This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for Vestibular Stimulation in Parkinson's Disease (NCT02703844), a randomized na study. The primary outcome is Change From Baseline in the Nonmotor Symptom Severity Scale (NMSS).
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT02703844 |
| Phase | NA |
| Condition | Parkinson's Disease |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Triple |
| Enrollment | 46 |
| Arms in the primary outcome | 2 |
| Lead sponsor | University of Kent |
| Results posted | Yes |
2. Clinical Question
Population
People with Parkinson’s disease were studied. For the primary outcome, the registry says results include participants who received at least one treatment with the study device and had a baseline NMSS assessment. The posted primary-outcome data include 23 participants in each arm.
Intervention
The active intervention was caloric vestibular stimulation. The arm description gives a schedule of 8 weeks, 7 days a week, twice daily for 19 minutes.
Comparator
The comparator was sham caloric vestibular stimulation, scheduled in the same manner as the active treatment: 8 weeks, 7 days a week, twice daily for 19 minutes. The registry also describes later crossover from placebo to active treatment in an unblinded fashion, but the stated condition for crossover is incomplete in the registry record.
Primary question
The primary question was whether the change in NMSS total score from baseline to the end of treatment at week 12 differed between the active and sham groups. Baseline was defined as the average of two visits. The NMSS measures non-motor symptom severity and frequency across nine dimensions.
The posted group summaries describe change from baseline, not the absolute score at week 12. The active arm’s median change was -33.5, with a 95% confidence interval from -40.3 to -17.7. The placebo arm’s median change was -2.0, with a 95% confidence interval from -16.5 to 8.3. Negative change denotes a lower scale total relative to baseline. The registry reports an ANCOVA p-value of 0.009 for the comparison, but does not report an adjusted effect estimate or confidence interval for that comparison.
This question concerns a measured scale outcome over the study’s stated assessment period. It does not, by itself, answer whether the intervention affects every non-motor symptom, how durable any observed change might be, or whether results apply beyond the participants represented in the registry outcome data.
3. Trial Design
The registry describes a randomized, triple-masked, parallel-group study of caloric vestibular stimulation in Parkinson’s disease. The listed phase is NA. The University of Kent is named as sponsor. Enrollment is listed as 46, and the record is marked completed. The study start is recorded as 2016-03 and primary completion as 2017-10-09.
Participants were assigned to an active device treatment or sham device treatment. Both arm descriptions specify treatment over 8 weeks, 7 days a week, twice daily for 19 minutes. The sham description also notes that participants allocated to that arm would later cross over, in an unblinded fashion, to active treatment; the description ends before the crossover condition is complete. Details of the allocation procedure and the identities of the people covered by triple masking are not reported in the registry record.
The primary endpoint was change in NMSS total score at week 12 compared with the average of two baseline visits. The outcome description says the analysis included participants who received at least one device treatment and had a baseline NMSS assessment. There are 23 participants in each arm’s posted primary-outcome summary. These outcome-specific numbers should not be treated as a complete account of all enrollment, follow-up, or missing assessments.
The registry lists 14 outcome measures, including secondary measures of change in MDS-UPDRS Part II and Part III. Two secondary outcomes are posted, and one analysis is listed. The registry does not report the full statistical analysis plan, the rationale for the planned enrollment, or how the outcome measures were prioritized. The trial’s parallel-group description and later crossover note are both relevant: the primary comparison is reported for active and placebo groups, while a separate Cross-Over group appears in the adverse-event data.
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.
Randomized Controlled Portion
| Stage | Active | Placebo |
|---|---|---|
| STARTED | 23 | 23 |
| COMPLETED | 16 | 17 |
| NOT COMPLETED | 7 | 6 |
Cross-Over to Active
| Stage | Active | Placebo |
|---|---|---|
| STARTED | 0 | 14 |
| COMPLETED | 0 | 5 |
| NOT COMPLETED | 0 | 9 |
Analysis population for the primary outcome, as defined in the registry: data reported for this outcome includes anyone who received at least one treatment with the study device and had a baseline NMSS performed. Change scores were not normally distributed and therefore nonparametric tests were used and medians are reported
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.
| Characteristic | Unit | Active | Placebo | Total |
|---|---|---|---|---|
| Participants analysed | 23 | 23 | 46 | |
| Age, Continuous | years | 69.7 (SD 11.3) | 72.2 (SD 6.6) | 70.9 (SD 9.2) |
| Sex: Female, Male: Female | Participants | 11 | 5 | 16 |
| Sex: Female, Male: Male | Participants | 12 | 18 | 30 |
| Race and Ethnicity Not Collected | Participants | — | — | 0 |
| Non-Motor Symptom Scale (NMSS) Total Score | score on a scale | 126.2 (SD 29.8) | 117.7 (SD 32.6) | 122 (SD 31.2) |
| MDS-UPDRS Part II Score | score on a scale | 22.5 (SD 9.1) | 21.4 (SD 5.3) | 21.9 (SD 7.4) |
| MDS-UPDRS Part III Score | score on a scale | 49.3 (SD 17.6) | 45.4 (SD 16.3) | 47.4 (SD 16.9) |
6. Primary Endpoint
The registry lists the primary outcome as Change From Baseline in the Nonmotor Symptom Severity Scale (NMSS), measured over Change at end of treatment (week 12) relative to the average of two baseline visits.
| Element | Registry entry |
|---|---|
| Result type | Median |
| Measure of spread | 95% Confidence Interval |
| Unit | change in scale total score |
| Endpoint classification | not reduced to a single comparison |
The primary endpoint is change from baseline in the Nonmotor Symptom Severity Scale (NMSS) total score. The registry describes the NMSS as a 30-item, rater-based scale assessing the severity and frequency of non-motor symptoms across nine dimensions in people with Parkinson’s disease. Its score range is 0-360, with 0 indicating no symptom burden. A change score expresses how the total at the stated assessment differs from baseline; a negative score change therefore represents a lower total relative to baseline.
The endpoint timeframe is change at the end of treatment, identified as week 12, relative to the average of two baseline visits. Using an average of two baseline assessments is part of the registered definition of the comparison point; further details of how the two visits were scheduled or handled are not reported in the registry record.
A scale covering multiple dimensions can provide a structured way to assess symptoms that are not captured by a single symptom measure. The NMSS therefore represents a broad non-motor symptom outcome rather than a measure of one specific symptom. As a general measurement point, a total score condenses information across its items and dimensions, so it does not show which individual components account for a change.
The registry says the change scores were not normally distributed and that nonparametric tests were used, with medians reported. The posted arm summaries are median changes with 95% confidence intervals. The record does not specify the construction method for those intervals or provide an adjusted difference between groups with its own confidence interval.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The active arm’s median NMSS change was -33.5, with a 95% confidence interval from -40.3 to -17.7. The placebo arm’s median change was -2.0, with a 95% confidence interval from -16.5 to 8.3. Both summaries are changes relative to the average of two baseline visits, and both arm summaries include 23 participants. The active-arm interval is entirely below zero, whereas the placebo-arm interval includes zero. These intervals describe uncertainty around the arm summaries; they are not an interval for the difference between arms.
The registry reports an ANCOVA analysis for the active-versus-placebo comparison, with p = 0.009. This is evidence that the reported analysis found a between-group result inconsistent with its null hypothesis at conventional levels of statistical testing, but the registry does not report the ANCOVA contrast estimate or its confidence interval. The p-value does not quantify the size of the adjusted difference, and it is not the probability that the null hypothesis is true. It also does not establish the clinical importance of the observed scale changes.
The arm medians and their intervals should not be substituted for the ANCOVA result. The median summaries describe each group separately, while ANCOVA is a model-based comparison. The registry does not report the covariates used in that model, the estimated adjusted contrast, or an interval for the contrast. Consequently, the posted information supports describing the direction and uncertainty of the separate arm summaries and reporting the sponsor’s p-value, but not reconstructing the magnitude or precision of the adjusted comparison.
As a general statistical point, a confidence interval gives a range of values compatible with the data and the method used to construct it, under the method’s assumptions. Overlap or non-overlap between separate group intervals is not a substitute for a confidence interval or hypothesis test for the group contrast. Likewise, statistical significance does not establish that a difference is important to patients or that it persists after the measured period.
The reported primary analysis population is defined by receiving at least one device treatment and having a baseline NMSS assessment. That is not necessarily the same as everyone randomized. The registry does not report how many randomized participants lacked an eligible baseline or treatment, how missing week-12 data were handled, or whether the analysis followed an intention-to-treat principle. Those omissions limit assessment of potential selection or attrition effects.
The record also lists 14 outcome measures and two posted secondary outcomes, but does not report a multiplicity strategy. If several outcomes or analyses were considered, the interpretation of one p-value can depend on whether and how multiple testing was addressed. No adjustment should be assumed from the information in the registry record. The result is therefore best read as the registry’s reported primary comparison, with the uncertainty about analysis details kept in view.
9. Statistical Methods Explained
Why are medians reported?
The registry states that change scores were not normally distributed and that nonparametric tests were used; medians are reported. The median is the middle value when observations are ordered. As a general statistical point, it is less sensitive than the mean to unusually large or small values. That can make it a useful summary when a distribution is skewed or when the assumptions of methods based on normally distributed data are not suitable.
A median does not describe all features of a distribution. Two groups can have the same median but differ in spread or shape, and a change in the median does not say that every participant changed by that amount. The registry gives confidence intervals around the median summaries but does not report the interval-construction method or additional distribution summaries.
What does the ANCOVA result contribute?
The registry lists ANCOVA as the method for the active-versus-placebo analysis and reports p = 0.009. As a general statistical concept, analysis of covariance compares groups while accounting for specified covariates in a model. Which covariates were included, how the model was specified, and whether any assumptions were checked are not reported in the registry record. The reported method name alone does not supply those details.
The ANCOVA p-value addresses the comparison carried out by that model, not the separate question of whether either arm’s median change differs from zero. The registry does not give an ANCOVA effect estimate or confidence interval. An effect estimate would convey the modeled group contrast, while its confidence interval would communicate the precision of that estimate. Without these, the p-value alone cannot show the magnitude or precision of the adjusted comparison.
The record reports both nonparametric handling of the change scores and an ANCOVA comparison. It does not explain how those methods relate to each other, whether the ANCOVA was performed on a particular transformed or otherwise modeled outcome, or what model assumptions were used. Those details are not reported in the registry record, so the analysis cannot be independently reconstructed from the posted summaries.
How should the confidence intervals be read?
The active and placebo arm summaries each include a 95% confidence interval: -40.3 to -17.7 for active and -16.5 to 8.3 for placebo. In general, a confidence interval expresses uncertainty in an estimate according to a statistical procedure. It is not a range containing every participant’s change, and it is not a direct interval for the difference between groups. The registry does not report a confidence interval for the ANCOVA comparison.
As a general statistical point, the comparison of two estimates requires attention to their joint uncertainty. Separate arm intervals do not by themselves provide the uncertainty of a contrast. This is why the reported ANCOVA p-value and the arm-specific medians answer related but distinct questions.
What can planning illustrations tell readers?
Planning sample-size illustrations are based on assumptions and are not the same as observed trial results. The stated illustration assumptions use a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation. The effect to detect is defined as a stated fraction of the observed effect. For outcomes other than time-to-event outcomes, the quantities are participants per arm. Any such illustration is a planning scenario, not observed power, and it does not establish the probability that the trial detected a real effect.
As a general statistical point, planning calculations depend on assumptions about the effect to detect, variability, allocation, and testing procedure. If an observed effect is used to construct an illustration, the result is still a hypothetical planning calculation, not a retrospective measure of the trial’s success or an independent confirmation of its result. The registry does not report the trial’s own sample-size justification, so the posted illustration should not be mistaken for the original design rationale.
What do the posted analysis details leave open?
The registry lists one analysis, 14 outcome measures, and two posted secondary outcomes. The primary outcome’s analysis population is described as participants who received at least one treatment and had a baseline NMSS performed. The record does not give a complete participant flow, the missing-data procedure, the prespecified covariates for ANCOVA, or a multiplicity plan.
As a general statistical matter, missing outcomes can affect an analysis if the participants with observed follow-up differ systematically from those without it. The appropriate interpretation depends on how much data are missing and on the assumptions used to handle them. Because those details are not reported in the registry record, the extent to which missingness could influence the primary comparison cannot be assessed from the posted description.
Multiplicity concerns arise when more than one outcome or hypothesis is tested. As a general concept, conducting multiple tests can increase the chance of at least one apparently notable result if no adjustment or prespecified hierarchy is used. The registry does not state whether an adjustment, hierarchy, or other multiplicity approach was used. The listed p-value should therefore be reported as presented, without assuming that it was or was not adjusted.
10. Analyses Reported by the Sponsor
| Method | Estimate | 95% CI | P-value |
|---|---|---|---|
| ANCOVA | not reported | 0.009 |
The University of Kent is named as sponsor. The registry’s reported between-arm analysis is ANCOVA, with p = 0.009. The arm-level results are reported as medians and 95% confidence intervals because the change scores were not normally distributed and nonparametric tests were used. These are not interchangeable statistical summaries: the median results describe each arm, whereas the ANCOVA result concerns a modeled comparison.
The registry does not report the ANCOVA contrast estimate, its confidence interval, the model covariates, or enough analytical detail to reproduce that analysis from the arm summaries. The posted medians and intervals alone cannot establish the ANCOVA estimate. Accordingly, the sponsor’s reported p-value can be stated as the trial’s own comparison result, but its magnitude and precision cannot be checked against an adjusted estimate using the registry information.
No numerical recalculated between-arm estimate or p-value is reported in the registry record. The available arm summaries can be described as posted, but they do not provide a like-for-like recalculation of the ANCOVA model. The information in the registry record does not identify a discrepancy between a recalculated result and the sponsor’s analysis.
11. Multiplicity, Missing Data and Analysis Populations
The registry lists 14 outcome measures and two posted secondary outcomes, as well as one analysis. It reports the NMSS change as the primary outcome, with an ANCOVA p-value of 0.009. A complete multiplicity plan is not reported in the registry record. In particular, the record does not say whether testing was controlled across outcomes, whether a hierarchy was prespecified, or whether the reported p-value was adjusted.
As a general statistical point, multiplicity matters when several hypotheses are tested because the interpretation of an individual p-value depends partly on the broader testing plan. The listed outcome count alone does not establish how many tests were performed or what correction, if any, was used. The p-value should therefore be read as the reported result for the listed comparison, without inferring a multiplicity procedure.
The outcome description defines the primary analysis population as people who received at least one treatment with the device and had a baseline NMSS assessment. The posted primary summaries include 23 participants in each arm. The registry does not report a complete account of randomized participants with and without baseline and follow-up measurements, reasons for missing outcomes, or the method used to handle missing week-12 values.
It is also not reported in the registry record whether the analysis was intention-to-treat, per-protocol, or another prespecified population, beyond the stated treatment-and-baseline criterion. The record does not state whether imputation or sensitivity analyses were used, or whether the analysis population differed across outcomes. These gaps limit assessment of attrition and missing-data assumptions. They do not establish that missing data occurred or that any particular bias resulted; they mean that those issues cannot be evaluated fully from the registered description.
12. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| Active | 2 | 23 | 8.7% |
| Placebo | 1 | 23 | 4.3% |
| Cross-Over | 0 | 14 | 0.0% |
Deaths
| Arm | Deaths (participants affected of at risk) |
|---|---|
| Active | 0 of 23 |
| Placebo | 0 of 23 |
| Cross-Over | 0 of 14 |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 5% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| Active | 6 of 23 |
| Placebo | 4 of 23 |
| Cross-Over | 6 of 14 |
Most frequent serious adverse event terms
| Event term | Organ system | Active | Placebo | Cross-Over |
|---|---|---|---|---|
| Arm fracture | Musculoskeletal and connective tissue disorders | 0 of 23 | 1 of 23 | n/a of 14 |
| Knee Infection | Infections and infestations | 1 of 23 | 0 of 23 | n/a of 14 |
| Fainting Episode | Nervous system disorders | 1 of 23 | 0 of 23 | n/a of 14 |
Most frequent non-serious adverse event terms
| Event term | Organ system | Active | Placebo | Cross-Over |
|---|---|---|---|---|
| Fall | Nervous system disorders | 1 of 23 | 2 of 23 | 4 of 14 |
| Dizziness | Nervous system disorders | 3 of 23 | 0 of 23 | 1 of 14 |
| Chest Infection | Infections and infestations | 0 of 23 | 2 of 23 | 1 of 14 |
| Ear discomfort | Ear and labyrinth disorders | 2 of 23 | 0 of 23 | 0 of 14 |
Serious events with the highest arm-level rate: arm fracture (4.3%), knee infection (4.3%), fainting episode (4.3%).
The adverse-event data list serious events affecting 2 participants in the Active group, 1 in the Placebo group, and 0 in the Cross-Over group. The corresponding numbers at risk are 23, 23, and 14. The registry lists 0 deaths in each of the three groups. These are participant counts as presented, not estimates of comparative risk or evidence that events were caused by a treatment.
The registry separately reports other adverse events affecting 6 participants in Active, 4 in Placebo, and 6 in Cross-Over, with the same group-specific numbers at risk. The record does not explain the detailed classification rules for the serious and other categories, nor does it provide exposure-adjusted rates or a statistical comparison of these counts. The Cross-Over group is distinct in these safety data and should not be merged with either original group without further information.
Arm fracture, Knee Infection, and Fainting Episode are each listed among the top serious-event terms at 4.3%. The registry does not identify the group attribution for these terms in the listed summary, so they should not be assigned to an arm based on the information reported.
As a general statistical point, adverse-event counts from small groups can be imprecise, and simple counts do not establish causality. The posted figures describe reported events in the registry categories; they do not by themselves establish comparative safety.
13. What the Result Does, and Does Not, Mean
The result addresses one registered primary question: change in NMSS total score at week 12 relative to the average of two baseline visits. The active group’s median change was -33.5, and the placebo group’s was -2.0. The registry reports an ANCOVA comparison with p = 0.009. Together, these are the central posted statistical findings for the primary endpoint, but they do not provide an adjusted effect size or an interval for the between-group contrast.
The numerical summaries do not establish why scores changed, whether every participant experienced a change in the same direction, or which NMSS dimensions contributed most. The scale combines information across 30 items and nine dimensions. A total-score comparison cannot by itself identify effects on individual symptoms. The record also does not provide information about the clinical importance threshold for a change in total score.
The analysis concerns the stated assessment at week 12 and does not establish whether any observed difference continued later. The registry does not report follow-up beyond the specified primary endpoint timeframe in the information used for this description. The trial’s result should not be interpreted as a statement about outcomes outside the population, treatment schedule, comparator, and assessment period recorded.
Statistical evidence and clinical interpretation are related but distinct. As a general point, a p-value indicates how the observed result relates to a statistical null model under the analysis assumptions; it does not measure the probability that the treatment is effective, the size of a difference, or its importance to patients. The missing ANCOVA estimate and confidence interval prevent direct evaluation of the adjusted difference’s magnitude and precision from the registry results.
The result also needs to be read alongside the analysis-population definition, unreported missing-data details, and the absence of a reported multiplicity plan. These are limits on what can be concluded from the registry record, not proof that the analysis was incorrect. The finding is a trial result on a scale-based primary endpoint and should not be extended to unmeasured outcomes or broader claims.
14. Why This Trial Matters Statistically
This trial is statistically informative because its registry record shows several distinct layers of evidence that should not be collapsed into one number. The arm summaries are medians with confidence intervals, while the between-arm result is an ANCOVA p-value. Keeping those summaries separate helps prevent an arm-specific change from being mistaken for an adjusted treatment contrast.
The outcome description explicitly notes that change scores were not normally distributed and that nonparametric tests were used, with medians reported. That detail gives context for the descriptive summaries. At the same time, the registry lists ANCOVA for the group comparison without reporting its model covariates or an adjusted estimate. This illustrates why a method label and p-value alone do not provide a complete account of a statistical analysis.
The record also makes the analysis population visible: participants had to receive at least one device treatment and have a baseline NMSS assessment. That description matters when interpreting the 23 participants in each arm’s posted primary summary. Without participant flow and missing-outcome information, the extent to which these data represent all randomized participants cannot be determined from the registry record.
Finally, the registry lists multiple outcome measures but does not state a multiplicity strategy, and its safety reporting includes a separate Cross-Over group. Those details encourage careful reading of what each count and test actually represents. As a general statistical lesson, transparent effect estimates, confidence intervals, analysis-population definitions, and missing-data and multiplicity plans make trial findings easier to interpret than a p-value or headline result alone.
15. 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 does not report the ANCOVA effect estimate or its confidence interval. Consequently, the adjusted difference between groups and its precision cannot be read directly from the posted primary analysis. The group medians and their intervals are useful descriptions but do not supply that adjusted contrast.
The ANCOVA covariates, model specification, and assumption checks are not reported in the registry record. The record says change scores were not normally distributed and that nonparametric tests were used, but does not explain how that statement relates to the listed ANCOVA comparison. This limits independent assessment or reproduction of the analysis.
The primary-outcome population includes people with at least one device treatment and a baseline NMSS assessment. A complete flow from allocation through week-12 assessment, reasons for missing measurements, missing-data handling, and sensitivity analyses is not reported. The registry therefore does not establish how the available outcome participants compare with everyone randomized.
The record lists 14 outcome measures and two posted secondary outcomes, but does not report a multiplicity approach. It also does not report the trial’s sample-size justification, interim-analysis plan, or full statistical analysis plan. The planning illustrations described for the page are not a substitute for the trial’s original design rationale.
The arm description indicates a later unblinded crossover from placebo to active treatment but is incomplete about the condition for crossover. The safety data list a Cross-Over group, but the registry information does not provide enough detail to interpret its relationship to the primary comparison or to combine it with the original groups. Finally, the posted primary endpoint concerns the specified week-12 scale change; it cannot establish longer-term outcomes or changes in individual NMSS dimensions.
16. Related Tutorials
17. Related Statistical Calculators
18. Sources
- ClinicalTrials.gov: Vestibular Stimulation in Parkinson's Disease — NCT02703844.
- Registry record last updated 2022-10-31; status last verified 2022-10.
- Results first posted 2022-10-31.
- 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.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
19. Record Summary
NCT02703844, titled Vestibular Stimulation in Parkinson’s Disease, is listed as a completed randomized, triple-masked, parallel-group study sponsored by the University of Kent. Enrollment is 46. The study start is recorded as 2016-03, with primary completion on 2017-10-09. The registered intervention is caloric vestibular stimulation, compared with sham stimulation.
The primary endpoint is change in NMSS total score at week 12 relative to the average of two baseline visits. The posted median changes are -33.5 for Active and -2.0 for Placebo, with 23 participants in each arm. The registry reports an ANCOVA p-value of 0.009 but does not report its effect estimate or confidence interval.
The record lists 14 outcome measures and two posted secondary outcomes. It also reports serious and other adverse-event counts for Active, Placebo, and Cross-Over groups, with 0 deaths in each. The full analysis plan, multiplicity approach, and missing-data methods are not reported in the registry record.