Scaling Up: A Multi-Site Trial of e-SBI for Alcohol Use in Pregnancy: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Scaling Up: A Multi-Site Trial of e-SBI for Alcohol Use in Pregnancy (NCT04332172), a randomized na study. The primary outcome is Number of Participants Abstinent From Alcohol as Measured by Timeline FollowBack Calendar Recall.

2. Clinical Question

3. Trial Design

This was a randomized, single-masked, factorial trial, with enrollment recorded as 384. It is titled as a multi-site trial. The status is completed, and the listed start date is 2023-05-19; the listed primary completion date is 2025-08-06. The phase is recorded as not applicable. The sponsor was Michigan State University.

The six groups covered general information, general information plus tailored SMS, the baseline brief intervention, the baseline intervention plus SMS, the baseline intervention plus two booster sessions, and the baseline intervention plus boosters and SMS. General information included treatment as usual and referral resources. The baseline intervention was described as a brief technology-delivered intervention for alcohol use during pregnancy. Booster sessions were remote online sessions, each described as very brief and delivered using participants’ own mobile devices. The record does not state who was masked.

The primary outcome was the number of participants abstinent from alcohol, assessed by Timeline FollowBack Calendar Recall at 34 weeks gestation and defined as no self-report of alcohol use in the preceding 90 days. Fourteen participants did not complete that assessment and were excluded from the primary analyses. The secondary outcome was a healthy birth outcome, defined as a full-term live birth of normal birthweight with no days in the neonatal intensive care unit; it was measured at birth, up to 42 weeks gestation.

The registry reports 19 analyses. The original statistical analysis plan, the allocation procedure in detail, and the rationale for the planned enrollment are not reported in the registry record. The record does not describe an interim analysis plan.

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.

Baseline Intervention

Pregnancy Intervention

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: 14 participants did not complete the Timeline FollowBack Calendar Recall so they were not included in these analyses.

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: 11 people were randomized but did not complete study intervention and assessments due to pregnancy loss (n=8) and participant withdrawal (n=3).

Only the first 6 of 7 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Number of Participants Abstinent From Alcohol as Measured by Timeline FollowBack Calendar Recall, measured over 90 days prior to 34 weeks gestation.

The primary endpoint counts participants who reported no alcohol use in the 90 days prior to 34 weeks gestation, using Timeline FollowBack Calendar Recall. It is a binary outcome: each participant is classified as abstinent or not abstinent according to the stated definition. The reported group values are counts of participants, rather than measurements of the amount or frequency of alcohol use.

Using a defined assessment window and a specified recall instrument gives the trial a concrete outcome for comparing the randomized groups. The measure addresses reported abstinence during late pregnancy, rather than alcohol use across the entire pregnancy or after birth. The registry specifies the assessment time and definition, but does not report corroborating biological measures or further details about how the recall data were collected.

The analysis denominator is not identical across groups: the registry says 14 participants did not complete the Timeline FollowBack assessment and were excluded from these analyses. The posted results therefore describe the participants included in the endpoint analysis, not everyone enrolled. The registry does not report the reasons for non-completion or a method for incorporating missing assessments.

Abstinence is a useful binary summary for the stated research question, but it does not distinguish between different levels or patterns of alcohol use among participants who did not meet the abstinence definition. Nor does it establish that a reported change in this measure accounts for any birth outcome. Healthy birth outcome was posted separately as a secondary endpoint.

7. Results Posted in the Registry

• General Information + SMS: 79.7% (47/59) versus 68.9% (42/61) on control; risk difference 10.8 percentage points (95% CI -4.7 to 26.3), risk ratio 1.16 (0.94–1.43); unadjusted two-proportion z-test p = 0.176.
• Baseline Brief Intervention: 83.1% (49/59) versus 68.9% (42/61) on control; risk difference 14.2 percentage points (95% CI -0.9 to 29.3), risk ratio 1.21 (0.98–1.48); unadjusted two-proportion z-test p = 0.069.
• Baseline Brief Intervention + SMS: 73.0% (46/63) versus 68.9% (42/61) on control; risk difference 4.2 percentage points (95% CI -11.8 to 20.1), risk ratio 1.06 (0.85–1.33); unadjusted two-proportion z-test p = 0.610.
• Baseline Brief Intervention + 2 Booster Sessions: 77.2% (44/57) versus 68.9% (42/61) on control; risk difference 8.3 percentage points (95% CI -7.6 to 24.3), risk ratio 1.12 (0.90–1.40); unadjusted two-proportion z-test p = 0.308.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Number of Participants With a Healthy Birth Outcome (Full-term Live Birth of Normal Birthweight, With no Days in NICU)

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • General Information + SMS: 83.3% (45/54) versus 87.0% (47/54) on control; risk difference -3.7 percentage points (95% CI -17.1 to 9.7), risk ratio 0.96 (0.82–1.12); unadjusted two-proportion z-test p = 0.588.
  • Baseline Brief Intervention: 87.7% (50/57) versus 87.0% (47/54) on control; risk difference 0.7 percentage points (95% CI -11.7 to 13.0), risk ratio 1.01 (0.87–1.16); unadjusted two-proportion z-test p = 0.914.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between General Information + SMS and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Baseline Brief Intervention and the control. The interval gives the range of effects that remain compatible with the data.

The registry posts one secondary outcome: healthy birth outcome, defined as a full-term live birth of normal birthweight with no days in the neonatal intensive care unit. It also reports two logistic regression analyses across the listed groups, with odds ratios of 1.211 and 1.373. Their respective 95% confidence intervals are 0.567 to 2.588 and 0.615 to 3.064, with p-values of 0.439 and 0.520. The registry does not explain the distinction between these two analyses or provide fuller model details in the record.

The secondary results do not provide a clear statistical signal in the posted regression analyses: both confidence intervals include 1, the null value for an odds ratio. The endpoint is distinct from the primary abstinence measure and concerns birth outcome. It should not be treated as confirmation or refutation of the primary finding, because the outcomes measure different things and are assessed at different times.

The unadjusted recalculations for the posted secondary comparisons also have intervals that include zero. That broad pattern indicates uncertainty around the group differences on the risk-difference scale; it does not establish that all groups have the same birth-outcome distribution. The primary and secondary outcomes should be interpreted on their own definitions and scales.

Secondary results generally carry less weight for a primary trial conclusion because the primary endpoint is the principal outcome designated for the study question, while additional outcomes create more opportunities for chance findings. Here, the registry reports 19 analyses, and a multiplicity adjustment or a hierarchy for interpreting secondary analyses is not reported. Thus, the secondary findings are best treated as exploratory context rather than as decisive evidence about the primary question.

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

• Number of Participants With a Healthy Birth Outcome (Full-term Live Birth of Normal Birthweight, With no Days in NICU)

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is binary: participants either met the stated abstinence definition or did not. For each randomized group, the registry reports the number assessed and the number recorded as abstinent. The simple recalculations compare these group summaries without adjustment. They are presented as risk differences, which express the difference between the two groups’ proportions with the event. A positive risk difference means the event proportion in the named intervention group is higher than in the general-information group under the recalculation’s stated direction.

What does a confidence interval convey?

A confidence interval gives a range of estimates compatible with the data and the statistical method used. Wider intervals indicate less precision; narrower intervals indicate more precision. For a risk difference, zero represents no difference between the compared proportions. When the interval includes zero, the posted data do not clearly distinguish the estimate from no difference at the confidence level used. This does not prove that the groups are identical. Conversely, an interval excluding zero does not alone establish that an effect is clinically important or that it will reproduce in another study.

How should the p-values be read?

The recalculated p-values are unadjusted. In general, a p-value describes how compatible the observed data are with a specified null model, assuming that model and the analysis method apply. It is not the probability that the null hypothesis is true, and it does not measure the size or importance of an effect. Because multiple comparisons and analyses are reported, unadjusted p-values should be read cautiously. The registry does not report whether a multiplicity correction was part of the original plan.

Why are there different effect measures?

The recalculation reports risk differences, while the sponsor’s posted regression analyses report odds ratios. A risk difference compares event proportions directly. An odds ratio compares the odds of an event between groups. These measures use different scales and are not interchangeable; their numerical values should not be compared as if they represented the same quantity. The registry does not explain the event and reference-group coding for the sponsor’s odds ratios, so their direction should be interpreted using the sponsor’s analysis and coding rather than inferred from the recalculated risk differences.

What does the factorial label establish?

The record classifies the trial as factorial and lists groups that combine a baseline intervention, SMS, and booster sessions in different ways. This describes the design structure, but it does not show that the posted comparisons estimate the separate contribution of each component. A factorial analysis would ordinarily need to specify how component effects and any interactions are modeled. Those details are not reported in the registry record. The posted recalculations compare each listed group with general information and should be read as such, rather than as formal estimates of component effects.

Why does the analysis population matter?

The registry reports that 14 participants did not complete the primary Timeline FollowBack assessment and were not included in those analyses. The group-specific denominators therefore reflect the reported primary outcome data, not all enrolled participants. Excluding participants with missing outcome data can affect interpretation if missingness is related to the outcome or differs across groups. Reasons for missing assessments, imputation, sensitivity analyses, and an intention-to-treat analysis are not reported in the registry record.

The reported logistic regression analyses are the sponsor’s analyses; the recalculated risk differences use only the posted summary counts. The registry does not provide the full regression specification, covariates, handling of missingness, or any adjustment method. Where the results differ, the sponsor’s own analysis takes precedence rather than treating the unadjusted recalculation as a correction to it.

11. Confidence Intervals and Precision

Precision is reflected by the width of the confidence interval as well as the point estimate. For the recalculated primary comparisons, the risk-difference intervals are broad enough to include zero in four comparisons. For example, the interval for baseline intervention alone versus general information extends from -0.009 to 0.293. This range includes a small negative difference as well as positive differences, so the estimate is not tightly pinned down. The interval for baseline intervention plus boosters and SMS extends from 0.053 to 0.337 and does not include zero under this unadjusted calculation.

The interval for baseline intervention plus SMS is -0.118 to 0.201, spanning negative and positive differences. The intervals for general information plus SMS and baseline intervention plus boosters are -0.047 to 0.263 and -0.076 to 0.243, respectively. These ranges make clear that a point estimate alone does not capture uncertainty. In particular, a confidence interval that includes zero should not be described as proof of no effect.

The sponsor’s reported logistic regression comparison of general information with baseline intervention plus boosters and SMS gives an odds ratio of 0.284, with a 95% confidence interval from 0.108 to 0.748. The interval excludes 1, the null value for an odds ratio. The coding behind that comparison is not described, and the sponsor’s result differs from the unadjusted risk-difference recalculation. The sponsor’s own analysis therefore takes precedence.

Group sizes for the primary comparisons are modest, and the confidence intervals show that estimates remain uncertain. The record does not report a precision target or an original sample-size justification. Confidence intervals quantify statistical uncertainty under the method used; they do not incorporate all possible sources of uncertainty, such as measurement error, missing outcomes, or differences between trial participants and other populations.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The sample-size illustrations are expressed as participants per arm for the binary-outcome comparisons. They are not total event counts. Each row considers detecting the full observed effect or a stated fraction of that effect. For example, the planning scenarios include 100% of the observed effect, 75% of the observed effect, and 50% of the observed effect. The figures are illustrations computed from the observed effect, not the trial’s original design assumptions.

The listed planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is the stated fraction of the observed effect. The registry does not report the original trial’s sample-size assumptions or justification, so these figures should not be read as describing how enrollment was set.

The 80% power column gives the illustrative participants per arm for a design with an 80% chance of detecting the specified effect if that effect were present, under the planning assumptions. The 90% power column gives the corresponding illustration for 90% power. Greater planned power generally requires more participants for the same effect. Within the posted scenarios, smaller effects require more participants: the 50% effect scenario has larger per-arm figures than the 75% and 100% scenarios.

These values are planning illustrations, not observed power. Observed power is not a useful substitute for reading the estimated effect and its confidence interval after a trial. The illustrations also depend on the observed effect, which can be uncertain, particularly when group sizes are limited. They should therefore be treated as conditional planning examples, not as definitive enrollment recommendations.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.689; treatment proportion p₁ = 0.797; average p̄ = 0.743.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6183 = 1.2119.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6136 = 0.5164.
4. n = (1.2119 + 0.5164)² / (0.1081)² = 255.7, rounded up to 256 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 <- 0.688524590163934                          # control proportion
p1 <- 0.796610169491525                          # 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(4332172)
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 = 256 (80% power), 342 (90% power)
 75% of observed effect: n per arm = 471 (80% power), 630 (90% power)
 50% of observed effect: n per arm = 1,093 (80% power), 1,463 (90% power)
power.prop.test: 255.7 
Simulated power at n = 256 per arm: 0.797 (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.

14. Analyses Reported by the Sponsor

The registry reports logistic regression results from the sponsor alongside simple, unadjusted recalculations based on the posted counts. For the primary comparison of general information with baseline intervention plus two booster sessions and SMS, the sponsor reports an odds ratio of 0.284, a 95% confidence interval from 0.108 to 0.748, and p = 0.011. The unadjusted recalculation for the same groups reports a risk difference of 0.195, a confidence interval from 0.053 to 0.337, and p = 0.009.

These are different statistical measures and analyses. The registry does not report the logistic model specification, covariates, or the event and group coding for the odds ratio. The numerical direction of an odds ratio cannot be inferred from the risk difference without knowing that coding. The confidence intervals and p-values are also not identical, as expected when methods or adjustment differ.

Because the recalculation is a simple unadjusted analysis and the sponsor’s own analysis is reported in the registry, the sponsor’s analysis takes precedence when describing the trial’s formal result. The recalculation is useful for viewing the posted counts on a risk-difference scale, but it should not be presented as a replacement for the sponsor’s model or as evidence that the sponsor’s result is erroneous.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists 19 analyses, including multiple pairwise comparisons among the six primary-outcome groups and analyses of the secondary birth outcome. When many comparisons are examined, the chance of obtaining at least one apparently notable result by chance can rise if each is interpreted without accounting for the full set of analyses. This is why multiplicity matters when interpreting an isolated p-value.

The record does not report a multiplicity correction, a prespecified hierarchy of comparisons, or which comparisons were designated as confirmatory. It also does not provide the complete statistical analysis plan. As a result, the posted unadjusted p-values should not be treated as if they account for all the analyses performed. The result with p = 0.009 in the recalculation and the sponsor’s reported p = 0.011 warrant attention as reported findings, but their interpretation should take the wider set of analyses into account.

For the primary outcome, 14 participants did not complete the Timeline FollowBack Calendar Recall and were excluded from the analyses. The registry does not report why assessments were missing, whether missingness differed by randomized group, or whether any imputation or sensitivity analysis was used. The listed denominators vary by group, so the primary results describe the available outcome assessments rather than all enrolled participants.

The record does not report whether all randomized participants were analyzed according to their assigned group, or whether another analysis population was used. It also does not describe an interim analysis plan. These omissions limit assessment of how the analyses were selected and how missing outcomes were handled. They do not demonstrate that a particular analytic problem occurred; they mean the registry record does not provide enough detail to assess it fully.

16. Safety Results

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

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: pregnancy loss (3.2%), hospitalization (high blood pressure/preeclampsia) (1.6%), hospitalization (vasa previa) (1.6%).

The safety listing reports serious adverse events and other adverse events by group, with counts at risk and counts affected. Serious adverse events affected 1 participant in general information, 2 in general information plus SMS, 1 in baseline brief intervention, 1 in baseline brief intervention plus SMS, 4 in baseline brief intervention plus two booster sessions, and 1 in baseline brief intervention plus two booster sessions and SMS. The reported number of deaths was 0 in each group.

For other adverse events, the affected counts were 10, 11, 12, 11, 23, and 10 across those same groups, respectively. The numbers at risk are also listed by group in the safety record. These are descriptive counts; the registry does not provide a comparative statistical analysis of adverse events or state whether events were judged related to the intervention.

The listed serious-event terms include Pregnancy Loss at 3.2%, hospitalization for high blood pressure/preeclampsia at 1.6%, and hospitalization for vasa previa at 1.6%. The registry record does not provide further context for these term percentages. Counts should be read alongside the applicable group size and event definitions, and should not be interpreted as evidence of causation.

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

The primary outcome describes self-reported abstinence during the 90 days prior to 34 weeks gestation. The posted counts and comparisons show that abstinence was recorded in each randomized group and that the estimated differences versus general information varied across intervention combinations. The recalculated comparison for the combination of baseline intervention, boosters, and SMS has a confidence interval excluding zero; the other recalculated primary comparisons have intervals that include zero.

The sponsor’s logistic regression analysis also reports a result for that combined group versus general information, with an odds ratio of 0.284 and a 95% confidence interval from 0.108 to 0.748. Because the coding and full model details are not reported, the odds ratio’s direction should not be inferred from the risk-difference recalculation. The sponsor’s own analysis takes precedence over the unadjusted recalculation for the formal reported result.

These findings do not establish that every component is independently effective, that boosters or SMS add a particular amount to the baseline intervention, or that the components interact. The registry does not report a separate factorial component analysis. The results also do not establish that changes in reported abstinence caused changes in birth outcomes. The secondary endpoint is a separate measure, and its posted regression results have wide confidence intervals that include the null value.

Self-reported abstinence is not the same as a direct measure of alcohol exposure, and the registry does not report corroborating measurement. Missing primary assessments were excluded, and the record does not describe how those missing outcomes were handled beyond exclusion. The findings therefore apply most directly to the recorded analyses and their included participants.

Statistical evidence from a trial does not by itself establish clinical importance, generalizability, or the likely result in every care setting. The record gives no minimum clinically important difference or detailed information about participant characteristics and implementation context. These results should be understood as estimates with uncertainty from a completed randomized trial, not as medical advice or a guarantee of an individual outcome.

18. Why This Trial Matters Statistically

Statistically, this trial illustrates why a multi-arm factorial design and a binary primary outcome require careful interpretation. Six groups represent different combinations of a brief technology-delivered intervention, boosters, and tailored SMS. The group-to-group estimates do not automatically answer whether each component has an independent effect or whether combinations interact; those questions require analyses that reflect the factorial structure. A separate component-effect analysis is not reported in the registry record.

The trial also shows the value of reporting both effect estimates and uncertainty. The recalculated risk differences make the scale of the between-group contrasts visible, while confidence intervals show that several estimates remain compatible with a range of possibilities. A single point estimate or p-value would not communicate that uncertainty. The sponsor’s logistic regression results add another measure, but the different scales and incomplete model description mean the measures should not be conflated.

Multiplicity is relevant because the record lists 19 analyses. Without a reported adjustment or prespecified hierarchy, one statistically notable comparison should be interpreted in the context of all reported analyses. Similarly, the exclusion of participants who lacked the primary assessment matters for understanding which data support the estimate.

Finally, the secondary birth outcome is distinct from the primary self-reported abstinence endpoint. Keeping those outcomes separate avoids treating one as a direct substitute for the other. Together, the design and reporting make this trial a useful example of how analysis populations, outcome definitions, model choice, confidence intervals, and multiple comparisons shape interpretation of clinical trial results.

19. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

Several limitations specific to this trial are visible in the registry record. The primary endpoint relies on self-report through Timeline FollowBack Calendar Recall. The record defines the recall window and assessment time, but does not report corroborating measurement or additional details about the data collection process. Self-report can be affected by recall or reporting error, which would not be reflected in the statistical confidence intervals.

Fourteen participants did not complete the primary assessment and were excluded from those analyses. Reasons for non-completion and the relationship between missingness and outcome are not reported. The available-case results could differ from results under other assumptions about missing outcomes, but the registry does not report imputation or sensitivity analyses.

The posted primary comparisons are individual group comparisons against general information. Although the trial is labeled factorial, the record does not report a separate analysis of the main effects of SMS or boosters, or their interaction with the baseline intervention. Nor does it report the full logistic model specification, including any covariate adjustment. This limits evaluation of how the sponsor’s model relates to the randomized design.

The record lists 19 analyses but does not report a multiplicity strategy or identify a confirmatory hierarchy. This makes isolated unadjusted findings harder to weigh. The confidence intervals also reflect substantial uncertainty for several comparisons. Finally, the record does not provide detailed participant characteristics, site-specific results, adherence information, or reasons for variation in follow-up. These details are not reported in the registry record and limit assessment of applicability and implementation.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Scaling Up: A Multi-Site Trial of e-SBI for Alcohol Use in Pregnancy — NCT04332172.
• Registry record last updated 2026-10-14; status last verified 2026-09.
• Results first posted 2026-10-14.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

23. Record Summary

This completed randomized, single-masked factorial trial enrolled 384 participants and evaluated technology-delivered alcohol-use support during pregnancy across six groups. The primary endpoint was self-reported abstinence in the 90 days prior to 34 weeks gestation, assessed by Timeline FollowBack Calendar Recall. Fourteen participants without that assessment were excluded from the primary analyses.

The sponsor reported a logistic regression odds ratio of 0.284 for general information versus baseline intervention plus two booster sessions and SMS, with a 95% confidence interval from 0.108 to 0.748 and p = 0.011. The unadjusted recalculation for that comparison gives a risk difference of 0.195, with a confidence interval from 0.053 to 0.337. Because the analyses differ, the sponsor’s own result takes precedence. The registry also reports a healthy birth outcome as a secondary endpoint, but its regression results are imprecise. Multiplicity handling, the original sample-size rationale, and further analysis details are not reported in the registry record.