A Study of LY900014 in Participants With Type 2 Diabetes Mellitus: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of LY900014 in Participants With Type 2 Diabetes Mellitus (NCT03305822), a randomized phase 1 study. The primary outcome is Pharmacokinetics (PK): Insulin Lispro Area Under the Plasma Concentration Curve Zero to 10 Hours (AUC[0-10) Following Each Treatment Arm.

2. Clinical Question

The clinical question is how the measured concentration-time exposure following LY900014 compares descriptively with exposure following Humalog in this study population. The registry also identifies a pharmacodynamic outcome: total glucose infused over the duration of a clamp for each treatment arm. That outcome concerns the effect on glucose handling during the clamp, whereas the primary outcome concerns drug concentrations in plasma.

This is a pharmacokinetic and glucodynamic study, not a trial whose posted primary endpoint is a clinical event or a long-term measure of diabetes control. The study record describes single-dose administration in two study periods. It does not report a formal statistical comparison of the primary outcome, so the posted summaries do not establish a statistically tested difference between treatments. The record does not report a clinical benefit question or a treatment recommendation.

3. Trial Design

The registry identifies this as a phase 1, randomized, double-masked, crossover study. It enrolled 38 participants with type 2 diabetes mellitus and is marked completed. The recorded study start is 2017-10-11, and the primary completion date is 2017-12-14. The sponsor is Eli Lilly and Company.

In a crossover design, participants receive treatments in different study periods, so each participant can contribute measurements under more than one treatment. Here, the record describes a single subcutaneous dose of 15 Units of LY900014 in one period and a single subcutaneous dose of 15 U of insulin lispro (Humalog) in the other. The outcome summaries list 38 participants for each treatment. These matching counts are consistent with the crossover structure, but the record does not provide participant-level sequences or explain how treatment order was assigned.

Randomization means treatment assignment follows a chance-based process, while masking is intended to limit knowledge of treatment assignment during the study. The record labels the study randomized and double masked, but further details about the randomization procedure, who was masked, and how masking was maintained are not reported in the registry record. A washout interval, if any, is not reported in the registry record; neither are period-specific details or an account of possible carryover between periods.

The study description says screening was required within 28 days before the start of the study and that each participant’s total trial duration would be between 3 and 8 weeks. These are descriptions of the planned participant experience, not evidence about the duration completed by each participant. The registry record does not report the number who completed both periods, withdrawals, or reasons for discontinuation.

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.

Period 1

Washout

Period 2

Analysis population for the primary outcome, as defined in the registry: All participants who received at least 1 dose of study drug and have measurable insulin lispro concentrations. No statistical analysis was done, per protocol.

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

Population: All participants who received SC LY900014 or SC Humalog.

6. Primary Endpoint

The registry lists the primary outcome as Pharmacokinetics (PK): Insulin Lispro Area Under the Plasma Concentration Curve Zero to 10 Hours (AUC[0-10) Following Each Treatment Arm, measured over Predose, 5 minutes (mn), 10mn, 15mn, 20mn, 25mn, 30mn, 35mn, 40mn,45 mn,50mn, 55mn, 60mn, 70mn, 90mn, 120mn, 150mn, 180mn, 240mn, 300mn, 360mn, 420mn, 480mn, 540mn, and 600 minutes(10 hours) postdose.

The primary endpoint is insulin lispro area under the plasma concentration curve from zero to 10 hours after each treatment. Its unit is hour times picomol per liter (pmol*h/L), and the registry reports a geometric mean with geometric coefficient of variation. The stated measurement window runs from predose through 600 minutes (10 hours) postdose, with concentration assessments at the listed intervening time points.

Area under the concentration-time curve summarizes exposure across a measurement interval. General explanation: it captures both the level of measured concentration and how long that concentration is present; it is not simply a single concentration reading. The zero-to-10-hour interval defines the exposure window being summarized. Pharmacokinetic outcomes are used to describe how a body processes a drug, which matches the study’s stated purpose.

The registry lists geometric means of 1380 for LY900014 and 1280 for Humalog, with geometric coefficients of variation of 22% and 24%, respectively. These are descriptive summaries. The record explicitly states that no statistical analysis was done, per protocol. It does not report a confidence interval, a formal treatment comparison, or an inferential result for this endpoint. Therefore, the posted values describe the observed summaries but do not by themselves quantify the precision of a treatment contrast.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry reports geometric means and geometric coefficients of variation for the primary pharmacokinetic outcome, and states that no statistical analysis was done, per protocol. No inferential model, treatment-contrast estimate, confidence interval, or p-value is reported. The following explanations are general statistical context, not a description of an unreported analysis.

What does the area under the curve represent?

General explanation: concentrations measured at successive times form a concentration-time profile. The area under that profile over a specified interval summarizes exposure during that interval. A larger area can reflect higher concentrations, longer-lasting concentrations, or both. Here, the defined interval is zero to 10 hours postdose, and the stated unit is hour times picomol per liter (pmol*h/L). The registry lists measurement times from predose through 600 minutes (10 hours) postdose.

The endpoint is a pharmacokinetic measure. It characterizes drug exposure, rather than directly measuring a participant’s long-term glucose control, symptoms, or a clinical event. The study also names a pharmacodynamic outcome, total amount of glucose infused over the clamp duration. Pharmacokinetics concerns drug concentration over time; pharmacodynamics concerns the measured response associated with treatment. The registry does not include an analysis of the secondary outcome in the registry record for this record.

Why report a geometric mean?

General explanation: a geometric mean is calculated on a logarithmic scale and then expressed back on the original scale. It is often used for positive measurements that have a multiplicative pattern or are skewed, where the arithmetic mean may not represent a typical value as well. The registry identifies the primary summary as a geometric mean. It does not state the reason for selecting this summary or provide a distributional assessment.

A geometric coefficient of variation describes relative dispersion on a multiplicative scale. General explanation: it expresses variability in relation to the scale of the geometric mean, rather than as an absolute distance in the outcome’s measurement units. The reported values are 22% for LY900014 and 24% for Humalog. These summaries convey observed variability, but they are not confidence intervals and do not show the uncertainty around a treatment contrast.

What does crossover design change?

General explanation: in a crossover study, a participant receives more than one treatment in separate periods. When outcomes from the same participant are compared, statistical methods can account for within-person pairing. Such methods differ from an analysis that assumes every treatment observation comes from a different, unrelated participant. The registry identifies a crossover design and reports 38 participants under each treatment, but does not specify whether a paired analysis was planned or used. It explicitly says that no statistical analysis was done for the primary outcome.

Study period and treatment order can matter in a crossover design. A change over time or a lingering effect from a prior period can complicate comparisons if not addressed. General explanation: crossover protocols commonly need to consider period and carryover effects, but the record does not report whether these were assessed, what intervals separated treatments, or what analytic approach would have been used. Those details should not be inferred from the design label alone.

How should the posted summaries be read?

The reported geometric means are 1380 for LY900014 and 1280 for Humalog. They are descriptive values, not a reported treatment-effect estimate. General explanation: an inferential comparison would ordinarily specify an estimand or contrast and provide a method for quantifying uncertainty, such as a confidence interval or a hypothesis test. Neither is reported for this endpoint. The registry’s statement that no statistical analysis was done takes precedence over any attempt to interpret the summaries as an adjusted or statistically tested comparison.

There is no reported basis for reconstructing a treatment contrast or its uncertainty from the displayed summaries. In particular, the record does not provide participant-level measurements, the within-participant correlation, or a variance estimate for the paired contrast. A recalculation from group summaries would not reproduce a crossover analysis without those details. No recalculated result is reported in the registry record for this trial.

General explanation: over-dispersion occurs when observed variability is greater than a model assumes. If a model underestimates variability, intervals can be too narrow and tests can overstate evidence. That concept is relevant to statistical modeling generally, but the record does not report a model or evidence of over-dispersion in this study. It would be inappropriate to claim that over-dispersion was present or adjusted for.

What can planning calculations tell readers?

General explanation: sample-size calculations based on an observed effect are planning illustrations. They answer a hypothetical question about how many participants per arm might be needed under specified assumptions; they do not measure the probability that a completed study has found a true effect. They are not observed power. In a crossover study, the relationship between repeated measurements also matters to design and analysis. The registry record does not report the study’s own sample-size justification, assumptions used to choose enrollment, or a formal power calculation. Any planning illustration based on the observed effect should not be confused with the trial’s original design rationale.

Overall, the reported outcome is a descriptive exposure summary within a small, single-dose crossover study, and the registry explicitly reports no statistical analysis for it. The record’s numbers help describe what was observed, while the missing inferential methods and participant-level information limit conclusions about comparative precision.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists one primary pharmacokinetic outcome and one secondary pharmacodynamic outcome, for two outcome measures in total. It reports no statistical analyses. A multiplicity strategy—how to control the chance of false-positive findings when several outcomes or comparisons are assessed—is not reported in the registry record. Because the record does not report inferential tests, it also does not show whether any multiplicity adjustment was planned or applied.

The primary outcome population is described as participants who received at least 1 dose of study drug and had measurable insulin lispro concentrations. The outcome summaries list 38 participants for each treatment. The registry does not report a participant flow, the number screened or excluded, missing outcome counts, or whether every enrolled participant contributed usable measurements in both periods.

Methods for handling missing concentrations or incomplete treatment periods are not reported in the registry record. There is no reported imputation method, sensitivity analysis, or explanation of how measurements below an assay limit would have been handled. The record’s population definition indicates that measurable concentrations were required, but it does not explain how many participants were excluded for not meeting that condition.

The record states that no statistical analysis was done, per protocol. It does not report a statistical analysis plan, any interim analysis, or details on protocol deviations. The reasons for the absence of analysis and any prespecified rules for excluding data are not reported in the registry record. These gaps restrict evaluation of how outcome availability and analysis-population selection may have affected the descriptive results.

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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: coronary artery stenosis (2.6%).

The adverse-event summaries report serious adverse events for 0 of 38 participants in the 15 U LY900014 group and 1 of 38 participants in the Humalog group. Deaths are reported for 0 participants in each group. Other adverse events are reported for 4 of 38 participants with LY900014 and 2 of 38 participants with Humalog. These are counts as recorded; they do not establish that treatment caused an event or that the treatments differ in safety.

The registry lists coronary artery stenosis at 2.6% among the top serious adverse event terms. The group attribution and event-level details for this term are not reported in the registry record. The summary does not provide event onset, severity beyond the serious classification, relatedness assessments, or participant histories. No causal interpretation should be made from the term listing alone.

General explanation: serious adverse event counts summarize events meeting a serious classification, while other adverse-event counts are a separate category in the registry summary. A small study can describe events observed during its follow-up, but these counts do not by themselves provide a complete account of comparative safety. The record does not report a detailed safety analysis or longer-term follow-up findings.

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

The posted results describe a single-dose pharmacokinetic comparison in 38 participants with type 2 diabetes who were on insulin injection therapy. The reported primary endpoint summarizes insulin lispro exposure over zero to 10 hours after treatment. The geometric means are 1380 for LY900014 and 1280 for Humalog, and the corresponding geometric coefficients of variation are 22% and 24%. The registry states that no statistical analysis was done, per protocol.

These results do mean that the registry contains descriptive exposure summaries for both treatment periods and gives a measure of observed variability for each. They also show that the study was designed as a randomized, double-masked crossover study, with both treatments given as single subcutaneous doses in separate periods. The study record identifies a secondary clamp outcome, but does not provide a result or analysis for it in the information available for this record.

The results do not establish whether a treatment difference is statistically distinguishable from chance variation. No confidence interval, p-value, or formal treatment-contrast estimate is reported. The summaries also do not establish equivalence or noninferiority: those conclusions require a prespecified framework and an analysis capable of assessing the relevant margin, neither of which is reported in the registry record.

Nor do these pharmacokinetic summaries establish a difference in long-term blood sugar control, clinical events, or patient experience. They are measurements of drug concentration exposure over a defined postdose period. The record does not report the clinical implications of the summaries, a longer-term outcome, or follow-up beyond the study description.

Interpretation is further limited by absent details about the analysis, treatment sequence, period effects, carryover, missing data, and the number of participants completing both periods. The descriptive values can be read as reported, but they should not be extended beyond the outcome, population, and study conditions recorded.

13. Why This Trial Matters Statistically

This trial is statistically informative as an example of a crossover pharmacokinetic study where the registry posts descriptive summaries but explicitly reports no statistical analysis. Recognizing that distinction is important: a numerical contrast between two summaries is not automatically a treatment-effect estimate with known precision. Without an inferential method or uncertainty interval, readers cannot judge how precisely the contrast is estimated.

The crossover design also makes the unit of comparison important. Participants contribute measurements in separate treatment periods, so observations within a participant are related. General explanation: an analysis that accounts for this pairing can use information differently from an analysis treating the periods as independent groups. The registry identifies the design but does not report a model or paired treatment estimate. That combination illustrates why design labels alone do not disclose how data were analyzed.

The endpoint is area under the concentration-time curve, a measure of pharmacokinetic exposure. Its geometric mean and geometric coefficient of variation communicate the center and relative dispersion of the reported measurements. They do not replace an estimate of comparative uncertainty. The registry’s explicit statement that no statistical analysis was done prevents readers from treating these descriptive summaries as results of a formal hypothesis test.

The record also distinguishes primary pharmacokinetic measurement from a secondary pharmacodynamic outcome involving glucose infused during a clamp. Keeping those questions separate avoids interpreting drug exposure as if it directly answered a glucose-response or clinical-outcome question. The secondary outcome is listed, but its result is not reported in the registry record for this record.

For biostatistical interpretation, the key lessons are to identify the estimand, account for the crossover structure, distinguish descriptive dispersion from uncertainty in a contrast, and avoid claims that exceed the reported analysis. Here, those analytical details are not reported in the registry record, so the conclusions remain descriptive.

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 registry provides limited detail for evaluating the primary endpoint beyond its definition, summary type, measurement window, and descriptive values. It says no statistical analysis was done, per protocol. A statistical analysis plan, treatment contrast, confidence interval, and p-value are not reported in the registry record. Consequently, comparative precision and the role of within-participant pairing cannot be assessed from the reported summaries.

Important crossover details are also absent. The treatment sequence, allocation procedure, interval between periods, and evaluation of period or carryover effects are not reported in the registry record. Although 38 participants are listed for each treatment, the registry does not report whether all enrolled participants completed both periods or provide a flow of participant exclusions and withdrawals.

The outcome population required at least 1 dose and measurable insulin lispro concentrations. The record does not report how many enrolled participants met this definition, whether any measurements were missing, or how missing or unmeasurable values were handled. Details of the assay, including its measurement limits, are not reported in the registry record.

The results concern a single dose and a defined 10-hour pharmacokinetic window. The record does not provide a reported result for the secondary clamp outcome in the registry record for this trial, or longer-term clinical outcomes. The serious adverse event summary is brief and does not give event-level context or attribution for the listed coronary artery stenosis term. These omissions limit the scope of any interpretation beyond the descriptive information recorded.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: A Study of LY900014 in Participants With Type 2 Diabetes Mellitus — NCT03305822.
• Registry record last updated 2020-05-27; status last verified 2018-01.
• Results first posted 2020-05-27.
• 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

NCT03305822 is a completed phase 1 randomized, double-masked crossover study sponsored by Eli Lilly and Company. It enrolled 38 participants with type 2 diabetes mellitus. The recorded start date is 2017-10-11, and the primary completion date is 2017-12-14.

Participants received a single subcutaneous dose of 15 Units of LY900014 in one study period and a single subcutaneous dose of 15 U of insulin lispro (Humalog) in another. The primary outcome was insulin lispro area under the plasma concentration curve from zero to 10 hours. The registry reports geometric means of 1380 for LY900014 and 1280 for Humalog, with geometric coefficients of variation of 22% and 24%. It states that no statistical analysis was done, per protocol.

The record also lists total glucose infused over the clamp duration as a secondary outcome. The secondary result and analysis are not reported in the registry record. Adverse-event summaries report 0 serious adverse events and 0 deaths with LY900014, and 1 serious adverse event and 0 deaths with Humalog.