Leveraging Electronic Health Record (EHR) Tools to Reduce Health Disparities for Patients With Uncontrolled Hypertension: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Leveraging Electronic Health Record (EHR) Tools to Reduce Health Disparities for Patients With Uncontrolled Hypertension (NCT05030467), a randomized na study. The primary outcome is Change in Systolic Blood Pressure.

2. Clinical Question

The trial asks whether provider-facing tools integrated into the electronic health record can change systolic blood pressure among patients identified as eligible with uncontrolled hypertension. It also addresses the broader goal of improving hypertension management and reducing health disparities in racial and ethnic minorities. The registry describes a multicomponent intervention rather than a single tool: disparities dashboards, enhanced electronic decision support, and simplified self-monitoring orders and communication materials.

Randomization was at the clinic site level, not at the individual patient level. The comparison therefore evaluates the intervention strategy as implemented for providers in intervention clinics against usual care in control clinics. The posted primary result is a mean difference of -0.04 mmHg, with a 95% confidence interval from -0.39 to 0.31. The registry does not report which arm is subtracted from which in the contrast, so the sign alone does not identify which arm had the higher mean. The result concerns the specified blood-pressure measure; it does not by itself establish whether the intervention changed disparities or other aspects of hypertension management.

3. Trial Design

The registry describes a randomized, parallel, two-arm cluster trial. Clinics or sites were randomized, and providers at intervention sites received the EHR tools. The comparison group continued usual care, with access to tools already in clinical practice. Because treatment assignment occurred by site, the units of assignment and analysis are not interchangeable with a design that randomizes each patient independently.

The intervention, named REDUCE-BP, combined provider disparities dashboards, enhanced electronic decision support, and simplified self-monitoring orders and communication materials. Its stated aim was to improve blood pressure control and reduce health disparities in racial and ethnic minorities. The registry identifies the intervention type as behavioral and the masking status as none. The study is marked completed. Registry dates list a start of 2022-05-03 and primary completion of 2024-05-03. Enrollment is listed as 49210; the number analyzed for the primary endpoint is not reported in the registry record.

The primary endpoint is change in systolic blood pressure from identification to the end of the 12-month follow-up, using EHR values from ambulatory care encounters. The analysis population is described as patients identified as eligible. The posted analysis, however, is labeled as a mean difference in final values. The registry record does not report the statistical method used for that contrast, nor does it explain in further detail how the analysis of final values corresponds to the endpoint description of change.

Other listed outcomes concern well-controlled blood pressure, medication intensification, guideline-concordant medications, diastolic blood pressure change, and systolic blood pressure differences. Results for those outcomes are not reported in the registry record included for this analysis. The record does not specify the number of clinics or providers, the site-level allocation procedure, or additional details of the randomization process.

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: 24 clinics were randomized to intervention and control (1:1). Patients were identified from EHR data on the basis of their meeting eligibility criteria at the randomized clinics. Providers from these clinics were not directly enrolled in the study. Outcomes were measured at the patient level.

Overall Study

Analysis population for the primary outcome, as defined in the registry: Patients identified as eligible were included in the analysis

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: This is the number of patients who met eligibility criteria.

6. Primary Endpoint

The registry lists the primary outcome as Change in Systolic Blood Pressure, measured over 12 months.

The primary endpoint is change in systolic blood pressure, measured in mmHg, from identification to the end of the 12-month follow-up. Values come from ambulatory care encounters recorded in the EHR. Systolic blood pressure is a continuous measure, and the registry identifies the endpoint summary as a mean with standard deviation.

Using a continuous measure preserves information across the observed blood-pressure values rather than reducing the result to a threshold category. A mean difference can describe the average separation between groups on the measurement scale. This is useful for evaluating whether the intervention strategy is associated with a shift in blood pressure across the eligible patient population. It does not, on its own, describe how many patients met a control threshold or whether the intervention changed medication practices; those are separate outcomes listed in the registry.

The endpoint description concerns change from identification, while the reported analysis is labeled “Mean Difference (Final Values).” The record does not clarify whether baseline values were incorporated into the reported contrast or how the final-value comparison represents change. That distinction matters when interpreting the estimate, and the available record does not permit a more detailed reconstruction of the analysis. The posted estimate is -0.04 mmHg with a 95% confidence interval from -0.39 to 0.31. The registry does not report a p-value for this analysis.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary reported statistic is a mean difference for a continuous outcome. A mean summarizes the average measurement within a group, while a mean difference expresses the separation between group means on the original measurement scale. Here, the outcome is systolic blood pressure in mmHg. The estimate is -0.04, and the reported 95% confidence interval runs from -0.39 to 0.31. The registry identifies the endpoint summary as a mean with standard deviation, but the standard deviations themselves are not reported in the record.

What does the confidence interval add?

A point estimate gives one summary of the observed contrast, but it does not show how much uncertainty surrounds that summary. A confidence interval communicates precision under the analysis assumptions: narrower intervals indicate greater precision, and wider intervals indicate less. The interval reported for this result includes zero. Zero represents no difference on the contrast scale, but the interval also includes nonzero values on either side. The record does not provide a clinical-importance threshold, so the interval cannot be classified as excluding or including a meaningful effect on that basis.

A confidence interval is not a range containing a fixed proportion of individual patient measurements. It also is not a statement that the true effect has a specified probability of lying within this particular interval. Its interpretation depends on the statistical model and the assumptions used to calculate it. The statistical method and its assumptions are not reported in the registry record, so the interval can be described as posted but not independently evaluated from the information available.

Why does the cluster design matter?

In a cluster-randomized trial, groups such as clinics are assigned to study arms, and people within a group may have outcomes that are more alike than outcomes of people from different groups. This within-cluster similarity is often described by an intracluster correlation. When it is present, the effective information in the data is not equivalent to treating every patient as an independently randomized unit. A method that ignores clustering can understate uncertainty and make an estimate appear more precise than it is.

For that reason, the registry’s note that a simple individual-level re-analysis would be misleading is important. The sponsor’s own analysis is the only analysis shown. The method used to account for sites, the number of sites, and any model details are not reported in the registry record. Without those details, readers cannot assess the implementation of the cluster adjustment or reproduce the analysis. This does not invalidate the posted result; it limits what can be checked independently from the registry record.

What does the analysis label mean?

The endpoint description defines change in systolic blood pressure between identification and the end of the 12-month follow-up. The reported statistic is labeled as a mean difference in final values. A final-value comparison and a comparison of changes are related but are not necessarily the same analysis. Their relationship can depend on how baseline values were handled, whether the analysis adjusted for starting measurements, and how the outcome was defined in the model. Those details are not reported in the registry record. The safest reading is therefore to describe the result using its posted label and avoid treating it as a fully specified change-score or baseline-adjusted estimate.

The record also does not report a p-value for the primary result. A p-value, when reported, summarizes how unusual data at least as incompatible with a null hypothesis would be under specified model assumptions. It does not measure the size or importance of an effect. The confidence interval is available, but the missing p-value should not be inferred from the interval or reconstructed using assumptions not stated in the record.

How should planning assumptions be read?

The registry includes a planning illustration based on the observed effect, with a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation. It describes the effect to detect as a stated fraction of the observed effect. This is a planning exercise, not the trial’s own design rationale, and it is not observed power. Sample-size illustrations concern how a future study might be planned under stated assumptions; they do not alter the uncertainty in the trial’s observed result or show the probability that its hypothesis was true.

For this cluster-randomized study, the illustration’s assumptions should not be mistaken for a complete design specification. The record does not report that the trial was designed using those assumptions, nor does it provide the information needed to evaluate a cluster-specific planning calculation. Any displayed planning quantities should therefore be read only as illustrations under their stated assumptions, not as evidence about the trial’s achieved power.

What remains unreported?

The registry record does not report the model specification, the approach to missing outcome values, details of analysis-set construction beyond including patients identified as eligible, or whether the primary analysis followed an intention-to-treat principle. It also does not report handling of multiple outcomes or analyses, or whether there were interim analyses. These omissions do not establish that such procedures were absent; they mean that their use cannot be assessed from the registry record.

10. Confidence Intervals for Each Arm

No two-arm comparison is shown for this trial, but each arm's mean can be given with a t-based 95% confidence interval computed from the posted mean, spread and number of participants.

R code and output

arm <- c("Intervention Arm", "Control Arm")
n   <- c(24747, 24463)
m   <- c(-11, -11)
sp  <- c(19.2, 19.1)
sd  <- sp      # posted standard deviations
for (i in seq_along(n)) {
  se <- sd[i] / sqrt(n[i]); h <- qt(0.975, n[i] - 1) * se
  cat(sprintf("CARM|%s|%d|%.3f|%.3f|%.3f|%.3f\n", arm[i], n[i], m[i], sd[i], m[i] - h, m[i] + h))
}

CARM  Intervention Arm  24747  -11.000  19.200  -11.239  -10.761
CARM  Control Arm  24463  -11.000  19.100  -11.239  -10.761

11. Analyses Reported by the Sponsor

The result shown is the sponsor’s own analysis: a mean difference in final systolic blood pressure values of -0.04 mmHg, with a 95% confidence interval from -0.39 to 0.31. The registry notes that, because the trial is cluster randomized, a simple individual-level re-analysis would be misleading, so only the sponsor’s analysis is shown.

No recalculated estimate is reported for comparison with the sponsor’s result. There is therefore no independent numerical recalculation to use as a check on the posted contrast, and no discrepancy between two estimates can be assessed. The analysis method used by the sponsor is not reported in the registry record, so the model, its handling of clustering, and the details behind the confidence interval cannot be reconstructed from the available information.

The sponsor’s estimate should be read as the reported result rather than treated as a result confirmed by a separate calculation. Its confidence interval describes the reported precision, subject to the analysis assumptions. The registry does not report a p-value, and the arm ordering for the contrast is not stated. The difference between the endpoint description of change and the analysis label of final values also remains unexplained in the registry record.

12. Multiplicity, Missing Data and Analysis Populations

The registry lists the primary outcome as change in systolic blood pressure and also lists secondary outcomes covering blood-pressure control, medication intensification, guideline-concordant medications, diastolic blood pressure change, and systolic blood pressure differences. It reports six outcome measures and three analyses. The results of the listed secondary outcomes are not reported in the registry record included for this analysis.

When a trial evaluates multiple outcomes or makes multiple statistical comparisons, the chance of obtaining at least one apparently unusual result can increase if each comparison is interpreted without regard to the others. Statistical approaches can account for multiplicity, but the appropriate approach depends on which outcomes are confirmatory and how the analyses were planned. The registry record does not report whether a multiplicity adjustment was used, how outcomes were prioritized, or whether a hierarchy or other decision rule was specified. This absence is not evidence that no such plan existed; the approach is not reported in the registry record.

The primary analysis description says patients identified as eligible were included. More detail about the analysis population is not reported in the registry record. In particular, the record does not explain whether all eligible patients had usable EHR measurements, how patients without follow-up values were handled, or whether the analysis followed an intention-to-treat principle. Missing data procedures, including any imputation or sensitivity analyses, are not reported in the registry record.

These omissions matter because outcome availability and analysis choices can affect both the estimate and its uncertainty. EHR measurements may be recorded at encounters, and the endpoint description specifies ambulatory care values, but the registry does not describe how encounter timing, multiple measurements, or missing measurements were handled. It also does not report an interim analysis plan or stopping rules. Readers should not infer that these procedures were or were not used; the registry record does not provide those details.

13. 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 0.01% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: unrelated sae (0.1%).

The adverse-event record contains an intervention-arm entry with 24747 at risk, 24 affected under serious adverse events, 17 affected under other adverse events, and 3 deaths. It also lists “Unrelated SAE” as the top serious adverse-event term at 0.1%. These are the entries reported for the intervention arm; they should not be taken as a complete safety comparison between randomized groups.

The control arm is labeled “Not Collected for the Adverse Event Reporting,” and its at-risk and event fields are recorded as zero. Those zeros reflect the absence of collected control-arm adverse-event reporting in this record, not evidence that no adverse events or deaths occurred in the control arm. A comparison of event frequency between intervention and control cannot be made from these entries.

The registry does not define in further detail whether the affected counts refer to people, event episodes, or another reporting unit. The figures should therefore be quoted as recorded rather than converted into rates or interpreted as comparative risks. The listed term “Unrelated SAE” is a registry classification; it does not provide a complete account of causality assessment or establish how every event was judged. Safety interpretation is limited by the absence of control-arm collection and by the lack of additional event-level detail in the registry record.

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

The posted primary result is a mean difference in final systolic blood pressure values of -0.04 mmHg, with a 95% confidence interval from -0.39 to 0.31. This is a summary of the randomized-group comparison for the stated outcome and follow-up. The interval includes zero and covers values on either side, so the result does not establish a clear nonzero difference from the information reported. Because the registry does not state the arm ordering, the sign of the estimate cannot be translated into a direction favoring either group.

The estimate does not, by itself, establish whether EHR-based tools improved blood-pressure control, changed medication use, or reduced racial and ethnic disparities. Those are distinct outcomes or goals. Secondary outcomes are listed, but their results are not reported in the registry record. The record also does not report a threshold for a clinically important change, so the clinical importance of values within the interval cannot be determined from the reported information.

The result should not be read as proof that the intervention had no effect. An interval that includes zero is compatible with no difference, but also with nonzero differences within its bounds under the analysis assumptions. Conversely, the point estimate alone does not demonstrate a beneficial effect. Confidence intervals summarize uncertainty in an estimate; they do not settle questions about causality, clinical value, or fairness of outcomes without supporting analyses and context.

There are important limits to interpretation. Sites, not individual patients, were randomized, and the analysis method used to account for clustering is not reported. The endpoint description concerns change from identification, while the result is labeled as a difference in final values; the relationship is not clarified. The registry does not report a p-value, multiplicity strategy, missing-data approach, or detailed analysis population beyond saying eligible patients were included. These details affect how confidently the result can be independently evaluated.

15. Why This Trial Matters Statistically

This trial is statistically notable because it tests a provider-level intervention in a cluster-randomized setting. The tools were delivered through EHR workflows at intervention clinics, while control clinics continued usual care. That design is suited to studying an intervention implemented through providers and sites, but it requires analysis methods that respect the grouping created by site-level assignment.

The distinction between the unit randomized and the outcome measured is central. Blood pressure is measured for individual patients, but patients within the same clinic may share provider practices, workflows, and other influences. Statistical uncertainty must reflect that structure. The registry explicitly warns that a simple individual-level re-analysis would be misleading and presents only the sponsor’s analysis. The method used to address clustering is not reported in the registry record, making transparent reporting of the model and its assumptions especially relevant.

The trial also illustrates why a stated endpoint and a reported analysis label should be read carefully. The primary outcome is described as change from identification to the end of follow-up, but the reported analysis is a mean difference in final values. Without further detail, readers cannot determine whether the analysis adjusts for baseline or directly compares change. Clear specification of estimands and models helps distinguish these approaches and supports reproducibility.

Finally, the intervention’s stated goal includes reducing health disparities, while the posted primary result concerns mean systolic blood pressure. A mean comparison alone does not show whether disparities narrowed or widened. That question requires results that directly evaluate outcomes across relevant groups. The registry lists the broader goals and additional outcomes, but the corresponding results and details needed to assess disparity effects are not reported in the registry record. The trial therefore raises important design and reporting questions about cluster methods, endpoint interpretation, and the alignment between a study’s goals and its reported results.

16. 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.

Several features of the registry record limit interpretation of the primary estimate. The outcome is described as change in systolic blood pressure, but the analysis is labeled as a mean difference in final values. The record does not explain how these descriptions relate, whether baseline measurements were included in the model, or whether the reported contrast was adjusted. The arm ordering for the contrast is also not reported, preventing interpretation of the sign as favoring intervention or control.

The study randomized sites, but the number of sites and the statistical method used to account for clustering are not reported in the registry record. Those details are important for assessing the uncertainty around a cluster-randomized result. The record notes that a simple individual-level re-analysis would be misleading, but it does not provide sufficient model detail to reproduce the sponsor’s analysis. The p-value for the primary result is not reported.

The primary analysis population is described as patients identified as eligible, but the number included in the analysis and the handling of missing EHR measurements are not reported. The record does not describe whether all eligible patients contributed endpoint data, how multiple encounter values were handled, or whether sensitivity analyses were conducted. Details about multiplicity, interim analyses, and the prespecified statistical plan are also not reported in the registry record.

The registry lists secondary outcomes, but their results are not reported in the record used for this analysis. It therefore does not permit a complete assessment of the intervention’s effects on blood-pressure control, medication practices, or disparity-related outcomes. Adverse-event reporting is also incomplete for a randomized comparison: the control arm is identified as not collected for adverse-event reporting. These limitations describe what cannot be assessed from the registry record; they do not establish that the unreported procedures or analyses were not performed.

17. Related Tutorials

18. Related Statistical Calculators

19. Sources

• ClinicalTrials.gov: Leveraging Electronic Health Record (EHR) Tools to Reduce Health Disparities for Patients With Uncontrolled Hypertension — NCT05030467.
• Registry record last updated 2025-06-06; status last verified 2025-05.
• Results first posted 2025-06-06.
• 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.

20. Record Summary

The registry describes a completed, randomized, parallel, two-arm cluster trial of EHR tools intended to support providers managing patients with uncontrolled hypertension. Clinics were randomized, with intervention providers receiving dashboards, decision support, and simplified self-monitoring and communication tools; control providers continued usual care. Enrollment is listed as 49210.

The primary endpoint is change in systolic blood pressure over the 12-month follow-up using ambulatory care EHR values. The posted analysis is a mean difference in final values of -0.04 mmHg, with a 95% confidence interval from -0.39 to 0.31. The arm ordering is not reported, and the registry does not clarify the relationship between the change endpoint description and the final-value analysis label. The statistical method, p-value, missing-data approach, and multiplicity strategy are not reported in the registry record.

The registry lists additional blood-pressure and medication outcomes, but their results are not reported. Adverse-event reporting is available for the intervention arm; the control arm is marked as not collected. These reporting limits constrain interpretation of both efficacy and safety comparisons.