New Ways to Help Patients Improve Their Diabetes Control: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for New Ways to Help Patients Improve Their Diabetes Control (NCT01143870), a randomized na study. The primary outcome is Change in Percent of Glycosylated Hemoglobin 6 Months Following Enrollment.

2. Clinical Question

3. Trial Design

This was a randomized, parallel-group trial with three groups and no masking. The intervention was described as the Diabetes Report Card, which communicated diabetes control using either the standard hemoglobin A1C value, a letter grade, or a face expressing emotion. The trial was sponsored by the University of Pennsylvania and is listed as completed. The registry lists the phase as NA.

The registry reports enrollment of 177. The posted arm counts are 58 for control, 58 for letter grades, and 61 for faces. The registry record gives a start date of 2010-04 and a primary completion date of 2011-07. It does not report further details about the randomization procedure, allocation concealment, recruitment setting, or the timing and delivery of each communication beyond the descriptions of the groups.

The primary endpoint was change in percent of glycosylated hemoglobin 6 months following enrollment. The registry identifies the outcome as a continuous measure summarized by its mean and standard error. Three secondary outcomes are listed: change in patient understanding of disease state, the number of hemoglobin A1C values checked during the 6-month study period, and change in hemoglobin A1C over 12 months from enrollment. Results for those secondary outcomes are not reported in the registry record. The record also does not report the original sample-size justification or the trial’s planned statistical analysis in enough detail to establish how the reported comparisons relate to a prespecified analysis plan.

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 Change in Percent of Glycosylated Hemoglobin 6 Months Following Enrollment, measured over 6 months from enrollment.

The primary endpoint measures change in percent of glycosylated hemoglobin 6 months following enrollment. Hemoglobin A1C is described in the registry as the standard way clinicians assess a patient’s diabetes control. A change outcome compares a participant’s measure over time rather than reporting only a single value at follow-up. The registry’s outcome summary is a mean change for each group, with a standard error describing the uncertainty around each reported group mean.

The endpoint is relevant to the communication question because the intervention was intended to make feedback about glycemic control easier to understand, with improved control identified as a goal. The registry describes the primary endpoint as a measure of glycosylated hemoglobin, not as a direct measure of understanding or of diabetes-related complications. The registry record does not report the detailed calculation of change, the exact measurement schedule beyond the 6-month timeframe, or the approach to participants who did not have a follow-up measurement.

The posted group means are negative: -0.74 for control, -0.55 for letter grades, and -0.89 for faces. These are group-level summaries of the reported change outcome. They do not, by themselves, show how individual participants changed or how the intervention affected understanding. The registry lists understanding and other outcomes separately as secondary outcomes, but their results are not reported in the registry record.

7. Results Posted in the Registry

• Letter Grades: mean -0.55 versus -0.74 on control; difference 0.19 (95% CI -0.74 to 1.12), standardized effect d = 0.07; z-test p = 0.690.
• Faces: mean -0.89 versus -0.74 on control; difference -0.15 (95% CI -1.08 to 0.78), standardized effect d = -0.06; z-test p = 0.753.

8. Statistical Interpretation

9. Statistical Methods Explained

The posted primary outcome is continuous: change in percent of glycosylated hemoglobin. A continuous outcome can take a range of numerical values, so the registry summarizes each group using a mean rather than a count of participants with an event. The reported means are -0.74 for control, -0.55 for letter grades, and -0.89 for faces. Each mean is accompanied by a standard error, which describes the precision of the estimated group mean rather than the spread of individual measurements.

What does the standard error describe?

In general, the standard error reflects how much an estimated mean would vary across repeated samples under a statistical model. A smaller standard error generally corresponds to a more precise estimate of the mean. It is not the same as the standard deviation of individual participant outcomes. The registry reports standard errors of 0.37 for control, 0.3 for letter grades, and 0.3 for faces. It does not report the standard deviations of individual changes as the endpoint summary. A standard error alone therefore does not reveal the full distribution of observed participant values.

How should the mean differences be read?

The comparisons report differences in means between each intervention group and control. The letter-grade estimate is 0.19, calculated as the intervention-group mean relative to the control mean in the posted comparison. The faces estimate is -0.15. Positive and negative signs indicate opposite directions on the reported scale; they do not, by themselves, indicate clinical importance. The intervals for these comparisons are -0.744 to 1.124 for letter grades and -1.084 to 0.784 for faces. Because each interval includes zero, the posted estimates are compatible with no difference as well as with differences in either direction across the interval.

The posted p-values are unadjusted: 0.69 for letter grades and 0.753 for faces. In general, a p-value summarizes the degree of incompatibility between observed data and a null model, conditional on the assumptions used. It does not measure the probability that a treatment works, and a larger p-value does not demonstrate equivalence. The registry does not report the full analysis specification, so details such as the exact model, handling of baseline measurements, or adjustment variables are not reported in the registry record.

Why does over-dispersion matter?

In general, over-dispersion means that the variability in data is greater than a model assumes. It is particularly relevant when outcomes are counts or event rates, where a simple model may underestimate variability and produce intervals that are too narrow. The primary endpoint is a continuous change measure, and the registry reports group means and standard errors. It does not report an over-dispersion assessment or a model for count outcomes. That issue should not be presumed to have been assessed for this trial. The secondary outcome about the number of A1C values checked is a count-type outcome, but no result or analysis method for it is reported in the registry record.

What can randomization contribute?

In general, random assignment is intended to make comparison groups similar on average with respect to both measured and unmeasured characteristics, supporting interpretation of differences as associated with assignment. This trial is listed as randomized, but the registry does not report the random sequence method, concealment procedures, or whether the analysis followed assignment regardless of intervention receipt. The posted group sizes and outcome summaries alone cannot verify those procedures. Masking is listed as none; the record does not describe whether outcome assessment or data analysis was masked.

What is recalculated?

The comparison statistics are unadjusted re-calculations from posted summary statistics. They are mean differences with interval estimates and unadjusted p-values. Such calculations summarize the available group means and their uncertainty, but they cannot reproduce participant-level adjustments or alternative models that are not described in the registry. The record does not flag a discrepancy between the posted summaries and the recalculated comparisons, so there is no reported conflict requiring the sponsor’s own analysis to take precedence. The registry reports zero posted analyses, which means a fuller analysis output is not available in the record.

General statistical interpretation depends on the study design and analysis assumptions as well as the reported values. Here, the randomized parallel-group structure and continuous primary outcome inform the basic comparison, while missing details limit assessment of the exact inferential procedure. The record does not report a prespecified analysis plan, model diagnostics, adjustment for baseline A1C, or sensitivity analyses. Those details should not be inferred from the summary statistics.

10. Confidence Intervals and Precision

The interval estimates for the mean differences provide a range of values compatible with the posted data under the analysis assumptions. In general, a confidence interval expresses the uncertainty of an estimate through a procedure that would cover the target value at a stated rate over repeated use. The confidence level for the intervals displayed in the registry record is not reported, so a particular coverage level should not be assigned to them.

For letter grades compared with control, the estimated mean difference is 0.19 and the interval runs from -0.744 to 1.124. For faces compared with control, the estimated mean difference is -0.15 and the interval runs from -1.084 to 0.784. Both intervals include zero, and each spans values on both sides of zero. Thus, the estimates do not narrowly locate the mean difference in one direction. The posted results leave uncertainty about both the direction and magnitude of a possible difference.

The group standard errors are 0.37 for control, 0.3 for letter grades, and 0.3 for faces. Standard errors indicate precision of the means, while the comparison intervals address uncertainty in differences between groups. Neither measure describes all individual variation. The registry record does not report standard deviations for the primary endpoint, participant-level data, or a confidence level for the comparison intervals, limiting further assessment of precision.

Precision is distinct from statistical significance and from clinical importance. An interval that includes zero does not establish that the true difference is exactly zero; it indicates that the posted data do not isolate a nonzero mean difference within the reported interval. The record does not define a clinically important difference for this endpoint, so the clinical meaning of values across either interval is not reported in the registry record.

11. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning figures are participants per arm for each two-group comparison, not total participants across all trial groups and not total events. The planning illustration uses a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. For each comparison, it considers effects equal to 100%, 75%, or 50% of the observed effect. These are planning illustrations based on the observed effect, not the trial’s original design assumptions, which are not reported in the registry record.

The 80% power column gives the planned participants per arm for a design with an 80% chance of detecting the stated effect under the illustration’s assumptions, if that effect and model assumptions were correct. The 90% power column gives the planned participants per arm for a design with a 90% chance under those same assumptions. These are prospective planning probabilities, not probabilities that a result is true and not a measure of the power achieved by this completed trial. This is not observed power.

For letter grades compared with control, planning for 100% of the observed effect gives 2,862 participants per arm for 80% power and 3,831 for 90% power. For faces compared with control, those figures are 4,664 and 6,243 participants per arm. The additional rows show that planning for 75% or 50% of the observed effect requires more participants per arm than planning for the full observed effect. In general, smaller effects are harder to distinguish from random variation and therefore need more participants to achieve the same target power.

The observed effects are very small in the posted comparisons, so the corresponding planning illustrations yield large participant counts. These counts should not be interpreted as a recommendation or as evidence about the trial’s actual design. The registry does not report the original sample-size justification, and the planning figures use assumptions specified for the illustration rather than assumptions documented as part of the trial protocol.

12. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = 0.190; pooled standard deviation σ = 2.565.
2. n = 2 × (1.96 + 0.842)² × 2.565² / 0.190² = 2,861.3, rounded up to 2,862 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- 0.19                      # observed difference in means
sigma <- 2.56517055963146                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(1143870)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 2,862 (80% power), 3,831 (90% power)
 75% of observed effect: n per arm = 5,087 (80% power), 6,810 (90% power)
 50% of observed effect: n per arm = 11,446 (80% power), 15,322 (90% power)
power.t.test: 2862.3 
Simulated power at n = 2,862 per arm: 0.802 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

13. Multiplicity, Missing Data and Analysis Populations

The trial has three groups and the posted primary comparisons assess letter grades versus control and faces versus control. The reported p-values for these comparisons are unadjusted. The registry record does not report whether a multiplicity procedure was planned or whether these comparisons were designated in advance as a family for adjustment. When several hypotheses are tested, multiplicity can increase the chance of at least one false-positive result across the family; in general, adjustment methods address that concern by changing the inferential threshold or procedure. No such adjustment should be assumed from the posted values.

The registry lists three secondary outcomes: change in patient understanding of disease state, the number of hemoglobin A1C values checked during the 6-month study period, and change in hemoglobin A1C over 12 months from enrollment. Their results are not reported in the registry record. The record lists four outcome measures in total, but does not provide enough detail to establish how the primary and secondary outcomes were ordered in the analysis plan or how multiple outcomes were handled. It also reports zero analyses, so no additional analysis outputs are available in the record.

Missing-data handling is not reported in the registry record. The posted arm counts describe enrollment, but the number of participants contributing primary endpoint data is not separately stated. The record does not report whether any follow-up measurements were missing, how missing measurements were handled, or whether sensitivity analyses were performed. It also does not specify an analysis population such as all randomized participants or participants with observed follow-up values.

These omissions matter because outcome availability and analysis-population definitions can affect interpretation of group summaries. They do not establish that data were missing or that analyses were biased; those details are simply not reported in the registry record. The registry also does not report the original statistical analysis plan, model specification, or any prespecified approach to baseline adjustment. Therefore, the posted comparisons can be interpreted as unadjusted summary comparisons, but cannot be treated as a complete account of the trial’s planned or conducted analyses.

14. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

The registry reports zero serious adverse events in each group: control, letter grades, and faces. The numbers listed as at risk are 58 for control, 58 for letter grades, and 61 for faces. The registry also reports zero other adverse events in each group among those listed as at risk. These are the posted event counts, not evidence that no unrecorded events occurred.

The record does not report the definitions used for serious or other adverse events, the method or period of collection, or further details about event monitoring. The deaths fields contain no posted count, so deaths are not reported in the registry record; they should not be interpreted as zero. No serious adverse event terms are listed. The group figures are descriptive and do not establish whether adverse event ascertainment was complete or comparable across groups.

These safety entries concern the communication groups in this trial and should be read in the context of the registry’s limited reporting. They are not a comprehensive safety assessment, and they do not provide information about event severity beyond the serious-versus-other categories. No conclusion about medical risk should be drawn from these counts alone.

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

The posted primary outcome results show mean changes in glycosylated hemoglobin of -0.74 for control, -0.55 for letter grades, and -0.89 for faces at 6 months. The unadjusted estimated mean differences compared with control are 0.19 for letter grades and -0.15 for faces. Their interval estimates both include zero, and their unadjusted p-values are 0.69 and 0.753. Taken together, these summaries indicate that the registry’s unadjusted comparisons do not precisely distinguish either communication group from control on the primary outcome.

This does not establish that the groups had exactly the same mean change. Nor does it show that either communication format is superior or inferior. The interval estimates include a range of possible differences in either direction, and the registry does not report a threshold for a clinically important difference. The posted data therefore do not resolve how large a difference might matter for this endpoint.

The endpoint is glycosylated hemoglobin change, not patient understanding, behavior, or complications. Although the study’s stated goal concerned making feedback easier to understand and the registry lists understanding as a secondary outcome, no secondary results are reported. The primary outcome results cannot substitute for those unreported outcomes or establish whether the communication formats changed participants’ understanding.

Interpretation is also limited by details that are not reported in the registry record: missing-data handling, analysis population, the full analysis plan, and the confidence level for the interval estimates. The comparisons are unadjusted, and the record does not say whether any multiplicity adjustment was planned. These points do not invalidate the reported summaries, but they constrain how far they can be generalized beyond the specific posted comparisons.

The registry reports a completed randomized trial with three communication approaches and a primary measure at 6 months. The evidence available in the record is best understood as a set of group means and imprecise unadjusted comparisons. It does not provide a definitive account of all trial outcomes or a basis for medical advice.

16. Why This Trial Matters Statistically

This trial raises a statistical question that is easy to overlook: changing how a measurement is communicated is not the same as changing the measurement itself. The registry’s primary endpoint assesses mean change in glycosylated hemoglobin, while the study’s stated goal also concerns patient understanding. Those are related questions, but they require distinct outcome measures and separate interpretation. The registry lists understanding as a secondary outcome, yet its result is not reported.

The design also illustrates the importance of precision when effects are small. The two posted mean differences are close to zero, and both interval estimates include zero while spanning values in either direction. A single point estimate cannot convey that uncertainty. The interval estimates make clear that the data do not tightly determine the direction or size of a difference between either communication group and control.

Planning illustrations based on the observed effects produce large participants-per-arm figures. They show how a small target effect can require substantial enrollment under a specified statistical framework. These illustrations are not a reconstruction of the trial’s original design: the original sample-size assumptions are not reported in the registry record. Keeping design assumptions separate from post hoc planning calculations is essential for avoiding the mistaken impression that a completed trial was designed around the displayed figures.

Finally, the registry record has limited analysis detail. The listed comparisons are unadjusted, while the planned approach to multiple comparisons, missing measurements, and analysis populations is not reported. Statistical interpretation depends not only on the point estimates, but also on the uncertainty, the outcomes selected, and the analysis procedures. This case therefore shows why registry results should be read as a defined but incomplete record of evidence rather than as a full statistical report.

17. 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 registry record does not provide participant-level data or the number of participants contributing measurements to the primary outcome. Although the enrollment counts are reported for each group, the record does not state whether all enrolled participants had a 6-month outcome measurement. Missing-data frequency, reasons for missingness, imputation, and sensitivity analyses are not reported, so the completeness of the endpoint summaries cannot be assessed from the registry alone.

The record also does not report the original statistical analysis plan, detailed randomization procedures, allocation concealment, or the model used for the primary outcome. It lists no masking. The comparison results are unadjusted, but the record does not explain whether baseline values or other variables were considered in the sponsor’s analysis. The confidence level for the reported interval estimates is not stated.

The primary endpoint is reported as a mean change with standard error, but the registry does not provide the standard deviation or distribution of individual changes. It also does not report a threshold for a clinically important difference. Consequently, the size and practical interpretation of differences cannot be judged fully from these summaries. Secondary outcomes are listed without results, including patient understanding and 12-month change in hemoglobin A1C.

Finally, the registry’s adverse-event entries are limited to zero serious and zero other adverse events in each group; definitions, collection procedures, and death counts are not reported. These limitations describe gaps in the registry record, not proof that the trial failed to collect or analyze the information.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: New Ways to Help Patients Improve Their Diabetes Control — NCT01143870.
• Registry record last updated 2018-02-28; status last verified 2018-02.
• Results first posted 2018-02-28.
• 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.

21. Record Summary

NCT01143870 is a completed randomized, open-label, parallel-group trial sponsored by the University of Pennsylvania. It enrolled 177 participants with diabetes and compared standard hemoglobin A1C feedback with a letter-grade interpretation and a face expressing emotion. The registry lists the phase as NA, a start date of 2010-04, and a primary completion date of 2011-07.

The primary outcome was change in percent of glycosylated hemoglobin 6 months after enrollment. The posted means were -0.74 for control, -0.55 for letter grades, and -0.89 for faces. The unadjusted mean differences versus control were 0.19 and -0.15, respectively; both interval estimates included zero. Three secondary outcomes are listed, but their results are not reported in the registry record. Missing-data handling, the original sample-size justification, and the detailed analysis plan are also not reported.