Hyperglycemia and Cardiovascular Outcomes With Type 2 Diabetes: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Hyperglycemia and Cardiovascular Outcomes With Type 2 Diabetes (NCT00191282), a randomized phase 4 study. The primary outcome is Number of Participants Who Experienced a Primary Combined Outcome.

2. Clinical Question

The clinical question concerns two approaches to insulin treatment, distinguished by the glucose pattern they were intended to target: postprandial hyperglycemia versus fasting and interprandial hyperglycemia. The registry describes the study objective as demonstrating a difference between these strategies in time to the first event in a composite cardiovascular outcome.

The composite included cardiovascular death, nonfatal myocardial infarction, nonfatal stroke, coronary revascularization, and hospitalization for acute coronary syndromes. The primary outcome timeframe began at randomization and included an 18 month initial treatment period, with extended treatment follow-up up to 5.5 years. The primary outcome is described as adjudicated, but details of the adjudication procedures are not reported in the registry record.

The question is about a comparison between assigned treatment strategies, not a direct comparison of individual insulin products or glucose measurements. The registry record does not report the specific eligibility criteria, baseline characteristics, or treatment targets needed to characterize the participants and regimens in greater detail.

3. Trial Design

HEART2D is a completed, randomized, parallel-group phase 4 study. The registry lists enrollment of 1116 and no masking. The sponsor was Eli Lilly and Company. The study start and primary completion dates are recorded as 2002-10 and 2007-10.

Participants were assigned to one of two insulin strategies. The postprandial arm used premeal insulin lispro, with or without bedtime NPH. The fasting arm used NPH, insulin glargine, or human insulin 30/70. The registry lists insulin lispro, NPH, insulin glargine, and human insulin 30/70 among the interventions. The allocation ratio, sequence generation, concealment procedures, dose titration rules, and treatment targets are not reported in the registry record.

The primary outcome results list 557 participants in the postprandial group and 558 in the fasting group. The stated primary population was all randomized patients who took at least one dose of study drug. The record does not explain the difference between the listed enrollment and the arm denominators for this outcome, nor does it report how many randomized participants did not take a dose or were excluded from the analysis.

Follow-up for the primary outcome ran from randomization to the first event, during an 18 month initial treatment period and an extended treatment follow-up period up to 5.5 years. The study used a time-to-event approach in its reported log-rank analyses. Details of visit schedules, follow-up completeness, censoring rules, and the statistical analysis plan 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.

Before assignment: One patient was randomized but discontinued prior to treatment and is not included in any of the analyses. In the Participant Flow table, the term study outcome is used to differentiate it from serious adverse events (SAEs). By definition, the study outcomes were all SAEs as it was a cardiovascular outcome trial.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All randomized patients who took at least one dose of study drug

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 Number of Participants Who Experienced a Primary Combined Outcome, measured over Randomization (Day 0) until first occurrence of primary combined outcome (18 month initial treatment period, extended treatment follow-up period up to 5.5 years).

The primary endpoint was the number of participants who experienced a primary combined outcome. Its components were cardiovascular death, nonfatal myocardial infarction, nonfatal stroke, hospitalization for acute coronary syndromes, and coronary revascularization procedures planned after randomization. The primary objective focused on time until the first occurrence, from randomization through the stated follow-up.

A composite endpoint combines several event types into one outcome. In general, composites can capture a broader range of clinically important events and can increase the number of observed outcomes available for comparison. They also require careful interpretation: the components may differ in severity, frequency, and clinical meaning, and an overall result does not show which component chiefly contributes to the composite. Component-specific outcomes are listed among the secondary measures, but the registry record does not provide all of their results in the registry record for this analysis.

Using time to first event means that the primary comparison concerns how long participants remained without any component event, rather than counting every event experienced by each participant. This approach accounts for the timing of events and can incorporate follow-up that varies across participants, provided the analysis and censoring assumptions are appropriate. The registry record does not report the detailed censoring rules or the adjudication process.

7. Results Posted in the Registry

• Fasting: 32.4% (181/558) versus 31.2% (174/557) on control; risk difference 1.2 percentage points (95% CI -4.3 to 6.7), risk ratio 1.04 (0.87–1.23); unadjusted two-proportion z-test p = 0.668.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Number of Participants Who Experienced Death From Any Cause or Any One of the Primary Outcomes

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Fasting: 33.9% (189/558) versus 32.0% (178/557) on control; risk difference 1.9 percentage points (95% CI -3.6 to 7.4), risk ratio 1.06 (0.90–1.25); unadjusted two-proportion z-test p = 0.496.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Fasting and the control. The interval gives the range of effects that remain compatible with the data.

Number of Participants Who Experienced Any One of the Primary Outcomes Adjusted for Indicators of Metabolic Control

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Fasting: 32.4% (181/558) versus 31.2% (174/557) on control; risk difference 1.2 percentage points (95% CI -4.3 to 6.7), risk ratio 1.04 (0.87–1.23); unadjusted two-proportion z-test p = 0.668.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Fasting and the control. The interval gives the range of effects that remain compatible with the data.

Number of Participants Who Experienced Primary Outcomes Adjusted for Metabolic Control and Major Cardiovascular (CV) Risk Factors

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Fasting: 32.4% (181/558) versus 31.2% (174/557) on control; risk difference 1.2 percentage points (95% CI -4.3 to 6.7), risk ratio 1.04 (0.87–1.23); unadjusted two-proportion z-test p = 0.668.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Fasting and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes extend the primary question in several directions: a combined measure that adds death from any cause, analyses of the primary outcome adjusted for metabolic control, and analyses adjusted for metabolic control and major cardiovascular risk factors. Other listed outcomes separate types of cardiovascular events, including death, myocardial infarction, stroke, and hospitalization for acute coronary syndromes. The registry lists 25 secondary outcomes as posted, but the summarized details do not include numerical results for every listed outcome.

For the three secondary measures with reported log-rank results, the posted p-values are 0.715, 0.914, and 0.706. These results do not supply a clear signal of a difference between strategies, and they broadly accord with the primary comparison. They do not independently confirm the primary result or prove that the strategies are equivalent. The adjustment analyses also depend on model details and covariate definitions that are not reported in the registry record.

Secondary analyses generally carry less weight than a prespecified primary analysis because many outcomes and analytic choices create more opportunities for chance findings. In general, when multiple hypotheses are tested, the chance of at least one nominally unusual result can increase unless the analysis plan accounts for multiplicity. The registry record does not report whether a multiplicity adjustment or hierarchy was specified. Therefore, secondary findings should be regarded as supportive or exploratory rather than treated as separate definitive tests.

The secondary results are most useful as context for the breadth of outcomes considered and for whether they appear to align with the primary comparison. They should not be used to selectively elevate one outcome without a reported prespecified analysis strategy, complete outcome reporting, and adequate precision information.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Number of Participants Who Experienced Death From Any Cause or Any One of the Primary Outcomes
• Number of Participants Who Experienced Any One of the Primary Outcomes Adjusted for Indicators of Metabolic Control
• Number of Participants Who Experienced Primary Outcomes Adjusted for Metabolic Control and Major Cardiovascular (CV) Risk Factors
• Number of Participants Who Experienced Death From Any Cause
• Number of Participants Who Experienced Cardiovascular (CV) Death
• Number of Participants Who Experienced Myocardial Infarction (MI)
• Number of Participants Who Experienced Stroke
• Number of Participants Who Experienced Hospitalization for Acute Coronary Syndromes (HACS)

9. Statistical Interpretation

10. Statistical Methods Explained

The registry includes both a time-to-event analysis and counts of participants experiencing the primary outcome. These approaches describe related but distinct features of the data. The time-to-event method uses information about when an event occurred and the length of follow-up. The unadjusted recalculation uses the numbers of participants with an event and the group denominators, without modeling event timing.

What does a log-rank analysis test?

In general, the log-rank test compares the event experience of groups over follow-up, using the timing of events and the number of participants still at risk at relevant times. It is commonly used to compare time-to-event distributions. The registry reports a log-rank hazard ratio for the primary outcome, but does not provide the model specification, its assumptions, or the confidence-interval limits. The registry record also does not provide the event-time data needed to independently reproduce the analysis.

A hazard ratio summarizes relative event rates over time under the model used. In general, interpreting a single hazard ratio as a constant relative rate depends on assumptions about how the relative event experience behaves over time. The registry record does not report whether proportional hazards were assessed or whether alternative time-to-event analyses were planned. The posted HR of 0.98 and p = 0.866 are the sponsor's reported analysis; they should not be treated as a direct estimate of the difference between the arm-level event proportions.

What does the risk difference add?

The risk-difference recalculation compares the observed proportions with an event in each arm. It is an unadjusted summary based on the posted counts: 174 events among 557 participants in the postprandial group and 181 among 558 in the fasting group. The reported risk difference is 0.012, with a confidence interval from -0.043 to 0.067. This scale expresses an absolute difference in event probability over the summarized follow-up, while the hazard ratio describes relative time-to-event experience. The recalculation does not account for event timing or censoring and is not a substitute for the registry's time-to-event analysis.

In general, a confidence interval describes a range of values compatible with the data under the method used. It is not a guarantee that the true value lies in the interval, nor does it assign probabilities to individual values in the usual frequentist interpretation. The interval's inclusion of zero is consistent with no risk difference, while its inclusion of values in both directions reflects uncertainty about the direction and size of the difference.

Why distinguish the analyses?

The registry reports two kinds of results: participant counts for the primary outcome and a log-rank HR with a p-value. They answer different questions and may yield different numerical summaries without contradiction. The count-based result ignores event timing, whereas the time-to-event result uses it. The unadjusted recalculation is useful for understanding the observed arm-level event proportions and their absolute contrast. The sponsor analysis is the registry's reported inferential result for the time-to-event comparison.

The registry also lists secondary log-rank HR results, including analyses of a composite that adds death from any cause and analyses adjusted for metabolic control or cardiovascular risk factors. Adjustment means accounting for specified variables in an analysis, but the registry record does not describe which variables were used, how they were measured, or the model details. The adjusted HR values are reported, but their confidence-interval limits are not.

Methods such as randomization aim to make comparison groups comparable at baseline on average. In general, random assignment supports causal interpretation of a between-group contrast, subject to implementation, adherence, missing data, and analysis choices. The registry record identifies the trial as randomized but does not report enough detail to assess sequence generation, concealment, adherence, or the handling of missing observations. Thus the design label alone does not resolve those questions.

The study is described as having no masking. In general, lack of masking can affect behavior, co-interventions, reporting, or outcome assessment, although an adjudicated endpoint can be one safeguard against subjective classification. Details of who knew assignments and how outcome adjudication was conducted are not reported in the registry record. These limitations apply to interpretation of the posted methods rather than establishing that bias occurred.

11. Confidence Intervals and Precision

The unadjusted risk-difference estimate is 0.012, and its 95% confidence interval runs from -0.043 to 0.067. Because the interval crosses zero, it is compatible with no difference on this scale. Because it also extends in both directions, it is compatible with a range of possible group differences, not only the point estimate. The interval therefore indicates uncertainty about both the direction and size of the absolute contrast.

In general, an interval is more informative than a p-value alone because it shows the range of effect estimates that are compatible with the data and analysis. The interval does not establish that every value within it is equally plausible, and it does not indicate whether any particular difference would be clinically important. A threshold for clinical importance is not reported in the registry record.

The registry's time-to-event result is HR 0.98 with p = 0.866, and the record labels the analysis as having a 95% confidence level but does not report the confidence-interval limits. Without those limits, the precision of the sponsor's HR cannot be assessed directly from the registry record. The recalculated risk-difference interval should not be used as though it were the interval for the hazard ratio.

The arm counts are 174 of 557 and 181 of 558. These counts describe the posted primary outcome results, but the record does not provide participant-level data, event times, or additional information needed to examine precision under alternative assumptions. The reported estimates are thus best read with their stated methods and limitations in view.

12. 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 illustrations express required sample sizes as participants per arm for the binary participant-count outcomes. They are not total sample sizes. The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set to the stated fraction of the observed effect. These figures are planning illustrations based on the observed effect, not the trial's original design assumptions; the trial's own sample-size assumptions are not reported in the registry record.

The 80% power column describes the number of participants per arm that, under those assumptions and for the effect size being considered, would give an 80% chance of detecting that effect at the stated significance level if the assumptions hold. The 90% power column uses the same kind of planning target but aims for a 90% chance. These are design calculations, not guarantees about a future study's result.

The rows consider planning for 100%, 75%, and 50% of the observed effect. Smaller effects require more participants per arm because they are harder to distinguish from random variation. For the primary outcome, the illustration for 100% of the observed effect gives 23,716 participants per arm for 80% power and 31,748 participants per arm for 90% power. Planning for 75% or 50% of the observed effect requires larger numbers, as reflected in the posted figures.

These figures are not observed power. Observed power is not a useful substitute for reading the estimate and its confidence interval: the illustrations instead show how sample-size requirements change under particular planning assumptions. The record does not report that these calculations formed part of the original trial design.

13. 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, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.312; treatment proportion p₁ = 0.324; average p̄ = 0.318.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6588 = 1.2912.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6588 = 0.5544.
4. n = (1.2912 + 0.5544)² / (0.0120)² = 23,715.4, rounded up to 23,716 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
p0 <- 0.312387791741472                          # control proportion
p1 <- 0.324372759856631                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^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 (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(191282)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 23,716 (80% power), 31,748 (90% power)
 75% of observed effect: n per arm = 42,056 (80% power), 56,300 (90% power)
 50% of observed effect: n per arm = 94,386 (80% power), 126,355 (90% power)
power.prop.test: 23715.4 
Simulated power at n = 23,716 per arm: 0.800 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-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.

14. Analyses Reported by the Sponsor

The registry reports a sponsor log-rank analysis for the primary outcome with HR 0.98 and p = 0.866. The confidence-interval limits are not reported. This is the sponsor's time-to-event analysis and is the primary registry-reported inferential result for the outcome.

The recalculation uses the posted participant counts and gives an unadjusted risk difference of 0.012, a 95% confidence interval from -0.043 to 0.067, and p = 0.668. This calculation is not a reproduction of the sponsor's log-rank analysis: it does not incorporate the timing of events or censoring. The HR and the risk difference use different scales and methods, so their estimates and p-values need not match.

Both analyses fail to identify a statistically clear difference in the posted results, but neither establishes equivalence. For the sponsor's own analysis, the registry-reported HR and p-value take precedence over a simple unadjusted recalculation. The recalculation can still help describe the absolute contrast in the arm-level counts, provided it is kept distinct from the sponsor's time-to-event result.

The registry does not report the complete sponsor analysis plan, model details, or confidence limits for the HR. Consequently, the exact assumptions and precision of the sponsor analysis cannot be fully evaluated from the registry record.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists one primary combined outcome and numerous secondary outcomes, including event components and adjusted analyses. It does not report a multiplicity strategy, such as a testing hierarchy, adjustment procedure, or designation of particular secondary outcomes as confirmatory. In general, testing several outcomes can increase the chance of finding a nominally significant result by chance. Without a stated plan, p-values across secondary outcomes should not be interpreted as if each were the only hypothesis tested.

Missing-data handling is not reported in the registry record. The record does not describe how incomplete follow-up, withdrawal, loss to follow-up, or missing covariate values were handled. It also does not provide the number of participants with incomplete outcome status or the rules used to censor participants in the time-to-event analyses. These details matter because different approaches to incomplete information can affect estimates and their uncertainty.

The primary outcome population is described as all randomized patients who took at least one dose of study drug. This description does not by itself establish that the analysis followed a full intention-to-treat principle, because the registry record does not specify how participants who did not take study drug were treated or whether all randomized participants were included in each analysis. The number analyzed is reported by arm, but the record does not provide a participant flow account or explain exclusions.

The study is randomized and unmasked. The registry does not report allocation concealment, protocol deviations, adherence summaries, or detailed outcome-adjudication procedures. Nor does it report whether the analysis followed a prespecified statistical plan. These items limit assessment of bias and analytic flexibility. Their absence from the registry record should be distinguished from evidence that a particular problem occurred.

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

The posted results show similar observed counts for the primary combined outcome in the two groups: 174 participants in the postprandial arm and 181 in the fasting arm. The unadjusted risk-difference estimate is 0.012, with a 95% confidence interval from -0.043 to 0.067. The sponsor's log-rank analysis reports HR 0.98 and p = 0.866. Taken together, the reported analyses do not demonstrate a statistically clear difference between the strategies on the primary outcome.

That conclusion is narrower than saying the strategies are identical or interchangeable. A failure to detect a difference is not proof that no difference exists. The risk-difference interval includes values on both sides of zero, and the registry does not report an equivalence or noninferiority design, a prespecified margin, or the confidence limits for the sponsor's HR. The results therefore cannot establish equivalence within a defined clinically acceptable range.

The endpoint is a composite, so the overall comparison does not show whether its components had similar patterns. A composite event could arise from different components, and the registry's primary summary does not indicate the contribution of each component. Although component outcomes appear among the secondary measures, the registry entries do not provide all component results or their precision.

The result applies to the strategies as described in the registry, in the study population and follow-up framework recorded there. It does not directly compare isolated insulin products, establish the effect of a specific dose or target, or describe outcomes outside the studied setting. Detailed eligibility criteria and treatment implementation are not reported in the registry record.

These are statistical interpretations of the registry results, not medical advice. The evidence should be understood in light of the trial's reported estimates, the uncertainty around them, the unreported analysis details, and the limits of the outcome information available.

17. Why This Trial Matters Statistically

This trial presents a useful statistical distinction between a treatment-strategy comparison and the specific products included in each strategy. The randomized parallel-group design places the focus on assigned approaches to postprandial versus fasting and interprandial hyperglycemia. That distinction matters because an estimate for a strategy should not automatically be interpreted as an estimate for any one insulin product considered separately.

The primary outcome combines several cardiovascular events and is analyzed over time. The registry also posts participant counts and a time-to-event result. Comparing these summaries illustrates why an absolute difference in observed event proportions and a hazard ratio are not interchangeable: one summarizes arm-level event occurrence, while the other reflects event timing over follow-up.

The estimates also show why statistical significance and precision are different questions. The reported p-values do not indicate a statistically clear difference, while the confidence interval for the recalculated risk difference shows uncertainty spanning values in both directions. A nonsignificant test alone cannot establish equivalence, and a point estimate alone cannot show the range of effects compatible with the data.

Finally, the registry's limited reporting of confidence limits for the sponsor analysis, multiplicity procedures, missing-data handling, and model details highlights the importance of complete statistical reporting. Clear reporting lets readers distinguish a trial's planned analysis from a simple recalculation and assess what the data can support without overstating the conclusion.

18. 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 trial-specific details needed for a fuller appraisal are not reported in the registry record. These include eligibility criteria, baseline characteristics, allocation concealment, treatment targets, dose-adjustment procedures, adherence, and participant flow. The record lists no masking, but does not describe which study personnel or outcome assessors knew treatment assignments.

The primary outcome is a composite with components that may differ in clinical importance and frequency. The registry describes the components but does not provide the full component-level numerical results in the registry record for this analysis. The endpoint's adjudication is stated, but the adjudication process is not described. Follow-up extends from the initial 18 month treatment period up to 5.5 years, while details about follow-up completeness and censoring are not reported.

The sponsor's primary analysis reports an HR and p-value but no confidence-interval limits. The registry also does not provide model specifications, assessment of time-to-event assumptions, or a complete statistical analysis plan. The unadjusted risk-difference recalculation supplies an interval for a different estimand and cannot fill the missing HR interval.

Information on multiplicity, missing outcomes, and the treatment of participants who did not take a dose is not reported. The primary population is described as randomized patients who took at least one dose, but the record does not give a full analysis population account. Adverse-event counts and selected serious-event terms are listed, but the summarized details do not provide all denominators or enough detail to assess comparative safety comprehensively. These reporting limits constrain interpretation; they do not establish that a bias or data problem occurred.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Hyperglycemia and Cardiovascular Outcomes With Type 2 Diabetes — NCT00191282.
• Registry record last updated 2011-01-20; status last verified 2011-01.
• Results first posted 2009-08-04.
• 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.

22. Record Summary

HEART2D was a completed, randomized, parallel-group phase 4 trial sponsored by Eli Lilly and Company, with no masking listed in the registry. It compared a postprandial insulin strategy with a fasting strategy in patients with type 2 diabetes after acute myocardial infarction. The primary outcome was time to the first event in a combined cardiovascular endpoint, followed from randomization through an 18 month initial treatment period and extended follow-up up to 5.5 years.

The primary outcome counts were 174 of 557 for postprandial and 181 of 558 for fasting. The registry reports a log-rank HR of 0.98 with p = 0.866. The unadjusted risk-difference recalculation is 0.012, with a 95% confidence interval from -0.043 to 0.067. These results do not establish a clear between-group difference, and they do not establish equivalence. Important details about analysis methods, missing data, multiplicity, and outcome follow-up are not reported in the registry record.