Effectiveness of Two Self Assessment Tools: KIOS-Bipolar or eMoods: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Effectiveness of Two Self Assessment Tools: KIOS-Bipolar or eMoods (NCT02418910), a registered phase not stated study. The primary outcome is Comparison of Two Brief Assessment Measures: KIOS Bipolar or eMOODs.

2. Clinical Question

3. Trial Design

The registry describes a study titled as a randomized, open 52-week comparison of KIOS-Bipolar and eMoods. The structured record, however, does not state the allocation method, and its analysis metadata flags that this is not a randomized comparison. The title alone is not enough to establish how participants were assigned or whether the comparison supports a randomized interpretation. The registry describes the participants as bipolar outpatients in any clinical state.

Enrollment was 50. The sponsor was The University of Texas Health Science Center at San Antonio. The recorded start was 2015-10, the primary completion was 2017-04, and the status is COMPLETED. The listed arm description concerns bipolar outpatients in any clinical state; arm-specific descriptions explaining which participants used which tool are not reported in the registry record. Therefore, the record does not establish a like-for-like control comparison between the tools.

The primary outcome timeframe is 52 weeks. Its description concerns assessment completion on time, completion only after follow-up, and weeks not submitted. The registry lists one outcome measure and no analyses. It does not report the statistical plan, analysis population, or comparative estimates. The project summary also describes an earlier KIOS field trial involving twenty bipolar patients over twelve weeks. That background is distinct from the present study and should not be treated as a result of this 52-week comparison.

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 Comparison of Two Brief Assessment Measures: KIOS Bipolar or eMOODs, measured over 52 weeks.

The primary outcome concerns participant assessment behavior over 52 weeks. It distinguishes assessments completed on time from those completed only after follow-up, and also describes the number of weeks in which an assessment was not submitted. These categories address whether a self-assessment tool can be used consistently in the study’s follow-up context. They measure completion patterns, not change in bipolar symptoms, symptom severity, or clinical status.

Completion measures can be useful when evaluating a self-monitoring tool because the usefulness of collected self-reports depends in part on whether assessments are recorded. Separating on-time completion from completion after follow-up retains information about the support needed to obtain a report, rather than treating every completed assessment as equivalent. Weeks not submitted can also indicate gaps in the record. These are general measurement considerations; the registry record does not explain why the study selected this endpoint or how the categories were operationalized.

The stated unit is participants, while the description also refers to weeks not submitted. The record does not clarify how participant-level counts and week-level missingness were combined or analyzed. It does not report endpoint results, group-specific denominators, or a comparison estimate. Consequently, the endpoint definition indicates what the study intended to track but not whether completion differed between KIOS-Bipolar and eMoods.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry record lists a binary outcome kind and describes the meaning as participants with an event (counts). The primary endpoint’s written description, however, includes completion categories and the number of weeks not submitted. The record does not explain how those elements were converted into an analysis variable or which event definition was used. This limits what can be said about the statistical model or the interpretation of a reported count. No analysis results are listed.

What would a count represent?

In general, a count records how often a defined event occurs, such as a participant meeting a specified assessment-completion definition. A count needs a clear event definition, a time window, and a denominator or exposure context to be interpretable. For this outcome, the timeframe is 52 weeks, but the record does not specify whether the primary comparison used a participant-level category, a count of completed assessments, a count of missed weeks, or another derived measure. The description should therefore be read as an outline of intended measurement rather than a complete analysis specification.

When comparing counts, the analysis method depends on how the observations arise. A simple comparison of participant counts may differ from an analysis that uses repeated weekly observations, because repeated observations from one participant are related rather than independent. This is a general statistical consideration, not a claim about the method used in this study. The registry record does not report whether the analysis accounted for repeated assessments or within-participant dependence.

Why can over-dispersion matter?

For count data, a model may assume a particular relationship between the average count and its variability. Over-dispersion means that the observed variability is greater than the model assumes. If present and not addressed, uncertainty estimates can be too small, which can make evidence appear more precise than it is. This is a general explanation. The registry record does not report a count model, a dispersion assessment, or a method for handling over-dispersion, so it is not possible to say whether this issue arose or how it was addressed.

How should the planning illustrations be read?

The planning assumptions specify a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is described as a stated fraction of the observed effect. These assumptions define planning illustrations, not the trial’s own reported design justification. The record does not report a sample-size justification for the trial, and the planning values should not be described as observed power. A planning calculation asks what sample size might be needed under specified assumptions; it does not show the probability that the completed study detected an effect.

The planning note also distinguishes the way sample sizes are expressed: for this non-time-to-event setting, the relevant presentation is participants per arm for 80% power or another stated power level. The actual numerical planning results are not included in this record. The observed effect used as a planning reference is also not reported in the registry record, so the illustration cannot be independently interpreted from the record alone. Planning for a fraction of an observed effect is a sensitivity exercise around an assumed target, not evidence that such an effect was observed in the current comparison.

What does a confidence interval add?

In general, a confidence interval communicates the precision of an estimate. It helps distinguish an estimate that is compatible with a broad range of possible differences from one that is more tightly estimated. A significance test alone, even if reported, would not describe the range of effects compatible with the data. The registry record lists no comparative estimate, confidence interval, or test result for the primary outcome. It is therefore not possible to judge either the direction or precision of a tool comparison from the recorded results.

Interpretation also depends on the design and analysis population. The title uses the term randomized, but allocation is not stated in the structured information, and the record flags that the comparison is not randomized. The registry record does not report which participants were included in any analysis or how deviations from intended tool use were treated. These omissions prevent a detailed assessment of the estimand: the precise comparison the analysis was intended to estimate.

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("Bipolar Outpatients in Any Clinical State")
n      <- c(43)
events <- c(43)
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  Bipolar Outpatients in Any Clinical State  43  43  100.0  91.8  100.0  91.8  100.0  92.0  99.9

11. Multiplicity, Missing Data and Analysis Populations

The registry record lists one outcome measure and zero analyses. It identifies the assessment-completion outcome as primary, but does not report a statistical analysis plan, a prespecified hypothesis test, or a hierarchy of outcomes. It also does not state whether additional outcomes or comparisons were assessed outside the listed record. Accordingly, it is not possible to determine how multiplicity was considered.

In general, multiplicity arises when several hypotheses, outcomes, time points, or subgroup comparisons are examined. Testing many possibilities can increase the chance of finding an apparently unusual result by chance. Prespecifying a primary outcome and analysis can help readers distinguish confirmatory questions from exploratory ones. The registry record identifies a primary outcome but does not provide enough detail to determine whether further statistical comparisons were planned or how they would be interpreted.

The endpoint description includes on-time completion, completion only after follow-up, and weeks not submitted. These are related but distinct features of assessment completion. The record does not clarify whether they were separate outcomes, components of a combined outcome, or descriptive categories. Nor does it state whether an analysis adjusted for multiple comparisons among them. No inference about multiplicity control can be drawn from the outcome description alone.

Missing data are central to an endpoint that explicitly tracks weeks not submitted. The registry record does not report the extent or pattern of missing assessment information, reasons for non-submission, or any imputation or sensitivity analysis. In general, missingness can affect interpretation if participants with unrecorded assessments differ systematically from those with recorded assessments. The endpoint’s inclusion of missed weeks may describe missingness, but it does not by itself explain how missing observations were handled analytically.

The record also does not report an analysis population, such as which participants would be included according to assignment or actual tool use. With allocation not stated and no analyses listed, it is not possible to determine whether results were intended to reflect assignment, tool use, or another population. These omissions limit both reproducibility and assessment of possible bias.

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.

The registry safety record lists the group as bipolar outpatients in any clinical state. For serious adverse events, it reports 43 at risk and 0 affected. For other adverse events, it reports 43 at risk and 0 affected. Deaths are reported as 0, and no serious adverse-event terms are listed.

These are the counts recorded in the registry, not a comparison between KIOS-Bipolar and eMoods. The record does not provide tool-specific safety counts, an explanation of the risk denominator, or details about event collection and follow-up. It also does not report how adverse events were defined or whether all types of events were captured. The absence of recorded affected participants in these fields should be described as a registry finding, not as proof that no unrecorded events occurred.

Safety counts answer a different question from the primary endpoint, which concerns assessment completion. They do not indicate whether participants completed assessments on time or how many weeks were not submitted. Because the registry record does not report comparative outcome analyses, the safety figures cannot establish whether either tool performed better on the primary measure.

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

The record identifies a study of KIOS-Bipolar and eMoods, with assessment completion over 52 weeks as the primary topic. It describes the population as bipolar outpatients in any clinical state and records enrollment of 50. The endpoint distinguishes timely completion, completion only after follow-up, and weeks not submitted. Those details show what the study set out to track, but the registry record does not report the comparative results.

As a result, the record cannot establish whether completion patterns differed between the tools, the direction of any difference, or the precision of an estimate. No group-specific endpoint counts, effect estimate, confidence interval, or analysis result is listed. A completed study status does not resolve these gaps. Nor can the project description’s account of an earlier KIOS field trial be treated as evidence for the outcome of this comparison; it concerns a separate study and a different stated focus.

The title characterizes the study as randomized and open, but the structured allocation information is not stated and the record flags the comparison as not randomized. The result should therefore not be interpreted as demonstrating a causal difference between tools. Even if comparative counts were available, their interpretation would depend on how participants were assigned, which participants were analyzed, and how missing assessments were handled. Those details are not reported in the registry record.

The endpoint is about assessment behavior, not symptom improvement or treatment effect. Although the project description explains that KIOS was developed to support bipolar disorder management, the primary outcome listed for this study does not measure whether a participant’s health improved. The posted information supports a description of the study question and the limits of the registry record, but not a conclusion that either tool is more effective, more acceptable, or safer.

14. Why This Trial Matters Statistically

This trial is statistically interesting because its primary outcome focuses on whether self-assessments are recorded over a long follow-up period, including distinctions between timely completion and completion after follow-up. Those distinctions can matter when evaluating the practical quality of repeated self-report data. A completed assessment and an on-time assessment are not necessarily equivalent indicators of engagement, and weeks without submission can reveal gaps that a single overall completion label might obscure.

The endpoint also raises a measurement question: its description refers both to participants and to weeks not submitted. Clear specification of the unit of analysis is important when a study collects repeated information. Participant-level counts and repeated week-level observations have different statistical structures. The record does not explain how these components were combined or modeled, making the endpoint difficult to reproduce from its description alone.

The allocation information is another important feature. A title that describes randomization does not establish that assignment was randomized when the structured record leaves allocation unstated and flags the comparison as not randomized. This distinction matters because design determines how confidently observed differences can be attributed to the tools rather than to differences among participants or other factors.

Finally, the record illustrates the difference between study completion and reported evidence. The status is COMPLETED, but no analyses or comparative outcome results are listed. Without estimates and uncertainty measures, readers cannot determine the magnitude or precision of any difference. The statistical importance of the record is therefore as much about the missing elements needed for interpretation as about the intended comparison itself.

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 does not report comparative primary outcome results, despite the COMPLETED status. It lists zero analyses and provides no group-specific counts, effect estimate, uncertainty interval, or test result. This prevents evaluation of the direction, size, and precision of any difference in assessment completion.

Although the study title describes a randomized, open study, the allocation method is not stated in the structured record, which also flags the comparison as not randomized. The record does not provide arm-specific descriptions sufficient to establish how tool use was organized. This limits assessment of whether the comparison was balanced or whether an observed difference could be attributed to tool choice.

The endpoint description combines participant assessment categories with the number of weeks not submitted, but does not define their operational relationship or the analysis unit. The record also does not report the number of participants contributing endpoint data, the pattern or reasons for missing assessments, or the method used to handle missingness. An analysis population and any adjustment for repeated observations are not described.

The statistical analysis plan, sample-size justification, and any interim analyses are not reported in the registry record. The planning assumptions describe illustrations based on an observed effect, but the observed effect and numerical planning outputs are not listed. Safety reporting is limited to the recorded counts and does not explain the event definitions or collection procedures. These limitations are specific to the information available in this trial record and constrain both reproducibility and interpretation.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Effectiveness of Two Self Assessment Tools: KIOS-Bipolar or eMoods — NCT02418910.
• Registry record last updated 2019-04-01; status last verified 2018-08.
• Results first posted 2019-04-01.
• 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

NCT02418910 is a COMPLETED study sponsored by The University of Texas Health Science Center at San Antonio. Its title describes a randomized, open 52-week comparison of KIOS-Bipolar and eMoods for people with bipolar disorder. The recorded enrollment is 50; the start was 2015-10 and primary completion was 2017-04.

The primary outcome concerns assessment completion on time, completion only after follow-up, and weeks not submitted. The record lists one outcome measure and zero analyses. Although the title says randomized, allocation is not stated in the structured record, and the comparison is flagged as not randomized. Comparative outcome results and an analysis plan are not reported in the registry record.

Safety fields list 43 at risk and 0 affected for serious adverse events, 43 at risk and 0 affected for other adverse events, and 0 deaths. No serious adverse-event terms are listed. These figures do not provide a comparison between the tools.