mHealth for Antenatal Mental Health: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for mHealth for Antenatal Mental Health (NCT02516982), a registered phase not stated study. The primary outcome is Number of Participants Responding Affirmatively to at Least One Whooley Question According to Survey Layout.

2. Clinical Question

The clinical context is antenatal recognition of depression. The study description links its screening approach to NICE antenatal and postnatal mental health guidance, including Whooley questions followed by a validated screening instrument such as the EPDS. Its stated aim is to assess feasibility of implementing that approach with mobile technology, not to establish a clinical treatment effect.

The record also describes two smartphone assessment approaches. One involved repeated assessment days, with momentary and contextual questions in addition to retrospective assessment; the other involved retrospective assessment. These are distinct from the tablet-layout comparison and do not form part of the stated primary endpoint. The registry does not report a comparison of clinical outcomes between the smartphone groups.

3. Trial Design

The study is titled “Tablet Computers for Implementing NICE Antenatal Mental Health Guidelines - Feasibility Study” and is marked completed. The registry lists an enrollment of 946, a start date of 2015-10, and a primary completion date of 2018-02. Imperial College London is listed as the sponsor. These details describe the record; further information about recruitment progress or the timing of individual assessments is not reported in the registry record.

Four groups are described. The two tablet screening groups differed in survey layout: the scrolling layout presented all questions on a single screen, while the paging layout presented one question at a time. Both surveys included demographic questions, Whooley questions, and the EPDS. The primary endpoint was limited to these two groups and assessed once immediately after recruitment.

The other groups used an app installed on participants’ own smartphones. The retrospective-plus-momentary group was assigned a sampling protocol of 6 consecutive days, once a month for 6 months. The retrospective-assessment group was assigned one assessment day a month for 6 months. Further detail about the exact questions and completion requirements is not reported in the registry record.

The registry does not state an allocation method. The arms therefore should not be treated as a randomized comparison, and the record cautions that they are not like-for-like control groups. The reason for the allocation approach, any stratification, and whether recruitment or assessment procedures differed by site are not reported in the registry record. The secondary outcomes listed concern completion time and requests for technical assistance according to survey layout. The record lists 8 outcome measures and 7 analyses, but does not explain their relationship or provide a complete 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.

Overall Study

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 Number of Participants Responding Affirmatively to at Least One Whooley Question According to Survey Layout, measured over One-off assessment immediately after recruitment.

The primary endpoint is the number of participants responding affirmatively to at least one Whooley question, considered according to tablet survey layout. The registry specifies a one-off assessment immediately after recruitment and says the measure applies only to the screening groups using the scrolling and paging layouts.

This count represents a screening response, not a diagnosis. It records whether a participant gave an affirmative response to at least one of the Whooley questions. In the described screening approach, the Whooley questions are followed by a validated screening instrument such as the EPDS. The endpoint is therefore relevant to evaluating how the two layouts capture responses during an antenatal screening process, while not itself establishing depression or a clinical outcome.

Counting affirmative responses can describe the number of participants who trigger further screening under each layout. Interpreting a difference between groups would require the corresponding outcome counts, denominators, and a suitable comparison analysis. The primary outcome results and an effect estimate with its uncertainty are not reported in the registry record. The endpoint definition alone cannot show whether either layout changed response patterns or whether any observed difference would be precise.

The registry also lists secondary measures of time needed to complete the Whooley questions and EPDS, the proportion requesting technical assistance, and the number of assistance requests by type. These measures address usability and practical implementation, but their results are not reported in the registry record. They are separate from the primary count of affirmative responses.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

For a binary endpoint such as an affirmative response, the basic outcome is whether each participant meets the stated response definition. In this study, the definition is an affirmative answer to at least one Whooley question. A count can communicate how many participants met that definition in each layout group. To compare groups quantitatively, analysts generally also need the number assessed in each group and a method for expressing uncertainty around the comparison. The registry record does not report the primary results or the analysis used for this endpoint.

What does a confidence interval describe?

In general, a confidence interval is a range of values compatible with the observed data under a specified statistical model and procedure. It conveys precision as well as the direction and size of an estimated effect. A narrow interval indicates greater precision than a wide interval, while an interval spanning values that imply different practical interpretations may leave the result uncertain. The registry record does not report a confidence interval for the primary endpoint, so its precision cannot be assessed from the record.

Why does allocation matter?

In general, random allocation can help make comparison groups similar on average in factors that influence an outcome. When allocation is not randomized, or when its method is unknown, group differences may reflect both the intervention condition and differences in who entered each group. The registry explicitly notes that allocation was not stated and that the arms are not a like-for-like control comparison. It does not report whether or how analyses accounted for group differences. This limits causal interpretation of any contrast between layouts.

What do the planning assumptions mean?

The registry describes a sample-size planning illustration based on the observed effect. It specifies a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The intended effect to detect is described as a stated fraction of the observed effect. In general, a planning calculation asks what sample size would be needed to detect an assumed effect under chosen statistical assumptions. It is a hypothetical design exercise, not a measure of the trial’s observed ability to detect an effect. Sample-size illustrations should therefore not be described as observed power.

The record says that non-time-to-event planning values are expressed as participants per arm; time-to-event values, if applicable, are expressed as total events. No trial-specific sample-size justification is reported in the registry record. The planning assumptions do not establish that the completed study was designed around the primary endpoint, and the registry does not report the actual analysis plan or a planned effect size in the registry record describing the trial.

Why can over-dispersion matter?

In general, over-dispersion occurs when the observed variability in count data is greater than a statistical model assumes. If a model assumes that counts vary less than they actually do, standard errors may be underestimated and intervals may be too narrow. Analysts may use methods that allow extra variability when the data structure and model call for it. The primary endpoint in this record is a participant count, but the registry does not identify a count model, report dispersion, or state that over-dispersion was assessed. No claim about whether it was relevant to this analysis can be made from the registry record.

For a binary participant-level endpoint, a count model is not automatically required: a comparison can instead be expressed using proportions or another suitable measure, depending on the analysis plan and data structure. That is general statistical context, not a description of the method used in this study. The registry does not report whether the primary analysis used an unadjusted or adjusted method, how it handled any clustering by clinic, or whether any covariates were included.

What can the outcome schedule tell us?

The primary endpoint has a single stated assessment immediately after recruitment. The smartphone groups, by contrast, had repeated monthly assessment schedules described in the registry. Repeated measures can require methods that account for the fact that observations from the same participant may be related. In general, treating correlated observations as independent can misstate uncertainty. The registry does not provide a statistical model for the smartphone assessments, or explain how repeated observations were handled.

The registry lists three secondary measures related to survey completion time and technical assistance. A time outcome and a count of assistance requests may need different analytic approaches from a binary endpoint. The registry does not report the distributions, summary measures, or models used for these outcomes. Nor does it provide estimates or uncertainty for the secondary measures. Statistical interpretation of those outcomes is therefore not possible from the registry record alone.

Because the primary comparison is restricted to the two tablet groups, results from the app groups would address separate questions about assessment approaches. Combining the four groups into one simple comparison would not reflect the stated primary endpoint. The registry does not report a planned hierarchy of outcomes, prespecified contrasts, or an adjustment strategy for multiple testing. These methods details are necessary to understand how any collection of estimates should be interpreted.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("Screening - Scrolling Layout", "Screening - Paging Layout")
n      <- c(273, 257)
events <- c(62, 60)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  Screening - Scrolling Layout  273  62  22.7  17.9  28.1  18.1  28.0  18.1  28.0
ARM  Screening - Paging Layout  257  60  23.3  18.3  29.0  18.6  28.9  18.6  28.9

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 8 outcome measures and 7 analyses. It identifies the primary endpoint as the count of participants responding affirmatively to at least one Whooley question according to tablet survey layout, and lists secondary outcomes concerning survey completion time and technical assistance. However, the registry record does not report a full statistical analysis plan, a hierarchy among all outcomes, or which analyses were considered confirmatory.

When several outcomes or comparisons are tested, the chance of at least one apparently notable result can increase. In general, multiplicity procedures can control that risk, but the choice depends on which outcomes and comparisons were planned and how they relate. Whether any adjustment was applied is not reported in the registry record. The number of listed outcomes and analyses alone does not establish how multiplicity was handled.

Missing-data procedures are also not reported in the registry record. The record does not say how participants with incomplete questionnaires or missing follow-up assessments were treated, whether missingness was described by group, or whether sensitivity analyses were conducted. These details matter because the people with observed responses may differ from those without observed responses.

Analysis populations are not described. The registry record does not state whether analyses included everyone allocated to a group, only participants who completed an assessment, or another defined population. Since allocation itself is not stated, an intention-to-treat analysis cannot be assumed. The record also does not report how exclusions, protocol deviations, or repeated app assessments were handled. Without these details, the target population represented by any analysis cannot be established.

12. 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 non-serious adverse event terms

The registry safety entries list no serious events in any group: 0 for the screening scrolling layout, 0 for the screening paging layout, 0 for retrospective plus momentary assessment, and 0 for retrospective assessment. Deaths are also listed as 0 in each group. The corresponding numbers at risk are 273, 257, 174, and 175, respectively. These are group-specific registry entries and are not added together.

Zero recorded serious events means the registry lists none for those entries. It does not prove that no event of any kind occurred, and it does not show that the study could rule out uncommon harms. The record does not report the safety-event definitions, collection procedures, observation periods, or whether non-serious events were actively solicited. It also does not provide event narratives or a safety analysis.

Safety counts should be read separately from the primary endpoint, which concerns affirmative Whooley responses. The registry’s safety entries do not provide the missing primary comparison or its uncertainty. They also do not support a conclusion about the relative safety of the four approaches, particularly because allocation was not stated and the arms are not described as like-for-like control groups.

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

The record describes a study intended to assess the feasibility of using tablet and smartphone technology in antenatal mental health screening and monitoring. It defines a primary endpoint for comparing two tablet layouts: the number of participants who responded affirmatively to at least one Whooley question. It also describes app-based assessment schedules and secondary usability measures.

The primary endpoint results are not reported in the registry record. Accordingly, the record does not establish whether the scrolling and paging layouts produced different response counts, how large any difference was, or how precise a comparison might be. No conclusion about the relative performance of the layouts can be drawn from the endpoint definition alone.

The stated study aim concerns feasibility, not treatment efficacy. An affirmative screening response is not a diagnosis, and the listed primary measure does not by itself establish the presence or severity of depression. The registry record does not report clinical outcomes that would show whether screening or app use changed health, care, or treatment decisions.

The completed status and enrollment entry describe the study record, but do not answer questions about completeness of outcome reporting or statistical certainty. The record does not report how participants were allocated, how missing outcome data were handled, or which analysis population was used. Those omissions limit interpretation of any potential group comparison.

The safety entries list 0 serious events and 0 deaths in each group, but those entries should not be taken as proof that all risks were absent. The registry does not report enough detail about event collection or follow-up to support a broader safety conclusion. Overall, the record documents a feasibility question and its planned primary measure, while leaving the principal comparative results and key analysis details unreported.

14. Why This Trial Matters Statistically

This study raises a practical statistical question about comparing ways of presenting the same screening questions. In the tablet groups, the content included Whooley questions and the EPDS, while the display differed: all questions on a single screen versus one question at a time. A count of affirmative responses offers a defined endpoint for examining whether layout is associated with different response patterns. The record does not report the outcome results, so the comparison cannot be quantified.

The broader design also illustrates why feasibility outcomes and clinical outcomes should be kept distinct. Completion time and technical assistance can inform usability, while affirmative screening responses describe a screening measure. Neither category alone establishes whether a digital approach improves clinical outcomes. The registry’s stated focus on feasibility makes clear that implementation questions are central to the study description.

Statistically, the allocation caveat is important. The registry does not state how participants were allocated and explicitly cautions that the arms are not a like-for-like control comparison. A contrast between groups may therefore be difficult to interpret causally without information about how groups were formed and whether their characteristics differed. The record does not report adjustment methods or a full analysis plan.

The smartphone arms also represent different assessment schedules from the tablet comparison. Their repeated assessments raise distinct questions about participant burden and analysis of repeated observations. Keeping these questions separate avoids treating all four arms as if they addressed one common endpoint. The registry does not report results for the app assessments or describe how repeated observations were analysed.

Finally, this record shows the importance of complete outcome reporting. A clear endpoint definition is useful, but readers also need group results, measures of uncertainty, missing-data procedures, analysis populations, and multiplicity handling to judge evidence. Those details are not reported in the registry record, limiting statistical conclusions about the study.

15. 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 several trial-specific limitations unresolved. Most importantly, it defines the primary outcome but does not report its results. There is no primary effect estimate or measure of uncertainty in the record, so the comparative response pattern and its precision cannot be assessed.

Allocation is not stated. The record also cautions that the arms are not a like-for-like control comparison. Without a description of how participants entered groups or how groups differed at baseline, a between-group contrast cannot be interpreted as an isolated effect of survey layout. The record does not report adjustment for participant characteristics or recruitment setting.

The outcome and analysis descriptions are incomplete. The record lists 8 outcome measures and 7 analyses, but does not provide the statistical analysis plan, multiplicity approach, analysis populations, or missing-data procedures. It does not state how incomplete screening questionnaires or missed smartphone assessments were handled. These omissions limit assessment of risk of bias and the meaning of any analysis.

The primary endpoint is a screening response count, not a diagnostic or clinical outcome. The record does not report whether response patterns were confirmed against a diagnostic assessment, nor does it report later health outcomes. The secondary measures concern timing and technical assistance, but their results are not reported.

The feasibility context should also be kept in view. The registry describes tablet and app approaches, but does not report detailed measures of acceptability, adherence, or implementation success beyond the listed secondary outcomes. The safety entries list no serious events or deaths, but event definitions and follow-up details are absent. These limitations do not establish that the approaches were ineffective or unsafe; they describe what the registry record does not report.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: mHealth for Antenatal Mental Health — NCT02516982.
• Registry record last updated 2019-09-23; status last verified 2019-09.
• Results first posted 2019-09-23.
• 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.

19. Record Summary

NCT02516982 is a completed feasibility study sponsored by Imperial College London, with an enrollment of 946. The registry lists a start date of 2015-10 and a primary completion date of 2018-02. Its stated aim was to assess mobile technology for implementing antenatal mental health screening guidance and monitoring mood and depressive symptoms during pregnancy.

The primary endpoint counts participants responding affirmatively to at least one Whooley question in the scrolling and paging tablet-layout groups. The tablet surveys also included the EPDS. Two other groups used smartphone assessment schedules. Allocation was not stated, and the registry cautions that the arms are not a like-for-like control comparison.

The primary outcome results, statistical analysis plan, missing-data procedures, and analysis populations are not reported in the registry record. Safety entries list 0 serious events and 0 deaths in each group. These entries do not replace the unreported primary comparison or provide a complete account of safety.