Palliative Care for Persons With Late-stage Alzheimer's and Related Dementias and Their Caregivers: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Palliative Care for Persons With Late-stage Alzheimer's and Related Dementias and Their Caregivers (NCT04948866), a randomized na study. The primary outcome is Incidence of Hospital Transfers (Patients).

2. Clinical Question

3. Trial Design

The registry describes a randomized, double-masked, parallel-group trial of the ADRD-PC Program compared with a control condition. The sponsor was the University of North Carolina, Chapel Hill, and the registry lists enrollment of 884. The trial status is completed. Its phase is listed as not applicable.

The study concerned patient-caregiver dyads affected by late-stage ADRD. The intervention was dementia-specific palliative care delivered by hospital-based specialty interdisciplinary palliative care teams, caregiver education, and transitional care. Control participants received late-stage ADRD caregiver educational materials from the Alzheimer’s Association, while patients received usual hospital and post-acute care. The registry describes the program as triggered during hospitalization.

The primary outcome was the rate of hospital transfers during the 60 days after discharge from the index hospitalization. The analysis population recorded for this outcome was 221 patients in each group. The registry also lists secondary outcomes concerning symptom treatment and control, access to hospice and community-based palliative care, transition to nursing-home-level care, and documented discussion of dementia prognosis. Results for some, but not all, of those outcomes are posted in the record.

The registry record does not report the original sample-size justification, the detailed masking procedures, or the trial’s original statistical analysis plan. The recorded enrollment and the outcome-specific group counts describe different aspects of the trial; the registry record does not explain the relationship between them. Details about the number of study sites are not reported in the registry record’s structured trial 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

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 Incidence of Hospital Transfers (Patients), measured over 60 days post index hospital discharge.

The primary endpoint is the incidence rate of hospital transfers in the 60 days after discharge from the index hospitalization. It is defined as the number of emergency room visits plus hospital admissions divided by person-days of follow-up. The unit is events per person day at risk. This combines two types of hospital-based transfer events and relates their count to the time participants were observed and at risk.

Rates are useful when participants contribute different amounts of follow-up time: the denominator represents exposure time rather than treating every participant as if observed for an identical duration. In general, an event rate describes how often events occur relative to time at risk; it is not the same as the proportion of participants who experienced at least one event. The posted endpoint definition does not state that each participant could contribute only one transfer.

The registry gives group rates of 0.99 for the ADRD-PC Program and 1.12 for control. The comparison is expressed as a rate ratio, which relates the intervention-group rate to the control-group rate. A ratio below 1 means the posted intervention-group rate is numerically lower. That numerical direction alone does not settle how to interpret the result, particularly because the registry’s own analysis reports a different inferential result. The choice of hospital transfers as the primary outcome reflects the trial’s stated aim of reducing burdensome transfers, but the registry record does not report further rationale for selecting this endpoint.

7. Results Posted in the Registry

• Intervention Condition: ADRD-PC Program: 0.99 versus 1.12 (events per person day at risk) on control; rate ratio 0.88 (95% CI 0.86–0.90, approximate, Poisson model); 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.

Symptom Treatment (Patients)

• 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 deviations.
• Result:
  • Intervention Condition: ADRD-PC Program: mean 8.06 versus 6.49 on control; difference 1.57 (95% CI 1.16 to 1.98), standardized effect d = 0.71; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Intervention Condition: ADRD-PC Program than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Symptom Control for Physical Symptoms (Patients)

• 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 deviations.
• Result:
  • Intervention Condition: ADRD-PC Program: mean 21.00 versus 21.80 on control; difference -0.80 (95% CI -2.62 to 1.02), standardized effect d = -0.09; z-test p = 0.388.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Intervention Condition: ADRD-PC Program and the control. The interval gives the range of effects that remain compatible with the data.

Symptom Control for Neuropsychiatric Symptoms-Severity (Patients)

• 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 deviations.
• Result:
  • Intervention Condition: ADRD-PC Program: mean 21.60 versus 21.80 on control; difference -0.20 (95% CI -1.36 to 0.96), standardized effect d = -0.03; z-test p = 0.735.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Intervention Condition: ADRD-PC Program and the control. The interval gives the range of effects that remain compatible with the data.

The posted secondary outcomes cover patient symptom treatment and control, access to hospice and community-based palliative care, nursing-home-level transitions, and documentation of dementia prognosis. The record lists 13 secondary outcomes, while the outcome and analysis counts each include 14 measures or analyses. Results are posted for only some of the listed secondary outcomes, so the pattern cannot be treated as a complete account of every planned secondary result.

Across the three posted symptom-score comparisons, the unadjusted recalculation shows a difference in the symptom-treatment score, while the physical-symptom-control and neuropsychiatric-symptom-severity estimates are close to zero and their intervals include zero. This overall pattern does not establish that changes in symptom scores explain the primary hospital-transfer result. These are distinct outcomes, and the registry does not report a mediation analysis or other analysis linking them.

Secondary outcomes can add context about the intervention’s possible effects across different aspects of care, but their statistical evidence requires caution. The registry record does not report a multiplicity adjustment or a prespecified hierarchy for interpreting these outcomes. When many outcomes are tested, the chance of at least one apparently notable result can increase even if none reflects a true effect. This general issue is one reason unadjusted secondary results carry less weight than a clearly prespecified primary analysis.

The primary result itself has an important interpretive complication: the recalculated test and sponsor-reported test disagree. Secondary results do not resolve that disagreement. Their role is descriptive and supportive rather than a substitute for a coherent primary analysis, particularly when the analysis plan and handling of missing measurements are not reported. The remaining secondary outcome results, including their estimates and uncertainty, are not reported in the registry record.

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

• Symptom Treatment (Patients)
• Symptom Control for Physical Symptoms (Patients)
• Symptom Control for Neuropsychiatric Symptoms-Severity (Patients)
• Symptom Control for Neuropsychiatric Symptoms-Distress (Patients)
• Access to Hospice (Patients)
• Access to Community-based Palliative Care (Patients)
• Transition to Nursing Home Level of Care (Patients)
• Documented Discussion of Dementia Prognosis (Patients)

9. Statistical Interpretation

10. Statistical Methods Explained

The primary endpoint is a rate: the number of hospital transfers divided by person-days at risk. A rate analysis is suited to event counts observed over exposure time. General statistical explanation: Poisson regression is commonly used for count outcomes, and a model may account for differing exposure time when estimating rates. The registry reports that the sponsor used Poisson regression, but the model details, exposure-time specification, covariates, and variance assumptions are not reported in the registry record.

The recalculation uses the posted group rates and event and exposure summaries to express the comparison as a rate ratio. A rate ratio compares one group’s event rate with the other group’s rate. Values below 1 indicate that the numerator group’s rate is lower numerically; values above 1 indicate a higher rate. This is an unadjusted comparison and should not be mistaken for the sponsor’s reported model result. The registry’s reported primary test is a Poisson regression superiority analysis with p = 0.522, whereas the recalculated test gives p < 0.001. The record does not include enough analysis detail to explain the difference.

General statistical explanation: count data can vary more than a simple Poisson model expects, a feature called over-dispersion. If present and unaddressed, it can affect uncertainty estimates and hypothesis tests. The registry record does not state whether over-dispersion was assessed or how it was handled. No conclusion about that issue can be drawn from the posted summaries alone.

The posted secondary outcomes with continuous scores are compared using mean differences. A mean difference is the intervention-group mean minus the control-group mean, so a positive or negative value indicates direction relative to that ordering. The recalculations use the posted means, standard deviations, and group sizes. The registry lists two-sided t-tests for these outcomes. A t-test evaluates whether the data are compatible with equal group means under its assumptions. The record does not report further details about the assumptions assessed, any covariate adjustment, or how incomplete measurements were handled.

For symptom treatment, the recalculated mean difference is 1.57, with a confidence interval from 1.159 to 1.981 and p < 0.001. For physical symptom control, the recalculated mean difference is -0.8, with a confidence interval from -2.617 to 1.017 and p = 0.388. For neuropsychiatric symptom severity, it is -0.2, with a confidence interval from -1.36 to 0.960 and p = 0.735. The sponsor’s posted two-sided t-test p-values for the latter two outcomes are 0.402 and 0.803, respectively. These differences between posted tests and recalculations are modest in inferential direction, but the record does not report enough detail to reconcile each pair.

General statistical explanation: a confidence interval describes uncertainty around an estimated quantity under the method used. It is not a range that contains the true value with certainty, and its interpretation depends on the method and confidence level. The registry does not state a confidence level for the intervals used in the recalculations. A p-value and a confidence interval are related ways to express uncertainty, but neither provides a measure of clinical importance by itself.

The posted analyses are described as superiority tests. A superiority test evaluates whether the data support a difference in the specified direction or between groups, depending on the test definition. It is not evidence that every secondary outcome was tested under a single prespecified hierarchy. The registry record does not report a multiplicity plan, the prespecified ordering of outcomes, or a statistical analysis plan. It also does not report interim analyses, adjustment for baseline characteristics, or the exact population included in each sponsor analysis.

All recalculated comparisons are unadjusted comparisons based on posted summary statistics. They do not reproduce an analysis that may use participant-level data, covariate adjustment, or other modeling choices. General statistical explanation: unadjusted and adjusted estimates can differ because they answer slightly different statistical questions or use different assumptions. In this trial, the exact reason for the contrast in the primary results is not reported in the registry record, so it should not be inferred from the numbers alone.

11. Confidence Intervals and Precision

The primary rate-ratio recalculation is 0.884, with a confidence interval from 0.863 to 0.905. The interval lies below 1, the value corresponding to equal rates in the two groups. Under the recalculation’s assumptions, that interval indicates that the estimated relative rate is fairly tightly bounded around the posted estimate. The registry record does not state the confidence level for this interval, so a particular coverage percentage should not be attached to it.

The group rates themselves are 0.99 and 1.12 events per person day at risk. These are rounded registry values; the recalculated rate ratio uses the posted event and exposure summaries. The apparent precision of the interval should not be confused with certainty about the model or with agreement between analyses. In particular, the sponsor’s Poisson regression gives p = 0.522, unlike the recalculated p < 0.001. Statistical precision within one calculation does not resolve that disagreement.

For the secondary outcomes, the symptom-treatment mean-difference interval runs from 1.159 to 1.981. The physical-symptom-control interval runs from -2.617 to 1.017, and the neuropsychiatric-symptom-severity interval runs from -1.36 to 0.960. The latter two include zero, meaning that their recalculated estimates are compatible with no mean difference under the interval method. The symptom-treatment interval does not include zero, but its interpretation remains subject to the unadjusted, multiple-outcome context.

General statistical explanation: narrower intervals indicate greater precision under a given method, while wider intervals indicate more uncertainty about the estimate. Precision is influenced by the amount of information and the variability of the outcome, among other factors. An interval does not account for unreported biases, missing data, or analysis choices. The outcome-specific group sizes also differ for the secondary symptom measures, and the record does not explain the reasons for those differences.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning figures are participants per arm, not total events. The assumptions used for these illustrations are a two-sided significance level of 0.05, equal allocation between the two arms, a normal approximation, and an effect to detect equal to the stated fraction of the observed effect. The trial’s own original design assumptions and sample-size justification are not reported in the registry record.

The 80% power column gives the planned number of participants per arm that the illustration estimates would be needed to detect the selected effect with 80% power under those assumptions. The 90% power column gives the corresponding number per arm for 90% power. In plain terms, power is the chance that a study using the stated assumptions will detect an effect of the specified size if that effect is present. It is not a guarantee that an effect will be detected.

The rows labeled 100%, 75%, and 50% of the observed effect apply progressively smaller target effects. Smaller effects need more participants to detect under otherwise identical assumptions. For the primary rate comparison, the listed figures are 17 and 22 participants per arm for 100% of the observed effect, 29 and 39 for 75%, and 64 and 85 for 50%, for 80% and 90% power respectively. These are planning illustrations calculated from the observed effect, not the trial’s original design calculations.

These figures are not observed power. Observed power would not describe the uncertainty in the estimate as well as the effect estimate and its interval do. The planning values should be read as assumption-dependent illustrations, not evidence that the actual trial was designed with those targets or that its achieved results have a particular probability of success.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, λ₀ and λ₁ are event rates per patient-year, and T is follow-up per participant in years. This assumes Poisson counts and equal follow-up.

Worked calculation for this trial

1. Control rate λ₀ = 409.0800; treatment rate λ₁ = 361.5975 events per patient-year; follow-up T = 0.16 years.
2. n = (1.96 + 0.842)² × (1/409.0800 + 1/361.5975) / (0.16 × [ln(0.884)]²) = 16.4, rounded up to 17 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
lambda0 <- 409.08                      # control rate, events per patient-year
lambda1 <- 361.5975                      # treatment rate
tt      <- 0.164271047227926                       # follow-up per participant, years
n_per_arm <- function(power, frac = 1) {
  l1 <- lambda0 * (lambda1 / lambda0)^frac   # smaller effect = fraction of the observed log rate ratio
  ceiling(round((z_a + qnorm(power))^2 * (1 / lambda0 + 1 / l1) / (tt * log(l1 / lambda0)^2), 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Simulation 1: Poisson counts. Simulation 2: over-dispersed counts (variance = 1.5 x mean), analysed with the matching variance.
set.seed(4948866)
n    <- n_per_arm(0.80)
reps <- 20000
m0 <- lambda0 * n * tt; m1 <- lambda1 * n * tt
x0 <- rpois(reps, m0); x1 <- rpois(reps, m1)
z  <- log(x1 / x0) / sqrt(1 / x0 + 1 / x1)
cat(sprintf("Simulated power, Poisson counts, n = %s per arm: %.3f\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE)))
phi <- 1.5
y0 <- rnbinom(reps, mu = m0, size = m0 / (phi - 1)); y1 <- rnbinom(reps, mu = m1, size = m1 / (phi - 1))
z2 <- log(y1 / y0) / sqrt(phi * (1 / y0 + 1 / y1))
cat(sprintf("Simulated power, over-dispersed counts (phi = 1.5), same n: %.3f\n", mean(abs(z2) > z_a, na.rm = TRUE)))

Output from running the code

100% of observed effect: n per arm = 17 (80% power), 22 (90% power)
 75% of observed effect: n per arm = 29 (80% power), 39 (90% power)
 50% of observed effect: n per arm = 64 (80% power), 85 (90% power)
Simulated power, Poisson counts, n = 17 per arm: 0.819
Simulated power, over-dispersed counts (phi = 1.5), same n: 0.643

The first simulation draws Poisson counts and should land close to the target power. The second allows extra variability (variance 1.5 times the mean, a purely illustrative value), which lowers the power at the same sample size, so rate trials usually need more participants than the Poisson formula suggests.

14. Analyses Reported by the Sponsor

The registry’s own primary analysis is reported as a Poisson regression superiority test with p = 0.522. The unadjusted recalculation from the posted event and exposure summaries gives a rate ratio of 0.884, a confidence interval from 0.863 to 0.905, and p < 0.001. These results do not agree statistically.

The sponsor’s reported analysis and the recalculation should therefore be distinguished rather than combined into a single conclusion. The recalculation uses posted summary statistics and is not a reproduction of a participant-level model. The registry record does not provide enough detail about the sponsor’s model, analysis population, or other analysis choices to identify why the results differ. It would be inappropriate to assume a reason from the summaries alone.

The registry record does not flag this as a discrepancy. That status does not remove the visible difference between the reported p-value and the recalculated one. The sponsor’s analysis remains the registry’s reported analysis; the recalculation is a separate, unadjusted check. Readers should consider the reported model result alongside the posted group rates and should not treat the recalculated test as a replacement for the sponsor’s analysis.

For secondary outcomes, sponsor-reported t-test p-values are also listed. The record gives p < 0.001 for symptom treatment, p = 0.402 for physical symptom control, and p = 0.803 for neuropsychiatric symptom severity. The recalculated p-values are similar in broad direction, although not identical for the latter outcomes. Details sufficient to reconcile those differences are not reported in the registry record.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists 14 outcome measures and 14 analyses, with 13 secondary outcomes. It does not report whether the investigators adjusted for testing multiple outcomes, used a gatekeeping procedure, or specified a hierarchy for interpreting them. Without that information, the posted secondary p-values should be treated as unadjusted evidence rather than as results that control the overall chance of a false-positive finding across all outcomes.

The registry record does not report the amount or pattern of missing outcome data, reasons for missingness, or the method used to address missing measurements. It also does not describe whether the analyses used complete cases, imputation, or another approach. The outcome-specific group sizes for the posted secondary symptom measures are smaller than the recorded group sizes for the primary outcome; why the counts differ is not reported in the registry record.

The analyzed population and analysis-set definitions are not reported in the registry record. Although the trial is randomized, the record does not state whether the primary analysis followed an intention-to-treat principle, how participants were handled if they did not receive the assigned condition, or how post-randomization exclusions were managed. The registry also does not report whether participants who died or otherwise stopped contributing follow-up were handled in a specific way for the rate endpoint.

These omissions limit assessment of how robust the estimates are to missing data, outcome selection, and analysis decisions. General statistical explanation: multiplicity adjustment can reduce false-positive risk across a family of tests, often at the cost of reduced sensitivity for individual outcomes. Missing-data methods rely on assumptions about why outcomes are unavailable. When the methods and their assumptions are not reported, the impact cannot be evaluated from summary results alone. These limitations do not establish that the analyses were mishandled; they identify information not reported in the registry record.

16. 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.

Most frequent non-serious adverse event terms

The registry reports serious adverse events affecting 0 of 221 patients in the ADRD-PC Program group and 0 of 221 patients in the control group. It also reports 0 serious adverse events affecting caregivers in either group, with 221 caregivers at risk in each. These figures describe the registry’s recorded serious adverse event counts; they do not establish that no other harms occurred or that all events were captured.

Deaths are listed separately for patients: 41 in the intervention group and 40 in the control group. The registry records 0 caregiver deaths in each group. A death count should not be assumed to be the same classification as the serious adverse event counts, and the record does not provide details about cause, timing, or relationship to the intervention.

Other adverse events affecting caregivers are recorded for 24 people in the intervention group and 40 in the control group; the patient groups each have 0 recorded in this category. The registry record does not provide event descriptions in the posted summary. These counts are not, by themselves, evidence about whether the groups differ in risk: interpretation would require details of event definitions, ascertainment, follow-up, and the analysis used.

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

The posted data describe a randomized comparison of a dementia-specific palliative care program with caregiver educational materials and usual care. For hospital transfers during the 60 days after index discharge, the intervention-group rate is numerically lower than the control-group rate. The unadjusted recalculation gives a rate ratio below 1 and a confidence interval that lies below 1. The registry’s reported Poisson regression test, however, gives p = 0.522, which conflicts with the recalculated p < 0.001.

This result does not establish that the program caused a reduction in hospital transfers. The difference between the sponsor’s analysis and the recalculation is unresolved in the registry record, and the record lacks detailed information about the model and analysis population. A numerical difference in group summaries is not enough to determine causation or to explain the mechanisms that might produce it.

Nor does the primary endpoint capture every aspect of care or caregiver experience. The trial’s stated aims include symptom treatment and control, supportive services, nursing-home transitions, and caregiver communication, shared decision-making, and distress. The posted results include only some of the secondary outcomes, and the registry does not report the full analysis details needed to evaluate all of those aims.

General statistical explanation: a confidence interval expresses uncertainty under the calculation used, not certainty about an effect or protection against bias. A p-value does not quantify the size or practical importance of a result. The reported interval and p-value should be understood in relation to the underlying analysis and its assumptions, especially when another reported analysis gives a different result.

The registry record therefore supports a careful description of the observed rates and the competing inferential results, not a definitive claim about the program’s effectiveness. It also does not establish that a particular approach is appropriate for an individual patient or caregiver. The findings apply to the outcomes and follow-up described in the record, and their broader interpretation depends on design and analysis details that are not reported.

18. Why This Trial Matters Statistically

This trial addresses a statistical challenge common in clinical research on serious illness: the primary outcome is a count of events accumulated over follow-up, not simply a single measurement per participant. Expressing transfers per person-day at risk makes exposure time part of the endpoint and calls for careful modeling and reporting of the rate comparison.

The record also illustrates why readers should distinguish a group summary from a model-based test. The posted group rates and recalculated comparison indicate a lower rate in the intervention group, while the sponsor-reported Poisson regression gives p = 0.522. This is not a minor presentation detail. Without model specifications or analysis-population information, readers cannot determine why the conclusions diverge. Transparent reporting of those details is important for reproducibility and interpretation.

The design includes patients and caregivers and addresses multiple outcomes that reflect different aspects of care. That breadth can be valuable, but it also makes clear reporting of outcome priorities and multiplicity procedures important. The registry record does not report those procedures, nor does it provide results for every listed secondary outcome.

Finally, the trial’s completed status and its recorded enrollment do not by themselves establish the precision or strength of every outcome analysis. The primary outcome group counts and secondary outcome group counts differ, and the record does not explain why. A statistically careful reading therefore considers the endpoint definition, exposure time, analysis approach, uncertainty, missing-data reporting, and the relation between the sponsor’s own analysis and any summary-statistic recalculation.

19. 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.
• Rate comparisons use a Poisson approximation unless the sponsor posted confidence limits; real event counts are often over-dispersed.

The primary interpretation is limited by the contrast between the sponsor-reported Poisson regression p = 0.522 and the recalculated p < 0.001. The registry record does not report enough model detail to reconcile the results. The posted rate ratio and its interval should therefore not be read as a substitute for the sponsor’s analysis.

The record does not report the original sample-size calculation, analysis plan, confidence level for the posted recalculated intervals, or detailed specifications for the primary model. It also does not report the analysis population, adjustment for baseline characteristics, interim analyses, or how incomplete follow-up was handled. These omissions restrict assessment of assumptions and reproducibility.

For secondary outcomes, results are posted for only some listed measures, and the record does not describe a multiplicity strategy. Outcome-specific group sizes differ from those for the primary endpoint, with no explanation in the registry record. The missing-outcome count, reasons for missingness, and methods for handling unavailable measurements are also not reported. These gaps limit interpretation of both precision and potential selection effects.

The outcome measures address hospital transfers and selected symptom scores, but do not alone provide a full assessment of caregiver communication, shared decision-making, distress, or all supportive services named in the trial’s aims. The registry record does not report every listed secondary result. Safety summaries also lack event descriptions and details about ascertainment, timing, or relationship to the intervention. These limitations concern the scope and completeness of the posted record; they do not establish that unreported analyses or procedures were absent from the trial.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Palliative Care for Persons With Late-stage Alzheimer's and Related Dementias and Their Caregivers — NCT04948866.
• Registry record last updated 2025-12-16; status last verified 2025-02.
• Results first posted 2025-12-16.
• 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.

23. Record Summary

NCT04948866 is a completed randomized, double-masked, parallel-group trial sponsored by the University of North Carolina, Chapel Hill. It evaluates the ADRD-PC Program for people with late-stage Alzheimer’s disease and related dementias and their family caregivers, compared with Alzheimer’s Association caregiver materials and usual patient care.

The primary endpoint is hospital-transfer events per person-day at risk during the 60 days after index hospital discharge. The posted group rates are 0.99 for the intervention and 1.12 for control. The unadjusted recalculation gives a rate ratio of 0.884, with a confidence interval from 0.863 to 0.905 and p < 0.001; the registry’s reported Poisson regression gives p = 0.522. The record does not explain this difference. Some secondary symptom outcomes and adverse-event summaries are also posted, while several analysis and missing-data details are not reported in the registry record.