Enhancing Patient Ability to Understand and Utilize Complex Information Concerning Medication Self-management: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Enhancing Patient Ability to Understand and Utilize Complex Information Concerning Medication Self-management (NCT02820038), a randomized na study. The primary outcome is Percentage of Participants Classified as Having Made an Informed Decision at 6 Months.

2. Clinical Question

3. Trial Design

The registry describes a randomized, triple-masked, factorial study with four arms. It lists rheumatoid arthritis as the condition and reports enrollment of 286 participants. The study began in 2016-09, reached primary completion on 2018-12-31, and is recorded as completed. The phase is listed as not applicable. The registry identifies the University of North Carolina, Chapel Hill as the sponsor.

The four arms combine two information approaches with the presence or absence of the SMART Program. The information approaches were Medication Guides or comparable Consumer Medication Information, and Drug details Boxes. Each approach was used either by itself or together with SMART. The arm descriptions say approximately 75 participants would be assigned to each group; these are described as approximate assignments, not the actual arm sizes in the posted results. The record’s enrollment figure does not establish how many participants were allocated to each arm or analyzed in each outcome.

The brief study description says data would be collected primarily through self-administered, Internet-based surveys using REDCap, with follow-up for 6 months after baseline data collection. The primary endpoint is assessed at the Month 6 follow-up. Secondary outcomes include knowledge of risks and benefits, measures of gist reasoning, satisfaction with medication information, treatment satisfaction, arthritis self-efficacy, and medication adherence. Several of these are described as adjusted for baseline.

In general, a factorial design arranges interventions as combinations of factors, allowing researchers to examine effects associated with each factor and potentially whether the factors work differently in combination. The registry’s factorial designation and the four arm descriptions are consistent with that structure. However, the registry record does not report the planned estimands, the primary statistical contrast, whether an interaction was a focus, or the model used to estimate effects. Triple masking is recorded, but the record does not specify which groups of people were masked.

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:

Analysis population for the primary outcome, as defined in the registry: Includes 221 participants who had sufficient data to classify as either meeting or not meeting the criteria for informed decision-making at the 6-month follow-up. (Excludes 65 people who had missing data at the 6-month follow-up.)

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 Percentage of Participants Classified as Having Made an Informed Decision at 6 Months, measured over Month 6 Follow-up.

The primary endpoint is the percentage of participants classified as having made an informed decision at the 6-month follow-up. The registry defines an informed decision as a value-consistent choice based on accurate knowledge. Classification depends on knowledge about medication risks and benefits, a values score, and whether the participant is currently taking one or more disease-modifying antirheumatic drugs.

Under the registry definition, a participant must meet one of two combinations of criteria. One combination requires at least 85% correct on the knowledge items, a score of 6 or more on the values scale, and current use of one or more DMARDs. The other requires 85% correct on the knowledge items, a score of 5 or less on the values measure, and no current DMARD use. Anyone who does not meet either combination is classified as not having made an informed choice. The values scale is described as a 10-item scale developed by Fraenkel et al., with scores ranging from 0 to 10; lower scores reflect reluctance to use medications to control disease activity.

This combined endpoint is used to represent whether a decision aligns with both knowledge and stated values, rather than treating knowledge alone or medication use alone as sufficient. It is a classification based on several pieces of information, not a direct measurement of health outcomes. The registry record reports that 221 participants had sufficient data for classification and that 65 with missing 6-month data were excluded. It does not report the percentage classified as informed in any arm, so the endpoint’s observed results and precision are not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Statistical methods are easiest to understand by separating the outcome definition, the comparison being estimated, and the uncertainty around that estimate. The registry provides a detailed rule for classifying the primary outcome but does not report the model, contrast, or inferential results. The explanations below describe general concepts; they are not claims about an unreported analysis plan.

What kind of outcome is the primary endpoint?

The primary endpoint is a binary classification: each participant with sufficient information is categorized as having or not having made an informed decision. A group-level percentage expresses the share of classified participants in one category. For a comparison between groups, the analysis could report a difference in percentages, a ratio, or another contrast, but the registry record does not report which effect measure was used. No contrast should be inferred from the endpoint title alone.

The classification rule is composite in the ordinary sense that it combines several conditions. Knowledge is assessed against a threshold, the values scale is divided by a score threshold, and current DMARD use is part of the qualifying combinations. This makes the endpoint understandable as an operational definition, but it also means its result depends on each component and on the chosen cut points. The record does not report results for the components within each arm, so it is not possible to identify which elements most often determined classification.

Why does the comparison matter?

Random allocation can support a causal comparison when the groups are analyzed according to an appropriate plan and follow-up is handled carefully. The design description alone does not show what estimate was calculated. With four arms arranged across information format and SMART participation, possible questions include an overall contrast between formats, an overall contrast for training, a comparison of individual combinations, or whether training effects differ by format. These are distinct questions and may require distinct estimates. Which were prespecified or reported is not reported in the registry record.

In general, a factorial analysis may estimate the average contribution of each factor while accounting for the other factor. An interaction asks whether the effect associated with one factor differs depending on the level of another factor. If an interaction is present, an average main effect can obscure differences between combinations. These are general explanations of factorial analysis, not a description of the trial’s fitted model. The record labels the model factorial but does not report whether an interaction term was included, how factors were coded, or which comparison was primary.

How should uncertainty be read?

General: an estimate describes the observed effect under a defined analysis, while a confidence interval conveys the estimate’s statistical uncertainty. A narrower interval indicates greater precision under the model, and a wider interval indicates less precision. An interval spanning no difference would generally be compatible with a range of effects that includes no difference, though interpretation depends on the chosen effect measure and interval method. The registry record does not give an interval for the primary endpoint.

General: a p-value is not the probability that a treatment works, nor does it measure the size or importance of an effect. It describes how unusual the data would be under a specified null model and its assumptions. A small p-value does not resolve concerns about bias, missing outcomes, multiple analyses, or practical importance. Because no primary endpoint p-value is reported in the registry record, no significance judgment can be made from this record.

What does the planning information mean?

Planning calculations are not observed outcomes. The registry’s sample-size assumptions describe an illustration based on the observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; the effect to detect is defined as a fraction of the observed effect. These assumptions describe how a planning illustration is framed, not the original trial design or an observed power calculation. For outcomes other than time-to-event outcomes, the stated convention is participants per arm; time-to-event illustrations use a total event count. No numerical planning estimates are reported in the record, and none can be reconstructed from the assumptions alone.

General: planning sample sizes depend on the effect size selected, the variability or event rate assumed, the significance level, allocation, and the chosen power target. A calculation based on an observed effect can be especially sensitive to uncertainty in that effect estimate. Such a calculation is best understood as a planning scenario, not evidence that the study had a particular probability of detecting an effect. It is not observed power.

What can be learned from the secondary outcomes?

The registry lists 24 outcome measures and 70 analyses. The listed secondary outcomes include knowledge of DMARD risks and benefits, several measures of gist reasoning, satisfaction with medication information, overall treatment satisfaction, arthritis self-efficacy, and medication adherence. Some are described as 6-month values adjusted for baseline. Adjustment generally means an analysis accounts for a baseline measurement when estimating a follow-up outcome; this can improve precision when appropriately specified. The registry record does not give the adjusted estimates, covariates beyond baseline where applicable, or the models used.

The multiplicity of outcomes and analyses makes it important to know which questions were primary and whether adjustments or a hierarchy were used. Testing many hypotheses can increase the chance of at least one apparently notable result by chance if the tests are treated as independent confirmatory tests without adjustment. The registry record does not report a multiplicity procedure or a hierarchy for the 24 outcomes and 70 analyses. The number of outcomes and analyses alone does not reveal how the investigators interpreted them.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists 24 outcome measures and 70 analyses, including the primary endpoint and multiple secondary outcomes. A study with many outcomes or comparisons can raise multiplicity concerns: when numerous statistical tests are conducted, the chance of at least one result appearing unusual under a null model can increase. This general concern does not establish that a particular result is false or that a specific adjustment was required. The registry record does not report a multiplicity correction, a testing hierarchy, or which analyses were treated as confirmatory.

The primary endpoint population includes 221 participants with enough data to classify informed decision-making at 6 months. The registry says 65 people were excluded because of missing data at that follow-up. It does not report the reasons data were missing, the number missing in each arm, or whether missingness was related to baseline characteristics or outcomes. It also does not report whether missing values were handled using imputation, weighting, a sensitivity analysis, or another method. Those details are not reported in the registry record.

In general, analyzing participants with observed outcomes can be unbiased under particular assumptions about missingness, but those assumptions cannot be verified from the count alone. If the likelihood of missing data is related to the unobserved outcome, a complete-case result may differ from the result that would have been obtained with full follow-up. Sensitivity analyses can explore how conclusions change under alternative assumptions; whether any were conducted is not reported.

The record gives an enrollment figure but does not provide arm-specific denominators for the primary analysis or explain the analysis population beyond the sufficient-data classification rule. It does not report whether participants were analyzed according to their randomized assignment, whether any were excluded for other reasons, or how protocol deviations were treated. Those aspects of the analysis population are not reported in the registry record.

11. 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 lists serious adverse event counts of 0 in each arm. It also lists other adverse event counts of 0 and deaths of 0 in each arm. These are the posted counts for the four groups: Other CMI Only, Other CMI & SMART Program, Drug details Boxes Only, and Drug details Boxes & SMART Program. The record lists participants at risk separately for each group, but the zero event counts do not by themselves establish how safety was monitored or what events were included.

In general, a zero count means no event was recorded in the stated category for the reported group and observation period. It does not demonstrate that the underlying event risk is exactly zero, and it does not establish that every possible adverse experience was captured. The precision of a safety estimate depends on the number observed, the period of observation, event definitions, and ascertainment. The registry record does not describe the safety observation window, adverse event definitions, collection methods, or whether events were assessed in a standardized way.

The counts should therefore be read narrowly as registry-reported events, not as a comparison proving equivalence or absence of harm. The record does not report inferential comparisons for these safety counts. It also does not list serious adverse event terms. No broader safety conclusion is supported by the counts alone.

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

The study addresses an important statistical distinction: conveying information is not the same as ensuring that a person can use it to make a choice aligned with knowledge and values. The registry’s primary endpoint attempts to capture that combination by applying an explicit classification rule at the 6-month follow-up. The four-arm design pairs two medication-information formats with the presence or absence of gist reasoning training.

What the record does show is the planned structure, the endpoint definition, enrollment of 286 participants, and that 221 participants had sufficient follow-up information for the primary classification. It also reports that 65 people were excluded from that classification because of missing 6-month data. The record lists several secondary outcomes and zero reported serious adverse events, other adverse events, and deaths in each arm.

What the record does not show is the primary endpoint percentage in any group, the difference between groups, an uncertainty interval, or a significance test. Without those results, it cannot establish whether one information format was associated with more informed decisions than the other, whether SMART was associated with a change, or whether combining a format with training altered the outcome. The design can frame these comparisons, but the design is not itself evidence of an effect.

The endpoint should not be interpreted as a direct measure of health improvement, medication appropriateness, or long-term adherence. Its definition describes a choice classified as informed through knowledge, values, and current DMARD use. It does not directly establish that a particular choice was medically preferable or that participants experienced better clinical outcomes. The registry also lists medication adherence as a secondary outcome, but its results are not reported in the record.

Any statistical conclusion would need to take account of the missing follow-up information, the exact analysis population, the selected contrasts, and the multiple outcomes and analyses. Those details and the comparative estimates are not reported in the registry record. The appropriate interpretation is therefore limited to what the registry describes: a randomized factorial study of communication strategies and training, with a defined 6-month informed-decision endpoint and no posted primary comparative result in this record.

13. Why This Trial Matters Statistically

This trial is statistically notable because the question involves more than comparing two versions of written medication information. The four arms combine information format with a reasoning-training program, creating a factorial structure. That structure can address whether results are associated with the information approach, the training, or their combination, provided the comparisons and analysis are clearly specified and reported.

The primary endpoint also illustrates how a meaningful concept is converted into a measurable outcome. Informed decision-making is defined through knowledge, values, and medication status. The threshold-based rule allows participants to be classified consistently, but it makes interpretation dependent on the selected thresholds and the quality of each component measure. Reporting component results alongside the combined classification could help explain what drives the endpoint; such results are not reported in the registry record.

The distinction between enrollment and outcome availability matters as well. The registry records enrollment of 286 participants, while the primary classification includes 221 people and excludes 65 with missing 6-month data. Those figures underscore why an outcome denominator and a missing-data account are important when interpreting a percentage. Without arm-specific denominators, reasons for missingness, and analysis details, readers cannot assess how the observed classification population relates to random assignment.

Finally, the record lists 24 outcome measures and 70 analyses. In studies with numerous outcomes and comparisons, a clear account of the primary contrast and multiplicity strategy helps readers distinguish confirmatory findings from exploratory ones. The record does not report that strategy or any comparative primary results. Its statistical importance lies in the questions and measurement structure it sets up, while the evidence needed to answer those questions is not reported in the registry record.

14. 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 main limitation for interpreting this trial from the registry record is that comparative primary endpoint results are not reported. There are no arm-specific informed-decision percentages, effect estimates, confidence intervals, or p-values. The record therefore supports explanation of the endpoint and study structure, but not assessment of the observed comparative effect or its precision.

The primary analysis population is described as 221 participants with sufficient data for classification, with 65 people excluded because of missing 6-month follow-up data. The record does not explain why data were missing, how missingness was distributed across arms, or how missing outcomes were handled in the analysis. Without this information, the potential effect of incomplete follow-up on the outcome estimate cannot be assessed.

Although the study is identified as randomized, triple-masked, and factorial, the registry record does not specify the primary contrast, the statistical model, the handling of interactions, or which people were masked. It also does not report whether the primary analysis followed randomized assignment or how deviations from assignment were handled. These omissions limit assessment of how the design was translated into the analysis.

The registry lists many outcomes and analyses but does not report a multiplicity approach or a hierarchy of hypotheses. The listed outcome information also does not include results for the secondary measures. The safety record reports zero events in the listed categories, but definitions, observation periods, and ascertainment procedures are not reported. These limitations are specific to the information available in the registry record and prevent a fuller assessment of efficacy-related results and safety reporting.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Enhancing Patient Ability to Understand and Utilize Complex Information Concerning Medication Self-management — NCT02820038.
• Registry record last updated 2020-01-18; status last verified 2019-11.
• Results first posted 2020-01-18.
• 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.

18. Record Summary

The registry describes a completed randomized, triple-masked, factorial study of medication information approaches and gist reasoning training for people with rheumatoid arthritis. The four arms used Medication Guides or comparable Consumer Medication Information, or Drug details Boxes, either alone or with the SMART Program. Enrollment was 286, and the primary endpoint was the percentage classified as having made an informed decision at 6 months.

The registry reports that 221 participants had sufficient data for primary endpoint classification and that 65 people with missing follow-up data were excluded. It lists 24 outcome measures and 70 analyses, including secondary measures of knowledge, gist reasoning, satisfaction, self-efficacy, and medication adherence. Serious adverse events, other adverse events, and deaths are each recorded as 0 in every arm. The primary comparative percentages and statistical estimates are not reported in the registry record.