"Pilot Study of Armodafinil in Patients With Dementia With Lewy Bodies": Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for "Pilot Study of Armodafinil in Patients With Dementia With Lewy Bodies" (NCT01023672), a na phase 4 study. The primary outcome is Mean Initial Sleep Latency as Measured by the Maintenance of Wakefulness Test (MWT).

2. Clinical Question

Statistically, the key issue is that the study can describe participants’ test performance over time, but it cannot separate a change during treatment from changes that might also occur without treatment. Open-label assessment may also be affected by expectations or knowledge of treatment. The registry lists no randomized comparison, and its design does not support a like-for-like comparison between treatment and control arms. Any interpretation therefore needs to distinguish the observed within-group description from evidence of a treatment effect.

3. Trial Design

The registry describes an open-label, single-center pilot study of armodafinil in participants with dementia with Lewy bodies. It is listed as a phase 4 study, with a single-group model, allocation not applicable, and no masking. Enrollment was 20. The record lists Mayo Clinic as sponsor and marks the study completed. The recorded start is 2009-11 and the primary completion is 2012-03.

All participants were assigned to the same intervention: oral armodafinil daily at 150-250 mg. The study summary says treatment response was assessed over 12 weeks and explicitly states that there was no placebo medication. Thus, the study has no concurrent untreated or placebo group against which to compare the change in the primary outcome. Baseline testing can provide a reference for later measurement, but it does not serve the same purpose as random allocation to a control group.

The primary outcome is reported at baseline and 12 weeks. The endpoint population is described as per protocol. The registry states that three participants did not complete the study and did not complete this test: two had worsening disease and one died. The precise number of participants included in the endpoint analysis is not reported in the registry record, nor is a fuller account of how incomplete measurements were handled.

The record lists one outcome measure and one analysis. It does not describe an allocation sequence because allocation is not applicable, and the masking field is none. Detailed procedures for adherence monitoring, treatment discontinuation, visit windows, or the statistical analysis plan are not reported in the registry record. Those details should not be inferred from the single-group label or from the use of baseline and follow-up measurements.

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:

Analysis population for the primary outcome, as defined in the registry: Per Protocol; 3 subjects did not complete the study (2 had worsening disease, and 1 died) and did not complete this test.

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 Mean Initial Sleep Latency as Measured by the Maintenance of Wakefulness Test (MWT), measured over baseline, 12 weeks.

The primary endpoint is median initial sleep latency, measured in minutes with the Maintenance of Wakefulness Test. Sleep latency is the time taken to fall asleep during a test opportunity. The registry describes the test as an assessment of a participant’s ability to stay awake while sitting quietly in a dim room, with back and neck support and monitoring using measures also used in a standard overnight sleep study.

The test has four parts, each lasting up to 40 minutes if the participant remains awake for that long, with the parts spaced at 2 hour intervals through the day. The registry notes that in healthy people sleep latency may be approximately 30 minutes and that more than 97% of people take eight minutes or longer to fall asleep. It states that latency less than eight minutes is considered abnormal. These descriptions explain the test’s context; they do not provide the study participants’ results.

The outcome is recorded at baseline and 12 weeks, with the median and inter-quartile range specified as the summary statistics. A median describes the central position of the observed measurements and is less influenced by unusually high or low values than a mean. The inter-quartile range describes the spread of the middle portion of the measurements. The numerical medians and inter-quartile ranges at either assessment, and any numerical change between them, are not reported in the registry record. Therefore the endpoint definition is available, but its observed result cannot be described quantitatively.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry identifies a median and inter-quartile range for the primary outcome, measured at baseline and 12 weeks. It does not report the numerical summaries or the statistical test used. The explanations below describe general statistical ideas relevant to interpreting this kind of outcome; they are not claims about methods that the investigators used.

What does a median describe?

In general, the median is the middle value after observations are ordered. It can be useful when a distribution is skewed or includes unusually high or low observations, because such observations have less influence on the median than on the mean. The inter-quartile range describes the spread of the middle half of the observations. Together, these summaries convey a typical value and some information about variation, but they do not show every participant’s result or how individual results changed.

Here, the endpoint is explicitly summarized using the median and inter-quartile range. The actual values are not reported in the registry record. Without them, it is not possible to describe the observed center or spread at baseline or follow-up. Nor can those summaries alone show whether the same participants improved, worsened, or had no change. A group-level median at each time point and a summary of paired individual changes answer related but distinct questions.

Why does pairing matter?

In general, repeated measurements from the same person are related. A comparison that accounts for that pairing focuses on each participant’s change between assessments, rather than treating baseline and follow-up observations as if they came from separate, unrelated groups. The choice of method depends on the data distribution, the estimand, and the analysis plan. The registry does not report which test or model was used for the primary endpoint, whether a paired analysis was conducted, or whether the main comparison was based on change scores or separate time-point summaries. No particular analysis should be attributed to the study without that information.

There are also practical complications when some participants lack a follow-up measurement. A complete-case analysis uses only participants with the required observed data. Other approaches may use assumptions or models to incorporate incomplete records. In general, these methods do not automatically eliminate bias: their validity depends on why measurements are missing and on the assumptions made. The registry identifies three participants who did not complete the study and did not complete the test, but it does not report a missing-data method or the exact endpoint analysis count.

What can a confidence interval show?

In general, a confidence interval gives a range of values compatible with an estimate under the method’s assumptions and conveys statistical precision. A narrow interval indicates greater precision than a wide interval, but an interval does not correct for bias from an uncontrolled design, missing data, or measurement limitations. The registry does not report a confidence interval for the primary outcome. Without an estimate and its uncertainty, the size and precision of a possible change cannot be evaluated.

A p-value, when used, addresses compatibility between observed data and a specified null model under the assumptions of the statistical test. It does not measure the probability that a treatment works, and it does not by itself communicate the size or importance of an effect. The registry does not report a p-value for the primary endpoint or identify the test that would generate one. No inference about statistical significance can therefore be made from the endpoint description alone.

Why is an uncontrolled comparison limited?

In general, a concurrent control group helps distinguish changes associated with an intervention from changes that may occur over the same period for other reasons. Random allocation can help balance measured and unmeasured factors between groups, on average. This study is recorded as a single-group, open-label study and has no placebo group. A baseline measurement is useful for describing participants before treatment, but the baseline is not a concurrent control and cannot account for other changes over time.

Open-label status means that masking is absent in the registry record. In general, awareness of treatment can affect expectations, behavior, reporting, or assessment. The degree to which any such influence mattered in this study is not reported. The test includes structured conditions and physiologic monitoring, but its format alone does not establish that outcome assessment was free of all potential influences.

How should planning illustrations be read?

In general, a sample-size calculation is a planning exercise based on assumptions about the target effect, variability, significance level, allocation, and analysis. A calculation made from an observed effect is an illustration of what a future design might require under those assumptions; it is not an observed trial result and is not observed power. The registry’s stated planning assumptions use a two-sided significance level of 0.05, equal allocation between two arms, a normal approximation, and effects expressed as stated fractions of the observed effect. Those assumptions describe a hypothetical two-arm planning exercise, not this single-group study’s design. No planning count or observed endpoint effect is reported in the registry record.

General methods cannot fill gaps in the registered results. In particular, the endpoint definition does not reveal the outcome values, an unspecified analysis cannot be reconstructed, and a planning illustration cannot serve as evidence of what happened among participants. The appropriate statistical reading is therefore limited to the design, endpoint, reported completion information, and adverse-event counts that the registry records.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists one outcome measure and one analysis. The primary outcome is median Maintenance of Wakefulness Test sleep latency, assessed at baseline and 12 weeks. The record does not report a multiplicity adjustment, a hierarchy of outcomes, or a broader statistical analysis plan. It is therefore not possible to determine whether additional analyses were planned, how they were prioritized, or whether any adjustment for multiple testing was intended. The number of listed outcomes alone does not establish the complete set of analyses undertaken.

For the primary endpoint, the stated population is per protocol. The registry says that three participants did not complete the study and did not complete this test: two had worsening disease and one died. The exact number included in the endpoint analysis is not reported in the registry record. Nor does the record describe whether any measurements were missing for other reasons, whether participants who discontinued treatment were assessed, or whether any imputation or other missing-data method was used.

In general, complete-case results can be difficult to interpret when missingness is related to health status or outcome. The reasons recorded for noncompletion make it important to know how missing observations were treated, but the record does not provide enough information to assess the consequences. No assumption about the unobserved test results should be made.

The registry also does not report a protocol-level analysis population definition beyond the per-protocol label, an intention-to-treat analysis, sensitivity analyses, or the handling of departures from treatment. These omissions do not show that such analyses were or were not performed; they mean the details are not reported in the registry record.

11. 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 adverse-event summary lists 20 participants at risk for serious adverse events and zero participants affected. It lists 20 participants at risk for other adverse events and five participants affected. No serious adverse event terms are listed. These are counts in the registry record and should be read separately from the primary sleep-latency endpoint.

The death count in the adverse-event summary is not reported in the registry record. The primary endpoint description separately states that one participant died and did not complete the study or the test. The registry does not explain how that death was classified in the adverse-event summary, so the two entries should not be reconciled by assumption.

A count of zero participants affected by serious adverse events in this record does not establish that serious events could not occur, nor does it by itself establish comparative safety. The study has no control group, and the record gives no comparative safety estimate. The number of participants and the study duration also constrain how much can be learned about uncommon or delayed events. These are general considerations when interpreting safety counts from a small, uncontrolled study; they do not add event findings beyond those recorded.

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

This trial matters as a registered pilot study of daily armodafinil in people with dementia with Lewy bodies, using a structured test of the ability to remain awake as its primary outcome. It records treatment over 12 weeks and specifies baseline and follow-up assessment. That design provides a framework for studying a sleepiness-related measure in this population, but the posted record does not report the endpoint’s numerical values.

The study does not establish a controlled treatment effect. It is open label, single group, and has no placebo or concurrent comparator. If a within-group change had been reported, it would still need to be distinguished from changes over time, participant expectations, and other influences. Because the numerical primary results are not reported in the registry record, even the direction and size of the described outcome change cannot be assessed.

The result also should not be read as a general conclusion about all people with dementia with Lewy bodies. The record concerns 20 enrolled participants at a single center, with three participants not completing the study or the test. The per-protocol designation and lack of detail about missing-data handling limit assessment of how the observed group, if reported, would represent those enrolled.

The safety counts describe what the registry lists: zero participants affected by serious adverse events among 20 at risk, five affected by other adverse events among 20 at risk, and no serious event terms listed. The death count in the safety summary is not reported in the registry record, while the endpoint description says one participant died and did not complete the test. These data should be treated as descriptive counts, not as a comparative assessment.

In short, the record specifies a clinical question and a measurable primary endpoint, and documents incomplete follow-up and adverse-event counts. It does not report the primary endpoint estimates, their precision, or a controlled comparison. The record therefore cannot support a quantitative conclusion about change in sleep latency or a causal conclusion about the intervention.

13. Why This Trial Matters Statistically

This study illustrates several statistical issues that arise in small clinical studies. First, the endpoint is clearly defined: median sleep latency on a test designed to measure the ability to stay awake, at baseline and 12 weeks. A clearly specified measure gives readers a basis for understanding what was intended to be assessed. However, a defined outcome is not the same as a reported outcome. The numerical median and inter-quartile range are not reported in the registry record, so readers cannot assess the observed values or their spread.

Second, the design makes the distinction between measurement over time and comparison between treatment groups especially important. With a single group, change from baseline is descriptive. There is no concurrent control to help identify what might have happened over the same period without the intervention. That distinction is central to interpreting pilot studies and prevents a within-group pattern from being treated as a causal estimate.

Third, the record identifies three participants who did not complete the primary test, including two with worsening disease and one who died. In general, incomplete follow-up can affect both the size and representativeness of an analyzed group. The absence of a reported analysis count and missing-data method makes it difficult to judge how those losses were handled.

Finally, the record shows why planning illustrations must be kept separate from trial findings. Its stated assumptions describe a hypothetical two-arm calculation, although the registered study itself is single group. Such planning is not an observed result, and it cannot substitute for estimates and uncertainty from the actual participants. The statistical value of the record is therefore in clarifying the design and its interpretive limits, rather than in supporting a numerical conclusion about the primary endpoint.

14. Limitations

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

The central limitation is that numerical results for the primary endpoint are not reported in the registry record. Although the outcome, unit, summary statistic, and assessment times are specified, the medians and inter-quartile ranges are absent. There is no reported estimate of change, confidence interval, or p-value. Direction, magnitude, and precision therefore cannot be evaluated.

The single-group, open-label design has no concurrent control. Changes between baseline and 12 weeks cannot be attributed to treatment on the basis of this design alone. The registry records no masking, and does not report details about outcome assessor procedures or other care during follow-up. Such details should not be inferred.

Three participants did not complete the study and did not complete the primary test; two had worsening disease and one died. The endpoint is described as per protocol, but the exact analysis count and handling of missing observations are not reported in the registry record. The record also does not report a detailed statistical analysis plan, the specific test or model used, or sensitivity analyses. These gaps prevent assessment of analytic choices and the possible effect of missing data.

The study is a pilot at one center, with 20 enrolled participants and a 12-week treatment period. Its size and duration limit the scope of what can be learned, particularly for uncommon or later events. The adverse-event summary lists counts but offers no control-group comparison; the death field is not reported there, although the endpoint description notes one death among those who did not complete the test. The registry does not explain the difference in reporting context.

Finally, the planning assumptions describe a hypothetical two-arm exercise, not the study’s single-group design. They should not be interpreted as the trial’s own design justification or as evidence of observed power. The record does not report a sample-size justification or other protocol details that would support a fuller evaluation.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: "Pilot Study of Armodafinil in Patients With Dementia With Lewy Bodies" — NCT01023672.
• Registry record last updated 2013-09-05; status last verified 2013-06.
• Results first posted 2013-09-05.
• 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.

18. Record Summary

The registry describes a completed, phase 4, open-label, single-center pilot study of daily oral armodafinil in dementia with Lewy bodies. Enrollment was 20, and treatment was described as 150-250 mg daily over 12 weeks. The design is single group, with no placebo or concurrent control and no masking.

The primary outcome is median sleep latency in minutes on the Maintenance of Wakefulness Test, measured at baseline and 12 weeks. The registry specifies an inter-quartile range and a per-protocol population, but the endpoint values and analysis precision are not reported in the registry record. Three participants did not complete the study or test: two had worsening disease and one died.

The safety summary lists zero serious adverse events among 20 at risk, five participants with other adverse events among 20 at risk, and no serious event terms. The death count in that summary is not reported. Overall, the record defines the study and its endpoint but does not report the numerical primary results or a controlled estimate of treatment effect.