (PSQI)

• 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:
  • MS: Take Control: mean 7.71 versus 7.41 on control; difference 0.30 (95% CI -0.82 to 1.42), standardized effect d = 0.07; z-test p = 0.599.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between MS: Take Control and the control. The interval gives the range of effects that remain compatible with the data.

The secondary results show a mixed pattern rather than a consistent separation across measures. The self-efficacy comparison has a nominal unadjusted result that differs from zero, while the depression and sleep-quality comparisons do not show similarly clear evidence in their posted analyses. This pattern does not mirror the primary MFIS result, which also does not show a clear between-program difference. The registry lists additional secondary outcomes, including quality-of-life, physical-activity, and walking measures, but their numerical results are not described in the available endpoint summaries.

Secondary outcomes can help characterize whether the comparison has a broader pattern across domains, but they do not replace the primary endpoint. A result for one secondary measure should be interpreted as one part of a collection of analyses, especially when multiple measures are assessed. If each is tested without accounting for the number of comparisons, the chance of finding at least one apparently notable result by chance can increase. The registry record does not report a multiplicity adjustment or an ordered testing strategy.

In addition, the registry does not give clinically meaningful thresholds for these secondary scales, nor does it report adjusted estimates or a complete explanation of their analysis populations. Their estimates are therefore best treated as exploratory signals rather than decisive evidence. The overall pattern does not establish that one program is broadly better across fatigue, self-efficacy, mood, sleep, or other domains. The secondary findings should be considered alongside the primary result and the limited reporting of analysis details.

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

• Multiple Sclerosis Self Efficacy Scale
• Beck Depression Inventory II (BDI-II)
• SF-36
• Rapid Assessment of Physical Activity (RAPA)
• Pittsburgh Sleep Quality Index (PSQI)
• Timed 25 Foot Walk (T25-FW)

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is continuous and is reported as a mean with a standard deviation in each arm. A mean describes the average score, while a standard deviation describes how dispersed participant scores are around that average. For the primary endpoint, the posted means are 46.1 and 46.7, with standard deviations of 12.2 and 11.9. These summaries give a group-level picture; they do not show the distribution of individual scores or changes for individual participants.

What does the mean difference represent?

The recalculated contrast is expressed as the mean for MS: Take Control minus the mean for Fatigue: Take Control. It is 0.6 MFIS units. A positive value means the first-named group had the higher mean score in this contrast. Since the registry describes lower MFIS scores as indicating less fatigue, the direction of the point estimate is numerically toward a higher mean in MS: Take Control. The confidence interval, however, spans values on both sides of zero, so direction at the point estimate should not be confused with a definitive difference.

What does the unadjusted test address?

The unadjusted p-value of 0.713 evaluates how compatible the observed contrast is with a zero mean difference under the calculation's statistical assumptions. A large p-value does not establish that the null hypothesis is true, nor does it quantify the probability that the programs are the same. It is one component of the evidence and should be read alongside the estimated difference and its interval. The registry's reported analysis uses a paired t test and reports p = .64. The registry does not provide the paired estimate, its confidence interval, or details explaining the pairing.

Why can the analysis methods differ?

The simple recalculation uses the posted arm-level means, standard deviations, and sample sizes to estimate an unadjusted comparison of group means. A paired t test instead analyzes paired observations and their within-pair differences. That method depends on how observations were paired and on the variability of those paired differences. The registry identifies the sponsor's method but does not report the pairing structure or the associated summary statistics. The recalculation is therefore not a reconstruction of every feature of the sponsor's analysis.

There is no discrepancy flag for this trial. The recalculated result and the registry-reported p-value both leave substantial uncertainty about a between-program difference. Their numerical values are not identical, which is unsurprising when methods differ, but both analyses support a cautious reading rather than a firm claim of a difference.

Why does over-dispersion matter in general?

General statistical context: when participant outcomes vary widely, the standard error of a group contrast tends to be larger, all else equal, and the confidence interval tends to be wider. This makes it harder to distinguish a modest average difference from the background variation. The posted MFIS standard deviations describe variation among participants, but the registry does not report a separate over-dispersion model or an analysis that explicitly addresses over-dispersion. The endpoint is analyzed as a continuous measure, not as a count outcome.

The registry reports 109 participants in each arm for this endpoint. It does not report the primary outcome's missing-data method, whether all randomized participants were included in the analysis, or whether covariates were adjusted for in the sponsor's model. Those details matter when assessing the analysis population and whether the reported comparison follows the original randomization in a complete and balanced way.

General statistical context: random assignment helps balance measured and unmeasured characteristics across groups on average, but it does not guarantee identical groups in a particular trial. Analysis decisions, missing outcomes, and adherence can also affect how the randomized comparison is interpreted. The registry's posted summaries do not provide enough detail to assess these features fully.

11. Confidence Intervals and Precision

The primary mean difference is 0.6 MFIS units, with a confidence interval from -2.599 to 3.799. A confidence interval describes a range of values reasonably compatible with the data and the statistical method used. It is not a guarantee that the true difference lies within the particular interval, nor is it a range containing a specified proportion of individual participants' scores.

Because the interval includes zero, it is compatible with no mean difference. It also includes values indicating a lower mean in MS: Take Control and values indicating a higher mean in that arm, using the stated contrast direction. The interval therefore leaves uncertainty about both the direction and size of the between-program contrast. Its limits are more informative than the point estimate alone: the estimate is close to zero, but the data do not locate the difference with high precision.

The registry does not report a clinically important difference threshold for the MFIS. Without such a threshold, the interval cannot be classified from the registry alone as excluding or including all differences that participants or clinicians would consider meaningful. The instrument's described range and direction help interpret scores, but they do not supply that threshold.

General statistical context: interval width is affected by the amount of outcome variability, the number of analyzed participants, and the analysis method. The posted endpoint counts and standard deviations contribute to precision, while the sponsor's paired test may use information about within-pair variation that is not reported. The recalculated interval is an unadjusted interval based on arm-level summaries, so it should be understood as a simple summary rather than a substitute for a fully documented trial analysis.

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 for this continuous-outcome comparison. They are not total event counts. The planning illustration uses the observed mean difference as its reference effect and considers planning for 100%, 75%, and 50% of that observed effect. Each of those effect assumptions has estimates for 80% power and 90% power.

In plain terms, the 80% power column gives the estimated participants per arm for a design with an 80% chance of detecting the assumed effect at the stated significance level, if that effect is real and the planning assumptions hold. The 90% power column uses a higher probability of detecting that same assumed effect, which generally requires more participants. For either power target, planning for a smaller fraction of the observed effect requires more participants than planning for the full observed effect. This is because a smaller difference is harder to distinguish from random variation.

The listed planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is the stated fraction of the observed effect. These are planning illustrations based on the observed result, not the trial's own original design assumptions; those original assumptions are not reported in the registry record.

These estimates are not observed power. They do not tell how likely the completed trial was to detect an effect after the results were known. They instead illustrate sample sizes for hypothetical future planning under stated assumptions. Since the observed primary contrast is small, planning around a fraction of it yields particularly large estimates; the figures should not be treated as a recommended sample size without a justified target effect and a full design plan.

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 = 0.600; pooled standard deviation σ = 12.051.
2. n = 2 × (1.96 + 0.842)² × 12.051² / 0.600² = 6,332.5, rounded up to 6,333 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 <- 0.600000000000001                      # observed difference in means
sigma <- 12.0509335737942                     # 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(1918800)
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 = 6,333 (80% power), 8,478 (90% power)
 75% of observed effect: n per arm = 11,258 (80% power), 15,072 (90% power)
 50% of observed effect: n per arm = 25,331 (80% power), 33,910 (90% power)
power.t.test: 6333.5 
Simulated power at n = 6,333 per arm: 0.802 (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. Analyses Reported by the Sponsor

The registry reports the sponsor's primary analysis as a paired t test with p = .64. The arm-level recalculation estimates a mean difference of 0.6 MFIS units and gives an unadjusted p-value of 0.713, with a confidence interval from -2.599 to 3.799. Both results indicate that the posted data do not establish a clear difference between programs on the primary endpoint.

The estimates should not be expected to match exactly. The recalculation is an unadjusted comparison from the posted arm-level summaries, whereas a paired t test uses paired differences. The registry does not report how the pairing was defined, the paired mean difference, or the variability of paired differences. Consequently, the recalculation cannot reproduce that analysis from the recorded summaries alone.

The registry's own analysis is the trial's reported analysis. The recalculation provides a transparent summary based on the available group statistics, with its method and limitations made explicit. Because the discrepancy flag is false, there is no flagged conflict requiring the registry analysis to take precedence over a materially inconsistent recalculation. The difference in p-values does not change the broad interpretation: uncertainty remains, and neither result establishes equivalence.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists multiple secondary outcomes and reports several analyses. Testing multiple outcomes creates a multiplicity issue: even when no true differences exist, repeated testing can increase the likelihood that at least one result appears statistically notable by chance. General statistical context: prespecified approaches such as adjusting significance thresholds, controlling an error rate, or defining a hierarchy can help address this issue. The registry record does not report whether such a strategy was planned or used.

The listed secondary measures include self-efficacy, depression, health-related quality of life, physical activity, sleep quality, and walking performance. The registry record does not report a complete multiplicity plan, the rationale for prioritizing particular secondary outcomes, or whether the reported secondary analyses were prespecified. Accordingly, individual unadjusted secondary p-values should not be treated as confirmatory evidence in isolation.

For missing data, the endpoint results include fewer participants than the total enrollment, but the registry does not report why, how missing observations were handled, or whether any imputation or sensitivity analyses were conducted. It also does not report a participant flow by arm for the primary analysis or a comparison of participants with and without observed outcomes. These omissions limit assessment of potential bias from incomplete follow-up.

The registry does not state whether the analysis followed an intention-to-treat population, a per-protocol population, or another defined analysis set. Nor does it report how adherence to the group programs was incorporated into analysis. Randomization supports a valid comparison when participants are analyzed appropriately according to assignment, but the record does not provide enough information to verify the analysis population or determine the effect of deviations from assigned treatment. The original statistical analysis plan and any interim-analysis procedures are not reported in the registry record.

16. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: ms related (0.9%), surgery (0.9%), fracture from a fall (0.9%), hospitalization for mental health (0.9%), cancer (0.9%).

The registry reports serious adverse events affecting 12 participants in Fatigue: Take Control and 10 in MS: Take Control; each arm lists 109 participants at risk. It also reports other adverse events affecting 86 and 82 participants, respectively, with 109 at risk in each arm. These are participant counts, not event rates per unit of follow-up, and the posted summary does not give a statistical comparison between arms.

The listed serious-event terms include MS related, surgery, fracture from a fall, hospitalization for mental health, cancer, and constipation requiring hospitalization, each reported as 0.9%. The record does not provide enough detail in this summary to determine the number or timing of every event by category, the relationship of events to either program, or whether categories overlap.

Deaths are not reported in the registry record. These adverse-event figures describe what was recorded in the registry; they do not establish that an event was caused by a program or that the programs have equivalent safety profiles. Differences in event counts should be interpreted cautiously without event definitions, follow-up details, and a prespecified safety analysis.

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

The primary comparison shows similar observed average MFIS scores at 4 months, with a small positive point estimate for the contrast of MS: Take Control minus Fatigue: Take Control. The confidence interval crosses zero and includes differences in either direction. The registry's reported paired analysis and the unadjusted recalculation both fail to establish a clear difference in the primary outcome.

This does not demonstrate that the programs are identical. A failure to find clear statistical evidence of a difference is not an equivalence result. Equivalence requires a prespecified range of differences considered unimportant and an analysis designed to assess whether the entire plausible range falls within that margin. No such margin is reported in the registry record.

Nor does this comparison determine whether either program improves fatigue compared with no program or another form of care. Both randomized groups received active education programs. Baseline-to-follow-up change estimates and a no-program comparison are not reported in the registry record. The result is specifically about the relative mean MFIS scores between these two programs at the posted timeframe.

The secondary outcomes offer a mixed set of results and should be viewed cautiously because there are multiple outcomes and no multiplicity plan is reported. The record also does not report enough information about missing outcomes, analysis populations, adherence, or clinically important thresholds to resolve all questions about the practical meaning of the estimates.

In general, statistical results describe uncertainty in group-level data; they do not predict an individual participant's outcome. These findings should be read as a summary of the registry's recorded comparison, not as medical advice or a treatment recommendation.

18. Why This Trial Matters Statistically

This trial illustrates a common challenge in clinical research: a randomized comparison can produce a point estimate close to zero without establishing that the programs are equivalent. The confidence interval is essential because it conveys uncertainty around the estimate and makes clear that the data allow a range of possible contrasts. Reading only the observed means or a p-value would provide a less complete account.

It also shows why active comparators require careful framing. The study asks whether two group education programs differ on the selected outcome, not whether education as a whole has an effect relative to no intervention. That distinction defines what the randomization can address and prevents the comparison from being interpreted more broadly than its design supports.

The record also highlights the importance of transparent statistical reporting. The sponsor's paired t test is identified, while the information needed to reconstruct that paired analysis is not reported. A separate unadjusted calculation from arm-level summaries is informative but answers a methodologically different version of the question. Clear reporting of the estimand, analysis population, missing-data approach, and prespecified plan would make the evidence easier to evaluate.

Finally, the secondary outcomes and planning illustrations emphasize two broader principles. Multiple outcome tests need a stated interpretation plan, and sample-size planning depends on a justified target effect rather than simply repeating an observed estimate. Together, these issues make the trial useful for understanding how estimates, intervals, analysis methods, and registry reporting shape statistical interpretation.

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 record does not report the original sample-size calculation or the assumptions used to plan enrollment. The brief summary describes an intention to enroll up to 600 participants, while recorded enrollment is 282; the reason for that difference is not reported. The record also does not explain why the primary endpoint results include 109 participants per arm or how participants not represented in those results were handled.

The sponsor's primary analysis is identified as a paired t test, but the pairing structure, paired estimate, confidence interval, and variability of paired differences are not reported. The simple recalculation uses arm-level summaries and cannot reproduce the paired analysis. The registry does not provide the full statistical analysis plan, covariate-adjustment details, interim-analysis information, or a clear definition of the analysis population.

Missing-data methods, participant flow by arm, adherence, and sensitivity analyses are not reported in the registry record. The record also does not state a clinically important difference threshold for the MFIS or provide enough context to determine the clinical importance of the observed contrast. For secondary endpoints, the multiplicity strategy and complete analysis details are not reported.

These reporting gaps do not change the posted group means, but they limit how fully the estimates can be evaluated. The recorded comparison is between two active education programs, with the primary assessment at 4 months; conclusions should remain within that scope.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Randomized Trial of Two Formal Group Programs for Multiple Sclerosis — NCT01918800.
• Registry record last updated 2018-07-26; status last verified 2018-06.
• Results first posted 2017-03-13.
• 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 record describes a completed, randomized, single-masked, parallel-group trial comparing Fatigue: Take Control with MS: Take Control for people with multiple sclerosis. Total enrollment was 282. The primary outcome was the Modified Fatigue Impact Scale at 4 months.

For the primary endpoint, the registry posts 109 participants per arm, mean scores of 46.1 and 46.7, and an unadjusted mean difference of 0.6 units with a confidence interval from -2.599 to 3.799. The registry-reported paired t test has p = .64; the unadjusted recalculation has p = 0.713. Neither result establishes a clear between-program difference.

The registry lists secondary outcomes and adverse-event counts, but does not report several details needed for fuller assessment, including the original sample-size justification, missing-data methods, analysis population, and multiplicity plan.