A 12 Month Site Randomized Trial in Adults With Type 2 Diabetes Mellitus and History of Cardiovascular Disease: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A 12 Month Site Randomized Trial in Adults With Type 2 Diabetes Mellitus and History of Cardiovascular Disease (NCT03936660), a randomized na study. The primary outcome is Number of Participants Prescribed All 3 Groups of Recommended Evidence-based Medications.

2. Clinical Question

This is a question about a care-delivery strategy and prescribing, not a direct comparison of medications assigned to individual participants. The intervention sought to help clinics organize multidisciplinary care. The primary endpoint captures whether participants were prescribed the specified combination of guideline-recommended therapies at follow-up. It does not, by itself, establish whether participants took the medications, whether the care pathway was implemented as intended, or whether cardiovascular events changed.

The trial is described as randomized at the site or clinic level. Consequently, the intervention contrast is between clinics assigned to the education strategy and clinics assigned to guidelines alone. Patients cared for at the same clinic may have more similar care and outcomes than patients at different clinics. That feature matters to both the design and the statistical uncertainty around any estimated difference.

3. Trial Design

COORDINATE-Diabetes is described as a randomized, parallel-group, clinic-level trial of an educational intervention. The registry record lists the allocation as randomized, masking as none, and the model as parallel. Its title describes a 12-month trial. The primary outcome timeframe is up to 12 months, with medication status assessed at the last follow-up visit. The record lists enrollment as 1049 and status as completed.

The intervention clinics were provided existing clinical care guidelines and a multifaceted educational intervention intended to support an integrated, multidisciplinary care pathway. Control clinics were provided existing clinical care guidelines. Because clinics were randomized, the assigned strategy applied at the clinic level. The record does not report the number of participating clinics, how clinics were allocated, or whether recruitment and follow-up procedures differed by group.

The listed start date is 2019-07-25 and the primary completion date is 2022-12-30. The phase is listed as NA. The sponsor is Duke University. The registry record identifies type 2 diabetes mellitus and cardiovascular diseases as conditions relevant to the trial.

Cluster randomization has consequences for analysis. Participants within a clinic share the clinic’s assigned strategy, so their outcomes may not be statistically independent. A comparison that ignored the clinic structure could give misleadingly narrow uncertainty. The record specifically cautions that a simple individual-level re-analysis would be misleading and indicates that only the sponsor’s own analysis should be used. Details of the sponsor’s model, including how it accounts for clustering, are not reported in the registry record.

The record describes the intervention but does not report how consistently clinics delivered it, how many participants were exposed to each component, or whether the care pathway was adopted. Those implementation details would help distinguish the effect of offering the educational strategy from the effect of receiving or sustaining its intended components.

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: All participants for whom medication data was available at 6 or 12 months.

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 Prescribed All 3 Groups of Recommended Evidence-based Medications, measured over Up to 12 Months.

The primary endpoint concerns prescribing of all three groups of recommended evidence-based medications at the last follow-up visit. The registry title describes a number of participants prescribed all three groups, while its outcome description calls the measure a proportion. The specified population is participants for whom medication data was available at 6 or 12 months. The registry record does not report the group-specific counts, denominators, or resulting proportions.

The three groups are defined as high-intensity statins; ACEIs or ARBs, including ARNIs; and SGLT2 inhibitors and/or GLP-1RAs with proven cardiovascular benefit, or metformin monotherapy with hemoglobin A1c < 7%. The record names the eligible examples within these medication classes. The endpoint therefore represents a combined prescribing criterion rather than a measure of use of any single drug or class.

This type of endpoint is used to assess whether recommended treatments are being prescribed together in clinical practice. It is relevant to a clinic-level education strategy because the intervention aimed to support coordinated care. However, a prescribing measure is not the same as medication adherence or a patient health outcome. It does not establish that every prescription was filled or taken, and it does not directly measure changes in cardiovascular events, symptoms, or survival. The endpoint also depends on available medication data and on the registry’s rules for classifying each participant against the combined criterion.

The record does not report the primary outcome result, the definition of the analysis denominator beyond the stated availability of medication data, or how incomplete medication information was handled. Without those details and the arm comparison, the endpoint’s observed magnitude and precision cannot be assessed.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Why does clinic randomization affect the analysis?

In a cluster-randomized trial, groups such as clinics, rather than individual participants, are assigned to the intervention or comparator. Participants treated within a clinic may share clinicians, routines, and organizational features. Their outcomes can therefore be correlated. Statistical methods that recognize this dependence are needed to estimate uncertainty appropriately. If the correlation is ignored, the analysis may treat related observations as if they contributed more independent information than they do. The registry record explicitly cautions that a simple individual-level re-analysis would be misleading.

General statistical explanation: the effective information in a clustered study depends not only on the number of participants but also on how outcomes are distributed across clinics and how similar participants within a clinic are. This is why the number of clinics, the distribution of participants by clinic, and the method for accounting for clustering matter. The registry record does not report those design details or the sponsor’s model specifications, so the impact of clustering on the precision cannot be evaluated from the information recorded.

What does an estimate and confidence interval tell readers?

A treatment-effect estimate summarizes the observed contrast between assigned groups on a defined outcome. A confidence interval describes the uncertainty around that estimate under the analysis assumptions. A narrower interval indicates greater precision; a wider interval indicates less precision. An interval should be interpreted in relation to both the direction and size of the estimated difference and the range of values compatible with the data and model.

General statistical explanation: a confidence interval is not a guarantee that the true effect lies within its particular endpoints, and it does not show the probability that the intervention is beneficial. It is a tool for expressing how much uncertainty remains under a specified statistical procedure. For this outcome, the interval would also need to reflect clinic-level assignment and the participants included in the analysis. The registry record does not report a primary estimate or confidence interval, so the precision of the comparison is not available.

How should a binary prescribing outcome be understood?

The primary outcome is defined by whether a participant met a combined prescribing criterion at follow-up. The record describes the endpoint as a count in its title and as a proportion in its description. A count refers to how many participants met the criterion; a proportion relates that count to a defined denominator. Interpreting a proportion requires knowing which participants were eligible for inclusion and how missing medication information was treated. The record identifies the outcome population as participants with medication data available at 6 or 12 months, but does not report the arm-specific denominators or missing-data rules.

General statistical explanation: an absolute difference between proportions would express the contrast in the frequency of meeting the criterion between groups. A relative measure would express that contrast differently and can appear larger or smaller depending on the starting frequency. Neither should be inferred without the observed counts and a stated analysis. Because the endpoint combines multiple medication classes, a result would not identify which component contributed most to a group difference.

What are the limits of a sample-size planning illustration?

The record describes planning calculations based on the observed effect, with a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation. The effect to detect is described as a stated fraction of the observed effect. These are planning assumptions, not a second analysis of the trial outcome. The calculations are not observed power and cannot establish whether a result is reliable or clinically important.

General statistical explanation: sample-size calculations depend on assumptions about the effect a future study seeks to detect, the outcome’s variability or event frequency, the analysis procedure, and the chosen error rate. If a planned study is sized around an effect observed in an earlier trial, the planning target can inherit uncertainty from that estimate. Smaller target effects generally require more information to distinguish from random variation, all else being equal. That general relationship does not supply a participant requirement for this trial; no such calculation can be reconstructed without the stated effect values and other inputs.

For a cluster-randomized design, a planning calculation should also reflect the fact that assignment occurs by clinic. Equal allocation between arms does not mean that the numbers of participants or clinics are necessarily equal in the completed trial. The record’s stated normal approximation and equal-allocation assumptions are features of the planning illustration, not evidence about the realized allocation or the sponsor’s analysis. No numerical planning results are reported in the registry information available for this explanation.

Why distinguish statistical from clinical interpretation?

Statistical uncertainty concerns how precisely the data estimate a contrast under the analysis model. Clinical interpretation concerns what the measured contrast means for care and health. A statistically detectable difference in prescribing would not by itself prove improved health, while an imprecise estimate would not by itself establish absence of a meaningful difference. The outcome is prescribing at follow-up, so conclusions should remain specific to prescribing and the defined population. The registry record does not provide the numerical results needed to assess either the statistical precision or the magnitude of the observed contrast.

Multiplicity is another general consideration when a trial evaluates multiple outcomes. Testing many outcomes can increase the chance of observing at least one apparently unusual result by chance if each is considered separately. Whether and how to address that issue depends on the prespecified analysis plan and the role assigned to each outcome. The registry record lists secondary outcomes but does not report a multiplicity procedure or a detailed statistical plan. Therefore, no adjustment method or hierarchy should be assumed.

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("Intervention", "Control")
n      <- c(457, 588)
events <- c(173, 85)
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  Intervention  457  173  37.9  33.4  42.5  33.5  42.4  33.5  42.4
ARM  Control  588  85  14.5  11.7  17.6  11.8  17.5  11.8  17.5

11. Multiplicity, Missing Data and Analysis Populations

The registry record lists a primary prescribing outcome and secondary outcomes concerning recommended therapies, LDL-C, blood pressure, and HbA1c. It does not report a multiplicity strategy, such as a testing hierarchy or adjustment procedure. Without that information, the interpretation of several outcome comparisons as a family cannot be reconstructed. A finding among secondary outcomes, if one were reported, would need to be interpreted with the number and role of comparisons in mind.

The primary outcome population is described as all participants for whom medication data was available at 6 or 12 months. The record does not report the number of participants with available medication data by arm, reasons for missing data, or how missing values were handled. It also does not report whether any imputation or sensitivity analysis was used. The stated population therefore identifies who could contribute based on data availability, but does not establish how complete or comparable the available data were across groups.

The record does not describe the analysis population in terms such as intention-to-treat or per-protocol, nor does it give a detailed statistical analysis plan. Those labels should not be inferred from the randomized design or from the endpoint description. It also does not report whether participants were analyzed according to their clinic’s assigned group or how clinic-level clustering was incorporated into the sponsor’s analysis.

These omissions limit interpretation of both the result and its precision. In particular, available-case analysis can be affected if availability of medication data is related to group assignment or to the outcome. Whether that concern applies cannot be assessed from the registry information. The registry record cautions against a simple individual-level re-analysis, reinforcing the importance of using an analysis that respects clinic-level allocation.

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 serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: urinary tract infection (0.2%), hyperglycaemic hyperosmolar nonketotic syndrome (0.2%), acute kidney injury (0.2%).

The registry reports serious adverse event figures separately by group. For the intervention group, the serious-event at-risk count is 459 and the reported affected count is 3. For the control group, the at-risk count is 590 and the affected count is 0. These are the recorded counts; they are not a test of whether the groups differ, and they do not establish that the intervention caused or prevented an event.

The other adverse-event figures are 4 affected among 459 at risk in the intervention group and 1 affected among 590 at risk in the control group. The record separately lists deaths as 6 for the intervention group and 16 for the control group. Those death counts should not be combined with the serious-event counts or interpreted as evidence of a treatment effect without the relevant definitions, timing, follow-up, and statistical comparison. The registry record does not report a comparative safety estimate or its uncertainty.

The listed serious-event terms are urinary tract infection, hyperglycaemic hyperosmolar nonketotic syndrome, and acute kidney injury, each recorded as 0.2%. These terms describe reported event categories; the record does not provide enough information to infer their comparative frequency by group or their relationship to the intervention. Event counts can be sparse, and simple differences in counts do not account for follow-up time, clinic structure, or chance variation. Safety interpretation should therefore remain descriptive and tied to the recorded definitions and reporting period.

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

The trial addresses whether a clinic-level educational strategy, added to existing clinical care guidelines, is associated with prescribing of a specified combination of evidence-based medication groups at follow-up. That question is relevant to how care is organized for adults with type 2 diabetes and cardiovascular disease. The endpoint is aligned with the intervention’s aim of supporting an integrated care pathway, but it captures prescribing rather than medication use or a direct health outcome.

The registry record does not report a numerical primary-outcome comparison. There is therefore no basis in the available record for describing the size, direction, or precision of the difference between intervention and control clinics. This is not evidence that the strategies produced identical prescribing results. It means that the primary comparison cannot be interpreted from the outcome description alone.

Even with a reported prescribing result, the scope of inference would need to remain clear. A result would refer to the defined prescribing criterion, at the last follow-up visit, among participants with medication data available at 6 or 12 months. It would not by itself establish adherence, sustained treatment, implementation quality, or a change in cardiovascular outcomes. The record does not describe whether the educational components were delivered consistently or whether participants took the prescribed therapies.

The randomized clinic-level design is important to any comparison. Participants within a clinic share the assigned strategy, so analyses need to recognize the clustered structure. The registry record cautions that a simple individual-level re-analysis would be misleading and gives precedence to the sponsor’s own analysis. The numerical details of that analysis are not reported in the registry information.

Accordingly, the supported interpretation is limited: the record documents a completed, clinic-randomized trial and defines its principal prescribing question, but does not provide the primary numerical result needed to judge the observed comparison. It also does not provide the statistical plan, missing-data procedures, or multiplicity approach needed for a fuller assessment of how the result should be interpreted.

14. Why This Trial Matters Statistically

This trial raises a statistically important distinction between the level of randomization and the level at which outcomes are measured. The intervention was assigned to clinics, while prescribing outcomes were recorded for participants. In such a design, participant observations are grouped within clinics. Recognizing that grouping is essential when estimating the uncertainty around a comparison. The registry record expressly warns that a simple individual-level re-analysis would be misleading.

The primary outcome also illustrates the difference between a count and a proportion. The title refers to the number of participants prescribed all three medication groups, while the description calls the endpoint a proportion. A count alone does not indicate how common the outcome was unless its denominator is known. The record identifies medication-data availability as the basis for inclusion but does not report the arm-specific denominators or the numerical outcome results. This makes it especially important not to infer a rate from the enrollment figure.

In addition, the intervention is a care-delivery strategy rather than a single treatment assigned to individuals. Its intended pathway could affect several prescribing decisions and organizational processes. The combined endpoint evaluates whether a multi-part criterion was met, but does not reveal which medication group or care process might account for a difference. Secondary outcomes add measures of therapies and risk factors, yet their presence alone does not establish the results or the priority assigned to each comparison.

Finally, planning calculations and observed trial findings answer different questions. A sample-size illustration can show how a future study might be planned under assumptions about the target effect and analysis. It cannot substitute for a reported estimate, confidence interval, and analysis that respects the clustered design. Keeping these distinctions clear helps prevent overinterpretation of planning assumptions or outcome definitions as evidence about the trial’s comparative result.

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 principal limitation for interpreting this trial record is that the numerical primary-outcome comparison is not reported in the information available. There are no arm-specific primary counts or proportions, no between-group estimate, and no confidence interval or p-value. As a result, the observed magnitude, direction, and precision of the primary comparison cannot be assessed.

Important design and analysis details are also not reported in the registry record. These include the number of randomized clinics, the distribution of participants across clinics, the sponsor’s model specification, and the way clinic-level clustering was handled. The record cautions against an individual-level re-analysis, but does not provide the model details needed to evaluate the sponsor’s approach.

For the primary endpoint, the record describes the analysis population as participants with medication data available at 6 or 12 months. It does not report outcome-data completeness by group, reasons for unavailable data, or missing-data methods. The stated endpoint is described both as a number and a proportion, but group-specific denominators are not reported. This limits assessment of how the analysis population and missingness may affect interpretation.

The registry record also does not report a detailed statistical analysis plan, a multiplicity procedure, or whether any interim analyses occurred. Those features should not be assumed. The secondary outcome descriptions do not substitute for their numerical findings. In addition, prescribing at follow-up is not the same as medication adherence or clinical benefit, and the record does not describe intervention fidelity or uptake. These limitations restrict conclusions to the trial’s documented design and outcome definitions rather than its unreported comparative results.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A 12 Month Site Randomized Trial in Adults With Type 2 Diabetes Mellitus and History of Cardiovascular Disease — NCT03936660.
• Registry record last updated 2024-11-21; status last verified 2024-04.
• Results first posted 2024-11-21.
• 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

COORDINATE-Diabetes is a completed, randomized, parallel, unmasked, clinic-level trial of an educational intervention for adults with type 2 diabetes mellitus and cardiovascular disease. Intervention clinics received existing guidelines plus a multifaceted educational strategy; control clinics received existing guidelines. Enrollment is listed as 1049. The start date is 2019-07-25 and the primary completion date is 2022-12-30.

The primary outcome concerns prescribing all three groups of recommended evidence-based medications at the last follow-up visit, up to 12 months. The specified population is participants with medication data available at 6 or 12 months. The registry record does not report the numerical primary comparison, confidence interval, or p-value. It also does not report detailed methods for clustering, missing data, or multiplicity. The record cautions that simple individual-level re-analysis would be misleading. Serious adverse event and death counts are listed by group, but no comparative safety analysis is reported.