Implementation and Evaluation of a Diabetes Prevention Clinical Pathway in Primary Care: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Implementation and Evaluation of a Diabetes Prevention Clinical Pathway in Primary Care (NCT05265312), a non-randomized na study. The primary outcome is Number of PCP Visits Where Ppt Was Referred to Diabetes Prevention Program Within 30 Days of Visit.

2. Clinical Question

The trial asks whether implementing a diabetes prevention clinical pathway in an intervention clinic is associated with changes in primary care activity and patient engagement compared with routine care in a control clinic. The stated primary question concerns referrals to a Diabetes Prevention Program: the registry defines the endpoint as the number of primary care visits where a patient was referred within 30 days of the visit during the intervention period. The stated timeframe is 12 months.

The rationale in the registry describes gaps in prediabetes care and barriers reported in prior work by the investigators. These include clinician knowledge and perceptions about Diabetes Prevention Programs, staffing constraints, and patient concerns such as motivation, time, and resources. The pathway is intended to address some of these barriers. The registry states the investigators’ hypothesis that it will increase clinician screening and intervention and improve patient engagement. Those statements describe the rationale and hypothesis, not findings from this comparison.

3. Trial Design

This is a completed, parallel, nonrandomized study of an implementation pathway in primary care. One clinic is identified as the intervention clinic and one as the control clinic. Patients with prediabetes seen for routine care at the intervention clinic are described as receiving care in the setting where the START diabetes prevention clinical pathway was implemented; patients with prediabetes seen for routine care at the control clinic are described as the comparison group. The registry lists enrollment of 1910. It does not report the enrollment separately by clinic in the record.

Because allocation was nonrandomized, the clinics and their patients cannot be assumed to be comparable at the start of the comparison. Differences in referral counts could reflect the pathway, but could also reflect differences between clinics, their patients, clinicians, workflows, or other factors. The registry record does not report an adjustment strategy that would address such differences. The study is parallel in structure, meaning the two clinic groups are considered over the same broad study period rather than as a crossover. Masking is listed as none.

The listed intervention is the START diabetes prevention clinical pathway, classified as another type of intervention. The study status is completed. The registry gives a start date of 2022-05-31 and a primary completion date of 2023-06-02. The primary endpoint timeframe is 12 months. These registry dates identify the study record’s timing; they do not establish the timing of every patient’s follow-up or the exact data collection schedule.

The record identifies two additional outcomes: the number of participants completing glycemic lab testing among those with an order during the intervention period, and achieving a 5% or greater reduction in weight compared with baseline. It lists five outcome measures overall, but does not report the other outcome details in the record. The registry record does not report a detailed protocol, clinic selection process, or the basis for the sample size.

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

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 PCP Visits Where Ppt Was Referred to Diabetes Prevention Program Within 30 Days of Visit, measured over 12 months.

The primary endpoint counts primary care visits where a patient was referred to a Diabetes Prevention Program within 30 days of the visit during the intervention period. The unit is primary care visits, and the registry specifies a 12-month timeframe. The endpoint therefore describes referral activity associated with visits; it is not defined as the number of unique patients referred, the number who enrolled in a program, or the number who completed one.

A visit-based count can be useful when the implementation goal concerns whether clinicians make referrals during routine care. It captures an action in the care pathway that occurs before any later participation or health outcome. In general, a count endpoint records how often an event occurs in the defined observation period. Its interpretation depends on the event definition, the population or visits eligible to contribute, and how opportunity for the event is measured. For example, a count of referrals can be affected by the number of eligible visits as well as the tendency to refer at a visit.

The registry description does not state whether repeated visits by one patient contribute separately, how referrals are verified, or whether the analysis uses a denominator such as eligible visits. It also does not report an observed count or a comparison estimate. The endpoint definition alone therefore identifies what the study intended to measure, but does not establish the amount of referral activity or the direction or precision of any clinic difference.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The methods that can be described from the registry are limited to the study structure, intervention, endpoint definitions, and listed design characteristics. The record identifies a parallel, nonrandomized comparison, no masking, an enrollment of 1910, and an intervention clinic and control clinic. It does not report an analysis plan or comparative analysis results. The concepts below explain how such methods are generally understood; they should not be read as a claim that a particular unreported procedure was used.

What does a nonrandomized comparison imply?

In general, randomization assigns participants or groups by chance to study conditions, helping to reduce systematic differences between comparison groups. In a nonrandomized design, allocation is not determined by chance. Treatment setting may then be connected to characteristics that also relate to outcomes. For a clinic comparison, differences in patient populations, staffing, local practices, or visit patterns can be entangled with exposure to the pathway. This is why an observed clinic contrast does not by itself isolate a causal effect.

Statistical adjustment can sometimes account for measured differences, but it cannot guarantee removal of bias from unmeasured factors or poorly measured variables. The registry record does not report a covariate set, adjustment method, matching approach, or sensitivity analysis. It is therefore not possible to assess whether any such method was used or whether it would support a causal interpretation. The stated nonrandomized allocation should remain part of any interpretation of a reported comparison.

How should a count endpoint be understood?

In general, a count is a nonnegative total of events over a specified period or set of opportunities. The primary endpoint counts visits with a referral within 30 days. Counts can be analyzed directly, or they can be expressed relative to the amount of observation or number of opportunities, depending on the study question and data structure. A raw total answers how many events were observed in the defined group. A rate can answer how frequently events occurred relative to an exposure measure. These are not interchangeable summaries.

The registry labels the unit as primary care visits but does not state the analysis denominator or whether repeated visits by the same participant are included as separate opportunities. It also does not report how visits were linked to referrals or how the 30-day window was applied. Those details affect the meaning of the endpoint. Without a stated denominator and analysis method, a count alone would not show whether differences reflect referral behavior, different numbers of visits, or both.

In general, count data may vary more than a simple count model expects, a property called over-dispersion. When present, over-dispersion can make standard errors too small if the model assumes less variability than the data contain. Repeated observations from the same person or clustering within a clinic can also make observations statistically dependent. Appropriate analysis may need to account for these structures. The registry record does not report whether over-dispersion, repeated visits, or clustering were assessed, and it does not identify a count model.

What does an estimate of uncertainty add?

In general, a point estimate gives a single summary of a contrast, while a confidence interval communicates uncertainty around that estimate under the assumptions of the analysis. A narrow interval indicates greater precision than a wide interval, but an interval does not resolve design bias or prove that the model assumptions are correct. A significance test, when reported, addresses a particular null hypothesis under specified assumptions; it does not measure the size or practical importance of a difference.

No estimate, interval, or test result for the primary outcome is reported in the registry record. Therefore, no assessment of statistical precision or compatibility with a null value can be made from this record. General concepts about intervals and tests do not fill in the missing study results.

What can planning calculations say?

Sample-size calculations are planning tools. In general, they describe the amount of information a proposed design may need to detect a prespecified effect under assumptions about variability, allocation, and a statistical test. They are not observed outcomes and do not show what a completed study found. A planning calculation based on an observed effect is especially distinct from a prospective justification: it uses an effect estimate as an input for illustration rather than documenting the assumptions behind the original design.

The registry record includes planning assumptions for an illustration based on an observed effect, including a two-sided significance level of 0.05, equal allocation, and a normal approximation. It states that the target effect is a stated fraction of the observed effect. The record does not report numerical sample-size outputs in the information, and it does not identify these assumptions as the trial’s original design justification. It would be inappropriate to treat such an illustration as observed power or as evidence that the study could detect a particular effect.

The registry does not report the primary analysis method, estimand, or model specification. It also does not report a prespecified threshold for a meaningful difference. Those elements would be needed to connect the endpoint definition to a reproducible comparative analysis. Without them and without results, the appropriate statistical account is descriptive: the study planned a nonrandomized clinic comparison using a referral count, but the registry record does not report what the comparison found.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists five outcome measures and names the primary referral-count endpoint along with two secondary outcomes: glycemic lab testing completion among participants with an order during the intervention period, and achieving a 5% or greater reduction in weight compared with baseline. The record does not report the full set of outcome definitions in the information, nor does it report an analysis plan identifying which outcomes were confirmatory or how multiple comparisons would be handled.

In general, testing several outcomes can increase the chance that at least one result appears statistically notable by chance. A prespecified hierarchy, adjustment procedure, or clear distinction between primary and exploratory analyses can help readers interpret a collection of tests. The registry record does not report whether any such approach was planned. Since it lists zero analysis entries, there are no reported outcome-specific estimates or tests to assess.

The registry record also does not report missing-data counts, reasons for missingness, imputation procedures, or sensitivity analyses. In general, missing outcome data can change the composition of the analyzed group and may bias estimates when missingness is related to outcome or group membership. For the glycemic testing outcome, the stated population is participants with an order; the record does not describe how missing orders or unavailable test results were handled. For weight change, it does not report how baseline and follow-up measurements were defined or how missing measurements were treated.

Analysis populations are not described in the registry record. It is not reported whether analyses included all eligible patients, only those with available outcome data, or another defined group. These omissions prevent assessment of the denominators, missing-data impact, and multiplicity strategy.

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

The record establishes that a primary care pathway was evaluated through a parallel comparison of an intervention clinic and a control clinic, and that the primary endpoint was a count of visits involving a referral to a Diabetes Prevention Program within 30 days. It also records an enrollment of 1910 and a completed study status. These registry entries describe the study, not its outcome.

The registry record does not report the observed referral count, a between-clinic estimate, or uncertainty around a comparison. It therefore does not show whether the intervention clinic had more or fewer referrals, whether any difference was large or small, or whether the data were precise enough to distinguish a difference from chance variation. The stated hypothesis should not be interpreted as evidence that the pathway achieved its intended goals.

Even if an outcome estimate were available, a nonrandomized clinic comparison would require careful interpretation. A difference could be related to the pathway or to differences between the clinics and their patients. The record does not report an adjustment strategy or information sufficient to evaluate the extent of such differences. An association between clinic and outcome would not automatically establish that the pathway caused the association.

The endpoint is a referral action, not program attendance, program completion, diabetes incidence, or a direct health outcome. The secondary outcome descriptions refer to glycemic testing completion and weight reduction, but their results are not reported. The record therefore does not establish what happened after referral or whether patient-level outcomes changed.

In general, statistical significance and clinical or operational importance are distinct questions. A precise estimate may still represent a small effect, while an uncertain estimate may leave a range of possibilities. Here, no comparative estimate is reported, so neither precision nor magnitude can be evaluated. The appropriate conclusion is limited to the study’s design and measurement aims; conclusions about the pathway’s effects are not reported in the registry record.

12. Why This Trial Matters Statistically

Implementation studies ask how care processes operate in routine settings, not only whether an intervention can be delivered under tightly controlled conditions. This study focuses on a clinical pathway intended to address reported barriers to prediabetes care, and its primary endpoint concerns referral activity during primary care visits. That makes the measurement question relevant to the process of connecting patients with prevention programs.

Statistically, the setting also illustrates why design matters. A comparison between clinics can be informative, but nonrandomized allocation leaves open the possibility that clinic and patient characteristics differ alongside pathway exposure. Those differences affect what a simple comparison can support. Clear reporting of denominators, repeated-visit handling, adjustment methods, uncertainty, and missing data would help readers distinguish a change in referral activity from variation in visit volume or clinic composition. The registry record does not report those analytical details or outcome estimates.

The endpoint also highlights a broader measurement distinction: process measures and patient outcomes answer different questions. A referral count can describe an activity in the care pathway; it does not establish whether a patient participated in a program or experienced a later health outcome. The registry’s secondary outcomes point to other stages of care, but their results are not reported.

The study is therefore statistically notable as a nonrandomized evaluation of an implementation strategy with a visit-based primary measure, while the available registry information limits conclusions about effects. Its record emphasizes the need for transparent outcome reporting and a clearly described analysis that fits the design and data structure.

13. 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 trial-specific limitation for interpretation is nonrandomized allocation to an intervention clinic and a control clinic. The groups may differ in ways that influence referral activity, and the registry record does not report methods for measuring or addressing those differences. With clinics as the comparison settings, clinic-level practices and patient mix may also be difficult to separate from exposure to the pathway.

The endpoint is defined as a number of visits with a referral within 30 days, but the record does not report the eligible-visit denominator, whether repeat visits by the same patient count separately, or how referrals were ascertained. Those details are important for understanding what a difference in counts would represent. No primary outcome result, comparative estimate, confidence interval, or analysis entry is reported, so the magnitude and precision of the comparison cannot be assessed.

The registry lists enrollment of 1910 but does not report group-specific enrollment in the record. It does not report the analysis population, missing-data handling, covariate adjustment, or methods for repeated observations and clinic clustering. The record also does not provide the full details for all five outcome measures or describe a multiplicity strategy. These omissions limit reproducibility and prevent evaluation of how outcome selection and availability may affect interpretation.

The adverse-event fields list zero entries for the recorded event categories in both groups, but the record does not explain monitoring or reporting procedures. Those entries should not be taken as a complete account of safety data. The registry record also does not report a sample-size justification for the trial’s original design or the detailed protocol needed to assess whether the reported design and planned analyses were followed.

14. Related Tutorials

15. Related Statistical Calculators

16. Sources

• ClinicalTrials.gov: Implementation and Evaluation of a Diabetes Prevention Clinical Pathway in Primary Care — NCT05265312.
• Registry record last updated 2025-02-05; status last verified 2025-02.
• Results first posted 2024-06-25.
• 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.

17. Record Summary

The registry identifies NCT05265312 as a completed study titled “Implementation and Evaluation of a Diabetes Prevention Clinical Pathway in Primary Care,” sponsored by Johns Hopkins University. It is listed as a parallel, nonrandomized study with no masking and enrollment of 1910. The record gives a start date of 2022-05-31 and a primary completion date of 2023-06-02.

The intervention is the START diabetes prevention clinical pathway in an intervention clinic; the comparator is routine care for patients with prediabetes at a control clinic. The primary endpoint is the number of primary care visits involving referral to a Diabetes Prevention Program within 30 days during a 12-month timeframe. Listed secondary outcomes concern glycemic lab testing completion among participants with an order and achieving a 5% or greater reduction in weight compared with baseline.

The registry record does not report outcome results or analysis entries. It also does not report the analysis methods, group-specific enrollment, missing-data procedures, or a detailed account of the original sample-size justification. Consequently, the record describes the evaluation’s design and intended measurements but does not provide evidence from which to estimate the pathway’s comparative effect.