Advancing People of Color in Clinical Trials Now: Involvement in Trials Using a Patient-Centered Website: A Community-Engaged Approach: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Advancing People of Color in Clinical Trials Now: Involvement in Trials Using a Patient-Centered Website: A Community-Engaged Approach (NCT03243071), a randomized na study. The primary outcome is ACTNOW! Clinical Trial Knowledge Assessment Score.

2. Clinical Question

The study asks whether access to a culturally and linguistically tailored website about clinical trials is associated with participants’ knowledge and intended engagement with clinical research, compared with access to NYU’s standard trial participation website. The registry summary describes willingness to enroll, behavioral intent, likelihood of referring others, and health literacy as patient-centered outcomes assessed around exposure to the website. The primary outcome posted in the results is a clinical-trial knowledge score measured at baseline.

The registry summary also describes intended outcomes before and after exposure, but the posted primary endpoint is specifically the ACTNOW! Clinical Trial Knowledge Assessment Score at baseline. The reported mean comparisons therefore need to be read according to that endpoint and timeframe, rather than as evidence of a post-exposure difference in willingness, behavior, referral, or knowledge.

3. Trial Design

The study is registered as randomized, with parallel groups and no masking. Its stated purpose is to evaluate the comparative effectiveness of a culturally and linguistically tailored clinical-trial literacy website. The intervention group had access to the tailored website, while the control group had access to NYU’s standard trial participation website. Both interventions are classified as behavioral.

The registry lists enrollment of 112 and describes the study as completed. The registered start date is 2017-08-14, and the primary completion date is 2019-01-28. The phase is listed as NA, and recruitment is the listed condition. The sponsor is NYU Langone Health.

The brief summary describes the intervention and control groups as n=50 each. The primary outcome results, however, present four arm entries: two labeled as intervention pairs and two labeled as control pairs, with different analysis counts. The record explains that the analyzed numbers differ from participant-flow numbers because enrollment allowed for attrition and analyses used participants with complete data. It does not explain why the results are organized into these four pair entries, so they should not be treated as a single set of conventional two-arm summaries without qualification.

The primary outcome is the ACTNOW! Clinical Trial Knowledge Assessment Score, recorded at baseline. The registry summary discusses additional intended outcomes and measurement before and after website exposure, but the posted primary result is the baseline score. The registry record does not report a detailed statistical analysis plan or the original design assumptions for the trial. The planning illustrations associated with the posted contrasts are separate from the trial’s original design and use stated assumptions described with those illustrations.

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: The discrepancy in overall number of participants analyzed with numbers provided in the participant flow module lies with over enrollment of study subjects to account for attrition loss and analysis only being done for those participants who had complete data. The two arms (intervention and control groups) are further divided into pairs for statistical analysis purposes (paired t-tests).

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 ACTNOW! Clinical Trial Knowledge Assessment Score, measured over Baseline.

The primary endpoint is the ACTNOW! Clinical Trial Knowledge Assessment Score, a continuous measure of knowledge about clinical trials. The assessment contains 26 statements, and its total score ranges from 0-104. Higher scores indicate a higher level of knowledge. The registry reports the endpoint as a mean with standard deviation, which describes the average score and the spread of individual scores within each reported arm entry.

A knowledge assessment can quantify what participants know about clinical trials, providing an outcome distinct from whether they say they are willing to enroll or intend to refer another person. The registry summary describes those intentions as patient-centered outcomes, but the posted primary endpoint is knowledge rather than actual enrollment behavior. The score can therefore describe knowledge at the specified measurement time; it cannot on its own establish whether participants later enroll in trials or refer others.

The registered timeframe for the primary score is baseline. That timing is essential to interpretation: the posted means describe scores before the intended website exposure, not a measured post-exposure improvement. Although the registry summary discusses health literacy measured before and after tailored messages, the primary score result shown is specifically baseline. The registry record does not report a post-exposure primary score in these results.

7. Results Posted in the Registry

• Intervention Group (Pair 1): mean 93.89 versus 94.88 on control; difference -0.99 (95% CI -5.36 to 3.38), standardized effect d = -0.09; z-test p = 0.657.
• Intervention Group (Pair 2): mean 93.58 versus 94.88 on control; difference -1.30 (95% CI -5.97 to 3.37), standardized effect d = -0.12; z-test p = 0.586.
• Control Group (Pair 4): mean 97.88 versus 94.88 on control; difference 3.00 (95% CI -2.57 to 8.57), standardized effect d = 0.26; z-test p = 0.291.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is continuous, so its summaries are expressed as means and standard deviations. A mean gives the arithmetic average of scores in a group. A standard deviation describes how dispersed those scores are around the average. These measures summarize the observed score distribution but do not show every participant’s score or the shape of the distribution.

What does a mean difference represent?

A mean difference compares the average score in one arm entry with the average score in its named reference entry. The reported comparisons use the control group labeled Pair 3 as the reference for both intervention entries. The negative estimates for those two contrasts indicate that the intervention-entry means are lower than the Pair 3 control mean in the posted baseline summaries. The positive estimate comparing Control Group (Pair 4) with Control Group (Pair 3) indicates a higher mean for the former entry. These directions describe the posted baseline data, not a post-website change.

The reported standardized effect, d, expresses a mean difference in standard-deviation units. This can help describe the size of a difference relative to the variability of scores, whereas the mean difference remains in score units. The posted d values are -0.090 for Intervention Group (Pair 1) versus Control Group (Pair 3), -0.118 for Intervention Group (Pair 2) versus Control Group (Pair 3), and 0.257 for Control Group (Pair 4) versus Control Group (Pair 3). These are standardized descriptions of the same baseline contrasts; they do not change the endpoint’s timing or establish a causal effect.

What does a confidence interval add?

A confidence interval gives a range of values compatible with the observed data under the method used to calculate it. It expresses uncertainty around an estimate, rather than giving a guaranteed range for the true value. For the first intervention contrast, the estimate is -0.99 and the interval runs from -5.355 to 3.375. For the second, the estimate is -1.3 and the interval runs from -5.974 to 3.374. For the control-entry contrast, the estimate is 3.0 and the interval runs from -2.573 to 8.573. Each interval spans zero, so the estimates do not rule out no mean difference, nor do they exclude differences in either direction.

How should the p-values be read?

The posted p-values are unadjusted: 0.657, 0.586, and 0.291 for the respective comparisons. A p-value measures how unusual data at least as discrepant as those observed would be under a specified null model. It is not the probability that the null hypothesis is true, and it does not measure the probability that the result is due to chance. The registry record does not report a multiplicity adjustment. The p-values also do not account for uncertainty about the comparison structure, the timing of the endpoint, or the completeness of the underlying data.

The comparisons are identified as recalculations from posted summary statistics and are unadjusted. The record does not report the full analytic procedure, including a detailed model specification or covariate adjustments. Random allocation is part of the registered design, but the posted baseline summaries alone do not show the allocation process or demonstrate how it was implemented.

Why can over-dispersion matter?

In general, over-dispersion means that observed variability is greater than a statistical model expects. When that occurs, a model that understates variability can produce uncertainty estimates that are too narrow. This endpoint is continuous and is summarized with standard deviations; the posted results do not report an over-dispersion assessment or a model-based analysis addressing it. The concept is therefore general context, not a finding about this study.

The planning illustrations use a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. They scale the observed effect to consider planning for 100%, 75%, or 50% of that effect. Those assumptions apply to the illustrations, not necessarily to the trial design. The registry record does not report the original sample-size justification or the trial’s own design assumptions.

Finally, analysis counts differ across the posted arm entries, and the record says that analyses used participants with complete data after allowing for attrition. Complete-case analysis can be affected if people with missing scores differ systematically from those included. The registry record does not report a missing-data method or enough information to assess the impact of missingness. These methodological limits matter when interpreting both estimates and uncertainty.

10. Confidence Intervals and Precision

The confidence intervals provide a direct view of the precision of the posted mean differences. For Intervention Group (Pair 1) versus Control Group (Pair 3), the estimate is -0.99 and the interval extends from -5.355 to 3.375. For Intervention Group (Pair 2) versus Control Group (Pair 3), the estimate is -1.3 and the interval extends from -5.974 to 3.374. For Control Group (Pair 4) versus Control Group (Pair 3), the estimate is 3.0 and the interval extends from -2.573 to 8.573.

Each interval includes zero. Thus, these results do not distinguish a zero mean difference from differences in either direction at the confidence level represented by the intervals. The intervals are also wide enough to include nonzero differences, so a non-significant p-value should not be translated into proof of equality. The estimates are compatible with a range of possible baseline contrasts.

Precision is influenced by the variability of scores and the amount of analyzed information. The posted standard deviations range from 9.69 to 11.8 across the arm entries, while the analysis counts differ across entries. The corresponding standard errors are reported as 1.686 and 1.456 for the first contrast’s reference and intervention entries, 1.686 and 1.687 for the second, and 1.686 and 2.29 for the control-entry contrast. These values help explain why uncertainty differs among comparisons; they do not resolve why the registry reports separate pairs.

Most importantly, all of these estimates concern the baseline score. Their precision does not turn them into estimates of website impact after exposure. The registry record does not report a post-exposure primary score in the results.

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 events. They are illustrations based on the observed effect, not the trial’s original design. The stated assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; each scenario uses the stated fraction of the observed effect. The registry record does not report the original design assumptions or sample-size justification.

The 80% power column indicates the number of participants per arm that the planning calculation estimates would be needed to detect the specified effect under those assumptions, with an 80% chance of rejecting a zero-difference null model if that effect were true. The 90% power column gives the corresponding planning target for a 90% chance under the same assumptions. These are prospective planning concepts, not measures of how likely the trial’s findings are to be correct.

For the comparisons of the intervention entries with Control Group (Pair 3), planning for 100% of the observed effect gives 1,920 or 1,125 participants per arm for 80% power, and 2,570 or 1,506 for 90% power, respectively. Planning for 75% of the observed effect gives 3,413 or 2,000 participants per arm for 80% power, and 4,569 or 2,678 for 90% power. Planning for 50% of the observed effect gives 7,678 or 4,500 participants per arm for 80% power, and 10,279 or 6,024 for 90% power. The control-entry comparison has its own figures: 239 participants per arm for 80% power and 319 for 90% when planning for 100% of its observed effect.

Smaller effects require more participants to detect under the same assumptions, which is why the planning targets increase when the assumed effect is reduced. These values are not observed power; they are planning illustrations that depend on the estimated effects and stated calculation assumptions.

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 = -0.990; pooled standard deviation σ = 10.947.
2. n = 2 × (1.96 + 0.842)² × 10.947² / -0.990² = 1,919.4, rounded up to 1,920 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 <- -0.989999999999995                      # observed difference in means
sigma <- 10.9472324298126                     # 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(3243071)
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 = 1,920 (80% power), 2,570 (90% power)
 75% of observed effect: n per arm = 3,413 (80% power), 4,569 (90% power)
 50% of observed effect: n per arm = 7,678 (80% power), 10,279 (90% power)
power.t.test: 1920.4 
Simulated power at n = 1,920 per arm: 0.799 (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 lists 6 outcome measures, while the posted secondary results count is 0. The analysis listing also reports 0 analyses. The primary result described is the ACTNOW! Clinical Trial Knowledge Assessment Score at baseline. The summary discusses additional outcomes, including willingness to enroll, behavioral intent, likelihood of referral, and health literacy, but the registry results do not include posted estimates for those outcomes.

When investigators examine several outcomes or make several comparisons, the chance of obtaining at least one apparently notable result can increase. Statistical methods can address this multiplicity, for example by adjusting a decision threshold or defining a primary outcome in advance. The registry record does not report whether such a procedure was planned or used. The listed p-values are explicitly unadjusted, so they should not be treated as multiplicity-adjusted evidence.

The primary outcome record says that participant counts in analyses differ from participant-flow counts because enrollment allowed for attrition and only participants with complete data were analyzed. This indicates a complete-data analysis, but the registry record does not report how missing scores were handled beyond that statement. It does not report imputation, sensitivity analyses, or a comparison of participants with and without complete data. Missing-data assumptions and methods are therefore not reported in the registry record.

The registry identifies randomized allocation, but the reported results do not specify an intention-to-treat analysis population or another formal analysis population. The displayed analysis counts vary among the four arm entries. The record does not provide a participant-level flow explanation for each pair or clarify how those entries map onto the study’s two described groups. These details limit assessment of whether all randomized participants were retained in the analysis and how exclusions could affect the comparisons.

The registry record also does not report a prespecified statistical analysis plan, interim analyses, or methods for adjusting the comparisons for baseline covariates. Consequently, the p-values and intervals should be understood as unadjusted summaries, not as a complete account of the trial’s inferential procedures.

14. 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 serious adverse event summaries list 59 participants at risk in the intervention group and 53 at risk in the control group. Serious adverse events are recorded as 0 in each group. Other adverse events are also recorded as 0 in both groups, and deaths are recorded as 0 in each group.

These are reported event counts alongside the registry’s at-risk counts; they are not estimates of comparative risk. The registry record does not provide event descriptions, severity details beyond the serious-event classification, or additional information about how events were collected. The at-risk counts differ from the analysis counts shown for the knowledge score, so the safety and outcome summaries should not be assumed to use identical analysis populations.

A zero reported count means no event was entered in that category for the displayed group summary. It does not establish that risk is absent, and it does not show that the websites are equivalent in safety. The summaries are limited to the categories and counts recorded in the registry.

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

The posted data describe baseline clinical-trial knowledge scores among the arm entries. Their means are close in the two intervention entries and the first control entry, while the second control entry has a higher reported mean. The unadjusted mean-difference intervals for the three posted contrasts all include zero. These data therefore do not provide a precise indication of a nonzero baseline difference for any of the listed comparisons.

These findings do not answer whether the culturally tailored website changed knowledge after exposure. The primary endpoint timeframe is baseline, and the posted comparisons are baseline summaries. They also do not establish effects on willingness to enroll, behavioral intent, actual trial participation, or referring others. Those outcomes are discussed in the registry summary, but corresponding results are not posted in the results.

The comparisons should also be interpreted according to their labels. Two contrasts compare an intervention pair with Control Group (Pair 3). The third compares Control Group (Pair 4) with Control Group (Pair 3), so it is not an intervention-versus-control comparison. The record does not explain why results appear as these separate pair entries, which limits a simple interpretation as one overall two-arm comparison.

Non-significant p-values do not prove there is no difference, and intervals containing zero do not prove equivalence. They indicate that the posted data are compatible with a range of possible baseline differences, including zero. Nor do the planning figures establish that the trial was adequately or inadequately powered: they are planning illustrations calculated from observed effects, not observed power or the trial’s original sample-size design.

The registry lists the study as completed and reports no serious or other adverse events in the displayed safety summaries. Those records describe the event counts entered; they do not establish absence of risk. Overall, the results are best read as limited descriptive and unadjusted baseline comparisons, with important post-exposure and analysis details not reported in the registry record.

16. Why This Trial Matters Statistically

This trial raises a statistical question that is important in community-engaged recruitment research: how should a literacy intervention be evaluated when the goal includes both knowledge and intended participation? Those are related but distinct constructs. A knowledge score can quantify responses to clinical-trial statements, while willingness, behavioral intent, referral, and actual enrollment require their own definitions and outcome measurements.

The registry’s primary endpoint is measured at baseline, so timing governs what the reported comparison can say. Baseline scores can describe the groups before the intended website exposure, but they cannot quantify change caused by that exposure. A post-exposure contrast or a change-from-baseline analysis would address a different statistical question. The results described in the registry do not report such a primary outcome estimate.

The record also illustrates why confidence intervals deserve attention alongside p-values. The intervals span zero and allow differences in both directions. A result that does not meet a conventional significance threshold is not proof that the groups are alike; the range of compatible effects and the precision of the estimate matter. The posted counts and variability help contextualize that uncertainty, but the split pair structure makes the overall comparison difficult to summarize as a single two-arm result.

Finally, transparency about analysis populations, missing data, multiplicity, and original design assumptions is part of interpreting a randomized study. Here, the registry notes complete-data analyses and attrition but does not report a fuller missing-data strategy or a detailed statistical plan. That distinction matters when judging how far the posted results can support conclusions about the intended outcomes.

17. 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 main trial-specific limitation is that the primary outcome is listed at baseline. The registered purpose and summary concern exposure to websites and outcomes before and after exposure, but the posted primary score summaries do not show a post-exposure result or a change over time. The available primary outcome comparisons therefore cannot answer whether knowledge changed following website access.

The result structure is also difficult to interpret as a conventional two-group comparison. Results are separated into two intervention pair entries and two control pair entries, and the reported comparisons use Control Group (Pair 3) as the reference for both intervention entries, plus a comparison between two control entries. The registry record does not explain this structure or provide an overall intervention-versus-control estimate. The analysis counts differ across entries, and the record gives only a general explanation involving attrition and complete data.

Further limitations are that the displayed contrasts are unadjusted, the p-values are unadjusted, and multiplicity procedures are not reported in the registry record. A full statistical analysis plan, details on the randomized analysis population, and a missing-data method are also not reported in the registry record. The record does not provide a trial-specific sample-size justification; the available planning calculations are based on observed effects and are not the original design rationale.

The registry summary discusses several outcomes, but secondary results are not posted. This prevents assessment of the intended outcomes beyond the baseline knowledge score. The safety summaries record zero serious and other adverse events, but provide limited detail about event collection and do not establish comparative safety.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Advancing People of Color in Clinical Trials Now: Involvement in Trials Using a Patient-Centered Website: A Community-Engaged Approach — NCT03243071.
• Registry record last updated 2022-02-01; status last verified 2022-01.
• Results first posted 2022-02-01.
• 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.

21. Record Summary

NCT03243071 is a completed, randomized, unmasked, parallel-group behavioral study sponsored by NYU Langone Health. It compares access to a culturally tailored clinical-trial literacy website with access to NYU’s standard trial participation website. Enrollment is listed as 112; the registered start date is 2017-08-14 and the primary completion date is 2019-01-28.

The primary endpoint is the baseline ACTNOW! Clinical Trial Knowledge Assessment Score, a mean score ranging from 0-104, with higher scores indicating greater knowledge. Results are reported in four arm entries and include unadjusted mean contrasts with confidence intervals and p-values. The registry describes complete-data analyses after allowing for attrition, but does not report a detailed analysis plan, missing-data strategy, multiplicity procedure, or original sample-size assumptions.