Sublingual vs IV Atropine Bioavailability Study: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Sublingual vs IV Atropine Bioavailability Study (NCT04290039), a randomized phase 1 study. The primary outcome is Area Under the Curve to From Time Zero to Infinity (AUC_∞).

2. Clinical Question

3. Trial Design

This was a randomized, open-label, phase 1 study using a three-sequence, three-period crossover design. It enrolled 15 healthy adult volunteers and is listed as completed. The record gives a start date of 2020-01-04 and a primary completion date of 2020-02-02. The sponsor was the Biomedical Advanced Research and Development Authority.

Each participant was intended to contribute observations under different atropine dosing conditions across study periods, allowing within-participant pharmacokinetic measurements. The treatments described were sublingual atropine sulfate ophthalmic solution 1% USP at 0.5 mg and 1.0 mg, and 1.0 mg atropine sulfate injection given intravenously. The registry record identifies randomization and the crossover model but does not report the actual treatment sequences, allocation procedure, washout interval, or period-specific participant flow.

The primary endpoint was AUC from time zero to infinity, summarized by treatment as a geometric mean and geometric coefficient of variation. Blood samples were collected before dosing and at listed post-dose times through 8 hours during dosing visits on Days 1, 8, and 15. The record describes noncompartmental estimation for sublingual and intravenous routes and a linear mixed model for geometric ratios of least-square means.

The AUC analysis included evaluable participants with applicable pharmacokinetic samples. The reported treatment summaries use different participant counts by treatment, and the registry record does not explain all differences between those counts or the total enrollment. Masking was listed as none. Further details about the protocol, sample-size justification, treatment adherence, and handling of incomplete periods 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.

Period 1 (Visit 1; Day 1)

Period 2 (Visit 2; Day 8)

Period 3 (Visit 3; Day 15)

Analysis population for the primary outcome, as defined in the registry: The PK analysis population includes all subjects who were randomized, received at least 1 study drug dose, and have PK samples collected for the applicable period.

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: Safety population includes all participants who were randomized and received at least 1 study drug dose.

6. Primary Endpoint

The registry lists the primary outcome as Area Under the Curve to From Time Zero to Infinity (AUC_∞), measured over Pre-dose through 8 hours post-dose at Days 1, 8 and 15.

The primary endpoint, area under the curve from time zero to infinity (AUC∞), summarizes atropine concentration in plasma over time. It combines the measured concentration-time profile into a single measure of systemic exposure, extending from dosing through the estimated remainder of the concentration profile. The reported unit is min*ng/mL. Blood samples were collected before dose and at post-dose times extending through 8 hours at the dosing visits on Days 1, 8, and 15. Pharmacokinetic parameters were estimated for each subject and period using a noncompartmental method, with the route-specific approach described for sublingual dosing and IV infusion.

Because concentration measures are commonly analyzed on a multiplicative scale, the registry summarizes AUC as a geometric mean and reports geometric coefficients of variation. It also gives geometric ratios of least-square means and associated 90% confidence intervals from a linear mixed model. A ratio compares the model-estimated exposure for one treatment with that for another; a ratio below 1 indicates a lower geometric estimate for the first treatment named in that comparison.

AUC is useful for comparing the extent of exposure across administration routes and doses. It does not by itself establish clinical effectiveness, the timing of an effect, or safety in a patient population. The endpoint is a pharmacokinetic measure in healthy volunteers, and the registry record does not report a clinical outcome as the primary endpoint.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary analysis uses treatment-level geometric means and geometric ratios estimated with a linear mixed model. These methods are suited to pharmacokinetic outcomes that are often interpreted on a multiplicative scale. A geometric mean describes the center of a set of positive measurements after analysis on a logarithmic scale and transformation back to the original scale. The geometric coefficient of variation describes relative spread on that scale; it is not the same quantity as a standard deviation in the original units.

What does a geometric ratio mean?

A ratio expresses one treatment’s estimated exposure relative to another’s. A value of 1 represents equal model-estimated geometric exposure; a value below 1 means the numerator treatment has a lower estimate than the denominator treatment. The registry reports ratios of least-square means from a mixed model. Least-square means are model-based treatment estimates, rather than simple averages calculated only from the displayed arm summaries. In a crossover design, a mixed model can account for the repeated observations contributed by participants across treatment periods. The registry does not report the full model specification, including the exact fixed and random effects or any period and sequence terms.

The 90% confidence intervals describe uncertainty in the estimated ratios under the model. In general, a confidence interval is a range generated by a statistical procedure that, under its assumptions and repeated use, has a stated coverage property. It does not mean that there is a 90% probability that the fixed true ratio lies inside this particular interval. A confidence interval also does not describe the range of individual participant responses. A narrower interval indicates greater precision for the estimate than a wider interval, but precision alone does not determine clinical importance.

The reported AUC summaries and treatment ratios should not be treated as interchangeable. A geometric mean is a treatment-specific summary. A geometric ratio directly expresses a model-based comparison. The model may account for features of the crossover data that are not represented by a comparison of the group summaries alone. The registry reports the mixed-model ratios and intervals but does not provide sufficient detail to reproduce the model independently.

Why use a crossover design?

In general, a crossover design exposes each participant to more than one treatment and can use within-participant comparisons. This can reduce the influence of stable differences between participants when estimating treatment contrasts. It also requires attention to period effects, sequence, and possible carryover from an earlier treatment period. The registry identifies three sequences and three periods, but details of the assignment, washout, and assessment of carryover are not reported in the registry record. The reader therefore cannot evaluate those design features from the registry entry alone.

Pharmacokinetic parameters were estimated by a noncompartmental method, described for extravascular administration for sublingual doses and for IV infusion. In general, noncompartmental analysis summarizes concentration-time data without requiring a specified compartment structure for the body. The endpoint extends AUC from the observed sampling period to infinity, so the estimate includes an extrapolated component beyond the final sampled time. The registry does not report the fraction of AUC extrapolated or criteria for deciding whether individual estimates were adequate.

What do the secondary recalculations represent?

The secondary continuous-outcome comparisons include unadjusted recalculations from the posted summary statistics. A mean difference is the difference between the two displayed group means, and its confidence interval describes uncertainty under the recalculation’s assumptions. These are not the sponsor’s reported analyses: the registry lists no reported analyses for the posted xerostomia outcomes. The recalculated p-values are unadjusted, so they do not account for the number of outcomes or comparisons examined. Such results can be useful for describing the data, but they should not be read as confirmatory evidence.

General power calculations estimate the sample size that would be needed to detect an assumed effect under specified design assumptions. The planning illustrations use two-sided significance level 0.05, equal allocation between two arms, and a normal approximation, with effects set as stated fractions of the observed effects. For example, the illustrations give participants per arm for 80% power and participants per arm for 90% power. They are retrospective planning illustrations, not the trial’s actual design justification and not observed power. The registry record does not report the trial’s sample-size justification.

Over-dispersion is a general concern when outcome variability exceeds what a simple statistical model expects; it can make standard errors too small and confidence intervals too narrow if the model does not account for it. That concept is not the same as the geometric coefficient of variation reported for AUC. The registry does not provide enough information to assess model diagnostics, distributional fit, or alternative analyses. In general, diagnostic checks and sensitivity analyses help show whether conclusions are robust to modeling choices, but their conduct is not reported in the registry record.

10. Analyses Reported by the Sponsor

The primary AUC comparisons are reported as the sponsor’s mixed-model analyses. The registry gives geometric ratios of least-square means and 90% confidence intervals for low-dose sublingual versus high-dose sublingual treatment, low-dose sublingual versus intravenous treatment, and high-dose sublingual versus intravenous treatment. The reported ratios are 0.579, 0.349, and 0.603, respectively. The corresponding confidence intervals are reported in the registry, and each interval lies below 1.

No p-values are reported for these primary pharmacokinetic comparisons. The registry does not report a separate recalculated primary comparison, so the sponsor’s mixed-model estimates are the available inferential results for this endpoint. The treatment-level geometric means and coefficients of variation add descriptive context but do not replace the model-based contrasts.

The record flags no discrepancy between sponsor results and a recalculation. For the secondary continuous outcomes, recalculated comparisons are unadjusted calculations from posted summary statistics, and the registry lists no sponsor-reported analyses for those outcomes. They should therefore be distinguished from the sponsor’s primary mixed-model analysis rather than treated as a verification or replication of it. The registry record does not provide the full analysis code or model specification, so independent reproduction of the sponsor’s primary estimates is not possible from the listed results alone.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 12 outcome measures and 9 analyses, including the primary AUC endpoint and five posted secondary outcomes. It does not report a multiplicity adjustment, a prespecified testing hierarchy, or a statistical plan identifying which comparisons were intended to support confirmatory conclusions. In general, testing multiple outcomes or contrasts increases the chance of at least one apparently unusual result when each test is considered separately. Adjustment procedures or a prespecified hierarchy can address that issue, but whether either was planned or used is not reported in the registry record.

The primary pharmacokinetic comparisons have 90% confidence intervals, while no p-values are reported for those comparisons. The secondary continuous comparisons are unadjusted recalculations. Their p-values should not be interpreted as adjusted evidence across all endpoints. The record does not report whether additional analyses were undertaken but not posted, or whether the posted set represents every analysis specified in the study protocol.

The described pharmacokinetic analysis population includes randomized participants who received at least one dose and had pharmacokinetic samples collected for the applicable period. The AUC summaries include 11 low-dose sublingual participants, 14 high-dose sublingual participants, and 14 intravenous participants. These counts are not the same as the registry’s total enrollment of 15. The record does not explain the reasons for the analysis counts, provide a participant flow by period, or detail missing sample patterns.

Missing-data methods are not reported in the registry record. In particular, the record does not state whether missing measurements were imputed, whether participants with incomplete periods contributed available data, or whether sensitivity analyses assessed the effect of missing observations. These details matter because pharmacokinetic estimates depend on the samples available for each participant and period. The registry also does not report whether exclusions were determined before analysis or the number of participants excluded for each reason. These omissions limit assessment of how representative the analyzed observations are of all randomized participants.

12. 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.

Most frequent non-serious adverse event terms

The posted safety counts are descriptive and use treatment-specific numbers at risk. No serious adverse events were reported in any of the three treatment groups. The number at risk was 14 for low-dose sublingual treatment, 15 for high-dose sublingual treatment, and 14 for intravenous treatment. The registry reports no deaths.

Other treatment-emergent adverse events were reported for 1 participant in the low-dose sublingual group, 1 in the high-dose sublingual group, and 4 in the intravenous group. These counts should be interpreted in the context of the small groups and the study’s short observation period. They do not, by themselves, establish a difference in risk between treatments. The registry record does not provide event terms in the listed serious-event summary and does not report inferential safety comparisons.

Absence of reported serious events in this study is not evidence that serious events cannot occur. A small phase 1 study in healthy volunteers has limited ability to characterize uncommon harms or safety in people with toxic exposure. The counts describe what was reported in this trial, not a general safety conclusion.

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

The results describe pharmacokinetic exposure after two sublingual atropine doses and an intravenous dose in healthy adult volunteers. The sponsor-reported model estimates place each sublingual-to-comparator geometric ratio below 1, with 90% confidence intervals also below 1. This provides a quantitative comparison of AUC under the study conditions. The geometric mean AUC summaries show the observed treatment-level exposure estimates, while the mixed-model ratios provide the reported between-treatment contrasts.

These findings do not establish clinical effectiveness, clinical equivalence, or interchangeability of administration routes. A pharmacokinetic endpoint is not a patient-centered outcome, and the registry record does not report outcomes in people experiencing organophosphate or carbamate toxicity. The study population was healthy adult volunteers. The results therefore do not directly answer how either administration route affects symptoms, survival, or other clinical outcomes in a poisoned population.

The estimates also do not describe every individual’s exposure. Geometric means and confidence intervals summarize groups and model uncertainty; they do not show the full range of participant responses. The reported geometric coefficients of variation provide information about relative variability, but they do not remove uncertainty arising from the limited number of analyzed participants.

Finally, the registry record does not report enough detail to assess all design and analysis choices, including washout, sequence handling, missing-data methods, or the complete model specification. Those limitations do not erase the posted estimates, but they define what can reasonably be concluded from them. The most direct interpretation is a comparison of systemic atropine exposure in the study’s healthy-volunteer crossover setting, rather than a conclusion about treatment benefit or risk in clinical use.

14. Why This Trial Matters Statistically

This trial illustrates why route comparisons in pharmacokinetics are not simple comparisons of group averages. The crossover design allows observations from the same participant under multiple treatment conditions, while the mixed-model geometric ratios provide treatment contrasts on a multiplicative scale. Reading the treatment summaries alongside the model-based ratios helps distinguish descriptive exposure levels from estimated relative exposure.

It also shows the importance of keeping endpoint type clear. AUC describes plasma exposure, not clinical response. The study may be relevant to understanding how exposure differs by route and dose, but it cannot by itself determine clinical benefit or establish that one route can replace another in a patient population. That distinction is especially important when a registry condition refers to toxic effects but the actual participants are healthy volunteers.

The analysis highlights precision and reporting. Ratios are accompanied by 90% confidence intervals, allowing readers to assess the direction and uncertainty of the estimated contrasts rather than relying on a yes-or-no significance label. At the same time, the absence of p-values, incomplete model details, and unexplained analysis counts constrain independent assessment. The reported results are interpretable at a summary level, but not fully reproducible from the registry record.

Finally, the study shows why exploratory secondary findings and sample-size illustrations should be kept in their proper roles. Unadjusted secondary comparisons can suggest patterns but do not provide multiplicity-controlled confirmation. Retrospective planning illustrations can demonstrate how sample-size needs change under different assumed effects, but they do not reveal the trial’s original design rationale or its power. Keeping these distinctions explicit helps readers use the pharmacokinetic results without making claims that the design or analysis cannot support.

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.

The registry record does not describe the actual sequence assignments, washout interval, or any assessment of carryover. These details are relevant to a crossover study because treatment-period effects and residual exposure from a prior period can affect interpretation. The record identifies the design structure but does not provide enough information to assess how those issues were handled.

The primary analysis population varies by treatment, with 11 participants in the low-dose sublingual AUC summary and 14 in each of the other summaries, despite total enrollment of 15. The registry does not explain the differences or give a participant flow by study period. Missing-data handling, exclusion criteria beyond the stated analysis-population definition, and sensitivity analyses are not reported.

The registry reports mixed-model ratios but not the full model specification or diagnostics. It does not state whether the model included sequence, period, or other terms, nor does it describe checks of assumptions. The confidence intervals can be interpreted as reported model-based uncertainty, but the information needed to independently verify those estimates is limited.

Generalizability is also constrained by the study setting: the participants were healthy adult volunteers, and the primary outcome was pharmacokinetic exposure. The study does not report clinical outcomes in people with toxic exposure. The registry lists adverse-event counts but provides limited detail about event types and no inferential safety comparison. The sample-size justification and any interim analyses are not reported in the registry record. These limitations restrict conclusions about clinical effects, uncommon safety outcomes, and how the estimates apply beyond the study conditions.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Sublingual vs IV Atropine Bioavailability Study — NCT04290039.
• Registry record last updated 2023-06-22; status last verified 2023-06.
• Results first posted 2021-08-11.
• 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.

19. Record Summary

The registry describes a completed, randomized, open-label phase 1 crossover study of sublingual and intravenous atropine in healthy adult volunteers. Fifteen participants were enrolled. The study compared sublingual atropine sulfate ophthalmic solution 1% USP at 0.5 mg and 1.0 mg with 1.0 mg intravenous atropine sulfate injection, using a three-sequence, three-period design.

The primary endpoint was AUC from time zero to infinity, summarized with geometric means and geometric coefficients of variation. The sponsor reported mixed-model geometric ratios of least-square means with 90% confidence intervals for the three treatment comparisons. The registry also lists five secondary outcomes, including three xerostomia assessments and adverse-event outcomes. Interpretation is limited by small analysis counts, incomplete reporting of missing-data and model details, and the healthy-volunteer pharmacokinetic setting.