A Novel Transition Program to Reduce Disability After Stroke: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Novel Transition Program to Reduce Disability After Stroke (NCT03485820), a randomized na study. The primary outcome is Reintegration to Normal Living Index (RNLI).

2. Clinical Question

The trial addresses a transition in care: the move home after inpatient stroke rehabilitation. Its registry summary describes a comparison between an environmental-modifications program and stroke education. The formal arm names are COMPASS and Education Program, respectively. The primary outcome focuses on participants’ satisfaction with their participation in home and community life, rather than on a direct count of stroke events or a measure of survival.

The registry lists ischemic and hemorrhagic stroke as the conditions under study. It describes a randomized, parallel-group design and an enrollment of 185. The summary says that participants were divided between the two programs, but the registry record does not report the realized number allocated to each group for the primary analysis. The primary-outcome population is described as those included in the mixed model, with participants missing covariate data excluded.

The clinical question is therefore about the comparative outcome on a disability-related quality-of-life measure at the stated follow-up, in the analyzed population. The registry record does not report group means, a between-group estimate, uncertainty intervals, or a statistical test for that outcome. It is not possible to tell from the registry record whether the observed scores differed, or how precisely any difference was estimated.

3. Trial Design

The registry describes a randomized, single-masked, parallel-group trial. Participants were assigned to COMPASS or an Education Program. The study is marked completed, with enrollment of 185. Its phase is listed as NA. The registry record identifies Washington University School of Medicine as the sponsor.

The study concerns people returning home after inpatient rehabilitation following ischemic or hemorrhagic stroke. The short study summary characterizes the two approaches as stroke education and environmental modifications. The COMPASS arm description calls it a complex intervention combining two evidence-based treatment strategies at the transition from inpatient rehabilitation. It also describes occupational therapy home visits intended to remediate barriers in the home and community. The available arm description is incomplete, so the registry record does not report the full details of the intervention components or their delivery.

The Education Program description states that an occupational therapy practitioner delivers the program in accordance with named evidence-based educational guidelines. It specifies four 75-minute sessions and says participants determine the topic order. The description lists possible topics, but the full text is incomplete in the registry record. Details such as adherence, deviations from the intended programs, and the extent of intervention received by participants are not reported in the registry record.

The primary outcome is the Reintegration to Normal Living Index (RNLI), assessed at baseline and 12 months post-stroke. The registry describes a mixed-model analysis population that excludes participants with missing covariate data. It does not report the number in that analysis population or explain the model specification, including the covariates or the exact comparison estimated.

The record does not report the randomization method, allocation concealment, the identity of the person or people masked, or whether masking was maintained. It also does not report a sample-size justification or a detailed statistical analysis plan. These absences limit what can be concluded about design safeguards and planned analysis choices. The record’s design labels establish the broad structure of the comparison, but do not fill in these operational details.

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

Analysis population for the primary outcome, as defined in the registry: Analyzed are all those included in the mixed model (excludes those with missing covariate data).

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 Reintegration to Normal Living Index (RNLI), measured over Baseline and 12 months post-stroke.

The primary endpoint is the Reintegration to Normal Living Index (RNLI), a disability-related quality-of-life instrument intended to measure satisfaction with home and community participation. The registry states that the instrument has been validated in community-dwelling people with chronic conditions. It uses 11 items, each on a 10 point scale. The item sum is divided by 110 and multiplied by 100 to produce an adjusted score ranging from 0 to 100. Higher scores indicate greater reintegration to normal living.

The endpoint is relevant to a transition-home program because it concerns participation in everyday life in the home and community, the setting the programs aim to address. It is distinct from a measure of impairment alone: a score reflects the participant’s reported reintegration, as defined by the instrument, rather than directly measuring a particular physical function or counting adverse events.

The registry specifies assessment at baseline and 12 months post-stroke, and describes the primary result as a least-squares mean with standard error. A least-squares mean is a model-based mean; its interpretation depends on the model and adjustment choices. Those choices are not reported in the registry record. The record also does not report numerical RNLI estimates, a between-group contrast, a confidence interval, a p-value, or a threshold for a clinically important difference. The scale’s direction is interpretable, but the size and precision of any group comparison cannot be assessed from the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry identifies the primary outcome as a continuous measure and describes its summary as a least-squares mean with a standard error. It also says that a mixed model was used. These terms describe broad features of an analysis, but they do not reveal its full specification. The model formula, covariates, treatment contrast, assumptions, and approach to incomplete outcome data are not reported in the registry record.

What does a least-squares mean represent?

In general, a least-squares mean is a model-based estimate of a group mean. Unlike a simple arithmetic average, it may reflect adjustment for variables included in a statistical model. Its meaning depends on which variables were included, how they were coded, which participants contributed data, and how the model represented time and treatment. The registry record does not identify those details, so the degree or nature of adjustment cannot be described.

The listed endpoint is assessed at baseline and 12 months post-stroke. A model can use observations at more than one time point, but different models can target different quantities. For example, an analysis could focus on group differences at a particular assessment, changes from baseline, or a treatment-by-time comparison. These are general possibilities, not descriptions of the analysis used in this trial. The registry record does not state which contrast was estimated or how the two assessments entered the model.

What does the standard error tell a reader?

In general, a standard error describes the statistical uncertainty around an estimated quantity, reflecting how much the estimate would vary across repeated samples under the model and its assumptions. A smaller standard error indicates greater precision for that estimate; it is not the same as the spread of individual participants’ scores. The registry identifies standard error as the dispersion measure for the primary least-squares mean, but numerical standard errors are not reported in the registry record.

A confidence interval is a range constructed by a statistical procedure to express uncertainty around an estimate. In general, it helps readers assess which effect sizes are reasonably compatible with the data and assumptions. No confidence interval for the primary comparison is reported in the registry record. A p-value, if reported, would address a different question about compatibility with a specified null model; no primary comparison p-value is reported either. Without these quantities and the estimate itself, there is no basis for judging the statistical precision or strength of evidence for a treatment contrast.

Why does the mixed-model description need more detail?

In general, mixed models can accommodate data structures in which observations are related, such as repeated measurements from the same participant, and can represent relationships between outcomes and explanatory variables. The name of a model family alone does not identify the analysis. Details about fixed and random effects, covariance assumptions, time coding, covariate adjustment, and estimation are needed to understand exactly what it estimates. The registry record only says that a mixed model was used; the specific choices are not reported.

Model assumptions also matter. A continuous outcome analysis relies on assumptions concerning the relationship between predictors and the outcome and the behavior of residuals, among other features. In general, departures from assumptions can affect estimates, standard errors, or both. The registry record does not report diagnostic checks or sensitivity analyses, so these cannot be evaluated from the posted information.

How should missing data be considered?

The registry states that the primary analysis included those in the mixed model and excluded participants with missing covariate data. It does not state how many people had missing covariates, why data were missing, or how missing RNLI measurements were handled. These are different issues: an exclusion based on missing covariates does not by itself specify how incomplete outcome measurements were treated. In general, missing-data methods rely on assumptions about why information is absent. Without a description of the methods and the pattern of missingness, the robustness of the result to missing data cannot be assessed.

The registry’s analysis-population description also does not establish whether the model followed an intention-to-treat principle. That label concerns how randomized participants are retained in the analysis and how assignment is handled; it cannot be inferred from the phrase “included in the mixed model.” The number randomized to each group, the number analyzed, and the reasons for exclusions are not reported in the registry record.

What can sample-size planning illustrations establish?

Any sample-size planning illustration associated with the results should be read as a hypothetical calculation, not as an observed trial result or as evidence that the study had a particular probability of finding an effect. The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, a normal approximation, and an effect to detect expressed as a fraction of the observed effect. For continuous outcomes, such illustrations concern participants per arm for a stated power target; time-to-event calculations, in general, concern total events. The registry record does not report a sample-size justification for the original design, and the planning assumptions do not replace one.

Because numerical outcome estimates are not reported in the registry record, there is no observed effect available in this record on which to base a reader’s interpretation of a planning illustration. Such calculations cannot be used to recover the missing treatment estimate, establish observed power, or demonstrate that an analysis found or did not find a difference. They are planning exercises conditional on their assumptions.

The primary outcome is continuous, and its scale direction is clear. But a scale description and a model label do not supply the estimated contrast, uncertainty, or analysis details. Statistical interpretation requires those elements together: a clearly defined estimand, an estimate, an uncertainty measure, and enough information about the model and analysis population to understand their basis. Most of that quantitative reporting is not reported in the registry record.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists the RNLI as the primary outcome and five secondary outcome domains: the Stroke Impact Scale Activities of Daily Living domain, and the In-Home Occupational Performance Evaluation activity, performance, satisfaction, and barrier severity scores. It records six outcome measures and six analyses. The outcome listing does not, by itself, explain how the analyses were organized or whether they were intended as separate comparisons.

Multiplicity arises when several outcomes, time points, or statistical tests are examined. In general, as the number of tests increases, the chance of at least one nominally unusual result can increase if no adjustment is made. Trials may address this through a hierarchy of outcomes, a multiplicity adjustment, or a clearly defined emphasis on the primary endpoint. The statistical plan for multiplicity, including any adjustment or testing hierarchy, is not reported in the registry record. Numerical results for the primary and secondary outcomes are also not reported, so their patterns cannot be assessed.

The primary assessment timeframe includes baseline and 12 months post-stroke. The registry does not state how those measurements were used to define the primary contrast. It also does not report whether there were additional time points or other planned comparisons within the analyses. The registry indicates that the primary outcome has several categories or time points and is not reduced to one comparison; therefore, a single numerical primary comparison should not be assumed.

For missing data, the record says the primary mixed-model analysis excludes people with missing covariate data. It does not report the count excluded, the number with missing outcome values, reasons for missingness, or the method used for incomplete RNLI measurements. It also does not describe sensitivity analyses for missing data. The analysis population is thus partly defined, but its size and the consequences of exclusions cannot be determined. The extent to which missingness could affect the comparison is not reported in the registry record.

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 5% reporting threshold.

The registry reports safety counts for COMPASS and the Education Program. For serious adverse events, it lists 0 affected among 85 at risk in COMPASS and 0 affected among 100 at risk in the Education Program. For other adverse effects, it lists 0 affected among 85 at risk in COMPASS and 0 affected among 100 at risk in the Education Program. Deaths are listed as 0 in each group.

These are event counts with the stated at-risk denominators, not measures of the RNLI outcome. A zero count means that no events of the listed type are recorded for those denominators in the registry; it does not establish that the programs have no risk. The registry record does not report how adverse events were defined, how actively they were sought, the duration or completeness of safety follow-up, or whether any events were assessed as related to either program.

The safety denominators also should not be treated as the primary-outcome analysis population. The record separately defines the primary analysis population as participants included in the mixed model, excluding those with missing covariate data, and does not report its size. The safety figures are descriptive; without event definitions and follow-up details, their scope is limited.

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

The registry record describes a randomized comparison of COMPASS and an Education Program among people transitioning home after inpatient rehabilitation for stroke. It identifies the RNLI as the primary outcome, defines the scale direction, and gives the baseline and 12-month post-stroke assessment timeframe. These registry entries establish what the trial set out to examine and how the primary measure is intended to be interpreted.

The record does not report the numerical RNLI results. It therefore does not establish whether the groups had different scores, how large any difference may have been, or how precisely a difference was estimated. No primary effect estimate, confidence interval, standard error value, or p-value is reported. The absence of a reported estimate is not evidence that the groups were equal, and it is not evidence that one program was better. It means the registry record does not contain the numerical result needed to make that comparison.

The RNLI measures reported satisfaction with home and community participation, with higher adjusted scores indicating greater reintegration to normal living. It does not directly answer every question about recovery after stroke. The registry-listed secondary measures cover activities of daily living and several dimensions of in-home occupational performance, but their numerical results are not reported either. These measures would address related but distinct aspects of functioning and participation.

Even if a group difference were reported, its practical meaning would depend on the size and precision of the estimate, the scale’s interpretation in this population, and whether the difference is important to participants. The registry record does not state a threshold for a clinically important RNLI difference. It also does not report enough detail about the analysis model, missing data, or the analysis population’s size to independently assess those aspects of the result.

The safety counts report no listed serious or other adverse effects and no deaths in the recorded groups. These counts should be read within the reported denominators and not as proof of absence of risk. In short, the record supports a description of the trial design and outcome framework, but not a conclusion about comparative RNLI results or the overall balance of outcomes.

13. Why This Trial Matters Statistically

Statistically, this trial is notable for asking a question that spans a transition in care and an outcome focused on participation in everyday life. The RNLI is not simply a count of events: it is a scaled, disability-related quality-of-life measure reflecting satisfaction with home and community participation. Its interpretation depends on both the score direction and the context of the population being studied.

The design is described as randomized and parallel-group, which identifies a direct comparative framework. Randomization can support a fair comparison when allocation and follow-up are conducted appropriately, but the registry record does not report the allocation procedure, concealment details, or complete follow-up information. The design label alone cannot establish how well those safeguards were implemented.

The primary endpoint is assessed at baseline and 12 months post-stroke, and the analysis is described as a mixed model producing least-squares means with standard errors. This makes the distinction between a raw average and a model-based estimate important. Readers need the model specification, the contrast estimated, and the uncertainty around that estimate to interpret the analysis. Those numerical and methodological details are not reported in the registry record.

The record also illustrates why analysis populations and missingness matter. Overall enrollment is 185, whereas the primary analysis is defined by inclusion in a mixed model and excludes people with missing covariate data. Without the analyzed count and information about outcome missingness, readers cannot see how closely the analysis population corresponds to the randomized population.

Finally, the record lists multiple secondary outcomes as well as the primary endpoint. Clear reporting of the testing hierarchy and any multiplicity strategy would help readers distinguish the primary inference from additional exploratory comparisons, but these details are not reported. The design and outcome descriptions make the statistical question understandable; the absence of numerical results and key analysis details prevents assessment of the answer.

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 of the registry record is that it does not report numerical results for the primary endpoint or the listed secondary outcomes. Without group estimates and a comparative effect estimate, the direction, magnitude, and precision of the outcome findings cannot be evaluated. The record’s completed status does not itself supply those results.

The primary analysis population is described as participants included in the mixed model, excluding those with missing covariate data, but its size is not reported. The registry also does not report how many RNLI observations were missing, when they were missing, why they were missing, or how incomplete outcome data were handled. These omissions make it difficult to assess the relationship between enrollment and analysis, or the possible influence of missing data.

The statistical description is limited. The record identifies a mixed model, least-squares means, and standard errors, but does not report the model specification, the precise treatment contrast, covariates, assumptions, or diagnostic and sensitivity analyses. It also does not report the multiplicity strategy across the primary and secondary measures. These details are important for understanding the analysis and the scope of any statistical claims.

Intervention information is incomplete. The COMPASS arm is described in broad terms, and the Education Program description gives session length and topic-order information, but the full arm descriptions are not reported in the registry record. Details of adherence, exposure, implementation, and fidelity are also not reported. The registry does not state who was masked despite identifying the study as single-masked.

The safety record lists zero events in the stated groups, but event definitions, collection procedures, and follow-up completeness are not reported. These specific limitations restrict interpretation of both benefit-related outcomes and safety counts. They should not be replaced with assumptions about the protocol or analysis.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: A Novel Transition Program to Reduce Disability After Stroke — NCT03485820.
• Registry record last updated 2026-11-04; status last verified 2026-10.
• Results first posted 2025-03-05.
• 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 NCT03485820, “A Novel Transition Program to Reduce Disability After Stroke,” as a completed, randomized, single-masked, parallel-group study sponsored by Washington University School of Medicine. Enrollment is listed as 185. The population concerns ischemic and hemorrhagic stroke, with participants transitioning home after inpatient rehabilitation.

The comparison is COMPASS versus an Education Program. The primary outcome is the Reintegration to Normal Living Index, assessed at baseline and 12 months post-stroke. The scale is adjusted to range from 0 to 100, with higher scores indicating greater reintegration to normal living. The primary analysis is described as a mixed model and excludes participants with missing covariate data.

The registry record does not report numerical primary or secondary outcome results, the primary comparison estimate, or its uncertainty. It lists zero serious adverse events, zero other adverse effects, and zero deaths in both groups. Details of the statistical plan, analysis population size, missing outcome data, and full intervention delivery are not reported in the registry record.