Connect.Parkinson: Connecting Individuals With Parkinson Disease to Specialists in Their Homes: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Connect.Parkinson: Connecting Individuals With Parkinson Disease to Specialists in Their Homes (NCT02038959), a randomized na study. The primary outcome is Feasibility of Virtual Visits for Parkinson Disease.

2. Clinical Question

The study’s stated aims extend beyond feasibility. They include assessing whether home-based telemedicine could improve quality of life, enhance quality of care, save time, reduce travel, and decrease care-partner burden. Those aims describe questions the study set out to address; they are not themselves evidence that a change occurred.

The primary endpoint is specifically a count of participants who completed at least one virtual visit successfully. It therefore concerns whether the intervention could be delivered at least once, rather than the number of visits received, the quality of each visit, or a clinical change in Parkinson disease. The secondary outcomes include measures of health status, cognition, Parkinson disease symptoms, care experience, and provider-visit time. The posted record includes numeric results for only some of those secondary measures.

The comparison requires care in interpretation: the control arm was not assigned virtual visits, whereas the endpoint is defined around completing a virtual visit. A count of zero in that arm is therefore not a parallel measure of whether usual care could achieve the same kind of virtual-visit feasibility. The primary result is most directly read as evidence about successful delivery in the intervention arm, alongside the registry’s formal between-arm contrast.

3. Trial Design

Connect.Parkinson is recorded as a randomized, single-masked, parallel-group study. The registry classifies its phase as not applicable. The study was sponsored by the University of Rochester, had a recorded enrollment of 210, and is listed as completed. The record gives a study start of 2014-03 and primary completion of 2016-08.

The two groups were Usual Care and Educational Materials and Virtual Visits and Educational Materials. The brief summary characterizes the study as a randomized comparative effectiveness study. In the intervention arm, participants were to receive educational materials and specialty care through virtual visits. In the control arm, participants were to receive educational materials and continue usual care in their communities. The arm descriptions mention baseline assessment procedures, but the available descriptions are truncated and do not establish the full sequence or details of visits.

The primary feasibility measure covers one year and counts participants completing at least one virtual visit successfully. The intervention arm’s posted endpoint count is 95 among 97 participants; the control arm’s count is 0 among 98 participants. These are the endpoint counts and denominators recorded for the arms, not a statement about the full enrollment total or the analysis populations for every outcome.

Secondary measures include changes from baseline to one year in EQ-5D Index Value, Montreal Cognition Assessment, and MDS-UPDRS components, as well as care experience and time spent on a Parkinson disease provider visit. Numeric summary results are available for three secondary measures in the record used for this analysis. Additional secondary outcome titles are listed, but numeric results for them are not reported in the registry record used.

The record does not report who was masked, how randomization was generated or concealed, eligibility criteria, or the intended analysis population. It also does not report a detailed protocol or a statistical analysis plan. These omissions limit how completely the design and conduct can be assessed from the registry record alone.

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: 272 individuals were referred to 18 sites to be consented. Two hundred ten individuals with Parkinson disease enrolled in the study, and 195 were eventually randomized.

Overall Study

Reasons for not completing this period:

5. Baseline Characteristics

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

6. Primary Endpoint

The registry lists the primary outcome as Feasibility of Virtual Visits for Parkinson Disease, measured over One year.

The primary endpoint is the number of participants who complete at least one virtual visit successfully during one year. It is a binary participant-level event summarized as a count: a participant either met the stated criterion or did not. The endpoint does not count every visit, rate the quality of a visit, or measure a change in symptoms or quality of life.

Feasibility is a useful early question when a study tests a delivery approach such as specialty care brought into participants’ homes. Before judging clinical or service outcomes, it matters whether the approach can be carried out and whether participants can complete a visit. The registry’s study aims explicitly include demonstrating feasibility among people with Parkinson disease who have limited access to care.

The endpoint definition also sets limits on what its result can establish. Completing at least one visit indicates that the minimum criterion was met; it does not establish that all intended visits were completed, that the care was superior to usual care, or that a health outcome improved. Those are distinct questions. The registry lists separate secondary outcomes concerning quality of life, cognition, disease ratings, care experience, and provider-visit time.

For this endpoint, the recorded event count is 95 among 97 participants assigned to Virtual Visits and Educational Materials and 0 among 98 assigned to Usual Care and Educational Materials. Because the comparator was not assigned virtual visits, its zero count is part of the formal comparison but is not a like-for-like test of virtual-visit delivery in both groups. The registry record does not report a more detailed definition of what qualified a visit as successful.

7. Results Posted in the Registry

• Virtual Visits and Educational Materials: 97.9% (95/97) versus 0.0% (0/98) on control; risk difference 97.9 percentage points (95% CI 95.1 to 100.8); unadjusted two-proportion z-test p = <0.001.

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.

Change in EQ-5D Index Value

• 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 errors.
• Result:
  • Virtual Visits and Educational Materials: mean -0.03 versus -0.04 on control; difference 0.01 (95% CI -0.02 to 0.04), standardized effect d = 0.13; z-test p = 0.373.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Virtual Visits and Educational Materials and the control. The interval gives the range of effects that remain compatible with the data.

Change in Montreal Cognition Assessment

• 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 errors.
• Result:
  • Virtual Visits and Educational Materials: mean 0.65 versus 0.21 on control; difference 0.44 (95% CI -0.25 to 1.13), standardized effect d = 0.19; z-test p = 0.212.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Virtual Visits and Educational Materials and the control. The interval gives the range of effects that remain compatible with the data.

Change in Parkinson Disease Rating Scale (MDS-UPDRS) Part IA

• 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 errors.
• Result:
  • Virtual Visits and Educational Materials: mean -0.85 versus -1.01 on control; difference 0.16 (95% CI -0.69 to 1.00), standardized effect d = 0.05; z-test p = 0.715.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Virtual Visits and Educational Materials and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes broaden the study’s focus beyond whether a virtual visit could be completed. They cover changes in a quality-of-life measure, cognition, and Parkinson disease ratings, alongside listed measures of care experience and visit time. These outcomes address aspects of health and care that the primary feasibility endpoint does not capture. The registry record includes numeric results for three continuous secondary measures; numerical results for the other listed secondary outcomes are not reported in the registry record used.

Across the three numeric summaries, the estimated changes and between-group contrasts are not accompanied by unadjusted p-values below 0.05, and each confidence interval includes zero. Taken together, this is not a consistent pattern of clear between-group separation in these measures. That reading should remain cautious: the outcomes are different constructs, the estimates have uncertainty, and a failure to show a clear difference is not evidence that the groups were equivalent.

These secondary results do not change the meaning of the primary endpoint. A successful virtual visit is a delivery milestone, while quality of life, cognition, symptoms, care experience, and provider-visit time are separate outcomes. A feasibility result cannot stand in for evidence about those outcomes; conversely, uncertain secondary estimates do not undo the recorded completion of visits in the intervention group.

Unadjusted secondary results carry less weight when several outcomes are assessed because each comparison is considered separately, without accounting for the broader set of tests. If many outcomes are examined, chance findings become more likely across the set. The registry record does not report a multiplicity plan, a hierarchy of outcomes, or the analysis plan needed to determine how these results were intended to be interpreted together. The secondary results are therefore best treated as exploratory evidence rather than definitive conclusions about benefit or lack of benefit.

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

• Change in EQ-5D Index Value
• Change in Montreal Cognition Assessment
• Change in Parkinson Disease Rating Scale (MDS-UPDRS) Part IA
• Change in Patient Assessment of Chronic Illness Care
• Minutes Spend on Last Parkinson's Disease Provider Visit
• Change in Parkinson Disease Rating Scale (MDS-UPDRS) Part IB
• Change in Parkinson Disease Rating Scale (MDS-UPDRS) Part 2
• Change in Parkinson Disease Rating Scale (MDS-UPDRS) Part 3

9. Statistical Interpretation

10. Statistical Methods Explained

What does a risk difference express?

For a binary endpoint, the risk in a group is the proportion of participants with the event. A risk difference expresses the first group’s event proportion minus the comparator group’s event proportion. It is an absolute contrast on the probability scale, rather than a relative measure such as a risk ratio. In this study, the event is completing at least one virtual visit successfully, and the registry reports an unadjusted risk difference of 0.979.

The meaning of that contrast depends on the endpoint and the group assignments. Virtual visits were assigned in the intervention group, while the comparator continued usual care with educational materials. Since completing a virtual visit is the event being counted, the contrast captures the intervention-specific delivery outcome as well as the randomized group distinction. It should not be interpreted as a generic comparison of two groups with equal opportunity to complete a virtual visit.

How should the p-value be read?

The posted unadjusted p-value for the primary comparison is <0.001. In general, a p-value measures how incompatible the observed data are with a specified null model, under the assumptions of the analysis. It is not the probability that the null hypothesis is true, nor does it measure the size or practical importance of an effect. Here, the p-value accompanies a very large difference in recorded completion, but the endpoint’s intervention-specific nature remains central to interpretation.

The registry does not report an adjusted model for the primary result. The comparison is therefore described as unadjusted, and the record does not state whether covariates, stratification variables, or other features were included in a sponsor analysis.

What do the secondary mean differences represent?

The continuous secondary outcomes are presented as mean changes from baseline to one year in each group, together with a mean difference between groups. A mean difference is an absolute contrast in the average change as recorded on the outcome’s scale. It retains the outcome’s own units; it is not automatically a measure of clinical importance.

The arm-level spread values for these outcomes are labeled standard errors. A standard error describes the estimated uncertainty in an arm’s sample mean, not the variation among individual participant measurements. The registry’s posted comparison intervals and p-values are unadjusted. No reported sponsor analyses are included for these secondary results, so the record does not specify whether another model or adjustment was used in a separate analysis.

A confidence interval is a range of values compatible with the data under the interval’s statistical procedure and assumptions. For a mean difference, an interval crossing zero indicates that the data are compatible with no difference as well as with differences in either direction at the confidence level used. It does not prove that the true difference is zero. The posted intervals for the three numeric secondary summaries all cross zero.

Why does over-dispersion matter?

For count outcomes, a simple statistical model may assume that the variability is tied closely to the average count. Over-dispersion means the observed variation is greater than that model allows. If it is ignored, standard errors may be too small and confidence intervals too narrow, making statistical evidence appear stronger than it is. This is a general consideration for count data; the primary endpoint is a participant count with a binary event status, not a repeated-event count. The registry record does not report an over-dispersion assessment or a model for repeated visits.

How are the planning illustrations different from trial power?

The continuous-outcome planning illustrations are calculations based on the observed effect, under assumptions stated in the registry record: a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. They show participants per arm for 80% power and participants per arm for 90% power at the stated fractions of the observed effect. Such values are planning illustrations, not observed power and not evidence that the trial was designed to detect those effects. The registry record does not report the trial’s own sample-size justification or design assumptions.

Planning based on an observed effect is especially uncertain when the estimate comes from a study with limited information. The observed effect may differ from the underlying effect, and a planning exercise based on it inherits that uncertainty. Smaller fractions of the observed effect require larger planned group sizes under the assumptions used in the illustrations. Those calculations do not establish that an effect of any particular size is clinically important.

What does the record say about the statistical plan?

The record gives the trial’s randomized, single-masked, parallel design and posts endpoint summaries, but it does not report the randomization procedure, masking details, analysis populations, missing-data method, model specification, or a multiplicity strategy. It also does not report interim analyses or a detailed statistical analysis plan. These are not details that can be inferred from the reported summary statistics. As a result, interpretation should remain tied to the endpoint definitions and numerical summaries actually recorded.

11. Confidence Intervals and Precision

The primary risk-difference estimate is 0.979, with a 95% confidence interval from 0.951 to 1.008. The interval is narrow relative to the magnitude of the estimate, indicating that the posted comparison is numerically precise under the calculation used. It extends slightly above 1, the upper bound for a probability difference. This can occur with some approximate interval procedures; it does not mean a probability difference greater than the maximum possible is literally attainable. The registry record does not name the interval method.

Precision is not the same as relevance or broad generalizability. The endpoint is specific to successful completion of at least one virtual visit, and the intervention group alone was assigned the virtual-visit service. Thus, the narrow interval quantifies uncertainty in the recorded contrast under the stated comparison, but does not resolve the limitation created by the endpoint’s relationship to treatment assignment.

For the continuous secondary outcomes, each posted 95% interval crosses zero. Those intervals communicate uncertainty about the between-group mean differences: values on either side of no difference remain compatible with the data under the interval method. The record does not report a clinically important threshold for these measures, so the intervals cannot be used to decide whether all plausible differences are too small to matter.

The secondary arm summaries report standard errors, and the comparisons include pooled standard deviations and standardized effect summaries. These quantities describe variability and scale in different ways. A standard error concerns uncertainty in an estimated mean; a standard deviation concerns spread in individual observations; a standardized effect expresses a mean contrast relative to a variability measure. None eliminates uncertainty in the comparison.

Intervals and p-values also do not account automatically for multiple outcomes. Since a multiplicity adjustment is not reported in the registry record, the posted secondary intervals and p-values should be read as outcome-by-outcome, unadjusted results rather than as a family-wise assessment. The registry does not report whether the analyses were prespecified or whether all eligible observations contributed to each estimate.

12. Multiplicity, Missing Data and Analysis Populations

The record lists multiple secondary outcomes, but it does not report a method for controlling multiplicity, such as an adjustment to p-values or confidence intervals, a prespecified outcome hierarchy, or a defined strategy for deciding which outcomes were confirmatory. The three numeric secondary comparisons are consequently presented as unadjusted results. Their p-values address each comparison individually and do not account for the possibility of examining several outcomes.

This matters because a set of separate tests has a greater chance of producing at least one apparently notable result by chance than a single prespecified test. The record does not report whether the secondary outcomes were prespecified in a protocol, whether additional outcomes were measured but not posted, or how the posted secondary results were selected. No conclusion about selective reporting can be drawn from the available information alone.

Missing-data handling is also not reported in the registry record. The number of participants with numeric results differs by outcome and arm, but the record does not explain why observations are unavailable, whether they were excluded, or how missing values were addressed. The analysis population for each outcome is not described as intention-to-treat, per-protocol, or another definition. The posted denominators should therefore be read as the numbers attached to the particular summaries, not as evidence of a specified analysis population.

The registry record does not report interim analyses, stopping rules, a detailed statistical analysis plan, or a sample-size justification for the trial. Their absence from the record is not proof that such procedures did not exist; it means their details are not reported in the registry record. Without them, it is not possible to assess how the posted tests relate to planned analyses or whether any changes to the analysis approach occurred.

13. 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.

The registry records zero serious adverse events among 98 participants at risk in the Usual Care and Educational Materials group and zero among 97 participants at risk in the Virtual Visits and Educational Materials group. It also records zero other adverse events among the participants at risk in each group. These are the adverse-event counts reported for the study; they should not be read as proof that no adverse events occurred outside the collection and reporting represented in the registry.

The record does not report the adverse-event collection period, definitions, ascertainment procedures, or details of events other than the posted counts. Death counts are not reported in the registry record. The zero counts therefore describe the safety entries available, but do not establish that the intervention has no risks or permit a detailed comparison of safety profiles. Interpretation should remain limited to what was recorded.

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

The clearest result concerns feasibility as narrowly defined: the registry records 95 participants in the virtual-visit group as having completed at least one virtual visit successfully during one year. That is direct information about whether the study’s minimum visit-completion criterion was met in that group. It does not report how many visits each participant completed beyond that threshold or whether all intended visits took place.

The formal comparison with usual care and educational materials is statistically strong, but the outcome is tied to the assigned intervention. Since the comparator did not receive virtual visits, the control arm’s zero event count is not a test of virtual-visit completion under comparable access. The contrast should therefore not be interpreted as evidence that virtual specialty care is superior on a clinical outcome.

The study description identifies broader aims, including quality of life, quality of care, time, travel, and care-partner burden. The primary endpoint does not measure these outcomes. The numeric secondary results available in the registry do not establish clear differences on the continuous measures reported, and the remaining listed secondary outcomes lack numerical results in the record used. Neither the feasibility result nor the secondary findings support a broad claim about clinical effectiveness.

Absence of a statistically clear secondary difference does not establish no effect. The estimates have uncertainty, the record does not report a clinically important threshold, and the analysis and missing-data procedures are not described. Likewise, a precise estimate for the feasibility contrast cannot answer questions it was not designed to measure.

The findings are thus most informative as a record of whether virtual visits could be completed under the study’s intervention conditions. They should not be used alone to conclude that this delivery model improves symptoms, quality of life, or other outcomes, nor that it is equivalent or inferior to usual care on those outcomes.

15. Why This Trial Matters Statistically

This trial illustrates why statistical interpretation begins with the endpoint definition, not the p-value. The primary measure counts participants who completed at least one virtual visit, while only one group was assigned virtual visits. A large risk difference and small p-value can accurately describe the recorded contrast yet still answer a narrower question than whether telemedicine improves health or care.

It also distinguishes feasibility from effectiveness. Establishing that a service can be delivered is an important question, particularly when a study aims to reach people with limited access to specialty care. But feasibility is not interchangeable with better outcomes. The trial lists separate measures for health status, cognition, disease ratings, care experience, and provider-visit time, and those require their own interpretation.

The record also shows how uncertainty and reporting completeness shape conclusions. The primary interval is precise for the stated contrast, while the available continuous secondary intervals include zero. Yet the absence of reported analysis populations, missing-data procedures, multiplicity methods, and a detailed analysis plan limits a full assessment of the secondary evidence. Statistical significance cannot fill those reporting gaps.

Finally, planning illustrations based on observed effects are not a retrospective measure of whether the trial had adequate power. They are conditional calculations under stated assumptions. Separating those calculations from observed findings helps prevent planning numbers from being presented as trial results or as proof that a non-significant finding is uninformative.

16. 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 primary endpoint’s structure is an important limitation for interpreting the between-arm contrast. Successful virtual-visit completion was relevant to the group assigned virtual visits; the comparator was not assigned that intervention. The result therefore provides a direct feasibility count for the intervention arm, but the randomized comparison does not represent two groups with the same opportunity to complete a virtual visit.

The registry’s available arm descriptions are incomplete, and the record does not give full eligibility criteria, the randomization procedure, concealment details, or who was masked. It also does not report the precise definition of a successful visit beyond the endpoint wording. These omissions prevent a detailed assessment of implementation and risk of bias from the registry record alone.

For secondary outcomes, numerical results are available for only some of the listed measures in this record. The analysis populations, reasons for differing outcome denominators, and missing-data methods are not reported. The record also does not report a multiplicity strategy or detailed statistical analysis plan. As a result, secondary findings should be treated cautiously, particularly when considering them collectively.

The registry does not report a sample-size justification for the trial or an interim-analysis plan. Planning illustrations based on observed effects do not resolve this limitation, because they are not the trial’s original design rationale and are not observed power. The generalizability of the findings is also difficult to assess from the available information because detailed eligibility criteria and participant characteristics are not reported in the record used.

17. Related Tutorials

18. Related Statistical Calculators

19. Sources

• ClinicalTrials.gov: Connect.Parkinson: Connecting Individuals With Parkinson Disease to Specialists in Their Homes — NCT02038959.
• Registry record last updated 2017-01-27; status last verified 2016-12.
• Results first posted 2017-01-27.
• 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.

20. Record Summary

NCT02038959, Connect.Parkinson, is a completed randomized, single-masked, parallel-group comparative effectiveness study sponsored by the University of Rochester. It compared usual care and educational materials with usual care, educational materials, and specialty care delivered through virtual visits into participants’ homes. The registry records enrollment of 210, a study start of 2014-03, and primary completion of 2016-08.

The one-year primary endpoint counted participants who completed at least one virtual visit successfully. The record lists 95 of 97 participants in the virtual-visit group and 0 of 98 in the comparator group, with an unadjusted risk difference of 0.979 and p < 0.001. The endpoint is a feasibility measure, not a direct clinical outcome. Numeric summaries for three continuous secondary outcomes are also recorded; each unadjusted confidence interval includes zero. Detailed analysis, missing-data, and multiplicity methods are not reported in the registry record.