MedStar Diabetes Pathway Chart Reviews: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for MedStar Diabetes Pathway Chart Reviews (NCT03442595), a registered phase not stated study. The primary outcome is Percent Change in A1C.

2. Clinical Question

3. Trial Design

This is a completed chart-review study of the MedStar Diabetes Pathway, registered as NCT03442595. Its purpose was to continue reviewing clinical outcomes as the pathway expanded, support continuous quality improvement, and compare patients receiving pathway care with patients receiving standard diabetes care through a primary care provider.

The registry lists an enrollment of 826 and gives the study dates as 2016-07-01 to 2018-12-31. The intervention is described as intensive diabetes education and medication management. Cases were patients with uncontrolled type 2 diabetes mellitus participating in the pathway. Matched controls had uncontrolled type 2 diabetes, matched cases on 5 criteria, and received standard care with a MedStar provider. The criteria used for matching are not reported in the registry record.

The record identifies the comparison as nonrandomized; allocation details are not stated. Therefore, the groups should not be treated as if assignment to pathway or standard care was random. Matching can make groups more comparable on the characteristics used for matching, but it does not establish comparability on characteristics that were not matched or measured. Details of how records were selected, how matching was implemented, and how differences between groups were handled analytically are not reported in the registry record.

The primary outcome was percent change in A1C, described as a mean with standard deviation over 12-16 weeks. The registry also lists six secondary outcomes concerning emergency room visits, hospitalizations, and their composite at 30 and 90 days. Although the record marks the study completed, its primary outcome estimate and inferential results are not reported in the registry record.

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: Hemoglobin A1C was measured at baseline and at 3 months in both groups.

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: A total of 460 patients started the diabetes boot camp (cases) during the study period, but only 366 completed it and were included in the results analysis.

6. Primary Endpoint

The registry lists the primary outcome as Percent Change in A1C, measured over 12-16 weeks.

The primary endpoint is percent change in A1C, reported as a mean with standard deviation. The registry describes the outcome as change in Hemoglobin A1C from baseline to 12-16 weeks for cases and controls, and says A1C was measured at baseline and at 3 months in both groups. The endpoint therefore concerns change over follow-up, rather than a single measurement at baseline.

A continuous measure can retain information about the range of observed values, rather than reducing results to a category. In general, a group mean describes the average outcome, while a standard deviation describes how spread out individual measurements are around that mean. A comparison of group changes could address whether average change differed between pathway cases and matched controls. The registry record does not report the group means, standard deviations, estimated difference, or confidence interval, so the size and precision of any observed comparison cannot be assessed from the record.

The registry's descriptions use both 12-16 weeks and 3 months to characterize follow-up. The endpoint is not described as a clinical threshold or a proportion reaching a target. Further details about the calculation of percent change, including the exact formula, are not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is continuous: percent change in A1C. The registry says the summary measure is the mean and the measure of dispersion is the standard deviation. The record does not give the actual means or standard deviations, nor does it describe the exact formula used to calculate percent change.

What do the mean and standard deviation describe?

In general, the mean is the arithmetic average of the observed values in a group. It provides a concise description of the group's central outcome, but it does not show how individual values vary. The standard deviation describes the spread of values around the mean. A small or large standard deviation should be interpreted in relation to the outcome's scale and context; no such values are reported for this study.

A mean change within each group and a comparison between groups answer different questions. The within-group mean describes the average change among the people included in that group. A between-group contrast addresses how those average changes compare. In a nonrandomized study, a difference between groups may reflect differences in patient characteristics or care as well as any relationship with pathway participation. The registry record does not report a between-group estimate or explain whether it adjusted for other characteristics.

What would a confidence interval add?

In general, a confidence interval gives a range of values compatible with the observed data under the statistical model used. It communicates uncertainty about an estimated quantity, such as a mean difference. A narrower interval indicates greater precision under that model, while a wider interval indicates less precision. It does not remove bias from a nonrandomized comparison, and it does not guarantee that the range contains the true value. No confidence interval for the primary outcome is reported in the registry record.

Why does the study design affect statistical interpretation?

Random assignment, when used, can balance measured and unmeasured characteristics on average across groups. This record explicitly describes a nonrandomized comparison. Matching can align cases and controls on selected characteristics, but the record does not identify the 5 matching criteria, describe the matching procedure, or state whether matching was accounted for in the analysis. In general, failing to account for the structure created by matching can affect uncertainty estimates. Whether that occurred in this analysis is not reported in the registry record.

The record labels one analysis and lists seven outcome measures: the primary A1C outcome and six secondary outcomes relating to emergency room visits and hospitalizations. A count of analyses does not explain the statistical model, the outcome-specific estimand, or the handling of multiple outcomes. Those details are not reported. For event outcomes, an incidence risk ratio is named in one secondary-outcome title, but the numerical estimate and method are not given.

How should sample-size planning be understood?

The stated planning assumptions describe illustrations based on the observed effect, a two-sided significance level of 0.05, equal allocation, and a normal approximation. The effect to detect is described as a stated fraction of the observed effect. For time-to-event tables, the assumption specifies total events; other planning results are described as participants per arm. The numerical planning results are not reported in the registry record.

In general, a planning calculation estimates the sample size or event count that would be needed under specified assumptions. It is not a description of how many participants were analyzed, and it is not a measure of the study's observed power. Because the record does not state the effect estimate used for planning or give the resulting sample sizes, the illustrations cannot be independently assessed from the registry record. They should not be treated as evidence that the study could reliably detect a particular effect.

More broadly, statistical interpretation depends on both an estimate and its uncertainty, as well as a clear account of the analysis population and model. The registry record supplies the outcome definitions and some design context, but does not report the primary numerical estimate, its precision, the model, or the analysis population.

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("Cases", "Matched Controls")
n   <- c(366, 366)
m   <- c(-3.06, -1.44)
sp  <- c(1.98, 2.11)
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  Cases  366  -3.060  1.980  -3.264  -2.856
CARM  Matched Controls  366  -1.440  2.110  -1.657  -1.223

11. Multiplicity, Missing Data and Analysis Populations

The registry lists seven outcome measures: percent change in A1C as the primary outcome, plus six secondary outcomes concerning emergency room visits, hospitalizations, and their composites at 30 and 90 days. It also lists one analysis. The record does not explain how the analysis count relates to the outcomes or whether results were tested separately, combined, or evaluated under a prespecified hierarchy.

When several outcomes are examined, the chance of obtaining at least one apparently notable result can increase if each is assessed without accounting for the number of comparisons. In general, multiplicity procedures can address this issue, but they may also change how estimates or p-values are interpreted. Whether any multiplicity adjustment or other strategy was planned or used is not reported in the registry record. No outcome-specific results are reported for assessing this issue.

The record does not describe missing-data frequency, reasons for missing measurements, or methods used to handle missingness. It does not state how many participants had baseline and follow-up A1C available, whether incomplete records were excluded, or whether any imputation was performed. The analysis population is not reported in the registry record. Thus, the listed enrollment of 826 cannot be assumed to be the number included in the primary analysis. Details about exclusions, follow-up completeness, and the handling of missing data are needed to judge how representative the analyzed records may be of enrolled participants.

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 1% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: chest pain requiring hospitalization (0.4%), worsening pre-existing heart failure (0.2%), new diagnosis of breast cancer (0.2%), deep vein thrombosis requiring intervention (0.2%).

For cases, the registry lists 460 at risk for serious adverse events and 5 affected, and 460 at risk for other adverse events and 3 affected. It lists 0 deaths. These are the figures recorded for the cases group; corresponding control-group adverse-event figures are not reported in the registry record.

The listed serious adverse-event terms and percentages are chest pain requiring hospitalization, 0.4%; worsening pre-existing heart failure, 0.2%; new diagnosis of breast cancer, 0.2%; and deep vein thrombosis requiring intervention, 0.2%. The record labels these as the top serious terms. It does not give further detail about how terms were selected or whether one person could appear under more than one term.

These safety figures describe what was recorded, not whether the pathway caused an event. The registry record does not report an assessment of attribution or a comparative safety analysis. The at-risk counts and affected counts should be read as registry summaries, not as a substitute for details on event timing, severity, ascertainment, or group-level comparison.

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

The record establishes that a completed chart-review comparison was conducted to examine the MedStar Diabetes Pathway and standard diabetes care. It identifies percent change in A1C at 12-16 weeks as the primary outcome and lists emergency room visits and hospitalizations as secondary outcomes. It also describes the cases and matched controls and reports some adverse-event summaries for cases.

However, the primary outcome values are not reported in the registry record. There is no numerical estimate of mean change, no comparison between groups, and no measure of precision such as a confidence interval. The secondary-outcome estimates are also not reported. Therefore, the record does not support a conclusion about the size or direction of the observed outcome differences, or how uncertain those differences were.

Even if numerical differences were available, the nonrandomized design would require care in interpreting them. The registry states that controls were matched on 5 criteria but does not identify them. Matching does not ensure balance on all factors related to pathway participation and outcomes. In general, observed group differences in a chart review may reflect both care and underlying differences between groups.

The listed enrollment of 826 is not the same as a reported analysis sample. The record does not state the number analyzed for A1C, the completeness of follow-up, or how missing observations were treated. Nor does it report enough detail to assess the analysis model or any approach to multiple outcomes. The completed status and planning assumptions do not fill those gaps. The most defensible reading is that the registry describes the study's intent, design, and selected safety counts, while leaving its principal comparative results and their statistical reliability unreported.

14. Why This Trial Matters Statistically

This study illustrates a common statistical challenge in clinical quality-improvement work: comparing outcomes during program expansion with outcomes under usual care using existing clinical records. Such reviews can describe practice in real care settings and can help identify questions for further evaluation. Their interpretation depends on how patients entered each group, what information was recorded, and how differences between groups were addressed.

Here, the registry identifies a matched comparison, but the criteria used for matching and the analysis approach are not reported. That distinction matters because matching is not equivalent to randomization. A comparison may remain affected by differences in patient characteristics, care settings, or record completeness that were not addressed by matching or analysis.

The continuous primary endpoint also highlights the importance of reporting an estimate with its uncertainty. Naming a mean and standard deviation as the planned summary does not show the observed result. Without group values, a between-group estimate, and a precision measure, readers cannot judge how large or stable the comparison is. The absence of the primary results is therefore a substantive reporting limitation, not merely a formatting detail.

Finally, multiple clinical outcomes and incomplete records raise questions about multiplicity and analysis populations. The registry lists seven outcomes but does not describe how they were handled together. It does not report missing-data methods or the number analyzed. These omissions limit reproducibility and make it difficult to distinguish a precise comparison from one that is vulnerable to selection or chance. The record is useful for identifying the evaluation framework, but not for drawing a statistical conclusion about comparative outcomes.

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.

Several limitations are specific to the information recorded for this study. First, the primary outcome results are absent: there are no group-level A1C changes, standard deviations, between-group estimate, confidence interval, or p-value in the registry record. The same is true of the listed secondary outcomes. This prevents assessment of the observed effects and their precision.

Second, the comparison is nonrandomized. Although controls were matched to cases on 5 criteria, those criteria and the matching procedure are not reported. The record does not state whether matching was incorporated into the statistical analysis or whether additional adjustment was performed. Residual differences between groups therefore cannot be assessed from the recorded methods.

Third, enrollment of 826 does not identify the number in each group or the number contributing to each outcome. Follow-up completeness, exclusions, missing-data handling, and analysis populations are not reported. These details are particularly relevant to an endpoint requiring baseline and follow-up measurements.

Fourth, the registry lists seven outcome measures and one analysis but does not describe the statistical models, multiplicity approach, or outcome-specific analysis plan. The exact calculation of percent change in A1C is also not reported. Finally, safety figures are reported for cases, without corresponding control figures or details on event attribution. Each limitation narrows what can be concluded from the registry record, even though the study status is completed.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: MedStar Diabetes Pathway Chart Reviews — NCT03442595.
• Registry record last updated 2020-10-20; status last verified 2020-10.
• Results first posted 2020-10-20.
• 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

NCT03442595, MedStar Diabetes Pathway Chart Reviews, is a completed chart-review study sponsored by Medstar Health Research Institute. The registry lists enrollment of 826 and study dates from 2016-07-01 to 2018-12-31. It compares patients with uncontrolled type 2 diabetes participating in the MedStar Diabetes Pathway with matched patients receiving standard diabetes care from a MedStar provider.

The primary outcome is mean percent change in A1C over 12-16 weeks. Six secondary outcomes concern emergency room visits, hospitalizations, and composites at 30 and 90 days. Numerical results for these outcomes, analysis-population details, and the statistical model are not reported in the registry record. The available case safety summary lists 5 affected by serious adverse events, 3 by other adverse events, and 0 deaths, among 460 at risk in each category.