Care of Persons With Dementia in Their Environments (COPE) in Programs of All-Inclusive Care of the Elderly (PACE): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Care of Persons With Dementia in Their Environments (COPE) in Programs of All-Inclusive Care of the Elderly (PACE) (NCT04165213), a randomized na study. The primary outcome is Determinants of Implementation Behavior Questionnaire (Clinicians).

2. Clinical Question

3. Trial Design

This was a randomized, parallel-group study of COPE training in PACE organizations. The registry classifies masking as triple, but does not report further details about who was masked or how masking was maintained. The listed phase is not applicable. The sponsor is the University of Pennsylvania, and the record lists the study as completed. Its recorded start date is 2021-06-28 and its primary completion date is 2024-05-31.

The protocol description separates the project into three phases. The first was to develop online COPE training and an automated approach to fidelity monitoring. The second was to randomize ten PACE organizations: five to traditional instructor-led training and five to online training. This phase was intended to assess staff competency and fidelity to COPE principles and protocols. The third was to evaluate participant outcomes by clinician training type. The protocol describes five family dyads at each organization, with dyads followed for 4 months. These are design intentions; the registry’s overall enrollment is 85.

The study includes clinician, caregiver, and person-with-dementia groups associated with the two training approaches. The primary outcome is a clinician measure, while listed secondary outcomes cover neuropsychiatric symptoms, caregiver function and upset, caregiver burden and competence, quality of life, and social networks. The record lists nine outcome measures but reports no analysis estimates. The protocol’s description of a non-inferiority analysis is incomplete in the registry record; the non-inferiority margin and analysis details are not reported in the registry record.

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: Prior to assignment, the study team worked with Trinity Health PACE to ensure at least 10 PACE sites would enroll in the study.

Clinician Enrollment

Reasons for not completing this period:

Caregiver Enrollment

Reasons for not completing this period:

PLWD Enrollment

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: These outcome measures were collected from clinicians in the control and intervention study arms at post-training baseline (just after training) and follow-up (after implementing COPE to dyads). The means reported are change scores between those timepoints.

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.

Only the first 6 of 7 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Determinants of Implementation Behavior Questionnaire (Clinicians), measured over Post-training baseline, 14 months.

The primary endpoint is the clinician score on the Determinants of Implementation Behavior Questionnaire (DIBQ). The instrument contains 93 items assessing 18 domains. For this study, the reported focus was on the knowledge and skills domains, which the record says have demonstrated good discriminant validity and are highly correlated in recent studies. The record describes each domain’s total score as ranging from 0 to 10, with higher scores indicating more positive knowledge and skills in COPE.

The endpoint is reported as a mean change score between the post-training baseline assessment and follow-up after COPE implementation. The registry’s timeframe label is “Post-training baseline, 14 months.” This makes the outcome relevant to implementation: it concerns clinicians’ reported knowledge and skills after training and subsequent use of COPE, rather than a direct measure of participant health outcomes.

A change score can describe how a measured score differs between assessments, but its interpretation depends on the actual estimates, their variability, and the comparison between training groups. The registry record does not report the group means, standard deviations, confidence intervals, or a test of the primary outcome. It therefore does not establish the direction, magnitude, or precision of any difference between training approaches.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Statistical methods are best understood in relation to the design and the type of outcome. General explanations below describe what common methods mean; they do not assert that the registry record used those methods. The actual statistical analysis plan and details of the completed analyses are not reported in the registry record.

What does a mean change score represent?

The primary endpoint is described as a mean change score between the post-training baseline assessment and follow-up. In general, a change score summarizes within-person differences across measurement occasions, and a group mean summarizes those differences across people. It can be useful for describing whether scores tended to move upward or downward. But interpretation also depends on the spread of individual scores and on how change compares across randomized groups. The registry record does not report the primary group estimates or their variability, so the observed pattern cannot be reconstructed from the record.

The DIBQ outcome focuses on knowledge and skills, with higher scores described as positive. This direction helps explain what an increase would mean on the measure, but it does not tell whether an increase occurred. Nor does the stated score range establish what size of change would matter in practice. A meaningful-change threshold is not reported in the registry record.

Why does organization-level randomization matter?

In this study, PACE organizations were assigned to training approaches. That is cluster randomization: the assignment unit is an organization, while outcomes are collected from clinicians and from caregivers and people with dementia associated with the organizations. In general, observations within one organization can be more alike than observations from different organizations because they share staff, routines, and setting. Such within-organization similarity is called clustering.

When clustering is present, an analysis that treats every person as fully independent may understate uncertainty. General approaches include models that represent organization-level variation or methods that adjust standard errors for clustering. The appropriate approach depends on the design, outcome, and data. The registry record does not report the analysis model, whether clustering was accounted for, or how the number of organizations informed the uncertainty estimates.

What would a confidence interval add?

In general, a confidence interval describes the uncertainty around an estimated effect under a specified statistical procedure. A narrow interval indicates greater precision than a wide interval, while an interval spanning values consistent with different interpretations may leave the result uncertain. It should be considered alongside the estimated effect and the design, not as a substitute for them. No confidence interval for the primary outcome is reported in the registry record.

A p-value, when reported, describes how incompatible data are with a specified null model under the assumptions of the analysis. It does not measure the size or importance of an effect, and it does not establish equivalence. No primary-outcome p-value is reported in the registry record. Thus, neither statistical significance nor the strength of evidence for a difference between training approaches can be assessed from the posted results.

How does a non-inferiority question differ?

General non-inferiority analysis asks whether an intervention is not worse than a comparator by more than a prespecified margin. The margin defines the largest difference that would still be considered acceptably small for the particular question. A conclusion depends on the estimated between-group effect and its uncertainty relative to that margin. Merely describing an intended non-inferiority analysis does not establish non-inferiority. The protocol mentions this approach for dyad outcomes, but the margin, analysis details, and results are not reported in the registry record.

What are planning illustrations?

Sample-size illustrations based on an observed effect are planning exercises, not results about the trial’s power. In general, such illustrations explore how many participants per arm might be needed under specified assumptions and for a chosen target effect. They depend on assumptions such as the significance level, allocation, and statistical approximation. They do not establish that a completed study had a particular probability of detecting an effect, and they cannot repair missing outcome estimates or an unspecified analysis plan. The registry record does not report its own sample-size justification.

For a clustered design, planning also depends on organization-level allocation and within-organization similarity. A calculation that omits those features may not reflect the uncertainty of a cluster-randomized study. No cluster-adjusted planning assumptions are reported in the registry record. Any planning illustration should therefore be read as conditional on its stated assumptions rather than as a reconstruction of the trial’s original design rationale.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists nine outcome measures, including the clinician DIBQ primary outcome and secondary measures for people with dementia and caregivers. When a study assesses multiple outcomes, domains, groups, or follow-up times, the chance of at least one apparently notable result can increase if many tests are considered without a prespecified strategy. General approaches include identifying a primary comparison in advance, adjusting for multiple testing, or treating secondary findings as exploratory. The registry record does not report a multiplicity strategy, a hierarchy among secondary outcomes, or which analyses were prespecified.

The primary outcome description identifies the knowledge and skills domains and a follow-up assessment, but the registry record does not provide numerical results or analysis estimates. It also does not report how the two domains were handled in analysis, whether separate comparisons were made, or how any multiplicity was addressed. The protocol describes additional participant outcomes, but their results and analysis methods are not reported in the registry record.

Missing-data handling is also not reported in the registry record. The record does not describe the extent or pattern of missing outcome observations, methods for handling missing scores, or whether any sensitivity analyses were conducted. In general, missingness can affect estimates when the availability of an outcome is related to participant characteristics or outcomes. The size and direction of any effect cannot be determined without information about the missing data and the analysis assumptions.

The analysis populations are not described in sufficient detail to determine whether results were intended to include everyone as randomized, only participants with observed outcomes, or another defined group. The record gives enrollment and group descriptions but does not state analysis-set definitions. It also does not report exclusions, adherence-based analyses, or how clinician and dyad observations were linked to organization assignment. These omissions limit assessment of the relationship between the randomization and any reported estimate.

11. Safety Results

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

Deaths

Other (non-serious) adverse events

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

The adverse-event entries list zero serious adverse events affecting participants in each of the six reported groups. They also list zero other adverse events affecting participants in each group. The registry gives separate numbers at risk for clinicians, caregivers, and people with dementia; these group-specific counts should be read as the denominators for their respective entries, not combined into a new total.

The record separately reports 2 deaths in the group of people with dementia receiving COPE from instructor-trained clinicians and 2 deaths in the group receiving COPE from module-trained clinicians. It lists no deaths in the clinician or caregiver groups. A death entry is distinct from the serious-adverse-event-affected count, which is zero in both person-with-dementia groups. The record does not report causes, timing, relatedness to the intervention, or adjudication details. It is therefore not possible to attribute these deaths to COPE or to compare risk from these entries alone.

These are registry counts, not a formal comparative safety analysis. The record does not describe how adverse events were collected, the observation periods for these counts, or whether all events were included. Zero affected counts in the posted entries do not establish that no events occurred outside the recorded categories or reporting process.

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

The record describes a randomized comparison of online COPE module training with instructor-led COPE training in PACE organizations. It identifies clinician implementation behavior as the primary outcome and lists outcomes concerning caregivers and people with dementia. These details establish the study question and the intended measurement framework, but they do not provide numerical evidence about the comparative result.

The registry does not report the primary-outcome means, standard deviations, confidence intervals, or analysis estimates. It therefore does not show whether DIBQ knowledge or skills changed after implementation, whether changes differed by training type, or how precise an estimated difference might be. The protocol’s stated interest in whether online training is the same or better than face-to-face training is an aim, not a finding. Similarly, the mention of non-inferiority does not establish that non-inferiority was demonstrated.

The record also does not report numerical results for its listed secondary outcomes. No conclusion about participant symptoms, caregiver burden, caregiver competence, quality of life, or social networks can be drawn from their listing alone. The reported safety entries show no adverse events affecting participants in the listed categories, and deaths are separately recorded for both person-with-dementia groups. The registry does not provide enough detail to interpret these entries causally.

In general, a randomized design can support a comparative estimate when the analysis respects the assignment structure and adequately handles missing outcomes. Here, important analysis details, including clustering methods and missing-data handling, are not reported in the registry record. This does not prove that the study’s analyses were inadequate; it means those features cannot be evaluated from this record.

The appropriate reading is therefore descriptive and limited: the registry documents a completed randomized implementation study and its planned measures, but does not post the estimates required to judge comparative effects. Absence of numerical results in the registry should not be mistaken for evidence of no effect, just as a stated study objective should not be mistaken for evidence of benefit.

13. Why This Trial Matters Statistically

This study is statistically notable because training was assigned at the PACE-organization level while outcomes were collected from clinicians, caregivers, and people with dementia. That design links an implementation question to multiple levels of observation. It also makes the organization-level allocation central to analysis: observations within a site may not be independent, and appropriate uncertainty estimates generally need to reflect the clustering.

The study also distinguishes implementation outcomes from participant outcomes. The primary measure concerns clinicians’ knowledge and skills after training and implementation, while the other listed measures address caregivers and people with dementia. These outcome types answer different questions. A change in clinician-reported implementation behavior is not itself a direct estimate of a participant outcome, and participant measures should be interpreted on their own scales and in their own populations.

The planned comparison is relevant to an implementation setting because it contrasts online training with instructor-led training rather than comparing COPE with no training. The design description also includes development of online training and a fidelity-monitoring approach. Evaluating training delivery, clinician implementation, and participant outcomes across phases can raise distinct statistical questions; the record does not report the analyses used to address them.

Finally, this record illustrates why a clear outcome definition is not the same as a reported result. It provides detail about the DIBQ domains, score direction, and assessment timing, but no numerical estimates or uncertainty measures. Without those results, readers can understand what the study set out to measure but cannot assess the magnitude or precision of the comparison. That distinction is essential when interpreting registry records.

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.

A central limitation for interpreting this trial record is the absence of numerical primary-outcome results and analysis estimates. Although the primary measure, score direction, and assessment timing are described, the record does not provide group means, standard deviations, confidence intervals, or a comparative test. The clinical magnitude and precision of the primary comparison therefore cannot be evaluated.

The registry describes randomization of PACE organizations but does not report how clustering was handled in the analysis. This is important because the organization is the allocation unit while outcomes are collected from people within organizations. The analysis model, variance estimation method, and treatment of the limited number of participating organizations are not reported in the registry record.

The record does not report the non-inferiority margin or the analysis details needed to interpret the protocol’s non-inferiority aim. It also does not provide a sample-size justification, an account of interim analyses, or a detailed statistical analysis plan. These items are not reported in the registry record and should not be inferred from the design description.

Missing-data methods, analysis-population definitions, and outcome-specific follow-up completeness are likewise not reported. The record does not explain how missing clinician, caregiver, or participant scores were handled, or whether any sensitivity analyses were performed. In addition, the adverse-event entries do not describe event collection methods, timing, or attribution. These limitations concern what can be assessed from the registry record; they do not establish that the underlying study lacked such procedures.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Care of Persons With Dementia in Their Environments (COPE) in Programs of All-Inclusive Care of the Elderly (PACE) — NCT04165213.
• Registry record last updated 2025-06-26; status last verified 2025-06.
• Results first posted 2025-06-26.
• 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 record describes NCT04165213, a completed randomized, parallel-group study of COPE training in PACE settings, sponsored by the University of Pennsylvania. It compares online module training for clinicians with instructor-led COPE training, with PACE organizations as the randomization unit. The record lists enrollment of 85 and describes planned clinician, caregiver, and person-with-dementia assessments.

The primary outcome is the clinician DIBQ knowledge and skills change score between post-training baseline and follow-up after implementation. Secondary outcomes cover neuropsychiatric symptoms, caregiver function and upset, burden, competence, quality of life, and social networks. The registry does not report numerical outcome estimates or statistical comparisons. Safety entries list zero serious and other adverse events affecting participants in all reported groups and separately record 2 deaths in each person-with-dementia group. Attribution and timing are not reported in the registry record.