Stroke Ready: A Stroke Preparedness Brief Intervention: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Stroke Ready: A Stroke Preparedness Brief Intervention (NCT03831451), a randomized na study. The primary outcome is Change in Behavioral Intent to Call 911.

2. Clinical Question

3. Trial Design

The registry describes Stroke Ready as a randomized controlled behavioral intervention trial with a crossover model and no masking. Its purpose was to assess a brief intervention intended to increase stroke preparedness. The two named interventions were the Stroke Preparedness Intervention and the Healthy Lifestyle Intervention. The record lists the study as completed, with a start date of 2019-03-04 and a primary completion date of 2019-04-23. The sponsor was the University of Michigan.

The Stroke Preparedness Intervention was delivered in person by a research team member and focused on recognizing stroke symptoms and the importance of calling 911. The Healthy Lifestyle Intervention was also delivered in person by a research team member and focused on stroke risk reduction, using an educational intervention based on patient materials from the American Heart Association. The preparedness intervention was designed with input from the community where it was tested.

The registry reports enrollment of 129. It lists an arm sample size of 74 for the Stroke Preparedness Intervention and 55 for the Healthy Lifestyle Intervention. These are the posted arm counts; the record does not explain how those counts relate to crossover periods or analysis populations. Details such as randomization procedures, intervention sequence, any washout period, and the timing between crossover interventions are not reported in the registry record. Phase is listed as not applicable. The record describes one outcome measure and does not list a sponsor-posted statistical analysis.

4. Participant Flow and Accounting

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

Overall Study

Analysis population for the primary outcome, as defined in the registry: Mean changes from pre-intervention to post-intervention

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 Change in Behavioral Intent to Call 911, measured over Immediately (up to 10 minutes) before the brief intervention and then immediately (up to 10 minutes) after completing the brief intervention.

The primary endpoint is titled “Change in Behavioral Intent to Call 911.” It measures change in a self-administered survey score from before to after the brief intervention. The Stroke Action test was modified based on community input. Scores range from 0-12 for stroke; a higher score means greater behavioral intent to call 911 for a stroke, while a lower score means less intent. The outcome is expressed in units on a scale, and the posted summary is the mean change from pre-intervention to post-intervention.

The endpoint was assessed immediately before the intervention and immediately after completing it, with the post-intervention timing described as up to 10 minutes after completion. An intent measure captures a stated behavioral response rather than directly documenting what a person would do during an actual stroke. That distinction matters when interpreting the result: the endpoint concerns reported intent on a survey, not observed emergency behavior or a clinical outcome.

Using a continuous score retains variation in the reported degree of intent rather than reducing responses to a yes-or-no category. A pre-to-post change focuses on how scores shift over the brief intervention period. The registry record does not report additional details about score validation, the handling of incomplete surveys, or a threshold for interpreting a score change as meaningful.

7. Results Posted in the Registry

• Healthy Lifestyle Intervention: mean 0.40 versus 2.10 on control; difference -1.70 (95% CI -8.53 to 5.13), standardized effect d = -0.08; z-test p = 0.625.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary endpoint is continuous: it is a mean change in score rather than a count of events or a time-to-event outcome. A mean summarizes the central level of a measure across observations. A pre-to-post change describes how the score differs between the assessment before the intervention and the assessment after it. The posted arm summaries are mean changes, accompanied by standard errors.

What does the standard error describe?

A standard error describes the uncertainty in an estimated mean under the assumptions of the method used to calculate it. It is not the same as the spread of individual participants’ scores. A smaller standard error generally indicates a more precisely estimated mean, while a larger one indicates greater uncertainty. The registry reports standard errors of 2.7 for the Stroke Preparedness Intervention and 2.2 for the Healthy Lifestyle Intervention. It does not report participant-level score distributions or the method used to calculate those standard errors.

How should the mean difference be read?

The posted contrast is Healthy Lifestyle Intervention minus Stroke Preparedness Intervention. Its estimate is -1.7 units. With this direction, a negative estimate means the posted mean change is lower in the Healthy Lifestyle Intervention group than in the Stroke Preparedness Intervention group. The contrast is an unadjusted recalculation from posted summary statistics. The registry record lists no sponsor-posted analysis, so it does not establish whether the sponsor used a different estimator or model.

In a randomized parallel-group comparison, an unadjusted difference between group means can summarize the contrast in outcomes. A crossover design can require additional care because participants may contribute data under more than one intervention. A paired analysis can account for observations from the same participant, while a model may also account for period or sequence effects. These are general statistical considerations, not descriptions of methods known to have been used in this trial. The record does not report how the crossover observations were analyzed or whether the posted comparison accounts for pairing, order, or period.

What do the interval and p value contribute?

An interval estimate conveys uncertainty around an estimated contrast by identifying a range of values compatible with the data and the method used to calculate the interval. It is more informative than a point estimate alone because it shows how much precision the estimate has. The posted interval extends from -8.526 to 5.126. The confidence level and calculation method are not reported in the registry record, so the interval should not be assigned a confidence percentage that the record does not state.

The unadjusted p value is 0.625. In general, a p value describes how unusual data at least as incompatible with a specified null model would be if that model and its assumptions were true. It does not measure the size or importance of an effect, and it does not give the probability that the null hypothesis is true. The registry record does not report a sponsor analysis plan or a decision threshold for the primary analysis. The listed p value should therefore be interpreted as a descriptive, unadjusted recalculation rather than a confirmed sponsor-reported test.

Why can crossover details matter?

In a crossover study, outcomes may depend not only on which intervention a participant receives but also on the order and timing of interventions. General concerns include within-participant correlation, period effects, and possible carryover from an earlier intervention. Appropriate methods depend on the actual design and data structure. The registry record identifies the model as crossover but does not report sequence, washout, period-specific results, carryover assessment, or an analysis model. These missing details limit the ability to assess whether the posted arm summaries capture the design’s dependencies.

The outcome is based on a modified survey scale, and the posted summaries concern mean changes. The record does not report how missing pre-intervention or post-intervention responses were handled, whether analyses followed an intention-to-treat principle, or which participants were included in the primary analysis. These are important components of a statistical analysis because they affect the data contributing to the estimate. Without them, the posted summary supports a description of reported values but not a full reconstruction of the analysis.

10. Confidence Intervals and Precision

The posted contrast is -1.7 units, with an interval from -8.526 to 5.126. The width of this interval indicates that the estimate is not highly precise: values spanning both negative and positive differences are compatible with the reported interval. The interval includes the possibility of a contrast close to zero as well as contrasts in either direction. The registry record does not report the confidence level or the method used to calculate the interval, so its coverage properties cannot be specified from the record.

The arm-level standard errors are 2.7 for the Stroke Preparedness Intervention and 2.2 for the Healthy Lifestyle Intervention. These quantify uncertainty around the posted mean changes, not the spread of individual responses. The sample counts listed for the arms are 74 and 55, respectively. The record does not explain whether those counts represent the analysis population for the crossover comparison or how repeated observations were handled. That uncertainty in the analysis population and design-specific method is relevant to interpreting precision.

Precision is distinct from the direction or practical importance of an estimate. A relatively broad interval does not establish that the interventions have the same effect; it indicates that the posted data and calculation leave substantial uncertainty about the contrast. The endpoint is a survey measure of stated intent assessed immediately before and after the intervention, and the registry does not report a threshold for a meaningful score change. Thus, statistical uncertainty can be described, but the record does not define how much change would be important in practice.

11. 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 event counts. They are illustrations based on the observed effect, not the original trial design. The stated assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each planning scenario uses a stated fraction of the observed effect as the effect to detect.

The 80% power column gives the estimated participants per arm for a future study to have 80% power under those assumptions and for the effect used in that scenario. The 90% power column gives the corresponding per-arm planning estimate for 90% power. Power is the probability of detecting an effect of the specified size under the assumed model and significance level, if that effect is present. These are planning illustrations, not observed power for the completed trial.

For 100% of the observed effect, the estimates are 2,300 participants per arm for 80% power and 3,078 participants per arm for 90% power. For 75% of the observed effect, they are 4,088 and 5,472 participants per arm. For 50% of the observed effect, they are 9,197 and 12,312 participants per arm. The pattern reflects a general planning principle: smaller effects are harder to distinguish from random variation and require more participants to reach a chosen power target. These estimates depend on the stated assumptions and should not be read as a recommendation or as the trial’s original sample-size justification; that original justification is not reported in the registry record.

12. 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 = -1.700; pooled standard deviation σ = 20.574.
2. n = 2 × (1.96 + 0.842)² × 20.574² / -1.700² = 2,299.1, rounded up to 2,300 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 <- -1.7                      # observed difference in means
sigma <- 20.5735439013656                     # 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(3831451)
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 = 2,300 (80% power), 3,078 (90% power)
 75% of observed effect: n per arm = 4,088 (80% power), 5,472 (90% power)
 50% of observed effect: n per arm = 9,197 (80% power), 12,312 (90% power)
power.t.test: 2300.1 
Simulated power at n = 2,300 per arm: 0.802 (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.

13. Multiplicity, Missing Data and Analysis Populations

The registry record lists one outcome measure and no posted secondary outcomes. It does not report whether the investigators assessed additional outcomes that were not posted, or whether multiple time points, subgroups, or alternative analyses were considered. A multiplicity plan, including any adjustment for testing multiple hypotheses, is not reported in the registry record. Without that information, it is not possible to assess whether the posted result was selected from a wider set of analyses or how any such selection was handled.

The record lists no sponsor-posted statistical analyses. It does not report a prespecified analysis plan, a hypothesis-testing hierarchy, or a primary analysis population. The p value and interval accompanying the comparison are unadjusted recalculations from the posted summary statistics, rather than a sponsor-reported analysis. The record does not describe whether the analysis used all randomized participants, participants who completed both interventions, or another population.

Missing-data handling is also not reported in the registry record. There is no description of incomplete surveys, withdrawals, exclusions, imputation, or sensitivity analyses. For a pre-to-post measure, missingness at either assessment can affect which changes are available to summarize. In a crossover design, missing observations in one intervention period may also affect the comparison structure. These are general considerations; the record does not establish that missing responses occurred or how they were treated.

Other details relevant to interpretation—including randomization procedures, intervention sequence, crossover period handling, and whether within-participant dependence was modeled—are not reported in the registry record. These gaps do not alter the posted summary values, but they prevent a full appraisal of the analysis choices and their potential influence.

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

The posted results describe immediate change in a self-reported measure of intent to call 911 for stroke. The mean change was 2.1 units for the Stroke Preparedness Intervention and 0.4 units for the Healthy Lifestyle Intervention. The unadjusted recalculated contrast, expressed as Healthy Lifestyle Intervention minus Stroke Preparedness Intervention, was -1.7 units, with an interval from -8.526 to 5.126 and an unadjusted p value of 0.625.

These results do not establish that the interventions are equivalent, nor do they establish that one is superior. The interval includes values in both directions and is broad, leaving uncertainty about the size and direction of the contrast. The registry record does not report a threshold for a meaningful difference on the scale, so statistical estimates cannot be translated into a judgment about practical importance from this record alone.

The endpoint is stated intent measured on a modified self-administered survey. It is not a record of calls made during a stroke, observed recognition of symptoms, or a clinical outcome. The assessment was scheduled immediately before and after the intervention, with the post-intervention measure taken immediately after completion, up to 10 minutes later. The registry record does not report longer-term follow-up, so it cannot show whether any change in stated intent lasted.

The registry identifies the study as randomized and crossover, but does not report the crossover sequence, analysis model, or handling of within-participant observations. The posted comparison is an unadjusted recalculation, and no sponsor statistical analysis is listed. These limitations mean the estimates should be interpreted as a statistical description of the posted summaries, rather than as a complete account of the trial’s analysis or as a basis for broader claims about real-world preparedness.

15. Why This Trial Matters Statistically

This trial illustrates a statistical distinction that is important in behavioral research: a change in a survey measure is not the same as a change in observed behavior. The primary endpoint records stated intent to call 911 for stroke. It is measured on a scale and summarized as mean pre-to-post change. That makes the outcome suitable for describing shifts in reported intent, but it does not directly measure what participants would do in an emergency.

The crossover design also highlights why design details matter to interpretation. If participants contribute observations under more than one intervention, those observations may be related. Statistical methods can account for that dependence, but the registry record does not say whether or how it was addressed. The posted unadjusted comparison is therefore informative as a summary of the reported arm statistics, while leaving important design-specific analytical questions unanswered.

Finally, the interval emphasizes that a point estimate alone does not convey how precisely a contrast has been estimated. The reported limits encompass a range of possible differences, while the registry provides no stated threshold for a meaningful score change. The planning illustrations further show, in general terms, that detecting smaller effects requires larger per-arm samples under the listed assumptions. Together, these features make the record useful for examining outcome definition, uncertainty, and design transparency without treating a short-term survey result as a direct measure of clinical or emergency behavior.

16. 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 registry record offers limited information for assessing who participated. It lists stroke as the condition and reports enrollment, but detailed eligibility criteria and participant characteristics are not reported. The arm counts are listed, but their relationship to crossover periods and the analysis population is not explained. Randomization procedures, allocation concealment, intervention order, and any washout period are also not reported.

The endpoint was a community-modified self-administered survey score. The record describes the score range and direction but does not report validation details, response distributions, or a threshold for a meaningful change. The immediate timing of assessment captures short-term reported intent; follow-up persistence is not reported. No secondary outcomes are posted, and the record does not report measures of actual emergency response behavior.

The registry lists no sponsor-posted statistical analyses. The calculation method for the interval and standard errors, the confidence level, the primary analysis population, and any adjustment for crossover structure are not reported. Missing-data procedures, protocol deviations, subgroup analyses, and sensitivity analyses are likewise not reported. These omissions limit the ability to determine how robust the posted estimates are to analysis choices.

The record lists zero serious adverse events, other adverse events, and deaths in each intervention group. These recorded counts do not, by themselves, establish that no unrecorded events occurred or characterize the completeness of adverse-event ascertainment. The registry does not report further detail on monitoring or event definitions.

17. Related Tutorials

18. Related Statistical Calculators

19. Sources

• ClinicalTrials.gov: Stroke Ready: A Stroke Preparedness Brief Intervention — NCT03831451.
• Registry record last updated 2019-12-13; status last verified 2019-12.
• Results first posted 2019-11-20.
• 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.

20. Record Summary

Stroke Ready: A Stroke Preparedness Brief Intervention is listed as a completed randomized behavioral trial with a crossover model and no masking. The University of Michigan is the sponsor. The registry reports enrollment of 129 and identifies stroke as the condition. The study dates listed are 2019-03-04 for start and 2019-04-23 for primary completion.

The primary outcome is mean change in self-reported intent to call 911 for stroke, measured immediately before and after an intervention. The posted mean changes were 2.1 units for the Stroke Preparedness Intervention and 0.4 units for the Healthy Lifestyle Intervention. The unadjusted recalculated difference was -1.7 units, with an interval from -8.526 to 5.126 and an unadjusted p value of 0.625. The registry record lists no sponsor statistical analysis or secondary outcomes. Details about the crossover analysis, missing data, and analysis population are not reported.