Traditional Ethiopian Coffee Ceremony in a Rural Ethiopian Hospital to Increase Hospital-based Delivery Rates: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Traditional Ethiopian Coffee Ceremony in a Rural Ethiopian Hospital to Increase Hospital-based Delivery Rates (NCT04232137), a randomized na study. The primary outcome is Hospital-based Delivery Rate.

2. Clinical Question

The clinical question is about delivery location. The intervention was communicated during antenatal care, while the outcome concerned a later delivery. The record does not report additional questions about maternal or newborn health outcomes, or whether the ceremony changed other aspects of care. Interpretation should therefore stay focused on the hospital-based delivery outcome and the comparison as recorded.

3. Trial Design

The registry describes a randomized, parallel-group trial with no masking. The sponsor is Vanderbilt University Medical Center. The study status is completed, and the phase is recorded as NA. The registered enrollment is 439. The record lists a start date of 2015-04-08 and a primary completion date of 2015-12-15.

In the coffee ceremony arm, antenatal care patients were told at their first antenatal care visit that a postpartum ceremony would be organized for them and up to 4 relatives. In the comparator arm, patients were told they would not receive a postpartum ceremony. The registry calls this the current status quo. Because masking is listed as none and the intervention was communicated to patients, patients knew which offer applied to them. The record does not describe masking of outcome assessment.

The primary outcome is hospital-based delivery rate, described as the percentage of antenatal care patients who deliver in the hospital, with a timeframe of up to 9 months. The posted results express the arm outcomes as participant counts: 76.0 of 237.0 in the coffee ceremony arm and 53.0 of 167.0 in the no-coffee-ceremony arm. These outcome denominators do not account for the full registered enrollment, and the registry record does not explain the difference or describe which participants were included in the outcome analysis.

The record lists 1 secondary outcome, health center or health post-delivery rate. It records 0 analysis entries. The original sample-size assumptions, the planned statistical analysis, allocation procedures beyond randomization, and any interim analyses are not reported in the registry record. The posted comparison statistics can be described, but they do not fill in those unreported design details.

4. Participant Flow and Accounting

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

Overall Study

5. Baseline Characteristics

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

6. Primary Endpoint

The registry lists the primary outcome as Hospital-based Delivery Rate, measured over Up to 9 months.

The primary endpoint is hospital-based delivery rate. The registry describes it as the percentage of antenatal care patients who deliver in the hospital and gives a timeframe of up to 9 months. The posted outcome values are counts of participants with the event, rather than percentages: 76.0 of 237.0 for the coffee ceremony arm and 53.0 of 167.0 for the no-coffee-ceremony arm. The registry does not provide a more detailed definition of the event or the population included in this endpoint.

Delivery location is relevant to the trial’s stated question because the intervention was intended to encourage patients to deliver at the hospital. The endpoint therefore captures whether the delivery occurred at that site. It does not, by itself, measure the quality or safety of care, maternal or newborn health, whether a patient attended the ceremony, or the reasons for a particular delivery location. Those outcomes and explanations are not reported as part of the primary endpoint.

As a binary outcome, each participant is represented according to whether hospital-based delivery was recorded. A comparison of the proportions can describe the difference between randomized groups, while uncertainty estimates indicate how compatible the data are with a range of possible group differences. The registry also lists health center or health post-delivery rate as a secondary outcome, but its result is not included in the posted outcome figures described for the primary comparison.

7. Results Posted in the Registry

• NO Coffee Ceremony: 31.7% (53/167) versus 32.1% (76/237) on control; risk difference -0.3 percentage points (95% CI -9.6 to 8.9), risk ratio 0.99 (0.74–1.32); unadjusted two-proportion z-test p = 0.944.

8. Statistical Interpretation

9. Statistical Methods Explained

What does the risk difference describe?

A risk difference compares the event proportions between groups by subtracting one group’s risk from the other’s. It expresses an absolute difference in the probability of the event. In this record, the reported risk difference is -0.003 for the comparison of no coffee ceremony with coffee ceremony. The negative sign indicates that the estimated event proportion for the comparison arm is lower under the direction used for this contrast. Its confidence interval, from -0.096 to 0.089, includes zero, so the data are compatible with a range of absolute differences that includes no difference as well as differences in either direction.

The direction of a contrast depends on which arm is subtracted from which. The comparison identifies coffee ceremony as the reference through p0 and no coffee ceremony through p1. Reading the sign alongside that ordering avoids describing a negative estimate as an effect of the intervention without specifying the contrast.

How do risk ratios and odds ratios differ?

A risk ratio compares the event probability in one group with the event probability in another. A value of 1 represents equal probabilities under the contrast. The posted risk ratio is 0.990, with a 95% confidence interval from 0.741 to 1.322. Because the interval includes 1, it includes the value corresponding to no relative difference.

An odds ratio compares the odds of an event rather than the event probabilities themselves. An odds ratio of 1 represents equal odds. The posted odds ratio is 0.985, with a 95% confidence interval from 0.644 to 1.506; this interval also includes 1. Odds and risks are related but not interchangeable, and the odds ratio should not be read as a risk difference. Each measure gives a different scale for describing the same binary outcome comparison.

What does the p-value add?

The posted unadjusted p-value is 0.944. In general, a p-value describes how unusual the observed data, or data more extreme, would be under a specified null model. It is not the probability that the null hypothesis is true, nor does it measure the probability that the intervention works. A large p-value does not demonstrate equivalence or prove the absence of an effect. The registry record does not report a detailed testing procedure, so the p-value should be read as the posted unadjusted result rather than as evidence about unreported analytical choices.

Confidence intervals complement p-values by showing a range of estimates compatible with the data and the method used to construct the interval. They make the uncertainty and possible magnitude of the effect more visible. The interval for the risk difference includes zero; the intervals for the risk ratio and odds ratio include 1. These are the respective values representing no difference on each scale.

Why can the outcome be analyzed as binary?

A binary endpoint classifies each participant as having or not having the event. Here, the event is hospital-based delivery. Counts of participants with the event and the group denominators form the basis for proportions and group comparisons. The posted results give the event counts and denominators for each arm. The registry record does not give details about outcome ascertainment, missing outcomes, or the exact analysis population.

General statistical methods for binary outcomes can compare proportions directly or use a model to estimate relative effects. The posted summary includes an absolute risk difference, risk ratio, and odds ratio, but the registry does not identify a model or give an analysis specification. It is therefore appropriate to interpret the reported estimates and their intervals without attributing them to an unreported modeling method.

Why does over-dispersion matter?

Over-dispersion is a general concern when observed variation is greater than a statistical model assumes. If it is ignored in a setting where it is present, uncertainty estimates can be too narrow and tests can appear more decisive than warranted. This is a general statistical consideration, not a reported feature of this trial. The endpoint is recorded as a binary participant count, and the registry does not report an over-dispersion assessment or a model for which that issue was evaluated.

What can be said about randomization and adjustment?

Randomization is intended to create groups that are comparable on average, including with respect to factors that may affect outcomes. It does not guarantee that every realized group will be identical, and it does not replace transparent reporting of the analysis. The registry records randomized allocation and a parallel model, but does not report sequence generation, concealment procedures, stratification, or a covariate adjustment plan. The available comparison is unadjusted; the record does not include analysis entries describing additional methods.

10. Confidence Intervals and Precision

Precision concerns how narrowly the data locate the effect estimate. The risk difference is -0.003, and its 95% confidence interval ranges from -0.096 to 0.089. This interval includes zero and extends in both directions. It therefore does not pin down a single direction or a narrowly defined magnitude for the group difference. The posted risk ratio is 0.990, with a 95% confidence interval from 0.741 to 1.322, while the odds ratio is 0.985, with a 95% confidence interval from 0.644 to 1.506. Each relative-effect interval includes 1, the value corresponding to no relative difference.

A confidence interval is not a range that assigns a specific probability to the true effect after the data have been observed. In general, it describes the results of a procedure that would cover the true parameter at the stated rate over repeated use under its assumptions. The interval’s width conveys uncertainty, and its location conveys the effects that are compatible with the data under that method. The registry record does not report the interval-construction method.

The posted estimates are close to the values representing no difference, but the intervals also allow effects away from those values. This combination is important: the results do not establish equivalence, and they do not identify a precise benefit or harm. The outcome denominators are 237.0 and 167.0, and the record does not explain how they relate to registered enrollment of 439. The absence of a reported analysis population and missing-data account adds uncertainty about the basis of the endpoint estimates. Statistical precision should therefore be judged from the intervals while keeping the unreported analysis details in view.

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. The posted planning illustration uses the observed effect as its reference and presents scenarios for 100% of the observed effect, 75% of the observed effect, and 50% of the observed effect. For each scenario, it gives participants per arm for 80% power and participants per arm for 90% power. These columns describe different target probabilities of detecting the specified effect under the illustration’s assumptions: the 90% column requires more participants per arm than the 80% column for each listed scenario.

The listed 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 the stated fraction of the observed effect. For 100% of the observed effect, the illustration gives 311,298 participants per arm for 80% power and 416,739 participants per arm for 90% power. For 75% of the observed effect, it gives 553,800 and 741,380 participants per arm, respectively. For 50% of the observed effect, it gives 1,246,906 and 1,669,253 participants per arm, respectively.

The pattern reflects a general planning principle: smaller effects are harder to distinguish from random variation, so detecting them at the same power generally requires more participants. These figures are planning illustrations calculated from the observed effect, not the trial’s original design assumptions; those original assumptions are not reported in the registry record. This is not observed power. The numbers describe hypothetical planning scenarios and should not be interpreted as a measure of the trial’s chance of detecting an effect after its results are known.

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 = 0.321; treatment proportion p₁ = 0.317; average p̄ = 0.319.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6592 = 1.2919.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6592 = 0.5548.
4. n = (1.2919 + 0.5548)² / (-0.0033)² = 311,297.4, rounded up to 311,298 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 <- 0.320675105485232                          # control proportion
p1 <- 0.317365269461078                          # 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(4232137)
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 = 311,298 (80% power), 416,739 (90% power)
 75% of observed effect: n per arm = 553,800 (80% power), 741,380 (90% power)
 50% of observed effect: n per arm = 1,246,906 (80% power), 1,669,253 (90% power)
power.prop.test: 311297.4 
Simulated power at n = 311,298 per arm: 0.798 (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. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when several outcomes, comparisons, time points, or analyses are tested. Repeated testing can increase the chance of obtaining at least one apparently notable result by chance. A statistical plan may address this through a prespecified primary outcome, an adjustment, or a clearly defined hierarchy. The registry identifies hospital-based delivery rate as the primary outcome and lists health center or health post-delivery rate as a secondary outcome. It does not report a multiplicity strategy, the number of planned comparisons, or whether any adjustment was intended. The listed 0 analysis entries mean the record does not provide analysis entries describing such decisions.

Only the primary comparison statistics are included in the posted information described for this analysis. A result for the listed secondary outcome is not reported in the registry record. Without that result or a full analysis plan, it is not possible to assess how the secondary outcome was analyzed or whether the reported primary comparison was the only planned test.

Missing data can affect an outcome comparison if some randomized participants lack an observed endpoint. The registered enrollment is 439, while the posted outcome denominators are 237.0 and 167.0. The registry does not explain the relationship between those figures, report reasons for unavailable outcomes, or describe imputation or other missing-data methods. It is therefore not possible to determine from the record whether missing outcomes were absent, excluded, or handled in another way.

The record also does not report the analysis population, such as whether all randomized patients were analyzed in their assigned groups, or whether another population was used. This information matters because excluding participants after randomization can affect comparability. No conclusion about the actual handling of missing observations or the analysis population should be inferred from the posted counts alone. These items are not reported in the registry record.

14. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

The registry lists serious adverse events and other adverse events separately by arm. For the coffee ceremony arm, 0 serious adverse events are listed among 237 at risk, and 0 other adverse events are listed among 237 at risk. For the no-coffee-ceremony arm, 0 serious adverse events are listed among 167 at risk, and 0 other adverse events are listed among 167 at risk. No serious adverse event terms are listed.

These entries describe what the registry reports; they do not establish that no adverse events occurred outside the recorded reporting framework. The record does not provide additional details about how events were collected, the period of safety follow-up, or the definitions and ascertainment procedures used. Death counts are not reported in the registry record. The intervention is behavioral, but its classification does not remove the need to read safety reporting as a record of reported events rather than a complete account of every possible outcome.

The adverse-event figures are separate from the primary delivery endpoint. Zero listed events in these fields should not be interpreted as evidence about the effect on hospital-based delivery, nor should they be used to infer a broader safety profile beyond the registry’s reported categories.

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

The result describes a randomized comparison of an offer: antenatal care patients were told either that a postpartum coffee ceremony would be organized for them and up to 4 relatives, or that they would not receive one. The primary endpoint concerns hospital-based delivery. The posted counts are 76.0 of 237.0 in the coffee ceremony arm and 53.0 of 167.0 in the no-coffee-ceremony arm. The reported risk difference is -0.003, and its 95% confidence interval includes zero. The unadjusted p-value is 0.944.

These results do not establish that the groups had exactly the same hospital-based delivery rate. The confidence interval also does not identify a clear difference in either direction: it includes zero and extends to values on both sides of it. The risk ratio and odds ratio intervals likewise include their respective no-difference values. The data are therefore compatible with a range of effects, and the estimate is not precise enough to treat a small difference as a settled conclusion.

The primary outcome also does not tell whether participants attended a ceremony, why they chose a delivery location, or whether the intervention affected maternal or newborn health. The registry lists health center or health post-delivery rate as a secondary outcome, but its result is not reported in the registry record. No conclusions about that outcome or other unreported outcomes follow from the primary comparison.

Interpretation is further limited by unreported details. The registry does not explain why the outcome denominators do not account for the registered enrollment, describe missing outcome handling, or specify the analysis population. It also does not report the trial’s original sample-size assumptions or detailed statistical plan. The posted data should be read as the recorded outcome counts and unadjusted estimates, with these reporting gaps in mind. They do not justify a claim that the intervention works, does not work, or has a particular effect beyond the uncertainty represented by the estimates and intervals.

16. Why This Trial Matters Statistically

This trial is statistically instructive because it connects a randomized behavioral offer to a binary outcome measured later in care. The estimand is not simply whether a ceremony was held; the primary endpoint is whether antenatal care patients delivered in the hospital. That distinction helps separate the intervention assignment from the outcome the trial set out to compare.

The posted effect measures illustrate why a single p-value is not a complete interpretation. The unadjusted p-value is 0.944, while the risk-difference interval ranges from -0.096 to 0.089. The risk-ratio and odds-ratio intervals also cross their respective no-difference values. Together, the estimates and intervals convey both the location of the observed comparison and its uncertainty. They do not establish equivalence merely because the p-value is large.

The record also highlights the importance of denominator reporting. Registered enrollment is 439, while the primary outcome denominators are 237.0 and 167.0. Without an account of the analysis population and outcome availability, readers cannot fully assess how the endpoint comparison relates to the randomized enrollment. This is a reporting limitation rather than evidence of a particular problem in conduct or analysis.

Finally, planning calculations based on an observed effect can be useful as illustrations, but they are distinct from the assumptions used to design the trial and from power observed after results are known. The registry’s illustration makes clear how smaller target effects require larger planned sample sizes under its stated assumptions. Keeping these concepts separate supports a more careful reading of the trial’s estimates and the limits of what the record can establish.

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 registry record leaves important trial details unreported. It does not explain the relationship between registered enrollment of 439 and the primary outcome denominators of 237.0 and 167.0. Reasons for missing or unavailable outcomes, the analysis population, and any handling of missing data are not reported in the registry record. These gaps make it difficult to assess how completely the endpoint data represent the randomized groups.

The registry records randomization, no masking, and a parallel design, but does not report the sequence-generation method, allocation concealment, or masking of outcome assessment. It also does not provide a detailed statistical analysis plan, an adjusted analysis, or the original sample-size justification. The posted comparison is unadjusted, and the record lists 0 analysis entries, so the available estimates cannot be situated within a fuller account of planned analyses.

The intervention was an offer communicated to patients, not a reported measure of ceremony attendance. The record does not say whether assigned patients received or attended a ceremony, or whether the intended relatives participated. As a result, the posted comparison pertains to the assigned offer as described, not to a separately measured effect of attendance.

The primary endpoint captures hospital-based delivery and does not by itself describe maternal or newborn health, care quality, or reasons for delivery location. Although the registry lists health center or health post-delivery rate as a secondary outcome, a result for it is not reported in the registry record. Safety reporting is limited to the listed adverse-event fields, and details of event collection and follow-up are not reported. These limitations define what can be concluded from the posted record without implying that unreported procedures or outcomes were absent.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Traditional Ethiopian Coffee Ceremony in a Rural Ethiopian Hospital to Increase Hospital-based Delivery Rates — NCT04232137.
• Registry record last updated 2020-05-07; status last verified 2020-04.
• Results first posted 2020-05-07.
• 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

The registry record describes a completed, randomized, parallel-group trial with no masking, sponsored by Vanderbilt University Medical Center. Its registered enrollment is 439, with a start date of 2015-04-08 and a primary completion date of 2015-12-15. The intervention was an offer of a postpartum traditional Ethiopian coffee ceremony; the comparator was no ceremony.

The primary outcome was hospital-based delivery rate, described as the percentage of antenatal care patients who deliver in the hospital, with a timeframe of up to 9 months. The posted counts are 76.0 of 237.0 for the coffee ceremony arm and 53.0 of 167.0 for the no-coffee-ceremony arm. The unadjusted risk difference is -0.003, with a 95% confidence interval from -0.096 to 0.089 and an unadjusted p-value of 0.944.

The registry lists health center or health post-delivery rate as a secondary outcome and records 0 analysis entries. The original design assumptions, analysis population, missing-data handling, and result for the listed secondary outcome are not reported in the registry record.