Community-Based Communication Approaches for Blood Donation in Ghana: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Community-Based Communication Approaches for Blood Donation in Ghana (NCT04783077), a randomized na study. The primary outcome is Donation Attempt.

2. Clinical Question

The registry identifies the study as randomized, parallel, and unmasked. It lists 144 participants as the enrollment. The primary outcome results list participants and counts of attempts, rather than a continuous measure. The planned final assessment was at 6-months and 2 weeks from randomization; the follow-up window was later extended by 34 days for all participants after an interruption in WhatsApp messaging. These design features frame interpretation: the findings concern attempts during a defined follow-up period in a pilot setting, not a definitive estimate of the effect of every communication approach.

3. Trial Design

This was a randomized pilot study with a parallel-group design and no masking. The registry lists the study as completed, with an enrollment of 144. It identifies the sponsor as the University of Minnesota and gives the study setting as the catchment areas of the Southern Zonal Blood Centre in Ghana. The study’s objectives included developing a docudrama and a WhatsApp intervention, assessing the feasibility of a larger randomized trial, and evaluating participant views of the proposed approaches and control.

The pilot combined qualitative and quantitative work. Its described components included qualitative development of the docudrama, key informant interviews to guide intervention development, a randomized pilot trial comparing WhatsApp communication with control, and a final mixed-methods assessment using a quantitative survey and focus groups. The registry also names Docudrama as an intervention. However, the posted primary outcome groups include FRD: Control, VNRBD: Control, FRD: WhatsApp, and VNRBD: WhatsApp; the Docudrama arm has a primary outcome count of zero participants. The reason for that entry and the relation between the listed intervention arms and the randomized outcome groups are not reported in the registry record.

The primary endpoint concerned a donation attempt at any time during follow-up. The planned final assessment was at 6-months and 2 weeks from randomization. The local protocol IRB approval lapsed between 25 April 2022 and 23 May 2022, during which participants could not receive WhatsApp messages. In response, the final follow-up window was extended by 34 days for all participants so participants in the WhatsApp arm could receive the planned messages. Further operational details, including the original allocation ratio and the exact procedures used to assign participants, 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.

Overall Study

Analysis population for the primary outcome, as defined in the registry: The analyses followed a modified intention-to-treat (mITT) approach. Individuals who consented, were randomized and not withdrawn at screening were analyzed in the group to which they were originally assigned. Missing data from the mITT analysis was handled using multiple imputation via chained equations and included stratum, group assignment and baselines variable that met a minimum Kendall's tau of 0.2 with the primary outcome. No data was collected at 6 months 34 days for the Docudrama arm.

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: Randomization was into WhatsApp and Control was stratified by donation type (FRD/VNRBD). A separate group of participants will be invited to a single viewing of a filmed docudrama.

6. Primary Endpoint

The registry lists the primary outcome as Donation Attempt, measured over 6 Months and 34 days.

The primary endpoint was a blood donation attempt at any time during follow-up. It records whether participants attempted to donate, rather than the amount of blood collected or whether an attempt resulted in a completed donation. The reported outcome is a count of participants with an event in each listed group.

This endpoint fits the study’s stated focus on communication approaches intended to encourage blood donation among first-time donors. In a pilot, recording an attempt can also help assess whether the outcome can be collected during follow-up and whether the intervention and trial procedures are workable. Those feasibility aims are part of the study description; the registry does not report a separate threshold for judging feasibility or acceptability.

The scheduled final assessment was at 6-months and 2 weeks from randomization. Because participants could not receive WhatsApp messaging during the lapse in local protocol IRB approval between 25 April 2022 and 23 May 2022, the final follow-up window was extended by 34 days for all participants. The endpoint description says this change was intended to allow participants in the WhatsApp arm to receive all planned messages. Interpretation therefore concerns donation attempts over the stated follow-up window, including that extension. The registry record does not report a separate endpoint for completed donation or provide further detail on how an attempt was verified.

7. Results Posted in the Registry

• VNRBD: Control: 6.7% (2/30) versus 7.4% (2/27) on control; risk difference -0.7 percentage points (95% CI -14.1 to 12.6), risk ratio 0.90 (0.14–5.96); unadjusted two-proportion z-test p = 0.913.
• FRD:WhatsApp: 7.1% (2/28) versus 7.4% (2/27) on control; risk difference -0.3 percentage points (95% CI -14.0 to 13.5), risk ratio 0.96 (0.15–6.37); unadjusted two-proportion z-test p = 0.970.
• VNRBD: WhatsApp: 22.6% (7/31) versus 7.4% (2/27) on control; risk difference 15.2 percentage points (95% CI -2.6 to 32.9), risk ratio 3.05 (0.69–13.45); unadjusted two-proportion z-test p = 0.111.

8. Statistical Interpretation

9. Statistical Methods Explained

The reported primary outcome is binary at the participant level: a participant either had a donation attempt during follow-up or did not. The posted group summaries give the number of participants with an event and the group size. For the unadjusted comparisons, the risk difference expresses the difference between the observed event proportions in a comparison group and FRD: Control. A positive value indicates a larger observed proportion in the comparison group; a negative value indicates a smaller one. These re-calculations use the posted summary counts and are not adjusted for other participant characteristics.

What does the risk difference add?

A risk difference keeps the comparison on the scale of event proportions, which can make the direction and size of a contrast easier to interpret in the context of the observed groups. Its confidence interval describes the range of differences compatible with the data under the calculation’s assumptions. When the interval includes zero, the data do not rule out no difference. Here, each recalculated interval includes zero. The small event counts also mean that a few participants’ outcomes can have substantial influence on the estimated proportions and on the width of the interval.

How does the sponsor analysis differ?

The sponsor’s reported method is logistic regression, with an odds ratio as the parameter. An odds ratio compares the odds of the outcome between groups; it is not a risk difference and should not be read as a percentage-point change. The reported analysis gives an odds ratio of 2.28, a 95% confidence interval from 0.61 to 8.50, and p = 0.23. The analysis lists four groups and does not include Docudrama. The registry record does not report the model’s covariates, reference-category details beyond the group listing, or other implementation details. The three pairwise risk-difference re-calculations and the sponsor’s regression result therefore answer related but not identical statistical questions.

What does a confidence interval communicate?

In general, a confidence interval is a measure of statistical uncertainty around an estimate, reflecting the data and method used to calculate it. A narrower interval indicates greater precision on that scale; a wider interval indicates less precision. It does not give the probability that the true effect lies within the particular calculated interval. For a risk difference, zero represents no difference. For an odds ratio, the no-association value is the value at which the odds are equal across groups. The sponsor interval extends on both sides of that null value, while each recalculated risk-difference interval extends on both sides of zero.

How should the p-values be read?

A p-value describes how unusual the data, or data more extreme, would be under a specified null model and the analysis assumptions. It is not the probability that the null hypothesis is true, and it does not measure the size or clinical importance of a difference. The recalculated p-values are unadjusted: 0.913, 0.97, and 0.111 for the three listed comparisons. The sponsor reports p = 0.23 for the logistic-regression analysis. Those values do not establish equivalence or prove that there is no association. With few events and broad intervals, the estimates remain imprecise.

What does the modified intention-to-treat description mean?

The registry says the outcome analyses followed a modified intention-to-treat approach. It defines the analyzed population as individuals who consented, were randomized, and were not withdrawn at screening, analyzed in the group to which they were originally assigned. It also says missing data were handled using multiple imputation via chained equations. In general, multiple imputation uses a model to create plausible values for missing observations and combines the resulting analyses to account for uncertainty from missingness. The registry record does not report the imputation model, variables included, number of imputed datasets, or diagnostic checks. The precise handling of missing outcome data therefore cannot be assessed from the registry description alone.

Randomization is intended to support comparisons between assigned groups, while an unmasked design means participants and study personnel were not blinded. The registry does not report details about allocation concealment, adherence, contamination between groups, or outcome-assessor masking. Those omissions matter when assessing potential bias, but they do not establish that bias occurred. As this was a pilot combining intervention development, trial feasibility work, and participant feedback, the registry’s analysis should be interpreted in that context rather than as a fully specified confirmatory analysis.

10. Confidence Intervals and Precision

The recalculated risk-difference intervals are broad relative to their point estimates. For VNRBD: Control versus FRD: Control, the estimate is -0.007 and the interval runs from -0.141 to 0.126. For FRD: WhatsApp versus FRD: Control, the estimate is -0.003 and the interval runs from -0.14 to 0.135. For VNRBD: WhatsApp versus FRD: Control, the estimate is 0.152 and the interval runs from -0.025 to 0.329. Each interval includes zero, so each is compatible with no difference as well as with differences in either direction. The interval for the VNRBD: WhatsApp contrast also permits effects larger than its point estimate.

The sponsor’s logistic-regression estimate is an odds ratio of 2.28, with a 95% confidence interval from 0.61 to 8.50. This interval is wide and spans the no-association value. It therefore does not pin down the magnitude or direction of the association with much precision. Because this is an odds-ratio estimate, it should not be interpreted as a risk difference or as a direct change in the proportion attempting donation.

In general, interval width reflects both the amount of information in the data and the variability expected under the analysis. Here, the outcome counts are small in the posted groups, which limits the information available for estimating differences. The intervals are not evidence that all values inside them are equally plausible, and they do not show the probability that a particular effect is true. They communicate that the posted estimates are uncertain. The registry does not report a prespecified precision target for the pilot or a minimum meaningful difference against which to judge these intervals.

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.

Each planning row is an illustration based on a stated fraction of the observed effect for one of the posted control comparisons. The values are participants per arm, not total participants, and the planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The figures are not the original trial’s design requirements: the registry record does not report the assumptions used to plan the trial itself.

The 80% power column gives the illustrative number of participants per arm for a design with an 80% chance of detecting the stated effect under those assumptions, if that effect and the model assumptions hold. The 90% power column gives the corresponding per-arm number for a design with a 90% chance. These are prospective planning quantities, not a measure of how much power the completed pilot had. This is not observed power.

Rows that plan for 75% or 50% of the observed effect assume smaller effects than the row based on 100% of the observed effect. Smaller effects are harder to distinguish from random variation, so they require more participants. This pattern is visible in the listed planning figures: for the VNRBD: WhatsApp comparison, for example, the 100% row gives 86 participants per arm for 80% power and 115 for 90% power; the 50% row gives 271 and 362, respectively. Other comparisons produce much larger planning figures, reflecting their different observed estimates. Because these calculations draw on pilot estimates with few events, they should be treated as illustrations rather than settled design recommendations. The original sample-size justification and any other original design assumptions are 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, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.074; treatment proportion p₁ = 0.067; average p̄ = 0.070.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.3617 = 0.7089.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.3617 = 0.3044.
4. n = (0.7089 + 0.3044)² / (-0.0074)² = 18,714.5, rounded up to 18,715 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
p0 <- 0.0740740740740741                          # control proportion
p1 <- 0.0666666666666667                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^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 (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(4783077)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 18,715 (80% power), 25,053 (90% power)
 75% of observed effect: n per arm = 33,676 (80% power), 45,082 (90% power)
 50% of observed effect: n per arm = 76,679 (80% power), 102,651 (90% power)
power.prop.test: 18714.5 
Simulated power at n = 18,715 per arm: 0.802 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-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. Analyses Reported by the Sponsor

The sponsor’s own reported result is a logistic-regression odds ratio of 2.28, with a 95% confidence interval from 0.61 to 8.50 and p = 0.23. It includes four groups: FRD: Control, VNRBD: Control, FRD: WhatsApp, and VNRBD: WhatsApp. The registry record does not report the regression model details, such as covariates or the precise contrast represented by the odds ratio.

The recalculated results instead make three unadjusted risk-difference contrasts, each using FRD: Control as the comparison group. They are based on the posted event counts and do not reproduce the sponsor’s regression model. The two approaches use different effect measures and may address different comparisons, so their estimates should not be expected to match numerically. The recalculations are supplementary descriptions of the posted summaries, not a replacement for the sponsor’s analysis. The record does not flag a discrepancy between the posted results and these re-calculations, so the sponsor’s analysis remains the reported trial analysis.

Both the sponsor’s interval and the recalculated intervals indicate substantial uncertainty. The sponsor’s p-value of 0.23 is not proof of no association, and the recalculated p-values do not establish that the groups are equivalent. Neither analysis resolves the uncertainty that follows from the small number of recorded attempts.

14. Multiplicity, Missing Data and Analysis Populations

The registry lists one primary outcome measure and one reported analysis. The sponsor’s logistic-regression analysis covers four groups. The additional unadjusted re-calculations compare three listed groups with FRD: Control. The registry record does not report whether these comparisons were prespecified, how the sponsor selected the reported analysis, or whether any multiplicity procedure was planned or used. It also does not report interim analyses or stopping rules. Without those details, the extent to which the reported p-values account for multiple comparisons or repeated looks at the data cannot be determined. The unadjusted p-values should be read as unadjusted rather than as multiplicity-controlled evidence.

The registry describes a modified intention-to-treat population: people who consented, were randomized, and were not withdrawn at screening were analyzed as assigned. This definition is informative, but the record does not give the number excluded at screening, the number with missing primary outcome data, or a participant flow account connecting the stated enrollment to the analyzed groups. It reports that missing data were handled by multiple imputation via chained equations, but details of the imputation procedure are not reported in the registry record. In particular, the model specification, variables, number of imputed datasets, and sensitivity analyses are not described.

The registry also does not report a protocol-level statistical analysis plan or the original sample-size justification. The design assumptions used for the pilot, the intended analysis population for every outcome, and any prespecified thresholds for interpreting feasibility are not reported. These omissions limit assessment of whether analyses followed a prespecified plan and how missingness, multiple comparisons, and group inclusion were handled. They do not by themselves show that the analysis was inappropriate; they mark information that is unavailable in the registry record.

15. 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 5% reporting threshold.

The posted safety results list zero serious adverse events, zero other adverse events, and zero deaths in each group for which safety figures are reported. The groups are FRD Control, VNRBD Control, FRD WhatsApp, and VNRBD WhatsApp. For Docudrama, the safety results list 14 participants at risk and zero serious adverse events, zero other adverse events, and zero deaths. No serious adverse event terms are listed.

These entries describe what was reported in the registry, not a guarantee that no adverse events occurred outside the recorded categories or reporting period. The registry record does not provide further detail about safety-event definitions, how events were collected, follow-up completeness for safety, or whether the absence of reported events reflects systematic assessment. The primary outcome’s Docudrama entry lists zero participants, while the safety entry lists 14 at risk; the reason for that difference is not reported in the registry record. These safety figures should therefore be read as the registry’s posted counts, alongside that unresolved difference, rather than as a complete account of all possible harms.

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

The posted results describe donation attempts during follow-up in a randomized pilot of communication approaches for first-time blood donors in Ghana. The counts were low in the groups included in the primary outcome comparisons: the control groups each had 2 recorded attempts, FRD: WhatsApp had 2, and VNRBD: WhatsApp had 7. The recalculated contrasts have confidence intervals that include no difference, and the sponsor’s logistic-regression result also has a broad interval. The estimates are therefore imprecise and do not determine whether the approaches changed donation attempts.

These findings do not establish equivalence between groups. A result that is not statistically significant does not show that two approaches have the same outcome, particularly when the intervals allow a range of possible effects. Nor does the larger observed count in VNRBD: WhatsApp alone establish that WhatsApp increased attempts. The estimate is uncertain, and the registry describes this as a pilot study with intervention development and feasibility objectives.

The outcome is an attempt, not necessarily a completed blood donation. The registry does not report a separate primary endpoint for successful donation or a formal criterion for deciding whether the pilot was feasible. The listed Docudrama primary outcome group has zero participants, and the reason is not reported in the registry record. As a result, the posted primary outcome data do not support a direct outcome comparison involving Docudrama.

The results can inform discussion of the observed event counts, the uncertainty around the comparisons, and the scale of sample sizes implied by planning from pilot effects. They do not settle the effectiveness of the communication strategies, provide a precise estimate of their effects, or resolve how a future trial should be designed. Such conclusions would require information and evidence beyond the posted pilot results.

17. Why This Trial Matters Statistically

This trial is statistically informative because it illustrates the difference between a pilot’s descriptive findings and the evidence needed for a precise comparison. The study was designed not only to assess donation attempts but also to develop communication strategies and evaluate the feasibility of a larger randomized trial. Those aims help explain why the posted results should be considered preliminary rather than treated as a final estimate of an intervention’s effect.

The event counts are sparse, and the recalculated intervals are wide. This creates a practical lesson in uncertainty: a point estimate can appear to favor one group, while the data remain compatible with no difference and with effects in the opposite direction. The sponsor’s odds-ratio analysis is also imprecise. Reporting an estimate together with its confidence interval is therefore more informative than focusing on a p-value alone.

The planning illustrations also show how strongly a sample-size calculation can depend on the effect chosen for planning. A smaller target effect requires more participants per arm under the stated assumptions. Because these illustrations use effects observed in a pilot with few events, they are not a substitute for a prospective design decision or the trial’s original sample-size justification, which is not reported in the registry record.

Finally, the contrast between the primary outcome listing for Docudrama and its safety listing highlights the value of clear group accounting. When the analyzed population and arm-level outcome reporting are not fully reconciled, readers cannot determine from the registry alone why a group appears in one set of results but not another. Transparent participant flow and analysis-plan reporting would make the statistical evidence easier to evaluate.

18. 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 counts are small, and the recalculated confidence intervals are broad. This limits the precision of the group comparisons. The registry describes a modified intention-to-treat analysis and multiple imputation, but does not give the number of missing outcomes, imputation details, or sensitivity analyses. It is therefore not possible to assess from the registry record how sensitive the reported findings are to assumptions about missing data.

The primary outcome listing includes zero participants for Docudrama, while the safety results list 14 participants at risk in that group. The reason for the mismatch is not reported in the registry record. The sponsor’s reported analysis covers four groups and excludes Docudrama; the registry does not explain the group selection or give enough regression detail to reproduce the model. The FRD and VNRBD labels are not expanded in the record.

The study was unmasked. The registry does not report whether outcome assessment was blinded, how attempts were verified, or the degree to which participants received and engaged with their assigned messages. It also does not report details of allocation concealment, contamination, or adherence. These are unreported design details, not evidence that a particular bias occurred.

The registry describes a pilot with qualitative development, feasibility assessment, and participant feedback, but it does not report formal feasibility thresholds, a prespecified statistical analysis plan, the original sample-size justification, interim analyses, or multiplicity procedures. The follow-up extension is explained as a response to an interruption in messaging, but the record does not provide participant-level follow-up completeness. Together, these limitations mean that the posted results are most useful as preliminary outcome and planning information, not as a definitive estimate for all listed communication approaches.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Community-Based Communication Approaches for Blood Donation in Ghana — NCT04783077.
• Registry record last updated 2024-01-10; status last verified 2024-01.
• Results first posted 2024-01-10.
• 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.

22. Record Summary

NCT04783077 is a completed, randomized, parallel, unmasked pilot study sponsored by the University of Minnesota. The registry lists enrollment of 144 and a study population of first-time blood donors in catchment areas of the Southern Zonal Blood Centre in Ghana. Its aims included developing docudrama and WhatsApp communication approaches, assessing feasibility for a larger trial, and evaluating participant views.

The primary endpoint was a blood donation attempt at any time during follow-up. The posted group counts were 2 of 27 for FRD: Control, 2 of 30 for VNRBD: Control, 2 of 28 for FRD: WhatsApp, and 7 of 31 for VNRBD: WhatsApp. Docudrama is listed with zero primary outcome participants; the reason is not reported in the registry record. The sponsor reported a four-group logistic-regression odds ratio of 2.28, with a 95% confidence interval from 0.61 to 8.50 and p = 0.23. The intervals around the unadjusted re-calculations are also broad, so the posted comparisons remain uncertain.