Northern-Manhattan Hispanic Caregiver Intervention Effectiveness Study: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Northern-Manhattan Hispanic Caregiver Intervention Effectiveness Study (NCT02092987), a randomized na study. The primary outcome is Score on Geriatric Depression Scale.

2. Clinical Question

3. Trial Design

This was a randomized, double-masked, parallel-group trial, listed as completed. The registry record identifies Columbia University as the sponsor. The recorded start was 2014-03, the primary completion was 2016-09, and enrollment was 221. The record labels the phase as NA. These design labels describe the registry entry; details behind the masking procedures and allocation process are not reported in the registry record.

The two arms were the NYU Caregiver Intervention and REACH OUT. Each arm also received social work support and educational materials. The study description calls the comparison pragmatic and gives a total intervention period of 6 months. The primary endpoint was measured at 6 months. The stated target population was Hispanic caregivers of people with dementia in Northern Manhattan, and the investigators’ objective was to obtain comparative effectiveness information that could inform caregivers and health providers choosing between the interventions.

There were 110 participants in the NYU Caregiver Intervention arm and 111 in the REACH OUT arm in the posted primary outcome. Those outcome counts total the registry enrollment, but the record does not provide a detailed account of participant flow, withdrawals, missing outcome data, or the analysis population. It also does not report the original sample-size justification, recruitment targets beyond the brief description, or the random sequence and allocation concealment procedures. These gaps limit what can be inferred about how the trial was planned and how the posted outcome set relates to every enrolled participant.

The study description identifies depressive symptoms and caregiver burden as central aims. The posted primary endpoint is the Geriatric Depression Scale score, with additional patient-reported measures at 6 months. The intervention-specific details beyond the names of the counseling programs and shared supports are not reported in the registry record. The record’s double-masking label is reported as a design feature, but which groups of people were masked and how masking was maintained are not reported.

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 Score on Geriatric Depression Scale, measured over 6 months.

The primary endpoint was the score on the Geriatric Depression Scale (GDS) at 6 months. The registry describes the GDS as a 30-item self-report assessment used to identify depression in older people. Each question is answered yes or no, one point is assigned per answer, and the points are combined into a cumulative score. Higher scores indicate more depressive symptoms.

The registry’s scoring description classifies scores from 0-9 as normal, 10-19 as mildly depressed, and 20-30 as severely depressed. These are categories in the scale description; they should not be mistaken for evidence that an intervention caused a person to move between categories. The posted arm results are group means and standard deviations, so they describe average scores and their spread rather than the number of participants in each category.

Depressive symptoms were a central part of the study question, which asked whether one caregiver intervention was more effective than the other in alleviating symptoms and burden. A self-report scale provides a structured way to compare symptom scores between the randomized groups at the stated assessment time. It does not, by itself, diagnose a participant or show why a score differed. The primary result concerns the average GDS score at 6 months; it is not a direct measure of every aspect of caregiver wellbeing, nor does it report the primary burden outcome. Additional details about how the GDS was administered, its handling when items were unanswered, or its interpretation beyond the stated score categories are not reported in the registry record.

7. Results Posted in the Registry

• REACH OUT: mean 9.80 versus 9.57 on control; difference 0.23 (95% CI -1.66 to 2.12), standardized effect d = 0.03; z-test p = 0.812.

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.

Score on Patient Reported Outcomes Measurement Information System (PROMIS) Physical Health Scale

• 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:
  • REACH OUT: mean 28.68 versus 30.10 on control; difference -1.42 (95% CI -3.50 to 0.66), standardized effect d = -0.18; z-test p = 0.181.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between REACH OUT and the control. The interval gives the range of effects that remain compatible with the data.

Score on Perceived Stress Scale (PSS)

• 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:
  • REACH OUT: mean 14.95 versus 14.98 on control; difference -0.03 (95% CI -2.18 to 2.12), standardized effect d = -0.00; z-test p = 0.978.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between REACH OUT and the control. The interval gives the range of effects that remain compatible with the data.

Score on the PROMIS Short Form (SF) Depression Scale

• 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:
  • REACH OUT: mean 48.51 versus 48.01 on control; difference 0.50 (95% CI -2.22 to 3.22), standardized effect d = 0.05; z-test p = 0.718.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between REACH OUT and the control. The interval gives the range of effects that remain compatible with the data.

The posted secondary outcomes cover physical health, perceived stress, and depression-related measures, with the registry listing four secondary outcomes overall. The same general pattern as the primary GDS comparison is that the reported group means are close relative to their variability and the recalculated confidence intervals include no difference. The registry’s own mixed-model p-values are also reported for these measures. Taken together, the secondary results do not present a strong, consistent signal that separates the two interventions across the measures described.

These measures provide context around the primary result because they address related aspects of participants’ self-reported health and wellbeing. They are not interchangeable with the GDS endpoint, however: physical health, perceived stress, and other depression scales measure different constructs. Similar findings across related outcomes may be informative, but they do not establish that all dimensions of caregiver experience are unchanged or that the interventions have the same effect.

Unadjusted secondary results carry less weight when several outcomes are examined because each additional comparison creates another opportunity for a chance finding. General statistical practice may address this through a prespecified hierarchy or a multiplicity adjustment, but the registry record does not report such a strategy. The secondary analyses are also described as exploratory in the study summary. Their p-values should therefore be treated as supporting context rather than as independent confirmatory evidence.

The registry lists a Patient Health Questionnaire (PHQ-9) outcome among the secondary outcomes, but a corresponding posted result is not included in the secondary results. Details about the intended role of each secondary measure, how they were selected, and how missing data were handled are not reported in the registry record. No individual secondary result should override the primary endpoint simply because it appears more favorable or less favorable; interpretation should consider the overall pattern and the exploratory status of these analyses.

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

• Score on Patient Reported Outcomes Measurement Information System (PROMIS) Physical Health Scale
• Score on Perceived Stress Scale (PSS)
• Score on the PROMIS Short Form (SF) Depression Scale
• Total Score on Patient Health Questionnaire (PHQ-9)

9. Statistical Interpretation

10. Statistical Methods Explained

For a continuous outcome, a group mean describes the arithmetic average of the recorded scores. The standard deviation describes how spread out individual scores are around that average. In this trial, the registry reports means and standard deviations for GDS scores in each arm at 6 months. These summaries describe the observed outcome distributions at group level; they do not show the shape of those distributions or how individual participants’ scores changed over time.

The recalculated comparison is an unadjusted difference in means, with the REACH OUT mean contrasted against the NYU Caregiver Intervention mean. A positive estimate means the reported average was higher in REACH OUT; because higher GDS scores indicate more depressive symptoms, the direction of the numerical estimate is toward a higher average score in that arm. The estimate is 0.23 units. This is a descriptive direction, not evidence that the difference is real or that one intervention caused it.

A confidence interval gives a range of values compatible with the data and the statistical model used for the calculation. It reflects sampling uncertainty under that model. A 95% confidence interval from -1.665 to 2.125 includes zero, the value corresponding to no average difference. It also includes estimates in both directions. The interval therefore does not support a precise conclusion about the direction or size of the difference. A confidence interval is not a range in which 95% of individual participant scores fall, and it does not account for every possible source of bias or uncertainty.

The unadjusted p-value of 0.812 measures how unusual the observed result, or a more extreme result, would be under a null model of no average difference and the assumptions used for the calculation. A large p-value does not establish that the null model is true. It can reflect a small underlying effect, uncertainty in the estimate, or both. Statistical evidence and practical importance are separate questions; the registry does not report a threshold for a clinically important GDS difference.

The registry’s own reported method is a Mixed Models Analysis, with a p-value of 0.70 for the primary outcome. Mixed models are general statistical methods that can represent repeated measurements and account for correlation among observations from the same participant. In general, they can also incorporate multiple sources of variation and covariates, depending on the model specification. The registry record does not report the model formula, the time points included, the covariance structure, covariates, or the precise interaction term being tested. It is therefore not possible to reconstruct the sponsor’s analysis from the reported entry or to explain exactly how it differs from the unadjusted summary-statistic comparison.

In general, an unadjusted comparison uses the group summaries without accounting for other variables. An adjusted or model-based analysis may target a different quantity or use information across time points. As a result, p-values from different methods need not match, even when they concern the same broad endpoint. In this record, the mixed-model p-value is reported without an accompanying effect estimate or confidence interval. That means it gives a test result but does not show the estimated size of the difference or its precision. It should not be translated into a statement that the groups were identical.

The statistical kind of the primary outcome is continuous, and the posted arm summaries use means and standard deviations. General statistical methods often use these summaries to estimate a mean difference and its standard error. The standard error describes the expected sampling variability of an estimate under the calculation’s assumptions. It is not the standard deviation of participant scores. The recalculated confidence interval and p-value are based on the posted summary statistics and are unadjusted; they are not a substitute for individual-level analysis.

General explanation of over-dispersion: for count outcomes, over-dispersion means that observed variability exceeds what a simple count model expects. It can make standard errors too small if ignored. That concept is not directly the issue for this continuous GDS outcome, and the registry does not report a count model for the primary endpoint. Likewise, general methods for time-to-event outcomes do not apply to the posted mean-score comparison.

The registry reports random allocation, which is intended in general to reduce systematic differences between groups at baseline. Randomization does not guarantee identical groups in a particular sample, and the record does not provide baseline outcome or covariate summaries in the registry record. It also does not report whether the mixed-model analysis followed an intention-to-treat principle or how missing observations were handled. Those details matter because incomplete follow-up can affect the population represented by the analyzed scores.

The reported p-value for the sponsor’s model and the recalculated unadjusted p-value answer related but not necessarily identical statistical questions. The registry’s model result is the trial’s own reported analysis. The simple recalculation offers a transparent comparison based on posted arm summaries and should be read as a supplementary check, not as a replacement for the sponsor’s analysis. Neither result alone establishes clinical importance, equivalence, or an effect on outcomes not reported for this comparison.

11. Confidence Intervals and Precision

The primary unadjusted mean difference is 0.23 GDS units, and its 95% confidence interval extends from -1.665 to 2.125. This interval is fairly broad relative to the small point estimate: it includes zero as well as possible differences on either side. The evidence therefore does not pin down a single direction or a narrow range of plausible effects. The trial’s observed means are close, but the uncertainty means that similarity of the point estimates should not be confused with a precise finding of no difference.

In general, precision describes how narrowly an estimate is determined by the data and the method. A narrower confidence interval indicates less sampling uncertainty than a wider one, under the same assumptions. The interval does not capture all uncertainty, such as possible selection effects, measurement limitations, or departures from the model. Those issues cannot be quantified from the registry record.

The sponsor-reported mixed-model result gives a p-value of 0.70 but no estimate or confidence interval. Without those, the size and precision of the model-based effect cannot be assessed from the registry report. The unadjusted confidence interval is available for the summary-statistic comparison, but it should not be attributed to the mixed model.

The recorded arm sizes for the primary outcome are 110 and 111, and the standard deviations are 7.12 and 7.25. These details help describe the information contributing to the summary comparison, but they do not remove uncertainty about the effect. A more informative account of precision would include the model-based effect estimate and confidence interval, along with details of missing data and analysis populations; those are not reported in the registry record.

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 illustrations based on the observed effect, not the trial’s original sample-size calculation. For these continuous outcomes, the figures are participants per arm, not total events. The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is specified as a fraction of the observed effect.

The 80% power column gives the illustrative number of participants per arm for a hypothetical study to have an 80% chance of detecting the specified effect under those assumptions, if that effect were the true one and the model assumptions held. The 90% power column gives the corresponding planning figure for a higher chance of detection. Higher planned power generally requires more participants for the same assumed effect. The registry does not report the trial’s own original design assumptions or sample-size justification.

The rows describe planning for 100%, 75%, and 50% of the observed effect. A smaller target effect requires more participants because a smaller signal is harder to distinguish from sampling variation. This pattern is a general feature of planning calculations, not evidence that the smaller effects occurred in the trial. For the primary endpoint, the observed effect is particularly small, so calculations that scale down that effect can yield very large illustrative sample requirements. These numbers are not recommendations and do not establish what sample size would be appropriate for another study; that decision would require a justified target effect and design assumptions.

These figures are not observed power. Observed power calculated from the trial’s estimated effect would largely restate the result and its uncertainty, rather than provide an independent measure of evidential strength. The planning values should instead be read as conditional illustrations of the sample sizes associated with the listed target effects and assumptions. They do not explain whether the completed trial was adequately planned, because its original planning basis is not reported in the registry record.

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, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = 0.230; pooled standard deviation σ = 7.186.
2. n = 2 × (1.96 + 0.842)² × 7.186² / 0.230² = 15,321.7, rounded up to 15,322 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
delta <- 0.23                      # observed difference in means
sigma <- 7.18559079467326                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^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 = ",")))

# Check against R's built-in function (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(2092987)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 15,322 (80% power), 20,512 (90% power)
 75% of observed effect: n per arm = 27,239 (80% power), 36,465 (90% power)
 50% of observed effect: n per arm = 61,287 (80% power), 82,046 (90% power)
power.t.test: 15322.7 
Simulated power at n = 15,322 per arm: 0.801 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Analyses Reported by the Sponsor

The registry reports the trial’s own primary analysis as a Mixed Models Analysis, with a p-value of 0.70. The reported entry does not include the corresponding effect estimate or confidence interval. This is the sponsor-reported analysis and should be distinguished from the unadjusted recalculation using the posted arm means and standard deviations.

The recalculated unadjusted comparison gives a mean difference of 0.23 units, a 95% confidence interval from -1.665 to 2.125, and a p-value of 0.812. The difference between these p-values does not, by itself, indicate a contradiction: a mixed model and a simple comparison of arm summaries can use different information and assumptions. The registry record does not provide sufficient model details to explain the difference more specifically.

The recalculation is a supplementary, transparent summary of the posted means; it does not replace the sponsor’s analysis or reproduce the mixed model. Because the sponsor’s result lacks an effect estimate and interval, its magnitude and precision cannot be compared directly with those of the unadjusted calculation. Both reported results fail to provide clear evidence of a difference, but neither establishes that the interventions are equivalent.

15. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when researchers evaluate several outcomes, time points, or statistical comparisons. In general, as the number of tests increases, the chance of at least one apparently notable result arising by chance can also increase. The record describes a primary outcome and multiple secondary measures, including exploratory aims, but does not report a multiplicity strategy, such as a hierarchy of tests or an adjustment to the significance threshold. The effect of any such approach on the interpretation cannot therefore be assessed from the registry record.

Missing data can affect both the estimated group means and the validity of a comparison, particularly if the likelihood of having an observed score is related to the outcome or differs between groups. The registry reports 110 and 111 participants in the primary outcome arms, but does not provide participant flow, the number with missing measurements, reasons for missingness, or methods used to address incomplete data. The listed outcome counts should not be taken as proof that every enrolled participant had complete data at every stage.

The registry identifies a Mixed Models Analysis for the primary result, but does not report its data-handling assumptions or whether additional methods were used for missing scores. In general, model-based analyses can use available repeated measurements under particular assumptions; those assumptions cannot be checked from the information reported. Sensitivity analyses examining alternative missing-data assumptions are not reported in the registry record.

The analysis population is also not specified. The registry record does not state whether the sponsor analyzed participants according to randomized assignment, restricted analyses to participants who received an intervention, or used another population definition. It does not report protocol deviations or exclusions from the analysis. These omissions make it difficult to assess how closely the reported estimates preserve the benefits of randomization or how broadly they represent all enrolled participants.

These are reporting limitations, not evidence that the trial mishandled multiplicity or missing data. They mean that readers cannot determine the methods used or test the robustness of the reported findings from the registry record alone. A protocol or full analysis report would be needed to evaluate those aspects in detail.

16. 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 lists zero serious adverse events among 110 participants at risk in the NYU Caregiver Intervention arm and zero among 111 participants at risk in the REACH OUT arm. It also lists zero other adverse events among those at risk in each arm. These entries describe the events reported in the registry for the stated groups; they do not establish that no health events occurred outside the registry’s reporting definitions or observation process.

Deaths are not reported in the registry record. The record also does not provide serious adverse event terms, with no specific serious-event terms listed. Details such as the adverse-event collection period, event definitions, relatedness assessment, severity grading, or completeness of reporting are not reported. With no reported events in these categories, a comparative estimate of safety cannot be derived from the entries alone.

These figures should be read as registry reporting, not as a comprehensive assessment of risk. The absence of listed events is not proof of zero risk, and the small counts of participants at risk do not establish that the interventions have equivalent safety profiles. The information is limited to the adverse-event fields reported in the registry record.

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

The primary result answers a narrow question: how the mean GDS scores compared between the two randomized groups at 6 months in this study population. The observed means were 9.57 for the NYU Caregiver Intervention and 9.8 for REACH OUT. The unadjusted estimate and its confidence interval do not show a clear difference in average scores. This is a statement about the posted data and their precision, not a recommendation about which intervention a caregiver should choose.

The result does not show that the interventions are identical. A finding that does not clearly distinguish groups can arise when the true effects are similar, but it can also occur when the data are not precise enough to identify a difference. The confidence interval includes zero and spans values in both directions. The registry does not report a prespecified equivalence margin or an analysis designed to establish equivalence.

The result also does not answer every question raised by the study description. Although caregiver burden is part of the stated research aim, a primary burden result is not posted in the outcomes. The GDS score does not represent all aspects of caregiver burden, stress, physical health, or quality of life. The result is specific to the scale and follow-up time recorded and does not establish what happened at other times.

The registry describes a Hispanic caregiver population in Northern Manhattan. Applying the findings to caregivers in other settings or populations would require considering differences in context, participants, and intervention delivery. The record does not provide enough detail about those features to define the limits of generalizability more precisely.

Finally, the trial’s registry results alone do not establish mechanism. The investigators proposed that a family-centered counseling approach might be especially relevant in this cultural context, but the posted primary comparison does not test that explanation directly. The data should be interpreted as a comparison of the interventions as described, with shared social work support and educational materials, rather than proof of why any observed scores were similar or different.

18. Why This Trial Matters Statistically

This trial is statistically informative because it compares two active caregiver interventions rather than contrasting an intervention with no support. Both groups received social work support and educational materials, so the main contrast concerns the NYU Caregiver Intervention versus REACH OUT on top of those shared components. That design aligns with the practical question of choosing between the two approaches, while also making the estimated contrast specific to that supported implementation.

The primary outcome is a continuous symptom score, and the registry reports both group averages and variability. This makes it possible to distinguish the direction of the observed estimate from its uncertainty. The means are close, but the confidence interval is broad enough to include differences in either direction. That contrast illustrates why a point estimate alone is insufficient and why a non-significant result should not be treated as evidence of equivalence.

The registry’s sponsor-reported mixed-model p-value and the unadjusted summary-statistic comparison also illustrate why method labels and effect estimates matter. A p-value alone provides limited information about magnitude and precision. Reporting an estimate and confidence interval for the model-based result would make it easier to understand the sponsor’s analysis and compare it with the descriptive recalculation.

The study question is rooted in a particular population and cultural rationale, and the investigators explicitly framed family-centered counseling as a possible reason for differing effectiveness. The posted outcome does not establish that rationale, but it makes clear that an intervention comparison should be interpreted in relation to the population and delivery context studied. The registry’s gaps on missingness, analysis populations, multiplicity, and model specification are also reminders that transparent reporting is necessary for a full statistical assessment.

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.

The primary outcome information is limited to the 6-month GDS score. A result for caregiver burden, despite its place in the stated study question, is not included among the outcome results described in the record. The registry lists additional secondary measures, but not every listed measure has a posted result in the information. The full scope of the planned outcome assessment is therefore not available.

The sponsor’s primary mixed-model result is reported only as a p-value. An effect estimate, confidence interval, model specification, and details of the time structure are not reported in the registry record. This prevents a direct assessment of the model-based effect’s size and precision, and prevents reproduction of the sponsor’s analysis. The unadjusted comparison is based on group summary statistics and cannot account for adjustments that might have been made in the mixed model.

Participant flow, missingness, reasons for missing data, and analysis populations are not reported. The registry also does not provide the trial’s original sample-size justification or the assumptions underlying its design. These omissions make it difficult to judge whether the study was planned to detect a particular clinically important difference or how representative the analyzed participants were of those enrolled.

The multiplicity strategy is not reported, although several outcomes are listed and some aims are described as exploratory. The record does not report whether secondary analyses were adjusted for multiple comparisons or organized under a prespecified testing hierarchy. The safety entries list no adverse events in the reported categories, but event definitions, collection procedures, and completeness are not described.

Finally, the registry’s intervention descriptions are brief. The record does not report details of treatment fidelity, adherence, intervention dose by participant, or masking procedures. These limitations do not show that the trial was conducted inadequately; they define what cannot be evaluated from the registry information.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Northern-Manhattan Hispanic Caregiver Intervention Effectiveness Study — NCT02092987.
• Registry record last updated 2021-07-09; status last verified 2020-09.
• Results first posted 2021-07-09.
• 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

NCT02092987 is a completed randomized, double-masked, parallel-group trial sponsored by Columbia University. It enrolled 221 participants and compared the NYU Caregiver Intervention with REACH OUT among Hispanic caregivers of people with dementia in Northern Manhattan. Both arms received social work support and educational materials, and the intervention period was described as 6 months.

The primary outcome was GDS score at 6 months. Mean scores were 9.57 and 9.8, respectively; the unadjusted mean difference was 0.23, with a 95% confidence interval from -1.665 to 2.125 and p-value 0.812. The sponsor-reported mixed-model analysis had p = 0.70, without a posted effect estimate or confidence interval. The available results do not establish a clear difference or equivalence between the interventions. Important details about missing data, analysis populations, model specification, and the trial’s original sample-size planning are not reported in the registry record.