Epilepsy Adherence in Children and Technology (eACT): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Epilepsy Adherence in Children and Technology (eACT) (NCT03817229), a randomized na study. The primary outcome is Anti-seizure Medication Adherence Rate.

2. Clinical Question

The study addresses an adherence problem described in the registry rationale: children may not take antiseizure medication as prescribed, and adherence support for children with epilepsy is an area of need. The trial tested a digital, behavioral approach rather than a medication. Its listed intervention components include education microlearning sessions, automated digital reminders, a problem-solving mHealth module, and an individualized adherence feedback report.

The primary result is a comparison of group averages at a stated follow-up time. It does not by itself establish how an individual child’s adherence changed, whether any difference was important to families, or whether the intervention affected seizures or quality of life. Those outcomes are listed among the secondary measures, but the primary question is specifically about adherence at Month 5.

3. Trial Design

The registry lists eACT as a completed, randomized, double-masked, parallel trial, with an enrollment of 268. Its title is Epilepsy Adherence in Children and Technology, and its official title describes fostering medication adherence in children with epilepsy using mHealth technology. The registry lists no phase beyond “NA.” The sponsor is Children’s Hospital Medical Center, Cincinnati. The recorded start date is 2019-04-15 and the primary completion date is 2024-06-15.

The intervention description is consistent with an adaptive approach: the brief summary identifies a sequential, multiple assignment, randomized trial, or SMART. The Stage 1 primary endpoint is reported at Month 5. The registry describes education and automated reminders as the control condition. The primary treatment comparison is the group receiving education, reminders, and individualized adherence feedback. For later treatment pathways, some participants are described as remaining on that approach, while participants categorized as non-responsive either continued individualized feedback or had problem-solving sessions added.

The listed group labels distinguish these later pathways, while the primary outcome result uses the broader treatment label “Treatment Group (E+ADR+IAF).” The registry record does not report the detailed sequence and decision rules that governed assignment to later pathways. It also does not identify which people were masked, how masking was maintained, or whether the analysis followed an intention-to-treat population.

Enrollment is a trial-level registry figure, while the posted primary outcome group summaries give their own group sizes. These describe different aspects of the record and should not be treated as interchangeable. The registry lists 22 outcome measures and reports 21 secondary results as posted. No sponsor-reported analyses are listed for the secondary outcomes.

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.

Before assignment: Following recruitment, participants were in a run-in period of 2 months. All participants who demonstrated adherence \> 95% ended study participation at the end of the 2-month run-in period and were ineligible for randomization.

Intervention Period-Stage 1

Reasons for not completing this period:

Intervention Period-Stage 2

Reasons for not completing this period:

Post-Intervention

Reasons for not completing this period:

Follow-up 1

Reasons for not completing this period:

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 Anti-seizure Medication Adherence Rate, measured over Stage 1 (Month 5).

The primary endpoint is the Anti-seizure Medication Adherence Rate at Stage 1 (Month 5). It is based on electronic monitoring using Simplemed+ pillboxes or Adheretech bottles. The registry describes calculating a monthly adherence rate for Month 5, expressed on a scale from 0-100%, with higher scores indicating higher adherence.

An electronically monitored measure offers a way to quantify medication-taking behavior over the specified month, rather than relying only on a general impression of adherence. The registry record does not report the exact calculation rules for this rate, how device non-use or incomplete monitoring was handled, or how missing monitoring data were treated. Those details matter because the score depends on the recorded medication-taking information and the rules used to turn those records into a monthly rate.

The primary result is reported as a mean and standard deviation for each group. A mean gives the group’s average score; the standard deviation describes how spread out scores were around that average. These summaries describe group-level adherence at the specified time. They do not show the full distribution of individual scores or establish whether a particular score represents adequate adherence for an individual child.

Adherence is central to the trial’s stated rationale, which links taking antiseizure medication as prescribed with important health and quality-of-life concerns. The registry’s primary endpoint, however, measures adherence itself. It should not be interpreted as a direct measure of seizure control, safety, quality of life, or health care use.

7. Results Posted in the Registry

• Treatment Group (E+ADR+IAF): mean 77.23 versus 68.93 on control; difference 8.30 (95% CI -0.42 to 17.02), standardized effect d = 0.27; z-test p = 0.062.

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.

Antiseizure Medication Adherence Rate

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Treatment Group (E+ADR+IAF in Stage 1 and remained in Stage 2): mean 74.42 versus 58.80 on control; difference 15.62 (95% CI 3.74 to 27.50), standardized effect d = 0.49; z-test p = 0.010.
  • Treatment Group (E+ADR+IAF) non-responsive but continue IAF: mean 48.87 versus 58.80 on control; difference -9.93 (95% CI -21.79 to 1.93), standardized effect d = -0.30; z-test p = 0.101.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Treatment Group (E+ADR+IAF in Stage 1 and remained in Stage 2) than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Treatment Group (E+ADR+IAF) non-responsive but continue IAF and the control. The interval gives the range of effects that remain compatible with the data.

Global Assessment of Severity of Epilepsy

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Treatment Group (E+ADR+IAF in Stage 1 and remained in Stage 2): mean 0.25 versus 0.28 on control; difference -0.03 (95% CI -0.20 to 0.14), standardized effect d = -0.07; z-test p = 0.722.
  • Treatment Group (E+ADR+IAF) non-responsive but continue IAF: mean 0.33 versus 0.28 on control; difference 0.05 (95% CI -0.11 to 0.21), standardized effect d = 0.11; z-test p = 0.549.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Treatment Group (E+ADR+IAF in Stage 1 and remained in Stage 2) 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 Treatment Group (E+ADR+IAF) non-responsive but continue IAF and the control. The interval gives the range of effects that remain compatible with the data.

PedsQL Epilepsy Module-Impact (Parent Report)

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Treatment Group (E+ADR+IAF in Stage 1 and remained in Stage 2): mean 81.67 versus 78.76 on control; difference 2.91 (95% CI -4.09 to 9.91), standardized effect d = 0.16; z-test p = 0.415.
  • Treatment Group (E+ADR+IAF) non-responsive but continue IAF: mean 81.01 versus 78.76 on control; difference 2.25 (95% CI -5.77 to 10.27), standardized effect d = 0.11; z-test p = 0.583.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Treatment Group (E+ADR+IAF in Stage 1 and remained in Stage 2) 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 Treatment Group (E+ADR+IAF) non-responsive but continue IAF and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes broaden the assessment beyond the primary Month 5 adherence measure. They include later antiseizure medication adherence, global assessment of epilepsy severity, parent-reported impact on quality of life, and other seizure-related outcomes listed in the registry. The posted results cover different follow-up periods and, for later adherence, include groups defined by their treatment pathway and response status.

Across these measures, the pattern is not uniformly aligned in direction or precision. The later adherence comparisons include one treatment-pathway estimate with an interval excluding zero and another whose interval includes zero. The posted comparisons for epilepsy severity and parent-reported quality-of-life impact have intervals that include zero. This broad pattern does not establish a consistent effect across all domains, and the results do not replace the primary comparison of electronically monitored adherence at Month 5.

Secondary findings can help describe whether the primary result is accompanied by patterns in related outcomes, but they have a different evidential role. The registry lists 22 outcome measures, and 21 secondary results are posted. When many outcomes and comparisons are examined, the chance of seeing an apparently notable result by chance can increase. The record does not report a multiplicity adjustment or a hierarchy that would identify which secondary tests were confirmatory.

The registry also lists no sponsor-reported analyses for the secondary outcomes. The displayed summaries therefore should be treated as descriptive, unadjusted comparisons rather than definitive tests of separate treatment effects. Their interpretation is further limited by the adaptive treatment pathways: later groups reflect response and treatment history, so their comparisons may not have the same meaning as an uncomplicated comparison between groups randomized at the outset. The secondary results can provide context, but they do not settle whether the intervention strategy affected seizures, quality of life, or other clinical outcomes.

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

• Antiseizure Medication Adherence Rate
• Antiseizure Medication Adherence Rate
• Global Assessment of Severity of Epilepsy
• PedsQL Epilepsy Module-Impact (Parent Report)
• Seizure Severity Adapted for Children Scale-Total Score
• Seizure Freedom
• Antiseizure Medication Adherence Rate
• Global Assessment of Severity of Epilepsy

9. Statistical Interpretation

10. Statistical Methods Explained

The posted continuous outcomes are expressed as group means and standard deviations. A mean summarizes the average score, while a standard deviation describes how much scores vary around that average. These summaries are useful for a broad comparison, but they do not reveal whether scores are symmetric, whether a small number of observations are especially influential, or how many participants had particular score ranges. The registry record does not report individual-level distributions.

What does a mean difference represent?

A mean difference is the treatment-group mean minus the control-group mean for the stated comparison. For the primary endpoint, the reported estimate is 8.3. A positive estimate means the treatment group’s reported average was higher on this scale. The estimate is in score units on the adherence measure; it is not a relative percentage change. The scale is described as ranging from 0-100%, but a difference in group means should still be read as a difference in score points, not as a ratio.

The displayed primary estimate and its interval are unadjusted comparisons based on the reported group summaries. They do not incorporate any adjustment for baseline characteristics or other covariates. The registry record does not report a sponsor’s adjusted primary model, the analysis population, or a prespecified approach to missing outcome data. Since discrepancy is not flagged in the registry data, there is no reported conflict requiring the posted recalculation to be set aside; nevertheless, the sponsor’s own analysis would take precedence if its results and methods were available.

Why is there a confidence interval?

A confidence interval expresses uncertainty around an estimated group difference. In repeated samples under the assumptions of the method, intervals constructed this way would capture the target parameter at the stated confidence level at the corresponding rate. For this comparison, the reported 95% confidence interval runs from -0.415 to 17.02. Because zero lies within that interval, the data do not rule out no mean difference at this level of precision. The interval also includes positive values, so it does not establish that the true difference is exactly the observed estimate.

A confidence interval is not a range containing most individual participants’ scores. It concerns uncertainty about the group-level difference. It also does not tell whether a difference is clinically important unless a meaningful-difference threshold has been defined. Such a threshold is not reported in the registry record.

What does the p-value add?

The unadjusted p-value for the primary comparison is 0.062. A p-value describes how unusual results at least as incompatible with a specified null hypothesis would be under that hypothesis and the analysis assumptions. It is not the probability that the null hypothesis is true, nor is it a measure of the size or importance of a difference. The registry record does not report the sponsor’s testing threshold, whether the primary test was adjusted, or how it was incorporated into the SMART analysis. The estimate, interval, and analysis context are therefore needed to interpret the p-value.

Why can over-dispersion matter?

In general, over-dispersion means that the observed variability is greater than a statistical model expects. When a model underestimates variability, standard errors may be too small and confidence intervals too narrow, making evidence appear more precise than it is. The primary outcome is a bounded adherence score, and the posted standard deviations show appreciable variation. The registry record does not state that over-dispersion was assessed or that a particular model was used to account for it. This is a general consideration, not a claim that over-dispersion occurred in this analysis.

The reported group standard errors are 3.913 for control and 2.112 for the primary treatment group. A standard error describes the precision of an estimated group mean, not the spread of individual scores. It commonly becomes smaller when the estimate is based on more observations or when variability is lower. It should not be confused with the standard deviation reported alongside each mean.

How does the adaptive design affect interpretation?

A sequential, multiple assignment, randomized trial can assign participants to later treatment options according to how they respond earlier. The registry identifies later groups by whether participants remained in a treatment pathway or were classified as non-responsive and continued individualized feedback or had problem-solving sessions added. These later groups are therefore tied to treatment history and response status. Their means should not automatically be read as if they came from a single, simple comparison of groups randomized once at the start.

The registry record does not report the decision rules for response, the timing of later randomizations, the statistical method for evaluating treatment sequences, or whether a primary analysis accounted for the adaptive design. Consequently, the posted group summaries support descriptive comparisons, but do not show the full causal evaluation of treatment strategies that a SMART analysis might target.

What can planning calculations tell readers?

The sample-size illustrations use the observed effect as a planning reference, with scenarios for detecting the full observed effect and smaller fractions of it. They are not a reconstruction of the trial’s original design assumptions. The registry record does not report the original sample-size justification or its assumptions. Planning calculations depend on stated choices such as the significance level, allocation, and statistical approximation, and can change when those choices or the target effect change.

These illustrations are not observed power. Observed power is not a useful substitute for reading the estimated effect and its confidence interval: it is largely a transformation of the observed estimate and its uncertainty. The appropriate emphasis is on the size of the estimate, its precision, the analysis method, and whether the outcome is meaningful for the trial’s intended question.

11. Confidence Intervals and Precision

The primary mean difference is 8.3, with a 95% confidence interval from -0.415 to 17.02. This interval is wide enough to include a value slightly below zero and a range of positive differences. The point estimate alone therefore gives an incomplete picture: the interval communicates that the estimate is uncertain and that the posted data are compatible with more than one interpretation of the treatment contrast.

The group means were 68.93 in control and 77.23 in the treatment group. Their standard deviations were 35.65 and 27.29, respectively. These standard deviations describe variation among observed scores, whereas the confidence interval describes uncertainty about the mean difference. The two kinds of spread answer different questions. The registry record does not provide individual-level scores or a clinical threshold that would identify which values within the interval would matter to families.

Precision is influenced by the amount of information in each group and by outcome variability. The posted primary group sizes are 83.0 and 167.0, and the standard errors of the group means are 3.913 and 2.112. These figures help describe how precisely the separate group means were estimated; they do not, by themselves, capture every source of uncertainty in the trial design or analysis.

The interval is a summary-statistic comparison and should not be mistaken for a reported adjusted analysis. The registry record does not report the sponsor’s primary model, missing-data procedures, or analysis population, so the interval cannot answer how alternative analysis choices might affect precision. A narrower interval would indicate more precise estimation, but precision alone would not establish clinical importance or eliminate concerns about design and analysis.

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 planning figures are participants per arm for 80% power and participants per arm for 90% power. These are not total-event counts; the event-count convention applies to time-to-event calculations, and the outcomes are continuous. The planning assumptions listed in the registry are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each scenario uses the stated fraction of the observed effect as the effect to detect.

The 80% power column represents a planning target in which the assumed study would have an 80% chance of detecting the specified effect under the stated assumptions if that effect were present. The 90% power column uses a higher chance of detection as the design target. Higher power targets generally require more participants under the same assumptions. The number needed also increases when the target effect is smaller: planning for 75% of the observed effect or 50% of the observed effect therefore calls for larger participant counts than planning for the full observed effect.

These calculations are illustrations based on the observed estimate, not the trial’s own design calculation. The original sample-size justification and its assumptions are not reported in the registry record. The figures are not observed power. They should not be read as a judgment that the completed trial was adequately or inadequately sized; they show how hypothetical planning requirements vary with the assumed effect and power target. Real design work would also need to reflect the intended analysis, outcome distribution, allocation structure, follow-up, and any features of the adaptive design.

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, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = 8.300; pooled standard deviation σ = 30.310.
2. n = 2 × (1.96 + 0.842)² × 30.310² / 8.300² = 209.3, rounded up to 210 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- 8.3                      # observed difference in means
sigma <- 30.3104345388759                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(3817229)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 210 (80% power), 281 (90% power)
 75% of observed effect: n per arm = 373 (80% power), 499 (90% power)
 50% of observed effect: n per arm = 838 (80% power), 1,122 (90% power)
power.t.test: 210.3 
Simulated power at n = 210 per arm: 0.797 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when a study evaluates many outcomes, time points, treatment contrasts, or subgroups. Even when each individual test uses a conventional error rate, conducting many tests can increase the chance that at least one result appears notable by chance. The registry lists 22 outcome measures and reports 21 secondary results as posted. It does not report a multiplicity procedure, a prespecified hierarchy for secondary outcomes, or which secondary comparisons were intended to be confirmatory. Unadjusted secondary p-values should therefore be interpreted cautiously and not as a collection of independent confirmations.

The registry record lists no reported analyses for the secondary outcomes. For the primary result, the posted comparison is an unadjusted mean difference from group summaries; the sponsor’s full analysis method is not reported in the registry record. It is also not reported whether the primary testing procedure accounted for repeated assessments or the trial’s adaptive structure. These omissions make it difficult to determine how the displayed comparisons relate to a prespecified inferential plan.

Missing-data handling is not reported in the registry record. The posted group sizes vary across outcomes and follow-up periods, but those figures alone do not establish why outcome data are absent or whether missingness could bias the comparison. The record does not describe device failures, adherence-monitoring completeness, imputation, sensitivity analyses, or how participants without an outcome were handled.

The analysis population is also not reported. It is therefore not possible to tell from the registry summaries whether all randomized participants were analyzed as assigned, whether any participants were excluded, or whether a different population was used for a particular outcome. These details matter because the interpretation of randomized comparisons depends on how participants and missing observations are handled. None of these gaps proves that the analysis was biased; they limit what can be concluded from the registry record alone.

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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: suicide attempt (1.8%).

The registry reports serious adverse event figures for four groups with their own at-risk counts. The control group had 0 serious adverse events among 89 at risk. The group receiving E+ADR+IAF in Stage 1 and remaining in Stage 2 had 0 among 48 at risk. The non-responsive group that continued individualized adherence feedback had 1 among 56 at risk. The non-responsive group in which problem-solving was added had 0 among 63 at risk. The registry reports 0 deaths in each group.

The registry lists “Suicide Attempt” as a serious adverse event term at 1.8%. It does not report the denominator or additional context for that percentage in the posted adverse-event summary. The counts should be read as reported for the named groups and their at-risk populations, not combined into a single trial-wide estimate.

Other, non-serious adverse events are also listed: 35 among 89 at risk in control, 13 among 48 in the group remaining on E+ADR+IAF, 24 among 56 in the non-responsive group continuing individualized feedback, and 21 among 63 in the group with problem-solving added. These are descriptive counts. The registry record does not report a comparative safety analysis, attribution to intervention, or adjustment for exposure time. The figures therefore do not establish whether event rates differed between treatment strategies.

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

The primary result describes a higher observed mean adherence score for the group receiving education, reminders, and individualized feedback than for the control group receiving education and reminders. The estimated mean difference was 8.3, with a 95% confidence interval from -0.415 to 17.02. The estimate is compatible with a range of effects, including no difference, so it should be read as an uncertain group-level result rather than as a precise measure of what each participant experienced.

The outcome is electronically monitored adherence at Stage 1 (Month 5). It is not a direct measure of seizures, epilepsy severity, quality of life, or health care use. Those topics appear among the trial’s stated aims or secondary measures, but an adherence result cannot stand in for them. Nor does a group mean show how many participants took medication as prescribed, whether particular individuals changed their behavior, or whether an individual score was clinically adequate.

The intervention contrast bundles several components. Because both primary groups received education and automated reminders, the comparison concerns the addition of individualized adherence feedback to those shared components. It does not isolate the effect of education, reminders, feedback, or therapist-led problem-solving separately. The later adaptive pathways address additional treatment sequences, but their posted group summaries do not by themselves establish the causal effect of a full strategy.

The registry record does not report a threshold for a clinically important difference, the sponsor’s complete primary analysis, missing-data methods, or the analysis population. It is therefore not possible from the posted summaries alone to decide whether the estimated difference is important to patients or whether alternative analysis choices would materially change the inference. The results support a careful description of the observed comparison and its uncertainty, not a broad claim about treatment efficacy or individual outcomes.

17. Why This Trial Matters Statistically

Adherence is a behavioral outcome that can be difficult to measure consistently. eACT used electronic monitoring from medication containers to construct its primary outcome, allowing the trial to assess recorded medication-taking behavior rather than relying only on a general assessment. The registry’s description makes the measurement approach and its Month 5 time point explicit, which is important when interpreting what the primary result does and does not represent.

The study also illustrates why a treatment comparison depends on its comparator. Both primary groups received an education module and automated reminders; the contrast adds individualized adherence feedback. A higher treatment-group mean therefore pertains to that specific bundle of components against the stated control, not to digital support versus no support. This distinction helps avoid attributing the observed comparison to a component that was shared between groups.

The adaptive design raises questions beyond a single mean difference. Later treatment options depend on response status and treatment history, so evaluating the full strategy requires methods that reflect the sequence of assignment and decision-making. The registry provides descriptions of later pathways, but not the detailed analysis plan needed to understand how those pathways were formally compared. This is a useful reminder that group summaries can be informative while still falling short of answering every design-level question.

Finally, the confidence interval focuses attention on precision rather than on a binary reading of the p-value. The primary estimate is positive, but its interval includes zero and extends across a range of possible differences. Considering the estimate, interval, variability, and analysis limitations together gives a more statistically grounded interpretation than treating one statistic as a verdict. That approach is particularly relevant when a trial measures several outcomes and reports unadjusted secondary comparisons.

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 registry record does not report the exact eligibility criteria, participant age range, or detailed definition of new-onset epilepsy. The population can therefore be described only at the level stated in the record: children with epilepsy, with the rationale focused on new-onset epilepsy. The extent to which the results apply to other populations cannot be determined from these summaries.

Although the registry identifies double masking, it does not state who was masked or how masking was maintained for a behavioral mHealth intervention. It does not report the sequence rules or decision criteria for response in the adaptive design. Without those details, the later treatment-pathway comparisons are difficult to interpret as effects of a complete adaptive strategy.

The primary endpoint relies on electronic monitoring, but device completeness, data-processing rules, and handling of missing monitoring records are not reported in the registry record. Missing-data methods and the analysis population are also not reported. Differences in outcome group sizes across measures are visible in the posted summaries, but they do not reveal the reasons for those differences or whether missingness affected estimates.

The primary comparison is an unadjusted summary-statistic comparison, and the registry does not report the sponsor’s full primary analysis, covariate adjustment, or a multiplicity strategy. A clinical threshold for an important adherence difference is not reported. The posted secondary results span multiple outcome domains and follow-up periods, but no sponsor-reported secondary analyses are listed. Together, these gaps limit conclusions about precision, confirmatory evidence, and practical importance. They do not establish that the trial’s analyses were inappropriate; they identify details that cannot be evaluated from the registry record.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Epilepsy Adherence in Children and Technology (eACT) — NCT03817229.
• Registry record last updated 2025-10-30; status last verified 2025-09.
• Results first posted 2025-10-30.
• 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

Epilepsy Adherence in Children and Technology (eACT; NCT03817229) is listed as a completed, randomized, double-masked, parallel study sponsored by Children’s Hospital Medical Center, Cincinnati. The registry describes a sequential, multiple assignment randomized trial focused on mHealth support for medication adherence among children with epilepsy.

The primary endpoint is electronically monitored antiseizure medication adherence at Stage 1 (Month 5). The control condition included mHealth education and automated digital reminders; the primary treatment group also received individualized adherence feedback. The reported mean difference was 8.3, with a 95% confidence interval from -0.415 to 17.02 and an unadjusted p-value of 0.062.

The registry lists later adherence, epilepsy severity, quality-of-life, and seizure-related outcomes. Detailed sponsor-reported analyses, missing-data methods, the analysis population, and the original sample-size justification are not reported in the registry record.