Financial Incentives Telephone Education and Skills Trial in African Americans With Diabetes (FITEST): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Financial Incentives Telephone Education and Skills Trial in African Americans With Diabetes (FITEST) (NCT02722499), a randomized na study. The primary outcome is HbA1c.

2. Clinical Question

3. Trial Design

FITEST was a randomized, single-masked, parallel-group trial sponsored by the State University of New York at Buffalo. The registry lists the study as completed, with a start date of 2016-03 and primary completion in 2021-03. The phase is listed as NA. Enrollment was 60.

The three behavioral arms combined telephone-delivered diabetes education and skills training with high-, moderate-, or low-frequency financial incentives. The primary-outcome description states that 20 subjects were randomized to each arm. The final analysis included 18 participants in the high-frequency arm, 20 in the moderate-frequency arm, and 19 in the low-frequency arm. Reasons for the difference between randomized and analyzed counts are not reported in the registry record.

The primary outcome was HbA1c mean change from baseline to 3 months. The registry describes blood specimens collected at screening and 3 months. Resource Utilization and Cost was a secondary outcome over baseline to 3 months. There are no reported primary statistical analyses in the registry record, and no primary between-arm effect estimate or test is posted. Masking is described as single, but who was masked and how masking was implemented 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: 20 subjects randomized to each arm, but 18 included in final analysis in high frequency, 20 in moderate frequency and 19 in low frequency arms respectively

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 HbA1c, measured over Baseline to 3 months.

HbA1c was the primary endpoint, expressed as mean change from baseline to 3 months. The registry describes blood specimens collected at the screening visit and at 3 months, with those time points used to calculate the change. The trial’s stated aim was to test whether financial incentives added to telephone-delivered diabetes education and skills training could improve HbA1c levels in African Americans with type 2 diabetes.

As a continuous endpoint, HbA1c can be summarized by a mean change, which conveys the average direction and size of change among participants included in each arm’s analysis. The reported confidence interval indicates the uncertainty around that arm-specific estimate under the calculation used. It is not itself a comparison between arms.

The registry reports mean changes of -1.73 in the high-frequency arm, -1.73 in the moderate-frequency arm, and -1.25 in the low-frequency arm. Each is accompanied by a 95% confidence interval. These summaries describe change within each arm over the stated period. Without a reported between-arm analysis, they do not quantify whether differences in incentive frequency are associated with different changes.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is continuous, and the registry reports an arm-specific mean change with a 95% confidence interval. A mean summarizes the central value of a set of measurements. In this trial, the reported quantity is change from baseline to 3 months, rather than a single measurement at one time point. The estimates describe the analyzed participants in each arm.

What does a confidence interval convey?

A confidence interval expresses uncertainty around an estimated mean under the statistical procedure used to construct it. Narrower intervals indicate greater precision under that procedure; wider intervals indicate less precision. The intervals reported for the primary endpoint give a range of values compatible with the respective arm-specific estimates under the calculation’s assumptions. The registry does not describe the interval method in further detail.

A confidence interval around a mean change is not the same as a confidence interval around a difference between arms. Even when each arm has a confidence interval, a comparative question requires a contrast between arms and an estimate of uncertainty for that contrast. No primary comparison of that kind is reported in the registry record. No primary p-value is reported either.

Why is the randomized comparison important?

Random allocation is intended to make the groups comparable on average at the point of assignment, reducing the influence of systematic differences in treatment allocation. The relevant comparative question is whether outcomes differ across the incentive-frequency strategies, with the education and skills-training component described as common to the arms. Arm-level change estimates alone do not answer that question. The registry lists the study as randomized and parallel, but does not report the method used to generate the allocation sequence or conceal assignments.

For the secondary cost outcome, the posted comparisons are unadjusted recalculations from summary statistics. They compare the high-frequency arm with the moderate-frequency arm and with the low-frequency arm. A mean difference expresses the difference between the two group means in the stated order; its confidence interval reflects uncertainty in that contrast under the calculation used. The registry’s posted secondary comparison results are not reported analyses by the trial investigators. No adjustment for multiple comparisons is described, so the unadjusted results warrant cautious interpretation.

What do the planning illustrations mean?

The sample-size illustrations are planning calculations, not a report of the trial’s design justification and not observed power. The registry does not report the trial’s own sample-size justification. The illustrations use the observed effect as a starting point and show planning for 100%, 75%, and 50% of that observed effect. Their stated assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. For non-time-to-event outcomes, the displayed quantity is participants per arm.

These values are conditional on those assumptions and on the effect sizes used for illustration. They do not show the probability that this completed trial detected an effect, and they should not be read as the sample size actually planned or required by the trial protocol. In general, smaller effects require larger samples to detect with a given target power, all else equal. The illustrations are most useful as demonstrations of how sensitive planning can be to the effect one aims to detect.

What role does variability play?

Variability matters because a group mean is estimated less precisely when individual observations are more dispersed, all else equal. Greater variability generally makes it harder to distinguish a real group difference from random variation. For the secondary cost outcome, standard deviations are reported for each arm and a pooled standard deviation is used in the recalculations. Cost data can be dispersed, so a mean may be influenced by observations far from the center. The registry does not report further distributional analyses, and the recalculations should be understood as unadjusted summaries rather than a full model of cost data.

No primary model specification, covariate adjustment, missing-data method, or sensitivity analysis is reported in the registry record. The registry also does not identify an analysis population such as intention-to-treat for the primary endpoint. Those omissions prevent a fuller assessment of the inferential methods behind the posted arm-level summaries.

10. Multiplicity, Missing Data and Analysis Populations

The registry record does not report a multiplicity strategy for the primary and secondary outcomes or for comparisons among the three incentive-frequency arms. It does not describe a hierarchy of hypotheses, an adjustment to significance thresholds, or a rule for prioritizing results. The secondary pairwise comparisons are unadjusted recalculations, so they should be interpreted with particular caution if considered alongside other outcomes or contrasts. The significance level stated for the planning illustrations is an assumption used in those calculations; it is not evidence of the trial’s prespecified analysis threshold.

For the primary endpoint, the registry provides arm-specific means and confidence intervals but no reported between-arm analysis. It also reports no primary p-value. The absence of a posted test means the evidence needed to assess the randomized comparison is incomplete in the registry record, rather than that a particular comparison was statistically significant or nonsignificant.

The primary-outcome description says 20 subjects were randomized to each arm, while the final analysis included 18 in the high-frequency arm, 20 in the moderate-frequency arm, and 19 in the low-frequency arm. The reasons for the differences are not reported. Nor are the timing and extent of missing outcome measurements, the method used to address missing data, or any sensitivity analysis. The registry does not state whether the reported analysis followed an intention-to-treat principle or another analysis-population definition.

These gaps matter because the way participants with missing measurements are handled can affect both the estimate and its uncertainty. Without a description of the missing-data approach and analysis population, readers cannot determine whether the arm summaries represent all randomized participants, only those with observed measurements, or another defined group. The registry also does not report a statistical analysis plan that would clarify which methods and outcomes were prespecified.

11. 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% reporting threshold.

The registry lists 20 participants at risk in each arm and reports 0 serious adverse events, 0 other adverse events, and 0 deaths in each arm. No serious adverse-event terms are listed. These are the safety figures recorded for the groups during the trial; they do not establish that adverse events could not occur or that none went unrecorded.

The registry does not provide further detail about how adverse events were defined, collected, monitored, or attributed, or the period covered by the safety summaries. With the number of participants at risk listed for each arm, zero recorded events should be interpreted as an absence of reported events in this record, not as proof that the intervention carries no risk. The figures also do not provide a comparative safety analysis or quantify uncertainty around event rates.

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

The registry reports negative mean HbA1c changes from baseline to 3 months in all three incentive-frequency arms, with a 95% confidence interval for each arm-specific estimate. This describes the direction and uncertainty of the observed mean change among participants included in each arm’s final analysis. The results do not, on their own, show whether the high-, moderate-, or low-frequency incentive structure performed differently.

The trial was randomized, but the registry does not post a primary analysis comparing the randomized groups. The arm-level summaries therefore cannot be treated as evidence that one incentive frequency caused more change than another. Nor do they establish that the observed changes were caused by the study intervention as a whole; the posted summaries do not separate the effects of telephone education, skills training, incentives, and other influences over time.

The intervals describe statistical precision, not clinical importance. The registry does not report a threshold for a clinically meaningful HbA1c change, and an interval’s position relative to zero does not answer that question. The absence of a reported primary comparison also means the results cannot quantify the uncertainty around differences between incentive frequencies.

The secondary resource-use outcome adds descriptive context but does not fill the gap in primary comparative inference. Its recalculated comparisons are unadjusted, and the record does not describe a multiplicity strategy. Sample-size illustrations are planning aids based on stated assumptions, not evidence of the trial’s achieved power or its original design rationale.

Overall, the posted record supports a cautious description of arm-specific outcomes. It does not support a definitive conclusion about comparative efficacy, clinical importance, or the effect of changing incentive frequency. Missing-data methods and analysis-population definitions are also not reported, which further limits interpretation of the final-analysis summaries.

13. Why This Trial Matters Statistically

This trial illustrates a distinction that is central to clinical-trial interpretation: a change within an arm is not the same estimand as a difference between arms. The primary results report mean change and confidence intervals for each incentive-frequency group, while the clinically relevant randomized comparison would require a direct contrast. Because that primary contrast is not reported, the arm-level results cannot answer whether the incentive frequencies differed.

It also highlights why the analysis population and missing-data methods matter. The numbers included in the final analysis do not match the number randomized in every arm, but the registry does not explain the differences. Without knowing how missing observations were handled, readers cannot fully judge the representativeness or precision of the posted estimates.

The record also distinguishes trial results from planning illustrations. Sample-size figures based on observed effects can show how planning assumptions influence required enrollment, but they are not evidence of observed power and do not reveal the trial’s own design justification. Keeping those purposes separate prevents retrospective calculations from being mistaken for protocol decisions.

Finally, secondary unadjusted comparisons have a different evidential status from a prespecified primary analysis. When multiplicity handling and the analysis plan are not reported, secondary findings should be treated as exploratory. These features make FITEST a useful example of how transparent reporting of contrasts, uncertainty, missingness, and analysis plans is necessary to move from descriptive summaries to credible comparative inference.

14. 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 primary limitation for statistical interpretation is that the registry does not report a between-arm analysis for HbA1c. Arm-specific mean changes and confidence intervals do not provide an estimate or uncertainty interval for the randomized comparisons among incentive frequencies.

The final-analysis counts differ from the 20 participants randomized to each arm, and the reasons are not reported. Missing-data handling, the primary analysis population, and sensitivity analyses are also not reported. These gaps make it difficult to assess the impact of incomplete outcome data on the summaries.

The registry does not report the trial’s sample-size justification, primary statistical model, covariate-adjustment strategy, or statistical analysis plan. It also does not report how the single masking was implemented, the allocation-sequence method, or allocation concealment. Those details would help readers assess prespecification and possible sources of bias.

Only Resource Utilization and Cost is listed as a secondary outcome, and the posted pairwise comparisons are unadjusted recalculations from summary statistics rather than reported trial analyses. The record does not describe multiplicity handling or a distributional analysis for the cost outcome. Safety reporting gives event counts but no further detail on definitions, monitoring, or the period covered. These limitations concern what can be concluded from the registry record; they do not establish that the unreported procedures were absent from the full study.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Financial Incentives Telephone Education and Skills Trial in African Americans With Diabetes (FITEST) — NCT02722499.
• Registry record last updated 2024-07-12; status last verified 2024-07.
• Results first posted 2021-11-05.
• 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.

18. Record Summary

NCT02722499, FITEST, was a completed randomized, single-masked, parallel-group trial sponsored by the State University of New York at Buffalo. Enrollment was 60. The arms paired telephone-delivered diabetes education and skills training with high-, moderate-, or low-frequency financial incentives.

The primary endpoint was HbA1c mean change from baseline to 3 months. The registry reports arm-level means and 95% confidence intervals, but no primary between-arm analysis. The final analysis included 18, 20, and 19 participants in the high-, moderate-, and low-frequency arms, respectively; reasons for the differences from the randomized counts are not reported.

Resource Utilization and Cost was the listed secondary outcome. The record reports no serious adverse events, other adverse events, or deaths in the groups. The primary analysis plan, missing-data methods, multiplicity strategy, and trial sample-size justification are not reported in the registry record.