Using Practice Facilitation and Operationalizing Referral Information Technology to Increase DSMES Utilization: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Using Practice Facilitation and Operationalizing Referral Information Technology to Increase DSMES Utilization (NCT05472142), a randomized na study. The primary outcome is Diabetes Self-Management Education and Support (DSMES) Referrals.

2. Clinical Question

3. Trial Design

NCT05472142 is registered as a completed, randomized, parallel-group pilot and feasibility study. Its title identifies the approach as UP FOR IT: using practice facilitation and operationalizing referral information technology to increase DSMES utilization. The study is described as pragmatic and cluster randomized. Clinics, rather than individual patients, are the stated units assigned to the intervention or control cluster. The number of clinics and the method used to generate or implement the random allocation are not reported in the registry record.

The intervention cluster was to receive health information technology together with practice facilitation during a 12-month intervention. The control cluster was to receive health information technology alone and not take part in that combined intervention. The registry record lists both practice facilitation and health information technology as interventions. Masking is listed as none, and the trial model is parallel. The phase is listed as not applicable.

The recorded enrollment is 22. Because this is a cluster-randomized study involving clinic staff and patients, that total does not by itself describe how many clinics were randomized or how participants were distributed across clusters and arms. Those details are not reported in the registry record. The recorded start date is 2023-05-30, primary completion date is 2024-10-30, and status is completed.

The primary outcome is DSMES referrals, with assessments at baseline, month 9, and month 15. The secondary outcomes are DSMES attendance and the percent of patients with A1c >9%. The registry record lists three outcome measures but no analyses. Detailed procedures for implementation, follow-up, and handling of differences among clinics are not reported.

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: Study participants included clinic staff (n=22 total, consented) and patients (n=1,803, not consented). Clinic staff received the intervention and patients of clinics could be indirectly impacted by intervention. Clinic staff aggregated patient data up to the clinic-level prior to sharing with the study team. Patients were not consented as there was no interaction from members of the research team with patients and there was no use of identifiable private information for research purposes.

Total Participants

Reasons for not completing this period:

Clinic Patients

Reasons for not completing this period:

Clinic Staff

Analysis population for the primary outcome, as defined in the registry: Measure Analysis Population Description: The analysis population differs from the Overall population because the study has two distinct populations that, together, make up the Overall population - clinic staff and patients. This study-specific baseline measure was only captured on one of the two populations - patients.

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 Diabetes Self-Management Education and Support (DSMES) Referrals, measured over Baseline, End of active intervention (month 9), End of follow-up (month 15).

The primary endpoint is DSMES referrals. The registry record describes the outcome as a referral rate measured at baseline, at the end of the active intervention at month 9, and at the end of follow-up at month 15. It states that clinic electronic health record data will be used. The outcome is also classified as a count of participants, so the registry record does not fully clarify how the count is converted into, or distinguished from, the stated referral rate.

In general, a referral outcome is relevant to this study because the intervention is intended to address referral barriers and support clinician awareness and motivation to refer people with diabetes to DSMES. Measuring referrals at baseline and later time points can describe whether referral patterns change over the study period. A comparison between the clusters could help assess whether patterns differ for clinics receiving the combined approach and clinics receiving health information technology alone.

The registry record does not give the operational definition of a referral, such as which electronic health record entries count, how repeat referrals are handled, or how the rate denominator is defined. It also does not report numerical endpoint results. Consequently, the outcome’s direction, size, and precision cannot be assessed from the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Randomization is a method for assigning study units to comparison groups so that, in expectation, systematic differences are reduced. In this study, the registry describes random assignment at the clinic level. The assigned intervention is a package combining health information technology and practice facilitation; the comparison clinics receive health information technology alone. The registry record does not report how the random sequence was generated, how clinics were assigned, or whether any restrictions were used in the allocation.

Why the cluster design matters

General statistical explanation: Cluster randomization is used when an intervention is delivered to, or operates through, groups such as clinics. People served by a clinic can share care processes, staff, and local conditions. Their outcomes may therefore be correlated. Analyses that account for this dependence generally have different uncertainty than analyses that treat every participant as unrelated to every other participant. The registry record specifically cautions that simple individual-level re-analysis would be misleading. The number and size of clusters, the degree of within-clinic similarity, and the method used by the sponsor to account for clustering are not reported in the registry record.

What a parallel comparison represents

General statistical explanation: In a parallel design, the randomized groups are followed as separate groups rather than crossing over between interventions. Here, the registry describes a control cluster and an intervention cluster. The contrast of interest is the combined health information technology and practice facilitation approach versus health information technology alone. The record does not report whether the outcome analysis compares follow-up measurements directly, adjusts for baseline, models change over time, or uses another approach. No statistical analysis plan is reported in the registry record.

Counts, rates, and denominators

General statistical explanation: A count records how many events or participants meet an outcome definition. A rate expresses events relative to a specified denominator, which might reflect the number of people eligible for referral or another defined population. The registry describes the primary outcome as a referral rate, but classifies it as a participant count. It does not define the numerator, denominator, or rules for counting referrals. Those details are necessary to understand what the endpoint represents and how comparisons should be interpreted.

Uncertainty and over-dispersion

General statistical explanation: A confidence interval expresses uncertainty around an estimated effect under the assumptions of the statistical model. For count outcomes, a model may need to allow for variation greater than a simple count distribution would predict. This is often called over-dispersion. If it is present and ignored, uncertainty can be misrepresented. In a cluster-randomized study, dependence within clinics is another source of uncertainty that an analysis may need to address. The registry does not report an outcome model, a dispersion estimate, or confidence intervals, so it is not possible to assess how these issues were handled.

What can be concluded from the record

The record lists three outcome measures and zero analyses. It does not report numerical results, an analysis population for the primary comparison in sufficient detail, or model specifications. General statistical methods can explain why those details matter, but they cannot substitute for the missing trial analysis. No effect estimate, p-value, or interval should be inferred from the design description or enrollment total.

The registry also states that the primary outcome’s analysis population differs from the overall population because the study includes clinic staff and patients. It further says the baseline measure was captured on only one of those populations, but the relevant text is incomplete. This makes the exact population underlying that baseline measure unclear. The registry record does not report how the population difference was handled in an analysis.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary outcome of DSMES referrals and two secondary outcomes: DSMES attendance and the percent of patients with A1c >9%. It reports three outcome measures overall, but lists zero analyses. No multiplicity strategy is reported in the registry record. In general, when several outcomes or comparisons are tested, the probability of at least one chance-positive result can increase unless the analysis plan accounts for the set of tests. That general concern cannot be evaluated for this trial because no test results or prespecified adjustment approach are reported.

Missing-data methods are also not reported in the registry record. The entry does not describe how incomplete electronic health record measurements, participants without follow-up information, or clinics with incomplete outcome data would be handled. In general, the implications of missing data depend on how much is missing, why it is missing, and the assumptions used by the analysis. None of those details can be assessed from the registry entry.

The population description says that clinic staff and patients together make up the overall population, and that the primary outcome’s analysis population differs from the overall population. It also states that the study-specific baseline measure was captured on only one of these populations, but the description is truncated. The registry record does not identify which population contributed that baseline measure or explain how baseline and follow-up populations were aligned.

The record’s warning against simple individual-level re-analysis is particularly relevant to the analysis population: the randomized unit is the clinic, while outcomes concern participants and clinic records. The sponsor’s analysis method would need to reflect that structure, but no analysis is recorded. The available information therefore does not establish how missing observations, population definitions, or multiple outcomes were handled.

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

The registry record establishes that a completed pilot and feasibility cluster-randomized study was designed to compare health information technology plus practice facilitation with health information technology alone. Its primary outcome was DSMES referrals measured from clinic electronic health record data at baseline, month 9, and month 15. It also lists DSMES attendance and the percent of patients with A1c >9% as secondary outcomes.

Those design details do not establish whether the intervention changed referral patterns or DSMES utilization. No numerical outcome results or analyses are reported in the registry record. The study’s purpose, design, recorded enrollment, and completed status cannot be used to infer an effect. There is no reported effect estimate or measure of precision from which to judge the size or uncertainty of a between-cluster difference.

In general, a pilot and feasibility study can inform whether an approach and its measurement procedures are workable, but the label alone does not specify which feasibility criteria were evaluated or whether they were met. The registry description calls this a pilot and feasibility study, yet does not report feasibility outcomes or decision criteria. The record therefore does not support conclusions about implementation success or the practical feasibility of the program.

The study also cannot establish, from the available record, how well its findings would apply to clinics beyond those enrolled. The number of clinics, their characteristics, and details of participant selection are not reported. Nor does the record provide enough information to assess the definition of a referral or the denominator used for the stated referral rate. These limitations constrain interpretation even apart from the lack of results.

The appropriate reading is limited: the registry describes a randomized comparison and its intended measures, but it does not provide evidence with which to estimate the effects of that comparison. No conclusion about benefit, lack of benefit, or harm follows from the absence of posted analyses.

12. Why This Trial Matters Statistically

This study is statistically notable for randomizing clinics while measuring outcomes that concern referrals and patients. That structure connects an intervention delivered through clinic processes to participant-level or clinic-record outcomes. It requires attention to the relationship between the assignment unit and the measurement unit. The registry explicitly warns that an individual-level re-analysis would be misleading, a useful reminder that the unit of randomization is central to valid uncertainty estimates.

The comparison is also informative as a design question: the control clinics receive health information technology, while intervention clinics receive health information technology with practice facilitation. The contrast is therefore not the combined program against no support. It is the additional contribution of the combined approach relative to the technology component alone, as described in the record. Clear interpretation depends on preserving that distinction.

The timing of the primary outcome assessments allows the intended evaluation to distinguish baseline referral patterns from measurements at the end of the active intervention and the end of follow-up. In general, repeated outcome measurements can help characterize changes over time, but the analysis must account for the design and clearly define what constitutes a referral. The registry’s description of the measure as both a referral rate and a count of participants leaves that operational detail unresolved.

Finally, the record illustrates the limits of a trial registration without posted analyses. It documents the question, intervention structure, and measurement schedule, but not the results needed for inference. Without the sponsor’s analysis, outcome estimates, and uncertainty measures, the design can be discussed but the trial’s statistical findings cannot be evaluated.

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 most consequential limitation for interpreting this specific trial is that the registry record lists zero analyses and reports no numerical outcome results. There is no estimate for the primary referral outcome or either secondary outcome. This prevents evaluation of the direction, magnitude, and precision of any observed contrast.

The primary endpoint is not fully operationalized in the record. It is described as a referral rate but categorized as a count of participants. The numerator, denominator, eligibility rules, and handling of repeated referrals are not reported. Since the endpoint is based on electronic health record data, the record also does not explain which records or coding conventions define a referral.

Important cluster-design details are absent. The number of clinics, their distribution across groups, and the analysis method used to account for clustering are not reported. The registry notes that individual-level re-analysis would be misleading, but does not include the sponsor’s analysis. Enrollment of 22 does not resolve the uncertainty about the number of randomized clinics or participants per cluster.

The registry describes clinic staff and patients as two distinct populations, and says the primary outcome analysis population differs from the overall population. Its text about which population contributed the baseline measure is incomplete. No explanation is reported for aligning the population measured at baseline with the population measured at follow-up.

Finally, missing-data procedures, multiplicity handling, detailed eligibility criteria, and the statistical analysis plan are not reported in the registry record. These gaps do not prove that the study lacked such procedures; they mean that their use and implications cannot be assessed from the record.

14. Related Tutorials

15. Related Statistical Calculators

16. Sources

• ClinicalTrials.gov: Using Practice Facilitation and Operationalizing Referral Information Technology to Increase DSMES Utilization — NCT05472142.
• Registry record last updated 2025-12-09; status last verified 2025-11.
• Results first posted 2025-12-09.
• 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

NCT05472142 is a completed, randomized, parallel cluster trial described as a pilot and feasibility study. Clinics in the intervention cluster were to receive health information technology and practice facilitation; control clinics were to receive health information technology alone. The recorded enrollment is 22, and masking is listed as none.

The primary outcome is DSMES referrals, measured at baseline, month 9, and month 15 using clinic electronic health record data. Secondary outcomes are DSMES attendance and the percent of patients with A1c >9%. The registry lists three outcome measures and zero analyses. It reports no outcome estimates or statistical results, so the trial’s effects and their precision cannot be evaluated from the registry record.