Differing Completion Rates of DIABETES Education on Patient Reported Outcomes: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Differing Completion Rates of DIABETES Education on Patient Reported Outcomes (NCT06419907), a randomized na study. The primary outcome is Change in DSMQ Diabetes Self Management Questionnaire Measuring Diabetes Self Management Skills Pre and Post Intervention.

2. Clinical Question

The trial also measured diabetes distress and mental and physical health-related quality of life. Participants completed three validated patient-reported outcome measures at baseline and again after the intervention; the registry describes follow-up as 2-4 months after intervention, with a separate 3-4 months from baseline timeframe for the delayed-education group. The aim was feasibility and assessment of outcome patterns, not a definitive test of comparative effectiveness. The posted results should therefore be read as descriptive evidence from the observed participants, with limited precision and incomplete information about the analysis procedures.

3. Trial Design

This was a randomized, open-label, parallel-group feasibility study of a behavioral diabetes self-management education programme. The registry record lists an enrollment of 120 and identifies the study as completed. The study began on 2024-05-16, and primary completion was 2025-05-20. The four groups differed in the amount of the face-to-face Type 2 Live Well session offered during the study: full attendance, 60% attendance, 10% attendance, or delayed education without attending the session during the comparison period.

The planned session was six hours. Group 1 received 100% of the session; Group 2 received 60%, or 3 hours 36 minutes; Group 3 received 10%, or 36 minutes; and Group 4 did not attend a face-to-face session. The registry describes the study as randomized but does not report the sequence-generation procedure, allocation concealment, or how participants were assigned in practice. Masking was listed as none, which is consistent with participants knowing whether they attended an education session. Whether outcome assessors or analysts were masked is not reported in the registry record.

Participants completed measures of self-care, diabetes distress and quality of life before and after the intervention. The main follow-up timeframe was described as 2-4 months after intervention; the primary outcome record gives 3-4 months post baseline, and the delayed-education group was described as repeating measures 3-4 months from baseline. The posted primary outcome includes 30 participants in Group 1, 27 in Group 2, 28 in Group 3 and 28 in Group 4. The outcome description states that 3 participants were lost to follow up and 4 were removed as not per protocol. It is not reported in the registry record how these exclusions were distributed across groups or whether they were applied consistently to the secondary outcomes.

The record labels the study phase as not applicable and describes it as a feasibility study. No protocol-level sample-size justification or detailed statistical analysis plan is reported in the registry record. The posted results include descriptive changes by group and selected unadjusted pairwise summaries for secondary outcomes, but no reported analysis for the primary endpoint.

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: Patients with a recorded clinical diagnosis of type 2 diabetes aged ≥18 years who met local and national eligibility criteria for DSME who could provide written informed consent and were responsible for daily management of their diabetes. Exclusion criteria include: * Lack capacity to make an informed decision. * Type 1, type 3c, monogenic or gestational diabetes. * Received structured education for diabetes within last 12 months either online or face to face. * Require 1:1 education

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: 3 lost to follow up, 4 removed as not per protocol

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

Population: Adherence was assessed during the study by reviewing the number of dropouts during the study and non-adherence to the randomised intervention. Four participants were not included in the final analysis due to protocol deviations: three attended structured education during the trial to meet their medical needs, one of whom was of mixed ethnicity, and another completed 100% DSME instead of the randomised 60%. A further 3 were lost to follow up.

6. Primary Endpoint

The registry lists the primary outcome as Change in DSMQ Diabetes Self Management Questionnaire Measuring Diabetes Self Management Skills Pre and Post Intervention, measured over 3-4 months post baseline.

The primary endpoint was change in the Diabetes Self-Management Questionnaire (DSMQ) score from baseline to follow-up. The registry describes the questionnaire as a 20-item measure of diabetes self-management skills, with some items reverse-scored. Its total is calculated by adding item scores and dividing by 60. The described scale runs from 0 to 10, with a higher score indicating better self-management. Change was defined as follow-up score minus baseline score, so a positive change represents an increase in the score.

A patient-reported measure is used to capture participants’ own accounts of self-management activities rather than relying only on clinician observations or biomedical measurements. That makes the endpoint relevant to a programme intended to support day-to-day self-care. The registry describes the DSMQ as having good interrater reliability, but does not report supporting reliability estimates for this study or evidence about how participants’ scores relate to meaningful changes in their lives.

The primary outcome record reports the median change for each group and gives a standard deviation as the dispersion measure. A median describes the middle of the observed change scores, while a standard deviation describes spread around the mean; these measures summarize different features of a distribution. No confidence intervals, formal between-group estimates or reported primary statistical tests appear in the registry record. The group summaries can therefore describe the observed changes, but do not establish the precision of a difference between attendance levels.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does a change score represent?

A change score is the follow-up measurement minus the baseline measurement. For the DSMQ, the registry explicitly states this direction: follow-up minus baseline. Positive values therefore correspond to higher follow-up scores than baseline, and the instrument description says that higher scores indicate better self-management. A change score expresses within-participant movement on the outcome scale. Comparing the average or median changes across groups is a separate step, because individual changes vary and the group estimates have sampling uncertainty.

For the primary endpoint, the registry reports medians alongside standard deviations. The median is a rank-based measure of the center of observed values, while the standard deviation is a measure of dispersion around the mean. A median and standard deviation can both be informative, but they do not describe the distribution in the same way. The record does not explain why these two summaries were selected together, whether the distribution was examined, or what inferential method was intended for the primary comparison.

How should a confidence interval be read?

A confidence interval describes uncertainty around an estimated quantity under the assumptions of the analysis. For the posted secondary pairwise summaries, the estimated quantity is a mean difference between Group 2 or Group 3 and Group 1. An interval containing zero is compatible with no difference under the analysis model, as well as with differences in either direction represented by the interval. An interval that does not contain zero indicates that zero is outside that interval, but does not by itself show that the result is clinically important, robust to multiple testing, or free from bias. The registry record does not state a clinical importance threshold for these outcomes.

The estimate and its interval should be considered together. A point estimate is one best summary from the observed data; the interval conveys how uncertain that estimate is. Wide intervals can be consistent with meaningfully different possibilities. The secondary results include intervals of varying widths, so the precision differs across comparisons. No confidence intervals are reported for the primary endpoint.

What do the unadjusted p-values mean?

A p-value evaluates how compatible the observed data are with a specified null hypothesis under the assumptions of a statistical model. It is not the probability that the null hypothesis is true, and it does not measure the size or importance of an effect. The secondary comparison values are explicitly unadjusted. They do not account for the fact that multiple outcomes and group comparisons were considered. If many tests are examined, the chance of at least one apparently unusual result can increase. The registry record does not report whether the trial team planned a correction, prioritized a particular endpoint for inferential testing, or treated these tests as exploratory.

The posted secondary summaries include point estimates, confidence intervals and unadjusted p-values for comparisons against Group 1. They are not identified as the sponsor’s reported analyses, and the record lists no reported analyses for these outcomes. They should therefore be read as statistical summaries of selected comparisons, not as a substitute for a complete prespecified analysis. The primary endpoint has descriptive group summaries but no posted inferential comparison.

Why does variability matter?

Variability describes how far participant-level observations tend to spread. When outcome values vary substantially, a group estimate is less precise for a given sample size than it would be with less variation, all else equal. This is one reason that a modest study may yield wide intervals even when the group point estimates differ. For the secondary outcomes, the registry reports different measures of dispersion: standard errors for diabetes distress and standard deviations for the quality-of-life outcomes. A standard error describes uncertainty in an estimated mean, whereas a standard deviation describes spread among observations. They are not interchangeable, and the posted summaries should be interpreted using the measure identified for each outcome.

How should the planning illustrations be interpreted?

The planning illustrations use observed effect estimates to show hypothetical participants per arm for 80% or 90% power under stated assumptions: a two-sided significance level of 0.05, equal allocation between two arms, a normal approximation, and an effect set to a stated fraction of the observed effect. The illustrated fractions are 100%, 75% and 50% of the observed effect. These are planning scenarios, not the trial’s original sample-size calculation, and not observed power. They do not establish that the trial was adequately powered or that an outcome is important.

The illustrations vary considerably across outcomes and comparisons because they depend on the observed estimate and variability. For example, the posted illustration for the diabetes-distress comparison of Group 2 with Group 1 gives 2,761 participants per arm for 80% power when planning for 100% of the observed effect, while the Group 3 comparison gives 90 participants per arm under that same scenario. For mental health quality of life, the corresponding figures are 131 and 51 participants per arm. These values are sensitive to the original observed effects, which may be imprecise in a feasibility study; they should not be treated as reliable definitive design targets. The record does not report the original sample-size justification.

Planning for a smaller fraction of an observed effect increases the illustrated sample sizes. This reflects a general statistical principle: detecting a smaller difference against the same background variability usually requires more participants. The illustrations are pairwise and assume equal allocation between two arms. They do not constitute a design for the full four-group study, account for the trial’s missing participants, or specify adjustments for multiple outcomes. They are therefore useful only as conditional planning examples under the stated assumptions.

10. Multiplicity, Missing Data and Analysis Populations

The record does not report a statistical plan addressing multiplicity. This matters because several outcomes were measured and the posted secondary results include more than one comparison. The reported secondary p-values are unadjusted. Without a prespecified hierarchy, adjustment method or distinction between confirmatory and exploratory tests, a nominal result should not be interpreted as if it were the sole planned test. The record also does not report an overall test comparing all four groups for the secondary endpoints.

The primary endpoint is identified in the registry, but the record does not report a primary between-group analysis, confidence intervals or a test statistic. It also does not report a detailed analysis plan specifying how the median change and standard deviation would be used, which population would be analyzed, or how baseline measurements would enter the model. No interim analyses or stopping rules are reported in the registry record. Their absence from the record does not establish whether they were considered elsewhere.

The primary outcome description states that 3 participants were lost to follow up and 4 were removed as not per protocol. It does not report the reasons for loss to follow up, the timing of missingness, or the group distribution of exclusions. The posted primary group sizes are 30, 27, 28 and 28. The secondary summaries use the same group counts for their listed arms, but the record does not explain whether all outcomes were available for every participant or whether missing items within questionnaires were handled.

Missing-data methods are not reported in the registry record. It is not stated whether analyses used complete cases, imputation, or another approach. The record does not identify an intention-to-treat population or explain the per-protocol exclusions beyond describing 4 participants as not per protocol. Excluding participants after randomization can change the balance created by random assignment, depending on who is removed and why. Without reasons and a prespecified population definition, it is difficult to assess that risk.

These omissions limit reproducibility and interpretation. The posted summaries can be read as results for the participants included in each outcome summary, but the registry record does not provide enough detail to determine how missingness, exclusions, multiplicity or model choice affected the estimates.

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.

Most frequent serious adverse event terms

Serious events with the highest arm-level rate: none was deemed to be study related and all were exacerbations of existing health conditions. (10.0%).

The serious adverse event summaries list 3 affected participants among 30 at risk in Group 1, none among 27 in Group 2, none among 28 in Group 3, and 1 among 28 in Group 4. No non-serious adverse events were listed in any group, and no deaths were recorded. The registry states that none of the serious events was deemed study related and that all were exacerbations of existing health conditions.

These counts describe what was recorded among the participants at risk; they do not establish that the intervention caused or prevented events. A zero count in a group does not show that an event is impossible, especially in a feasibility study with these group sizes. The record does not provide event-level descriptions, timing, duration, severity beyond the serious classification, or the basis for judging relatedness. It also does not report a formal between-group safety analysis.

Safety information should therefore be read descriptively. The posted data indicate that serious events were recorded in Group 1 and Group 4, with none recorded in Groups 2 and 3, and that the listed events were considered unrelated to study participation. The counts are too limited to support broad comparative safety conclusions.

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

The posted results describe a randomized feasibility study of different attendance levels at a structured face-to-face diabetes education programme. They show group-level changes on patient-reported measures among the participants included in the outcome summaries. The primary DSMQ medians are positive in Groups 1, 2 and 3, and negative in Group 4. Those summaries are informative descriptions of the observed participants, but no formal primary comparison is reported. They therefore do not establish that one attendance level produced a larger change than another.

The secondary outcomes provide additional descriptive and selected pairwise information. Their estimates and intervals differ in direction and precision, and their unadjusted p-values must be considered in the context of multiple outcomes and comparisons. The registry record does not report clinically important thresholds, so the numerical changes alone cannot establish whether participants experienced a meaningful improvement or worsening. Nor can the posted secondary results answer the primary question in place of a primary between-group analysis.

The design is randomized, but randomization alone does not guarantee that all post-randomization analyses preserve the benefits of random assignment. The record reports participants lost to follow up and removed as not per protocol, without explaining their distribution or the approach to missing outcomes. The analysis population and missing-data methods are not reported in the registry record. These features leave uncertainty about how representative the analyzed participants are of all randomized participants.

The results do not establish that a particular completion rate is effective, ineffective, equivalent to another rate, or clinically preferable. In particular, a lack of a reported difference is not evidence of equivalence, and a nominal secondary p-value is not proof of a causal effect. The study’s feasibility focus, modest group sizes, selected comparisons and incomplete reporting of the analysis plan mean that the evidence is preliminary. Its value is in documenting observed outcome patterns and highlighting the need for clearly specified analyses and adequately planned comparative evaluation.

13. Why This Trial Matters Statistically

This study addresses a statistical challenge common in behavioral and feasibility research: the intervention is not simply present or absent, but varies by the amount of programme attendance. The four-group design distinguishes full attendance, partial completion, minimal attendance and delayed education. That structure can help researchers formulate questions about dose or completion, but it also requires a clear plan for which contrasts are primary and how the groups will be compared.

The registry identifies a primary patient-reported endpoint, yet the posted results provide group summaries without a reported primary inferential comparison. This distinction is important: descriptive summaries can show the distribution of observed outcomes, but do not by themselves quantify uncertainty in the between-group contrast that answers the main research question. A future definitive study would need a transparent primary analysis and a clear account of how missing observations and post-randomization exclusions are handled.

The posted secondary comparisons illustrate why effect estimates, intervals and p-values should be read together. A point estimate can look different from zero while remaining imprecise, and an unadjusted p-value does not address the multiplicity created by several outcomes and contrasts. The sample-size illustrations further show how planning depends on the observed effect and assumptions; they are not a substitute for a prospective design justification based on a clinically meaningful target.

As a feasibility study, the trial can contribute information about the design and measurement of a larger evaluation. However, the registry record does not report operational feasibility outcomes such as recruitment or attendance rates in a way that allows those aspects to be assessed from the posted results. The main statistical lesson is that a randomized structure is valuable, but its interpretive strength depends on prespecified contrasts, appropriate uncertainty estimates, transparent missing-data handling and careful control of confirmatory claims.

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 primary endpoint is reported using a median and standard deviation, but the registry does not explain the analysis rationale or provide a distributional assessment. No primary confidence intervals or formal between-group tests are posted. As a result, the primary outcome cannot be interpreted with the same inferential detail as the selected secondary summaries.

The outcome records describe follow-up using different time references: the study summary gives 2-4 months after intervention, the primary outcome specifies 3-4 months post baseline, and the delayed-education group is described as measured 3-4 months from baseline. The registry does not explain whether the exact assessment timing differed by group or how timing variation was handled analytically. Such variation may complicate comparisons if participants were assessed at different intervals.

There are incomplete details about participant flow. The primary outcome description reports 3 lost to follow up and 4 removed as not per protocol, but does not give group-specific reasons or clarify whether removal occurred before or after outcomes were assessed. Missing-data procedures and the analysis population are not reported. These limitations make it difficult to assess potential selection effects or reproduce the analyses.

The posted secondary comparisons cover only two intervention groups against Group 1. They do not include all possible group contrasts, an overall four-group test, or a posted comparison against delayed education. The registry does not report multiplicity handling, a prespecified statistical analysis plan, or clinically important difference thresholds. Consequently, exploratory secondary patterns should not be treated as confirmatory evidence.

The sample-size illustrations are based on observed effects and stated planning assumptions, rather than the trial’s own reported design justification. The record does not report a prospective target effect or rationale for enrollment. Finally, the open-label design means participants knew whether they attended education; patient-reported outcomes may be influenced by expectations or reporting behavior. The record does not describe methods to assess or reduce such influences. These limitations are specific to interpretation of the posted trial information and do not imply that unreported procedures were not conducted.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Differing Completion Rates of DIABETES Education on Patient Reported Outcomes — NCT06419907.
• Registry record last updated 2025-09-22; status last verified 2024-08.
• Results first posted 2025-09-22.
• 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

NCT06419907 is a completed randomized, open-label, parallel-group feasibility study sponsored by the University of Liverpool. It enrolled 120 participants and examined Type 2 Live Well, a face-to-face diabetes self-management education programme, across four attendance conditions: 100%, 60%, 10% and delayed education. The study began on 2024-05-16, with primary completion on 2025-05-20.

The primary endpoint was change in DSMQ self-management score at 3-4 months post baseline. Posted group summaries report median changes and standard deviations, but no formal primary between-group analysis. Three secondary outcomes cover diabetes distress and mental and physical health-related quality of life. Selected unadjusted pairwise summaries compare Group 2 and Group 3 with Group 1.

The record reports 3 participants lost to follow up and 4 removed as not per protocol, but does not describe the missing-data method or analysis population. Serious events were recorded in Group 1 and Group 4 and were judged unrelated to the study; no non-serious events or deaths were listed. Overall, the posted results are descriptive and exploratory, with important analysis details not reported in the registry record.