Mild Cognitive Impairment in Parkinson's Disease: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Mild Cognitive Impairment in Parkinson's Disease (NCT01519271), a randomized phase 4 study. The primary outcome is Alzheimer's Disease Cooperative Study- Clinical Global Impression Change (ADCS-CGIC).

2. Clinical Question

The clinical question concerns perceived global change in cognition and related functioning, rather than a single cognitive task. The registry describes mild cognitive impairment as a problem that can affect functioning, quality of life, mood, and caregivers, and notes that little was known about its treatment in Parkinson’s disease. The study therefore compared scores after each treatment phase using an assessment designed to rate change.

The scale runs from 1, very much improved, to 7, very much worse; a score of 4 represents no change. Consequently, a lower group mean indicates a more favorable direction on this scale. The posted means were 3.48 for the Exelon patch phase and 3.92 for the placebo phase. The comparison is a mean difference of -0.44, defined in the posted results as Exelon patch relative to placebo. Those values describe the observed summary scores; they do not, by themselves, establish how certain the treatment contrast is or how the change was distributed among individual participants.

The registry lists the Montreal Cognitive Assessment as a secondary outcome. The primary question addressed by the posted comparison is specifically about ADCS-CGIC scores at the end of each study phase. The registry record does not report a separate primary question about each domain of cognition or a responder comparison, even though its scale description defines a responder as a participant scoring 1 or 2.

3. Trial Design

This was a Phase IV, randomized, double-blind, placebo-controlled, crossover study of Exelon Patch (rivastigmine transdermal system) and placebo patches. It was conducted at a single site and sponsored by the University of Pennsylvania. The record gives a start of 2011-12, a primary completion of 2014-06, and a status of completed. Enrollment was 28.

A crossover design means participants receive more than one treatment phase, so the treatment contrast can involve outcomes from the same participants under different treatments. In this trial, the primary-outcome description explicitly says participants were exposed to one placebo phase and one Exelon patch phase, and that the data compare treatment groups. The primary measure was administered at the end of each phase. Two participants discontinued study participation over the course of the study. The posted group summaries list 26 participants for placebo and 28 for Exelon patch; those entries should be read in the context of the crossover design, not as evidence that the trial had two separately randomized parallel groups of those sizes.

Randomization describes assignment to a treatment sequence, while double-blinding indicates that treatment identity was concealed from participants and investigators as described in the official title. The registry’s masking field says quadruple. The record does not explain which four roles were masked or provide the sequence-allocation details. The order of treatment phases, any washout interval, and the handling of period or carryover effects are not reported in the registry record.

The registry describes the trial as 24 weeks in its brief summary, but does not report the duration of each treatment phase or further scheduling details. It lists two outcome measures and one posted analysis. A full statistical analysis plan, original sample-size justification, and details of any interim analysis are not reported in the registry record. The posted sample-size illustrations are derived from the observed effect and are not the trial’s original design assumptions.

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.

Before assignment: 48 participants assessed for eligibility, 28 randomized (17 did not meet inclusion/exclusion criteria and 3 refused to participate after screening)

Phase 1

Phase 2

Analysis population for the primary outcome, as defined in the registry: Please note this study utilized the crossover design (i.e. participants were exposed to two phases of treatment, one with placebo and one with Exelon patch). The data are comparing differences in treatment groups. Over the course of this study 2 participants discontinued study participation.

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 Alzheimer's Disease Cooperative Study- Clinical Global Impression Change (ADCS-CGIC), measured over The ADCS-CGIC will be administered at the end of each study phase..

The primary endpoint was the Alzheimer’s Disease Cooperative Study–Clinical Global Impression of Change (ADCS-CGIC), reported as a mean score with standard deviation. The registry says the measure is commonly used in dementia psychopharmacology studies to assess global change. It is not a baseline measure: the record says it is administered only at the end of each study phase because it evaluates change.

The seven-point scale ranges from 1, very much improved, through 4, no change, to 7, very much worse. A lower score therefore indicates a more favorable rating of change. The registry defines a responder as someone scoring 1 or 2, but the posted primary comparison is expressed as a difference in mean scores, not as a responder proportion. A mean can convey the overall location of scores while concealing variation in individual experiences or the share of participants at each response level.

The posted mean was 3.92 (standard deviation 0.94) for placebo and 3.48 (standard deviation 0.89) for the Exelon patch phase. This endpoint was selected to capture an overall clinical impression of change rather than performance on one isolated cognitive domain. The registry lists the Montreal Cognitive Assessment as a secondary outcome, but the primary result concerns ADCS-CGIC. Further details about who rated the measure, how ratings were combined, and the instrument’s administration procedures are not reported in the registry record.

7. Results Posted in the Registry

• Exelon Patch (Rivastigmine Transdermal System): mean 3.48 versus 3.92 on control; difference -0.44 (95% CI -0.93 to 0.05), standardized effect d = -0.48; z-test p = 0.078.

8. Statistical Interpretation

9. Statistical Methods Explained

The posted result is a comparison of mean scores on a continuous outcome. A mean is the arithmetic center of a set of observations, while a standard deviation describes how dispersed individual scores are around that center. These summaries help describe the groups’ score distributions, but they do not show the shape of each distribution or where every participant’s score fell. The scale is bounded from 1 to 7, so the values also need to be interpreted within that defined range.

What does a mean difference represent?

The comparison is defined as the Exelon patch mean minus the placebo mean. Its estimate is -0.44. A negative value means that the Exelon patch phase had a lower average score than the placebo phase. Because lower ADCS-CGIC scores indicate a more favorable global change rating, the direction of the estimate favors a lower score in the Exelon phase. The sign describes direction, not certainty. The interval and the design details are essential to interpreting the estimate.

What does an interval around an estimate tell us?

In general, a confidence interval gives a range of estimates compatible with the data under the statistical procedure used, and its width communicates precision. The posted interval extends from -0.929 to 0.049 when rounded to three decimal places. Since it crosses zero, values representing no mean difference are within that range. The registry record does not report the confidence level or the procedure used to construct the interval, so those details cannot be specified from the posted record. An interval is not a range that contains every participant’s outcome; it concerns uncertainty in an estimated contrast.

How should the p-value be understood?

The unadjusted p-value is 0.078. In general, a p-value describes how incompatible the observed data are with a specified null model, assuming that model and the statistical procedure apply. It is not the probability that the null hypothesis is true, and it does not measure clinical importance. The word unadjusted matters: the result is reported without a stated multiplicity adjustment. The registry record does not report the primary analysis plan or a trial-specific significance threshold, so the sample-size illustration’s 0.05 assumption should not be mistaken for a confirmed analysis-plan rule.

Why does the crossover design matter?

In general, crossover studies can compare treatment responses within participants, which may account for stable differences between people. They also require attention to treatment sequence and period, because outcomes can vary over time and the effect of one phase may potentially persist into another. These are general design considerations, not assertions about what occurred in this trial. The registry describes the placebo and Exelon phases, but does not report the analytic model, whether a paired comparison was used, or how period and carryover were considered. It also does not report the treatment sequence or a washout procedure. The posted arm counts and summary means do not supply those missing analysis details.

In general, a paired analysis and an analysis treating observations as independent use different information and can yield different uncertainty estimates. A within-participant analysis depends on the variation in participants’ differences across phases, which cannot be recovered from the two means and standard deviations alone. No paired-difference standard deviation is posted. It would therefore be inappropriate to infer a specific paired analysis from the crossover label or to reconstruct one from the group summaries.

What information is missing about the analysis?

The registry record reports one analysis and two outcome measures, with the Montreal Cognitive Assessment listed as secondary. It does not report the model specification, covariates, analysis population, missing-data method, or a method for handling repeated measurements. It also does not report an interim-analysis plan. These details matter because they describe which observations contribute to the estimate and how uncertainty is calculated. Their absence does not establish that no such methods were used; it means they are not reported in the registry record.

In general, multiple outcome or repeated statistical tests raise a multiplicity question: the more claims assessed, the more important it is to know whether hypotheses were prioritized or adjustments were planned. The record identifies the primary measure and one secondary outcome, but does not report a multiplicity strategy. The unadjusted p-value should therefore be read as the posted unadjusted result, not as evidence that all potential analyses were accounted for in a formal testing hierarchy.

Finally, the scale’s responder definition does not change the metric of the posted comparison. The result is a difference in means, not a comparison of the fraction scoring 1 or 2. In general, responder analyses answer a different question and can depend on a threshold chosen in advance. The registry record does not post a primary responder analysis or explain whether one was prespecified.

10. Confidence Intervals and Precision

The mean difference is -0.44, with a posted interval from -0.929 to 0.049 when stated to three decimal places. The interval is relatively broad compared with the point estimate and spans zero. It therefore represents uncertainty about the direction and size of the mean contrast: the data are compatible with a lower Exelon-phase mean, with no mean difference, and with a slightly higher mean within the interval’s upper end. These statements concern the estimated mean difference, not individual participants’ outcomes.

In general, an interval communicates precision as well as direction. A narrow interval confines compatible estimates more tightly, while a wide interval leaves greater uncertainty. Here, the interval’s inclusion of zero means that the posted estimate does not precisely distinguish a difference from no difference under the interval procedure. The registry record does not report the interval’s confidence level or its construction method, so neither should be assumed.

The group summaries are 3.92 with standard deviation 0.94 for placebo and 3.48 with standard deviation 0.89 for Exelon patch. The numbers of observations listed are 26 and 28, respectively. The trial is a crossover, but the record does not report whether the interval accounts for within-participant pairing. In general, the uncertainty for a paired treatment contrast depends on the variability of the within-person differences, not only on the separate group standard deviations. That variability is not reported.

The unadjusted p-value of 0.078 is another summary of the evidence under a statistical model, but it does not replace the interval or explain clinical relevance. The registry does not report a threshold for a clinically meaningful score difference. Thus, precision can be discussed in terms of the interval crossing zero, but clinical importance cannot be determined from the posted contrast alone.

11. 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 table and curve are planning illustrations based on the observed effect, not the trial’s original design assumptions. The figures are participants per arm, not total event counts. They assume a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; the effect to detect is a stated fraction of the observed effect. These assumptions are listed for the illustration and should not be attributed to the trial’s original design, whose sample-size assumptions are not reported in the registry record.

The 80% power column gives the illustrated number of participants per arm for a design with an 80% chance of detecting the assumed effect under the stated statistical assumptions, if that effect and model were correct. The 90% power column gives the corresponding planning figure for a 90% chance. These are prospective planning concepts: power concerns the chance of detecting a specified effect in a hypothetical design, not the probability that a study result is true.

The rows consider planning for 100%, 75%, and 50% of the observed effect. For 100% of the observed effect, the illustration gives 68 participants per arm for 80% power and 91 participants per arm for 90% power. For 75% of the observed effect, it gives 121 and 162 participants per arm. For 50% of the observed effect, it gives 272 and 364 participants per arm. Smaller effects require more participants under these assumptions, and targeting higher power also requires a larger planned sample.

These figures are not observed power for this completed crossover trial. They are illustrations tied to the observed contrast and the stated simplified assumptions. The illustration treats allocation as equal between two arms; the registry’s study design is crossover, so these figures should not be read as the actual trial’s design prescription or as a direct description of its analysis.

12. 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 = -0.440; pooled standard deviation σ = 0.914.
2. n = 2 × (1.96 + 0.842)² × 0.914² / -0.440² = 67.8, rounded up to 68 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 <- -0.44                      # observed difference in means
sigma <- 0.914379781390307                     # 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(1519271)
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 = 68 (80% power), 91 (90% power)
 75% of observed effect: n per arm = 121 (80% power), 162 (90% power)
 50% of observed effect: n per arm = 272 (80% power), 364 (90% power)
power.t.test: 68.8 
Simulated power at n = 68 per arm: 0.798 (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.

13. Multiplicity, Missing Data and Analysis Populations

The registry lists ADCS-CGIC as the primary outcome and the Montreal Cognitive Assessment as a secondary outcome. It reports two outcome measures and one analysis. These entries identify the outcomes and analysis count shown in the record, but do not explain the full set of analyses that may have been planned or performed. A prespecified testing hierarchy or adjustment for multiple outcomes is not reported in the registry record.

Multiplicity matters because testing several outcomes or comparisons can increase the chance of at least one apparently unusual result under a null model. In general, an analysis plan may address this by prioritizing a primary outcome, adjusting inference, or clearly distinguishing confirmatory from exploratory results. The registry identifies a primary measure, but does not report whether an adjustment or other multiplicity strategy was used. The posted p-value is unadjusted.

The handling of missing outcome data is also not reported. Two participants discontinued study participation. The record does not say whether their phase-end scores were missing, how missing scores were handled, or whether any imputation or sensitivity analysis was performed. It does not identify the analysis population, such as whether results were based on all randomized participants or those with usable measurements. The posted participant counts by treatment phase do not answer those questions on their own.

The record does not report the statistical model or whether the crossover comparison accounted for repeated observations from the same participant. Treatment sequence, period effects, carryover, and any washout procedure are not described. Interim analyses and stopping rules are not reported in the registry record. These are unknowns in the posted record, not evidence that such procedures were absent.

The primary endpoint description defines an ADCS-CGIC responder as a participant scoring 1 or 2, but the posted primary comparison is a mean difference. A threshold-based responder analysis would address a distinct question. Whether one was planned, and whether its results were analyzed, is not reported in the registry record. Likewise, no protocol or statistical analysis plan details are available in the record to establish additional prespecified contrasts or subgroup analyses.

14. 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 5% reporting threshold.

Most frequent non-serious adverse event terms

The registry reports serious adverse events affecting 0 of 26 participants at risk in the placebo group and 0 of 28 at risk in the Exelon patch group. It lists no serious adverse-event terms. The absence of listed serious events in these groups describes the posted counts; it does not establish that no other safety outcomes occurred or characterize how safety events were collected and adjudicated.

For other, non-serious adverse events, the record lists 9 affected participants among 26 at risk in the placebo group and 11 among 28 at risk in the Exelon patch group. These are counts of participants affected, not a comparison adjusted for exposure time or a measure of event severity. No percentages or rate comparisons are needed to read the posted counts.

Death counts are not reported in the registry record. The record does not give further details about event definitions, observation periods, withdrawals due to adverse events, or the methods used to assess treatment attribution. Because the study used a crossover design, safety summaries should also be read in the context of exposure to both phases; the registry’s group-level entries do not provide additional exposure details.

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

The posted primary result is a mean ADCS-CGIC score of 3.48 for the Exelon patch phase and 3.92 for the placebo phase, with an Exelon-minus-placebo mean difference of -0.44. Since lower scores indicate more favorable ratings, the point estimate is in the direction of a lower score for the Exelon phase. The interval extends from -0.929 to 0.049 when rounded to three decimal places and includes zero. Thus, the posted estimate is uncertain about whether the underlying mean contrast differs from zero.

This result does not establish that the treatment works or does not work. A p-value of 0.078 is posted without adjustment; it is not a measure of clinical benefit and does not express the probability that the treatment effect is real. The registry record does not report a threshold for a clinically meaningful difference on the mean score. It also does not post the proportion of participants who met the responder definition of scoring 1 or 2.

The endpoint is a global-change rating administered at the end of each phase. It can summarize an overall impression, but a mean does not show each participant’s rating or how many improved, stayed the same, or worsened. The registry lists the Montreal Cognitive Assessment as a secondary outcome, but the primary result is not a direct estimate for that separate measure.

The trial’s crossover design is relevant to interpretation because participants received both phases. The registry does not report enough analytic detail to determine how the posted comparison handled the repeated measurements, treatment sequence, period, or possible carryover. Missing-data handling and the analysis population are also not reported. Those omissions limit conclusions about the exact analysis and the scope of the estimate.

The safety entries report zero participants with serious adverse events in each treatment group, while other adverse events affected 9 placebo participants and 11 Exelon patch participants. Death counts are not reported in the registry record. These counts are descriptive and do not by themselves establish comparative safety, particularly without additional information about event definitions and exposure.

16. Why This Trial Matters Statistically

This trial illustrates why the estimated effect, its uncertainty, and the design details need to be read together. The point estimate is a difference in mean global-change scores, not a direct statement about each participant. Its negative direction corresponds to a lower, more favorable rating for the Exelon patch phase, but the interval crosses zero. The point estimate alone would conceal that uncertainty.

The crossover structure is another central statistical feature. Participants experienced both treatment phases, so the relationship between their scores across phases can affect the precision of a treatment contrast. The registry describes the crossover exposure but does not report whether the analysis used paired observations or how it addressed period and carryover. This distinction matters because the group means and standard deviations do not reveal the within-participant variability needed to reconstruct a paired comparison.

The endpoint also demonstrates that the metric determines what a result means. ADCS-CGIC is a seven-point measure of global change, and the primary result is expressed as a mean. The record separately defines a responder threshold, but does not present the main comparison as a responder analysis. A mean contrast and a comparison of responder counts answer different statistical questions.

Finally, registry reporting determines how much can be evaluated. The posted result includes sample summaries, an interval, and an unadjusted p-value, while important details such as the analysis population, missing-data method, multiplicity strategy, and model specification are not reported. Distinguishing reported results from unreported methods is essential to interpreting a small, completed clinical trial without overstating what its summary statistics establish.

17. 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 gives limited detail about the crossover implementation. It says participants received placebo and Exelon patch phases, but does not report the sequence allocation, the length of each phase, a washout interval, or how possible period and carryover effects were handled. The trial is described as 24 weeks, but further timing details are not reported.

The primary comparison is presented as a mean difference, but the analysis method is not reported. In particular, it is unclear from the registry record whether the comparison accounts for paired observations within participants. The posted group means and standard deviations do not provide the variation in each participant’s difference between phases. The reported counts of 26 and 28 should therefore be read as treatment-phase summaries, not as proof of independent parallel-group samples.

Two participants discontinued study participation, but the record does not state whether their endpoint measurements were available or how missing data were handled. The analysis population, treatment of incomplete crossover participation, and any sensitivity analyses are not reported. These unknowns constrain assessment of the estimate’s robustness.

The registry gives the ADCS-CGIC scale definition and a responder threshold, but no threshold for a clinically meaningful difference in the mean score. The primary result is a mean comparison, and no responder comparison is reported. The record lists a secondary MoCA outcome but does not provide a posted result for it in the registry record described. The multiplicity plan, detailed outcome administration procedures, and statistical analysis plan are not reported.

The posted safety information is also limited: it gives counts of participants affected by serious and other adverse events, but does not report death counts, event-level details, exposure duration by treatment, or further ascertainment methods. These specific reporting gaps limit what can be inferred from the record; they do not imply that the trial did not collect other information.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Mild Cognitive Impairment in Parkinson's Disease — NCT01519271.
• Registry record last updated 2017-04-07; status last verified 2017-02.
• Results first posted 2017-04-07.
• 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.

21. Record Summary

NCT01519271, titled “Mild Cognitive Impairment in Parkinson’s Disease,” is a completed Phase IV randomized, double-blind, placebo-controlled crossover study sponsored by the University of Pennsylvania. The official title describes a single-site study of Exelon Patch (rivastigmine transdermal system) for mild cognitive impairment in Parkinson’s disease. Enrollment was 28; the registry gives a start of 2011-12 and primary completion of 2014-06.

The primary endpoint was the end-of-phase ADCS-CGIC score. The posted means were 3.92 for placebo and 3.48 for Exelon patch, with a mean difference of -0.44 and an unadjusted p-value of 0.078. The posted interval spans zero. The registry lists MoCA as a secondary outcome and reports two outcome measures and one analysis.

The primary-outcome description notes that participants received both treatment phases and that two discontinued study participation. Analysis details, missing-data handling, and the trial’s original sample-size assumptions are not reported in the registry record. Serious adverse events affected 0 participants in either group; other adverse events affected 9 placebo participants and 11 Exelon patch participants. Death counts are not reported.