The SUCCEED Trial of Secondary Stroke Prevention: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for The SUCCEED Trial of Secondary Stroke Prevention (NCT01763203), a randomized na study. The primary outcome is Systolic Blood Pressure.

2. Clinical Question

3. Trial Design

SUCCEED was a randomized, parallel-group trial with single masking. It enrolled 487 participants and is recorded as completed. The study began on 2014-02-13, and its primary completion date is 2018-09-20. The registry lists the phase as NA. The identity of the people who were masked and the practical masking procedures are not reported in the registry record.

Participants were allocated to Care Management+Community Health Worker or usual care. The posted outcome summaries include 241.0 participants in the care-management group and 246.0 in usual care. The first arm is described as care management; the comparator arm is described as written materials. The registry describes the intervention as an outpatient clinic and community-based approach for stroke patients in the Los Angeles County safety net. It also states that the study aimed to measure the costs of running the intervention relative to usual care. Cost results are not reported in the registry record summarized by these outcome data.

The primary outcome was mean systolic blood pressure at 12 months. The record describes measurement at baseline and 3 months as well, but gives no corresponding numerical summaries in the posted primary outcome results. The registry lists secondary outcomes concerning dyslipidemia, glycated hemoglobin, inflammation, body mass index, waist circumference, physical activity, salt intake and soda intake. It records 20 outcome measures and 19 analyses. More detailed eligibility criteria, the recruitment process, allocation-concealment procedures, and the trial’s original sample-size justification are not reported in the registry record.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: Not applicable; all enrolled participants were assigned to an arm.

Overall Study

Reasons for not completing this period:

5. Baseline Characteristics

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

Population: Not applicable - analysis population is different from participants enrolled

6. Primary Endpoint

The registry lists the primary outcome as Systolic Blood Pressure, measured over 12 months.

Systolic blood pressure is a continuous measure recorded in mm Hg. The primary endpoint is the mean value at 12 months. The registry describes additional measurements at baseline and 3 months, which were intended to track blood pressure over the participation period; the primary time point remains 12 months.

Blood pressure is relevant to the study’s stated purpose because the intervention was designed to improve control of stroke risk factors among people with stroke. A continuous endpoint preserves the magnitude of recorded measurements rather than reducing each participant to a category. Comparing group means can therefore describe the average difference between the randomized groups at the specified follow-up time. The mean alone does not describe all participant outcomes: the standard deviation also indicates how dispersed measurements were within each group.

The registry reports the 12-month arm means and standard deviations. It does not provide a primary endpoint estimate or confidence interval from the sponsor’s mixed-model analysis. The summary-statistic recalculation gives an unadjusted mean difference, but it is not a substitute for the sponsor’s stated analysis. Details of how the primary endpoint was collected, how missing measurements were handled, and what covariates or repeated observations entered the model are not reported in the registry record.

7. Results Posted in the Registry

• Care Management+Community Health Worker: mean 133.10 versus 136.90 on control; difference -3.80 (95% CI -7.57 to -0.03), standardized effect d = -0.18; z-test p = 0.048.

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.

Dyslipidemia

• 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:
  • Care Management+Community Health Worker: mean 98.70 versus 104.10 on control; difference -5.40 (95% CI -14.02 to 3.22), standardized effect d = -0.11; z-test p = 0.219.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Care Management+Community Health Worker and the control. The interval gives the range of effects that remain compatible with the data.

Percentage of Glycated Hemoglobin (Hemoglobin A1C)

• 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:
  • Care Management+Community Health Worker: mean 6.50 versus 6.60 on control; difference -0.10 (95% CI -0.32 to 0.12), standardized effect d = -0.08; z-test p = 0.378.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Care Management+Community Health Worker and the control. The interval gives the range of effects that remain compatible with the data.

Body Mass Index

• 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:
  • Care Management+Community Health Worker: mean 29.50 versus 29.20 on control; difference 0.30 (95% CI -0.68 to 1.28), standardized effect d = 0.05; z-test p = 0.549.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Care Management+Community Health Worker and the control. The interval gives the range of effects that remain compatible with the data.

The registry lists a range of secondary risk-factor and behavior outcomes. Numerical summaries are posted for dyslipidemia, glycated hemoglobin and body mass index. Across these posted comparisons, the unadjusted intervals include no difference. This overall pattern does not demonstrate that the groups were equivalent; it means the available comparisons do not clearly separate the group means from no difference. Sponsor mixed-model p-values are reported for some, but not all, of the secondary outcomes with numerical summaries.

The secondary pattern does not overturn or resolve the primary result. The primary outcome is systolic blood pressure at 12 months, and its recalculated summary comparison conflicts with the sponsor’s mixed-model result. Secondary measures concern different outcomes, so they cannot establish what the primary model would have shown or explain its discrepancy. They provide context about other measured risk factors, rather than a substitute primary analysis.

Unadjusted secondary comparisons carry less weight because a study that evaluates many outcomes creates more opportunities for apparently notable results by chance. The registry does not report a multiplicity adjustment or a hierarchy for interpreting secondary outcomes. In addition, sponsor analyses are not reported for every numerical secondary result, and detailed methods for each outcome are not available. For these reasons, the secondary findings are best treated as supportive or exploratory, not as independent confirmation of a primary effect.

The registry identifies additional outcomes such as inflammation, waist circumference, physical activity, salt intake and soda intake, but detailed results for those measures are not reported in the registry record. No conclusion about their direction or precision should be inferred from the results for the outcomes with posted numerical summaries.

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

• Dyslipidemia
• Percentage of Glycated Hemoglobin (Hemoglobin A1C)
• Inflammation: C-reactive Protein
• Body Mass Index
• Waist Circumference
• Physical Activity
• Salt Intake
• Soda Intake

9. Statistical Interpretation

10. Statistical Methods Explained

The registry identifies the primary result as a mixed-model analysis and reports p = 0.73 for the comparison between Care Management+Community Health Worker and usual care. It does not report the model’s coefficient, confidence interval, covariates, covariance structure, handling of repeated measurements, or analysis population. Those details are needed to understand exactly how the sponsor’s result was estimated. They are not reported in the registry record.

What does a mean difference describe?

For a continuous outcome, a mean is the arithmetic average of recorded measurements in a group. The standard deviation describes how spread out individual measurements are around that average. A mean difference expresses the contrast between two group averages in the outcome’s original units. For systolic blood pressure, the unit is mm Hg. In the recalculation, the contrast is care management minus usual care, so a negative estimate indicates a lower posted mean in the care-management arm. This descriptive direction does not by itself establish a reliable group effect.

The recalculated estimate and interval use the posted arm summaries and are unadjusted. They do not account for the baseline or 3-month measurements, any covariates, or other features that may have been included in the sponsor’s model. No assumptions about the model’s details should be inferred from the term mixed model alone. The contrast between the recalculated result and the reported p-value is why the sponsor’s analysis takes precedence.

What does a p-value tell the reader?

In general, a p-value describes how compatible the observed data are with a specified null model, under the assumptions of the test. It is not the probability that the null hypothesis is true, and it does not measure the size or importance of an effect. The registry gives p = 0.73 for the sponsor’s primary analysis but does not provide the corresponding model-based estimate or interval. The unadjusted recalculation gives p = 0.048. These values arise from different analyses and should not be presented as interchangeable.

Why do intervals matter?

In general, a confidence interval gives a range of values compatible with data and an estimation procedure at a stated confidence level. Its width conveys precision: a wider interval indicates less precision than a narrower interval. The recalculated primary interval runs from -7.567 to -0.033 in the underlying result, with the upper endpoint very close to zero. The registry does not identify the confidence level for these recalculated intervals. The sponsor’s model-based confidence interval is not reported. An interval near zero is especially sensitive to differences in analytic approach, as the discrepancy in this record illustrates.

Why can a mixed model differ from an endpoint-only comparison?

In general, a mixed model can represent correlation among repeated measurements from the same participant and may incorporate measurements taken at multiple times. Such models can also include specified covariates. Whether and how these features were used in SUCCEED is not reported in the registry record. The endpoint-only recalculation uses the posted 12-month summaries, so it cannot reconstruct a model that may have used additional information or different assumptions.

Why does over-dispersion matter?

For continuous measures, variation among participants affects the uncertainty around a group mean and the precision of a comparison. Greater variation generally makes a given mean difference harder to estimate precisely. The posted standard deviations describe spread in each arm, while the recalculated comparison uses a pooled standard deviation. In general, over-dispersion is particularly important when a model assumes less variability than the data exhibit: standard errors may then be too small and statistical evidence overstated. The registry does not report an over-dispersion assessment for the sponsor’s analysis, and the outcome is continuous rather than a count outcome for which a count-model dispersion parameter would ordinarily be central.

What cannot be inferred from the analysis label?

The method label does not reveal whether the analysis followed intention-to-treat principles, which participants contributed to each model, how missing observations were addressed, or whether the model was prespecified. The registry record does not report these details. Randomization identifies the intended basis for comparing groups, but interpretation still depends on how the recorded data and missing values entered the analysis. Those uncertainties should remain explicit rather than being filled in from assumptions about common trial practice.

11. Confidence Intervals and Precision

The arm summaries give a mean of 133.1 mm Hg with a standard deviation of 20.5 for Care Management+Community Health Worker, and a mean of 136.9 mm Hg with a standard deviation of 21.9 for usual care. The recalculated unadjusted mean difference is -3.8 mm Hg. Its interval extends from -7.567 to just below zero. This range indicates that the unadjusted estimate is imprecise near the no-difference boundary: values close to no difference remain compatible with the summary-statistic calculation, as do more negative values within the interval.

The sponsor’s mixed-model p-value is 0.73, but the registry does not give a model-based estimate or interval. Consequently, the precision of the sponsor’s estimated treatment contrast cannot be assessed from the reported p-value alone. A p-value does not show the range of plausible effect sizes, and it does not explain why the sponsor’s analysis differs from the simple recalculation.

In general, an interval’s width reflects both variability and the amount of information available, along with the estimation method. The arm standard deviations show substantial dispersion relative to the difference between the posted means. The recalculated interval is therefore important context for the point estimate, but it remains an unadjusted calculation. The confidence level for these recalculated intervals is not reported in the registry record. The sponsor’s analysis should guide interpretation because the posted discrepancy indicates that the simple calculation does not reproduce the sponsor’s result.

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 for comparisons of continuous means, not total events. They are illustrations based on the observed unadjusted effect, not the original trial design assumptions. 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 set to the stated fraction of the observed effect. The trial’s own original sample-size justification and design assumptions are not reported in the registry record.

The 80% power column describes the planned number of participants per arm under which a study would have an 80% chance of detecting the specified effect if that effect and the planning assumptions were correct. The 90% power column uses a higher chance of detection, and therefore requires more participants per arm under the same assumptions. These are planning calculations; this is not observed power.

The rows for 100%, 75% and 50% of the observed effect illustrate how planning changes when the target effect is smaller than the effect estimated in this trial. Smaller effects require more participants to distinguish them from random variation at the same significance level and power target. The observed effect is itself uncertain, so these illustrations should not be read as a definitive sample-size recommendation for a future study. A future design would need a justified target effect and assumptions appropriate to its setting and analysis.

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 = -3.800; pooled standard deviation σ = 21.219.
2. n = 2 × (1.96 + 0.842)² × 21.219² / -3.800² = 489.5, rounded up to 490 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 <- -3.80000000000001                      # observed difference in means
sigma <- 21.2187647939611                     # 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(1763203)
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 = 490 (80% power), 656 (90% power)
 75% of observed effect: n per arm = 871 (80% power), 1,165 (90% power)
 50% of observed effect: n per arm = 1,958 (80% power), 2,621 (90% power)
power.t.test: 490.4 
Simulated power at n = 490 per arm: 0.796 (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 sponsor’s primary analysis is reported as a mixed-model analysis with p = 0.73. The simple recalculation from the posted 12-month arm means gives a mean difference of -3.8 mm Hg, an interval ending just below zero, and p = 0.048. The discrepancy flag indicates that these results do not agree. The sponsor’s own analysis takes precedence; the recalculation is an unadjusted comparison of posted summary statistics and cannot replace the reported model result.

The registry does not report the mixed-model estimate, its interval, or enough model details to reproduce the analysis. It is therefore not possible to identify from the registry record why the p-values differ. Potential analytic differences should not be guessed. The reported p-value is evidence about the sponsor’s stated model, while the recalculation is a separate descriptive check based on the arm summaries.

For interpretation, the reported result should be cited as the sponsor’s analysis and the recalculated result labeled as unadjusted. Neither the arm means nor the recalculated p-value should be represented as if they were the sponsor’s mixed-model estimate. The unresolved discrepancy limits how confidently the numerical summaries can be reconciled.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists 20 outcome measures and 19 analyses, including a primary endpoint and multiple secondary outcomes. Testing several outcomes can increase the chance of at least one apparently unusual result even when there is no underlying group difference. In general, multiplicity procedures address that problem by controlling how evidence is interpreted across a family of tests. Whether SUCCEED used an adjustment, a prespecified outcome hierarchy, or another approach to multiplicity is not reported in the registry record.

This matters particularly for secondary outcomes. Their unadjusted comparisons should not be treated as if each were the sole planned test. The registry’s reported p-values and the recalculated unadjusted comparisons do not establish that multiplicity was addressed. The record also does not state whether the secondary analyses were prespecified or exploratory.

Missing-data handling is not reported in the registry record. Although arm-level summaries list 241.0 and 246.0 participants, those counts do not by themselves establish that every randomized participant had a 12-month measurement or entered every analysis. The record does not describe the amount or reasons for missing measurements, imputation, sensitivity analyses, or assumptions about missingness. The baseline and 3-month measurement schedule is described, but the primary outcome result is for 12 months; the registry does not explain how earlier measurements were used in the mixed model.

The analysis population is also not specified. The registry does not state whether the sponsor used all randomized participants, a subset with observed measurements, or another population definition. The mixed-model label does not establish intention-to-treat analysis or a particular missing-data method. These omissions limit assessment of robustness and make it important not to infer an analysis strategy from standard practice.

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.

The posted safety figures record 0 serious adverse events among 241 participants at risk in the Care Management+Community Health Worker group and 0 among 246 participants at risk in usual care. The fields for other adverse events likewise record 0 affected participants in each group. The registry provides no serious adverse event terms in the posted list.

Deaths are recorded separately: 4 in the care-management group and 7 in usual care. The registry figures do not state whether these deaths were related to the intervention, whether they were classified as adverse events, or when they occurred. They should not be interpreted as evidence of causation or as a comparison of mortality risk without fuller information.

Zero affected participants in the posted adverse-event fields does not prove that no safety events occurred in the trial. The registry record does not explain the collection period, definitions, ascertainment process, or completeness of safety reporting. These counts describe what is posted and do not establish comparative safety.

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

The registry reports a randomized comparison of care management with a community health worker against usual care, with systolic blood pressure at 12 months as the primary outcome. The posted means are 133.1 mm Hg and 136.9 mm Hg, respectively. A simple unadjusted calculation gives a mean difference of -3.8 mm Hg, but the sponsor’s mixed-model analysis reports p = 0.73. Because the two analyses disagree, the sponsor’s result takes precedence and the summary-statistic calculation should be regarded as a separate check, not the definitive analysis.

The record does not establish that the groups had identical outcomes, nor does it establish a reliable primary effect based on the unadjusted recalculation. The sponsor’s p-value alone does not quantify the size or precision of its estimated difference, because the corresponding estimate and interval are not reported. The discrepancy and the missing model details leave important uncertainty about how the primary result relates to the displayed arm summaries.

The record also does not establish what happened for every secondary outcome, how missing observations were handled, or whether secondary testing was adjusted for multiplicity. The posted summaries concern group averages and do not describe individual responses. The trial’s stated purpose included measuring intervention costs, but cost findings are not reported in the outcome information.

Statistical findings from this trial do not, on their own, establish the practical value of the intervention or support an individual treatment decision. They describe the posted results and their limitations. Understanding the result fully would require the sponsor’s model estimate and interval, analysis specifications, missing-data information, and fuller reporting of the outcome and cost analyses.

18. Why This Trial Matters Statistically

SUCCEED is statistically informative because its registry record brings together a randomized comparison, a clinically relevant continuous primary outcome, and measurements planned across follow-up. The design makes the intended comparison clear: care management with a community health worker versus usual care for secondary stroke prevention in the Los Angeles County safety net. The registry also identifies a broader set of risk-factor and behavior outcomes, illustrating that trial interpretation extends beyond a single endpoint.

The most important statistical feature is the mismatch between the simple calculation from the posted 12-month means and the sponsor’s mixed-model result. This is a practical reminder that arm summaries do not necessarily reproduce a model-based analysis, particularly when the model’s details and the use of repeated measurements are not reported. It also shows why estimates, intervals, p-values and analysis methods should be read together rather than treating one value as a complete description.

The record highlights the value of reporting model estimates and confidence intervals alongside p-values. Without these, readers cannot judge the magnitude and precision of the sponsor’s primary comparison. Information about missing data, analysis populations and multiplicity would also help distinguish planned confirmatory analyses from exploratory findings. The planning illustrations demonstrate how uncertainty about effect size affects sample-size needs, but they are not a substitute for the trial’s original design rationale.

Overall, the statistical importance of the trial lies not only in its reported group summaries, but also in the need to reconcile those summaries with the sponsor’s analysis and to interpret multiple outcomes cautiously.

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.

Several trial-specific reporting gaps limit interpretation. The registry gives a sponsor mixed-model p-value for the primary outcome but does not report the model-based estimate or interval. It also does not describe model covariates, repeated-measure structure, analysis population, missing-data method, or sensitivity analyses. The substantial difference between the sponsor’s p-value and the simple unadjusted recalculation cannot be explained from the available record.

The primary endpoint is reported at 12 months, while measurement at baseline and 3 months is described without corresponding primary outcome summaries in the data. The record does not explain whether those earlier measurements entered the sponsor’s model. The original sample-size justification and design assumptions are not reported; the available planning figures are illustrations based on the observed effect rather than the trial’s own design.

Multiplicity procedures and the status of secondary analyses are not reported. The registry lists many outcomes, but numerical results are not available for every measure in the record. The sponsor’s analyses are also incomplete across the secondary outcomes with numerical summaries. These gaps make broad conclusions from secondary findings especially uncertain.

Finally, eligibility criteria, participant characteristics, recruitment details and the identity of masked parties are not reported in the registry record. The safety entries do not provide event terms or context for the deaths, and cost results are not reported. These limitations restrict assessment of generalizability, safety reporting and the study’s stated economic objective.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: The SUCCEED Trial of Secondary Stroke Prevention — NCT01763203.
• Registry record last updated 2019-11-26; status last verified 2019-11.
• Results first posted 2019-11-26.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

23. Record Summary

The registry record describes SUCCEED as a completed, randomized, parallel-group, single-masked trial enrolling 487 participants. It compared Care Management+Community Health Worker with usual care in a study of secondary stroke prevention among stroke patients in the Los Angeles County safety net.

The primary outcome was mean systolic blood pressure in mm Hg at 12 months. The sponsor’s mixed-model analysis reports p = 0.73. A simple unadjusted recalculation from the posted group summaries gives a mean difference of -3.8 mm Hg and p = 0.048; because these results conflict, the sponsor’s analysis takes precedence. The model estimate and interval, detailed analysis methods, missing-data handling, multiplicity procedures and original sample-size justification are not reported in the registry record.

Secondary outcomes cover additional risk factors and behaviors, but their results and analyses are not complete in the posted record. Serious adverse-event fields record 0 affected participants in each group, while deaths are recorded as 4 and 7; the registry does not provide further context to interpret those figures.