This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for A Study of 2 Different Formulations of Insulin Lispro in Healthy Participants (NCT02111083), a randomized phase 1 study. The primary outcome is Pharmacokinetic Parameter: Area Under the Serum Insulin Concentration Versus Time Curve From Zero to Time of Return to Baseline (AUC0-tlast).
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT02111083 |
| Phase | Phase 1 |
| Condition | Healthy Volunteers |
| Allocation | Randomized |
| Design model | Crossover |
| Masking | Double |
| Enrollment | 38 |
| Arms in the primary outcome | 2 |
| Lead sponsor | Eli Lilly and Company |
| Results posted | Yes |
2. Clinical Question
Population
Healthy participants were enrolled. The primary endpoint population was participants who had received at least one study treatment and had evaluable pharmacokinetic data. The registry record reports enrollment of 38 participants.
Intervention
Insulin Lispro A was the test formulation: 20 units at a strength of 200 units per milliliter, administered subcutaneously. Participants received it once in two of the four study periods.
Comparator
Insulin Lispro B was the reference formulation: 20 units at a strength of 100 units per milliliter, administered subcutaneously. Participants received it once in two of the four study periods.
Primary question
Did the two formulations produce similar serum insulin exposure, measured by the area under the serum insulin concentration versus time curve from dosing to return to baseline? The registry also describes comparing effects on blood sugar and assessing safety.
The central statistical question is about pharmacokinetic similarity, not whether one formulation is superior to the other. The posted primary comparison is a ratio of geometric least-squares means, with a 90% confidence interval. That interval describes the precision of the estimated ratio, but interpreting it as evidence of bioequivalence requires a prespecified acceptance range. Such a range is not reported in the registry record, so the posted estimate alone does not establish whether a formal bioequivalence criterion was met.
The study also lists pharmacodynamic measures related to glucose infusion and a pharmacokinetic measure of time to maximum observed serum concentration. These capture different features of insulin action and concentration over time. They do not replace the primary exposure endpoint, and results for all listed outcomes are not available in the information presented in the registry record.
3. Trial Design
This was a phase 1 randomized, double-masked crossover study in healthy participants. The registry record reports enrollment of 38 participants. In a crossover design, participants receive more than one treatment during the study, so comparisons can be made within participants as well as between formulations. Here, each participant was scheduled to receive four injections across four study periods: two doses of Insulin Lispro A and two doses of Insulin Lispro B.
Insulin Lispro A was the test formulation, given as 20 units at 200 units per milliliter, and Insulin Lispro B was the reference formulation, given as 20 units at 100 units per milliliter. Both were administered subcutaneously. The exact order of formulations, the randomization procedure, the length of each period, and any washout interval are not reported in the registry record. The masking description is double, but the registry record does not identify which groups of people were masked.
The primary endpoint was measured on Day 1, from predose through 8 hours after dosing in each period. Its analysis population included participants with at least one study treatment and evaluable pharmacokinetic data. The registry record does not report how many participants met that endpoint-specific population definition or whether all enrolled participants completed every period.
The study status is completed. Its recorded start is 2014-05 and its primary completion is 2014-08. The brief description says that participation was expected to last up to approximately 9 weeks for each participant. These design features frame the interpretation: the endpoint compares exposure under the study conditions, while the crossover structure means the comparison is not equivalent to a simple parallel-group trial. Details needed to assess the handling of period or sequence effects 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.
First Intervention (1 Day)
| Stage | Insulin Lispro Dosing Sequence TRTR | Insulin Lispro Dosing Sequence RTRT |
|---|---|---|
| STARTED | 19 | 19 |
| COMPLETED | 19 | 19 |
| NOT COMPLETED | 0 | 0 |
Interval Between Dosing (7-8 Days)
| Stage | Insulin Lispro Dosing Sequence TRTR | Insulin Lispro Dosing Sequence RTRT |
|---|---|---|
| STARTED | 19 | 19 |
| COMPLETED | 19 | 19 |
| NOT COMPLETED | 0 | 0 |
Second Intervention (1 Day)
| Stage | Insulin Lispro Dosing Sequence TRTR | Insulin Lispro Dosing Sequence RTRT |
|---|---|---|
| STARTED | 19 | 19 |
| COMPLETED | 19 | 19 |
| NOT COMPLETED | 0 | 0 |
Interval Between Dosing (7-8 Days)
| Stage | Insulin Lispro Dosing Sequence TRTR | Insulin Lispro Dosing Sequence RTRT |
|---|---|---|
| STARTED | 19 | 19 |
| COMPLETED | 19 | 19 |
| NOT COMPLETED | 0 | 0 |
Analysis population for the primary outcome, as defined in the registry: All participants who had at least one study treatment and had evaluable pharmacokinetic (PK) data.
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.
| Characteristic | Unit | Insulin Lispro Dosing Sequence TRTR | Insulin Lispro Dosing Sequence RTRT | Total |
|---|---|---|---|---|
| Participants analysed | 19 | 19 | 38 | |
| Age, Continuous | years | 32.9 (SD 6.1) | 30.4 (SD 8.2) | 31.7 (SD 7.2) |
| Sex: Female, Male: Female | Participants | 2 | 1 | 3 |
| Sex: Female, Male: Male | Participants | 17 | 18 | 35 |
| Ethnicity (NIH/OMB): Hispanic or Latino | Participants | 0 | 0 | 0 |
| Ethnicity (NIH/OMB): Not Hispanic or Latino | Participants | 19 | 19 | 38 |
| Ethnicity (NIH/OMB): Unknown or Not Reported | Participants | 0 | 0 | 0 |
| Region of Enrollment: Singapore | participants | 19 | 19 | 38 |
6. Primary Endpoint
The registry lists the primary outcome as Pharmacokinetic Parameter: Area Under the Serum Insulin Concentration Versus Time Curve From Zero to Time of Return to Baseline (AUC0-tlast), measured over Day 1, predose through 8 hours post dose in each period..
| Element | Registry entry |
|---|---|
| Result type | Geometric Mean |
| Measure of spread | Geometric Coefficient of Variation |
| Unit | picomole*hour/liter (pmol*h/L) |
| Endpoint classification | not reduced to a single comparison |
The primary endpoint is the area under the serum insulin concentration versus time curve from zero to the time insulin concentration returns to baseline, abbreviated AUC0-tlast. It is measured in picomole*hour/liter. Conceptually, an area under a concentration-time curve captures overall exposure across the measured interval: it reflects both the concentration reached and how long concentration remains elevated. The endpoint is pharmacokinetic, meaning it describes the concentration of insulin in the body over time rather than a direct clinical outcome.
The registry reports the endpoint using geometric means and a geometric coefficient of variation. Geometric summaries are often used for measurements that are positive and may vary proportionally, because comparisons can then be expressed as ratios. The posted inferential comparison is a ratio of geometric least-squares means, which represents the relative exposure estimate for the two formulations under the analysis model.
Exposure is useful when evaluating whether two formulations are handled similarly by the body. However, the meaning of a ratio depends on its uncertainty and on any prespecified acceptance limits. A confidence interval conveys uncertainty around the estimated ratio; formal bioequivalence interpretation also requires the relevant bounds. Those bounds and the detailed endpoint calculation rules are not reported in the registry record. The endpoint therefore informs the pharmacokinetic comparison but, by itself, does not establish a broader conclusion about blood sugar effects or safety.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The registry reports geometric mean AUC0-tlast values of 2200 for Insulin Lispro A and 2230 for Insulin Lispro B. The corresponding geometric coefficients of variation are 16% and 17%. The reported inferential comparison is a ratio of geometric least-squares means of 0.986, with a 90% confidence interval from 0.965 to 1.01. The analysis includes 38 participants for each formulation in the posted arm summaries.
The point estimate is close to 1, the value that would represent equal estimated exposure on a ratio scale. The confidence interval describes a range of values compatible with the data and the analysis, under the model assumptions. Its inclusion of 1 means that equal exposure is among the values compatible with this interval; it does not by itself demonstrate equivalence. Equivalence is assessed against prespecified margins, and those margins are not reported in the registry record. The result should therefore be read as an estimate with uncertainty, not as an automatic pass or fail against a formal bioequivalence criterion.
The registry reports no p-value for this comparison. A p-value is not necessary to describe the point estimate and confidence interval, and its absence does not make the interval uninterpretable. However, the registry record does not provide enough information to reconstruct all inferential details, including the full statistical model, the treatment of period and sequence, or the criteria used for formal equivalence assessment.
The endpoint population is defined as participants who received at least one study treatment and had evaluable pharmacokinetic data. The posted summaries list 38 participants for each formulation, but the registry record does not report additional participant-flow information or explain whether every participant contributed to every endpoint analysis. Because the study used a crossover design, the same participant could contribute observations under both formulations; the two arm summaries should not be interpreted as evidence of two distinct groups of 38 people.
The registry lists four secondary measures: time to maximum observed serum concentration, total amount of glucose infused, maximum glucose infusion rate, and time to maximum glucose infusion rate. Numerical results for these outcomes are not reported in the registry record in the information available for interpretation. Accordingly, the posted primary estimate addresses serum insulin exposure, while the other stated pharmacokinetic and pharmacodynamic questions remain unresolved by the numerical results.
Safety summaries report no serious adverse events among 38 participants at risk for each formulation. The record also lists other adverse events affecting 21 participants for Insulin Lispro A and 16 for Insulin Lispro B. These are descriptive counts; the registry record does not provide enough detail in this summary to infer causality, comparative risk, or the nature and severity of individual non-serious events. The study's completed status and posted summaries describe the available record, not a general conclusion beyond the trial population and measurements.
9. Statistical Methods Explained
Statistical methods help distinguish an estimated treatment contrast from the uncertainty around it, and help clarify which claims the data can support. The principal posted result is a ratio of geometric least-squares means for AUC0-tlast, accompanied by a 90% confidence interval. The registry record does not describe the full model or all analysis choices, so interpretation should remain tied to the quantities it reports.
What does the ratio represent?
A ratio compares one estimated quantity with another on a relative scale. A ratio of 1 corresponds to equal estimated values. The reported ratio is 0.986. This is the point estimate from the posted comparison, not a direct measure of the probability that the formulations are identical. Ratios are commonly used for positive pharmacokinetic measurements because proportional differences can be more meaningful than absolute differences when values vary in scale.
The arm-level summaries are geometric means, while the inferential result is described as a ratio of geometric least-squares means. A geometric mean summarizes values on a multiplicative scale, usually by working with logarithms and then returning to the original scale. A least-squares mean is a model-based adjusted mean. The registry record does not state the model terms used to obtain the adjusted estimates, so it is not possible to specify which factors, if any, were included in that adjustment.
How should the confidence interval be read?
The 90% confidence interval runs from 0.965 to 1.01. In general, a confidence interval is a method for expressing the precision of an estimate: intervals that cover a narrower range convey more precision than intervals covering a wider range, all else equal. A confidence interval is not a statement that a particular value has a specified probability of being true. Its interpretation depends on the analysis method and assumptions.
For a bioequivalence question, a confidence interval is ordinarily judged against prespecified limits for acceptable similarity. The registry record reports the interval but does not report those limits. Thus, the interval can be described and its relation to 1 noted, but a formal determination against acceptance bounds cannot be made from the registry information alone. The registry record also does not provide a p-value for the comparison.
Why use geometric summaries?
Pharmacokinetic measurements are often positive and can show proportional variation. Geometric summaries are suited to this type of scale because they describe a typical value multiplicatively rather than as an arithmetic average. The geometric coefficient of variation describes dispersion relative to the geometric mean. The posted coefficients are 16% for Insulin Lispro A and 17% for Insulin Lispro B. They convey variability in the summarized measurements, but they are not confidence intervals for the ratio and do not by themselves indicate whether the formulations are equivalent.
What does crossover design change?
In a crossover study, a participant receives multiple treatments at different times. This allows a treatment comparison to use information from the same participant under each formulation, which can reduce the influence of stable differences between participants. It also makes period, sequence, and possible carryover considerations relevant to the analysis. The registry record confirms a randomized crossover design with four study periods but does not report the treatment sequence, washout details, or how period and sequence effects were handled. Those details should not be inferred from the design label alone.
The posted summaries list 38 participants for each formulation. In a crossover study, these formulation-specific counts do not imply that there were separate groups of 38 participants. The primary population is defined by receipt of at least one treatment and evaluable pharmacokinetic data. The registry record does not report detailed participant flow or how missing or unevaluable measurements were handled.
What do the planning illustrations mean?
Planning illustrations based on an observed effect are not the same as the trial's original design justification, and they are not observed power. The stated planning framework uses a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation, with the effect to detect expressed as a fraction of the observed effect. These are planning assumptions, not evidence that the trial was designed around them. The original sample-size justification is not reported in the registry record. In addition, an illustration framed around two arms does not automatically capture the analysis structure of a crossover trial.
More generally, statistical power is a design-stage probability defined under assumed effects and variability. Once results have been observed, a post hoc sample-size illustration can show what a planning calculation would imply under stated assumptions, but it cannot establish the probability that the observed result is correct. It should not replace the estimate, confidence interval, endpoint definition, or study-design information when interpreting the posted comparison.
10. Analyses Reported by the Sponsor
| Method | Estimate | 95% CI | P-value |
|---|---|---|---|
| not stated | Ratio of geometric least square means 0.986 | 0.965 to 1.01 | not reported |
The registry's posted inferential result is a ratio of geometric least-squares means of 0.986, with a 90% confidence interval from 0.965 to 1.01. The record does not provide a separate recalculated ratio or confidence interval for comparison with that result. The arm summaries provide geometric mean values and geometric coefficients of variation, but those summaries alone do not reproduce a model-based least-squares comparison in a crossover design.
No disagreement between a sponsor analysis and a recalculation is identified in the registry record. The reported comparison should therefore be read as the available inferential analysis, rather than replaced by an estimate reconstructed from arm summaries. A simple unadjusted calculation from summary values would not necessarily match an analysis that accounts for the study design or uses participant-level data. The registry record does not report the full model specification, so the precise relationship between the model-based result and any potential unadjusted calculation cannot be established.
The reported interval is informative about the precision of the posted estimate, but the formal interpretation of bioequivalence also depends on acceptance limits. Those limits are not reported in the registry record. The sponsor's reported numerical result and the missing decision criteria are therefore distinct parts of the interpretation.
11. Multiplicity, Missing Data and Analysis Populations
The registry record lists seven outcome measures and seven analyses, including the primary pharmacokinetic endpoint and secondary pharmacokinetic and pharmacodynamic measures. Four secondary outcomes are identified: time of maximum observed serum concentration, total amount of glucose infused, maximum glucose infusion rate, and time of maximum glucose infusion rate. Numerical results for those secondary measures are not reported in the registry record in the information.
When several outcomes or comparisons are assessed, multiplicity can increase the chance of at least one apparently notable finding if each is interpreted independently. Statistical plans may address this through a hierarchy, adjustment, or other strategy. The registry record does not report whether multiplicity was addressed, whether secondary outcomes were formally tested, or whether a hierarchy of endpoints was prespecified. No adjustment should therefore be assumed.
The primary endpoint population is defined as all participants who had at least one study treatment and evaluable pharmacokinetic data. The posted arm summaries each list 38 participants. The registry record does not report a participant flow diagram, the number excluded from the endpoint analysis, or the reasons for any missing or unevaluable measurements. It also does not state whether missing observations were imputed, omitted from particular analyses, or handled using another method.
These gaps matter because missingness and multiplicity affect how confidently results across endpoints can be interpreted. The available record supports description of the posted primary estimate and its confidence interval, but does not permit assessment of the complete endpoint family, missing-data assumptions, or the consequences of any unreported analyses. Interim analyses and stopping rules are not reported in the registry record.
12. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| Insulin Lispro A | 0 | 38 | 0.0% |
| Insulin Lispro B | 0 | 38 | 0.0% |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 5% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| Insulin Lispro A | 21 of 38 |
| Insulin Lispro B | 16 of 38 |
Most frequent non-serious adverse event terms
| Event term | Organ system | Insulin Lispro A | Insulin Lispro B |
|---|---|---|---|
| Catheter site related reaction | General disorders | 7 of 38 | 5 of 38 |
| Infusion site pain | General disorders | 5 of 38 | 4 of 38 |
| Infusion site swelling | General disorders | 5 of 38 | 4 of 38 |
| Infusion site bruising | General disorders | 3 of 38 | 1 of 38 |
| Injection site bruising | General disorders | 3 of 38 | 1 of 38 |
| Application site erythema | General disorders | 2 of 38 | 1 of 38 |
| Pyrexia | General disorders | 2 of 38 | 1 of 38 |
| Vessel puncture site bruise | General disorders | 1 of 38 | 2 of 38 |
| Vessel puncture site pain | General disorders | 2 of 38 | 1 of 38 |
| Headache | Nervous system disorders | 2 of 38 | 1 of 38 |
The registry reports zero serious adverse events among 38 participants at risk for Insulin Lispro A and zero among 38 participants at risk for Insulin Lispro B. These are observed counts in the posted safety summary. They do not establish that serious events could not occur, nor do they provide a precise estimate of comparative safety. The registry record does not list serious adverse event terms.
The summary also reports other adverse events affecting 21 participants for Insulin Lispro A and 16 for Insulin Lispro B, with 38 participants at risk for each. The record does not provide event-level details in the information, so the counts cannot establish which events occurred, their severity, timing, or whether they were considered related to treatment. The terms used in the registry summary should not be treated as a causal comparison.
Deaths are not reported in the registry record. That absence should not be interpreted as a count of zero. The safety information is descriptive and should be read in the context of the study's healthy-participant population, crossover design, and posted observation period. It does not provide a basis for conclusions about safety in other populations or settings.
13. What the Result Does, and Does Not, Mean
The result provides a quantitative comparison of serum insulin exposure after the two formulations in this randomized, double-masked crossover study of healthy participants. The reported ratio of geometric least-squares means is 0.986, and its 90% confidence interval extends from 0.965 to 1.01. The point estimate is near 1, while the interval expresses uncertainty around that estimate. The arm-level geometric means are 2200 for Insulin Lispro A and 2230 for Insulin Lispro B.
These findings do not, on their own, establish formal bioequivalence. That interpretation requires specified acceptance limits, which are not reported in the registry record. Nor does a result for the primary pharmacokinetic endpoint answer every question about blood sugar response or safety. The registry lists pharmacodynamic outcomes involving glucose infusion, but numerical results for those outcomes are not reported in the information available in the registry record.
The findings also should not be generalized beyond the study population without additional evidence. The participants were healthy volunteers, and the study description concerns short-term treatment and measurement during study periods. The registry record does not report enough detail to assess every feature of the analysis, including treatment sequence, period effects, missing-data handling, or the statistical approach to multiple outcomes.
For safety, zero serious adverse events were recorded among 38 participants at risk for each formulation, while other adverse events were reported for 21 and 16 participants, respectively. These counts are descriptive, and the record does not supply event-level details needed to interpret causality or compare event types. The appropriate statistical conclusion is therefore bounded: the record reports a near-unity exposure ratio with a stated confidence interval, alongside limited descriptive safety counts, but does not report the criteria and details required for broader claims.
14. Why This Trial Matters Statistically
This trial is statistically informative because it illustrates the distinction between a point estimate, its uncertainty, and a formal decision rule. The estimated ratio of geometric least-squares means gives a concise relative comparison of serum insulin exposure. The 90% confidence interval adds information about precision. Neither quantity, taken alone, says whether a prespecified bioequivalence standard has been satisfied; that requires acceptance bounds, which are not reported in the registry record.
The crossover design also matters for reading the summaries. Participants received both formulations across study periods, so the comparison is not a conventional parallel-group comparison between two independent sets of participants. Formulation-specific summaries remain useful descriptions, but the analysis may depend on within-participant information and on how period and sequence features are handled. Those analytic details are not reported in the registry record.
The study also makes clear why an endpoint's statistical meaning depends on what it measures. AUC0-tlast summarizes serum insulin exposure over time. Glucose infusion measures describe pharmacodynamic response in the study setting. A result for one type of endpoint should not be substituted for evidence about another. Although the registry lists several secondary measures, numerical results for them are not reported in the information.
Finally, the safety counts show the limits of descriptive reporting. Counts of participants with events provide an overview, but without event details, attribution, and uncertainty estimates, they cannot support a detailed comparative safety assessment. Taken together, the record offers a focused example of how to interpret a formulation comparison while keeping design, endpoint, precision, and reporting limitations distinct.
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 trial-specific limitations constrain interpretation of the posted results. The primary endpoint is defined as AUC0-tlast, but the registry record does not provide detailed rules for determining return to baseline, the analysis model, or the handling of concentrations that were missing or unevaluable. The endpoint population is described, yet participant flow and the number of exclusions from each analysis are not reported in the registry record.
The record identifies a randomized, double-masked crossover design with four study periods, but it does not report the treatment sequence, period lengths, washout interval, or procedures used to address period, sequence, or carryover effects. These details are relevant to assessing how the crossover comparison was implemented and analyzed. The identity of the parties masked is also not reported.
The reported 90% confidence interval can be interpreted as an uncertainty interval for the ratio, but the registry record does not report bioequivalence margins or a full statistical analysis plan. It therefore does not allow a reader to determine whether the interval met a prespecified decision rule. The record also does not report how multiplicity was handled across the seven listed outcome measures or whether interim analyses were conducted.
Numerical results for the listed secondary outcomes are not reported in the information available in the registry record. Safety reporting includes counts but no serious-event terms, and deaths are not reported. Detailed event descriptions and attribution are also absent. The study enrolled healthy participants, so these findings do not by themselves describe results in other populations. These limitations concern the information available in the registry record and do not establish that the unreported procedures or analyses were absent from the conduct of the study.
16. Related Tutorials
17. Related Statistical Calculators
18. Sources
- ClinicalTrials.gov: A Study of 2 Different Formulations of Insulin Lispro in Healthy Participants — NCT02111083.
- Registry record last updated 2015-05-21; status last verified 2015-05.
- Results first posted 2015-05-21.
- 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.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
19. Record Summary
This completed phase 1 study enrolled 38 healthy participants and used a randomized, double-masked crossover design. Participants received Insulin Lispro A, the 200 units per milliliter test formulation, and Insulin Lispro B, the 100 units per milliliter reference formulation, across four study periods.
The primary endpoint was AUC0-tlast, measured from predose through 8 hours post dose in each period. The reported ratio of geometric least-squares means was 0.986, with a 90% confidence interval from 0.965 to 1.01. The registry record does not report the acceptance limits needed to interpret this result as a formal bioequivalence decision.
The registry lists four secondary outcomes, but their numerical results are not reported in the information available in the registry record. Serious adverse events were reported for zero participants in each formulation summary; other adverse events affected 21 participants for Insulin Lispro A and 16 for Insulin Lispro B. Event details and deaths are not reported in the registry record.