Efficacy & Safety of Inhaled Insulin in Type 1 Diabetes: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Efficacy & Safety of Inhaled Insulin in Type 1 Diabetes (NCT00308308), a randomized phase 3 study. The primary outcome is Compare the Mean Change From Baseline to Week 52 in HbA1c.

2. Clinical Question

The central question was whether the change in HbA1c from baseline to Week 52 differed between a regimen using inhaled Technosphere insulin with basal insulin glargine and a regimen using rapid-acting insulin aspart with basal insulin glargine. The registry identifies type 1 diabetes as the condition under study and describes the primary outcome as a mean change in HbA1c.

The trial was designed as a Phase 3, randomized, parallel-group comparison and was open-label. Its official title describes a 52-week treatment period followed by a 4-week follow-up. The registry record does not report a separate primary follow-up outcome, so the primary comparison is the change through Week 52. Secondary measures included weight, fasting plasma glucose, achievement of a specified HbA1c threshold, and hypoglycemia outcomes. Those measures add context but do not change the primary question.

3. Trial Design

NCT00308308 was a Phase 3, randomized, controlled, parallel-group clinical trial sponsored by Mannkind Corporation. The registry lists 589 participants as enrolled and describes the study as completed. It records a start in 2006-02 and primary completion in 2008-07. The trial was open-label, meaning masking was listed as none.

The two arms compared distinct prandial insulin approaches while both included basal insulin glargine. One arm received Technosphere insulin with insulin glargine; the active comparator arm received rapid-acting insulin aspart with insulin glargine. The official title describes inhaled Technosphere/insulin in the first regimen and subcutaneous basal and prandial insulin in the comparator regimen. Allocation was randomized, but the allocation ratio and the randomization procedure are not reported in the registry record.

The primary outcome was mean change in HbA1c from baseline to Week 52, expressed as a percentage measure. The stated primary analysis population was intention to treat with last observation carried forward. The arm-level primary results include 277 participants for the TI plus glargine group and 262 for the aspart plus glargine group. These analysis counts should not be read as the full enrollment, nor assumed to apply to every secondary outcome: the reported analysis populations vary across outcomes.

The registry lists seven secondary outcomes, including changes in weight and fasting plasma glucose, the number achieving a Week 52 HbA1c level less than or equal to 7.0%, and hypoglycemia incidence and event rates. Numerical secondary results are reported for weight, fasting plasma glucose, and the HbA1c threshold outcome. Numerical results for the listed hypoglycemia outcomes are not reported in the registry record. Further details of the original sample-size rationale, allocation procedures, interim monitoring, and prespecified 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: 3 week Screening Period prior to randomization - 1420 Screened / 598 Eligible of which 589 were randomized - 822 screen failures 24 Subjects randomized but never dosed

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Intention to treat (ITT) with Last Observation Carried Forward (LOCF)

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: ITT Population (589 were randomized - 24 Subjects randomized but never dosed)

6. Primary Endpoint

The registry lists the primary outcome as Compare the Mean Change From Baseline to Week 52 in HbA1c, measured over Baseline to Week 52.

The primary endpoint measures the average change in HbA1c between baseline and Week 52. HbA1c is a continuous measure, and the registry reports least-squares means for each arm, with standard errors as the measure of dispersion. A change score places the result in terms of how the measured level moved over the stated period rather than simply comparing final values. The unit is recorded as Percentage.

Comparing average change can be useful because the trial question concerns whether the groups experienced different mean shifts during follow-up. A least-squares mean is a model-based group mean; it is not necessarily the same as a raw arithmetic average. The registry identifies an intention-to-treat population and last observation carried forward for the primary endpoint, but does not give the full model specification or explain the covariates used to obtain those means.

The posted arm estimates are -0.13 for TI plus insulin glargine and -0.37 for insulin aspart plus insulin glargine. The unadjusted recalculation expresses the contrast as the aspart-plus-glargine change minus the TI-plus-glargine change, yielding a mean difference of -0.24. The sign therefore depends on which group is subtracted from which; the magnitude describes the contrast between the two reported changes. The endpoint alone does not establish why the groups differed or whether a difference is clinically important. The registry does not report a clinical-importance threshold for this outcome.

7. Results Posted in the Registry

• Insulin Aspart + Insulin Glargine: mean -0.37 versus -0.13 on control; difference -0.24 (95% CI -0.40 to -0.08), standardized effect d = -0.25; z-test p = 0.004.

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.

Change From Baseline in Weight to Week 52

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard errors.
• Result:
  • Insulin Aspart + Insulin Glargine: mean 1.40 versus -0.50 on control; difference 1.90 (95% CI 1.04 to 2.76), standardized effect d = 0.44; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Insulin Aspart + Insulin Glargine than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Change From Baseline in Fasting Plasma Glucose to Week 52

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard errors.
• Result:
  • Insulin Aspart + Insulin Glargine: mean -25.70 versus -46.50 on control; difference 20.80 (95% CI 5.77 to 35.83), standardized effect d = 0.28; z-test p = 0.007.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Insulin Aspart + Insulin Glargine than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Number of Subjects Achieving Week 52 HbA1c Levels Less Than or Equal to 7.0%

• 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:
  • Insulin Aspart + Insulin Glargine: 16.0% (35/219) versus 16.3% (33/202) on control; risk difference -0.4 percentage points (95% CI -7.4 to 6.7), risk ratio 0.98 (0.63–1.51); unadjusted two-proportion z-test p = 0.921.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Insulin Aspart + Insulin Glargine and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes span continuous measures, a threshold-based HbA1c measure, and hypoglycemia outcomes. Together they broaden the description of the trial beyond the primary mean HbA1c change, but they do not all answer the same question. The registry reports numerical analyses for change in weight, change in fasting plasma glucose, and the number achieving a Week 52 HbA1c level less than or equal to 7.0%. The corresponding analysis counts vary by outcome and are smaller than the full enrollment in some analyses.

At a broad level, the posted secondary results do not form a single uniform pattern: the continuous measures show different group contrasts, while the threshold outcome has a small unadjusted estimated difference and a confidence interval spanning zero. These observations should be considered alongside, rather than substituted for, the primary mean-change result. A threshold outcome discards information about the degree of change among participants on either side of the cutoff, whereas a continuous outcome retains the measured scale.

The registry also lists incidence and event-rate measures for total and severe hypoglycemia. Their numerical results are not reported in the registry record. As a result, the secondary record is not a complete numerical account of every listed endpoint, and no overall safety or efficacy pattern should be inferred from the listed outcome names alone.

Unadjusted secondary results carry less weight as confirmatory evidence because multiple outcomes create more opportunities for chance findings. The registry does not report whether or how multiplicity was controlled. In addition, secondary analyses may use different populations and methods, and their estimates may not have been the primary basis for the trial's conclusions. Their role is therefore supportive and descriptive unless a prespecified testing strategy is documented.

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

• Change From Baseline in Weight to Week 52
• Change From Baseline in Fasting Plasma Glucose to Week 52
• Number of Subjects Achieving Week 52 HbA1c Levels Less Than or Equal to 7.0%
• Incidence of Total Hypoglycemia
• Incidence of Severe Hypoglycemia
• Total Hypoglycemia Event Rate
• Severe Hypoglycemia Event Rate

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is continuous: each participant's HbA1c change can take a range of values, and the group result is expressed as a mean. The registry reports least-squares means and standard errors. A standard error describes the sampling uncertainty in an estimated mean; it is not the spread of individual participant values. The original standard deviations are not posted for the primary endpoint, so the recalculation uses an approximate standard deviation derived from the reported standard errors and sample sizes. That makes the recalculated standardized effect and planning illustrations approximate rather than exact reconstructions of participant-level data.

What does the mean difference represent?

A mean difference compares the average outcome in one group with the average in another. Its sign depends on the order of subtraction. For the primary unadjusted recalculation, the reported contrast is the aspart-plus-glargine mean change minus the TI-plus-glargine mean change. The resulting estimate is -0.24. Reversing the order would reverse the sign but not change the size of the contrast. The sponsor's ANCOVA estimate is positive 0.24, consistent with expressing the comparison in the opposite direction.

The sponsor's method is listed as ANCOVA, or analysis of covariance. In general, ANCOVA compares group outcomes while adjusting for one or more covariates. Adjustment can improve precision and account for baseline differences, depending on the model and its assumptions. The registry does not identify the covariates or full model specification for the sponsor's primary analysis. Accordingly, the sponsor's adjusted result and the simple recalculation from posted summaries should be understood as related but not necessarily identical statistical estimands.

How should a p-value be read?

A p-value describes how unusual data at least as incompatible with a specified null model would be if that model and the analysis assumptions were correct. It is not the probability that the null hypothesis is true, and it does not measure the size or practical importance of an effect. The primary unadjusted recalculation reports p = 0.004. The sponsor's primary ANCOVA record does not include a p-value. These are not two p-values from an identical analysis; the registry identifies the sponsor result as an ANCOVA analysis and the recalculation as an unadjusted contrast from posted summaries.

A p-value is also sensitive to sample size and the variability of measurements. A small p-value can accompany a modest estimate when precision is high, while an estimate of similar size may be uncertain in a smaller or more variable dataset. For that reason, interpretation should consider the estimated difference and confidence interval alongside the test result.

Why do the analysis methods matter?

The primary results are least-squares means from an intention-to-treat population using last observation carried forward. Intention-to-treat generally aims to analyze participants according to their randomized group, preserving the randomized comparison. The registry's stated population is ITT, but the reported arm counts for this outcome are 277 and 262, while total enrollment is 589. The available record does not explain the relation between all randomized participants and the included endpoint observations.

Last observation carried forward replaces a missing later measurement with the participant's last observed measurement. As a general statistical approach, this method assumes that carrying the earlier value forward is a reasonable representation of the missing follow-up value. That assumption may be implausible when the outcome can change over time, and the method can understate uncertainty because the imputed value is treated as though it were observed. These are general limitations of the method; the registry does not report the amount or pattern of missing primary endpoint data, or sensitivity analyses using other missing-data methods.

Over-dispersion is relevant mainly to count outcomes, where the observed variation may exceed what a basic count model assumes. If present, it can make standard errors too small and confidence intervals too narrow unless the analysis accommodates it. The registry lists hypoglycemia event-rate outcomes, but the numerical event results and analysis details needed to evaluate dispersion are not reported. This general issue cannot be assessed from the available registry information.

What can the recalculation establish?

The simple recalculation uses the published summary means, standard errors, and analysis counts. It offers a transparent, unadjusted comparison and supports an approximate confidence interval and p-value. It does not reproduce the sponsor's ANCOVA, recover participant-level distributions, adjust for covariates, or account for any analysis features not documented in the registry record. The sponsor's own analysis should be read as the reported adjusted result; the recalculation is a separate statistical reading of the posted summaries, not a replacement for the trial's analysis.

11. Confidence Intervals and Precision

The sponsor's primary ANCOVA reports a mean difference of 0.24 with a 95% confidence interval from 0.08 to 0.40. The interval is expressed in the sponsor analysis's contrast direction, opposite to the aspart-minus-TI direction used for the unadjusted recalculation. A confidence interval gives a range of values compatible with the data and the method at the stated confidence level. It does not mean there is a 95% probability that this particular interval contains the true effect; that probability interpretation is not the usual frequentist meaning.

The unadjusted recalculation gives a mean difference of -0.24. Its confidence limits are reported with more than three decimal places in the statistical output, so they are not repeated. The sponsor's interval is the registry's reported primary interval and reflects ANCOVA rather than the unadjusted summary-statistic calculation. These intervals should not be treated as identical simply because the point estimates have the same magnitude.

The sponsor interval lies on one side of zero in its contrast direction, while its width indicates that the exact size remains uncertain. Whether that range is narrow enough for a particular clinical or non-inferiority conclusion depends on a prespecified margin and a clinical-importance threshold. Neither is reported in the registry record. The primary report's non-inferiority or equivalence label therefore cannot be fully evaluated from the confidence limits alone.

Confidence intervals address sampling uncertainty under a specified model. They do not account automatically for every source of uncertainty, such as missing-data assumptions, selection of outcomes, model misspecification, or multiplicity. In this trial, open-label conduct and the unreported details of the analysis plan also limit how fully the precision and robustness of the primary estimate can be assessed.

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 figures are illustrations based on the observed effect, not the trial's original design calculation. For these continuous outcomes, each figure is the number of participants per arm, not a combined total and not a count of events. The planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each planning scenario uses the stated fraction of the observed effect as the effect to detect. The original trial's sample-size assumptions are not reported in the registry record.

The 80% power column describes the number of participants per arm that the illustration estimates would be needed for an 80% chance of detecting the specified effect under the assumptions, if that effect and model were correct. The 90% power column uses a higher target chance of detection and therefore generally calls for more participants per arm. Neither column describes what the completed trial actually achieved.

Rows that plan for a smaller fraction of the observed effect require more participants per arm. A smaller target difference is harder to distinguish from random variation, all else equal. This is why planning for 75% or 50% of the observed effect produces larger participant requirements than planning for 100% of it. For an outcome with a very small observed difference, the corresponding illustration can become extremely large; such figures are sensitive to the estimated effect and should not be read as a practical recommendation.

These values are not observed power. They are forward-looking planning illustrations conditional on the observed estimate and stated assumptions. They do not show the probability that the trial's result is true, nor do they compensate for unreported design details, missing data, multiple outcomes, or departures from the normal approximation.

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, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = -0.240; pooled standard deviation σ = 0.960.
2. n = 2 × (1.96 + 0.842)² × 0.960² / -0.240² = 251.3, rounded up to 252 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
delta <- -0.24                      # observed difference in means
sigma <- 0.9603132926872                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^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 (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(308308)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 252 (80% power), 337 (90% power)
 75% of observed effect: n per arm = 447 (80% power), 599 (90% power)
 50% of observed effect: n per arm = 1,006 (80% power), 1,346 (90% power)
power.t.test: 252.3 
Simulated power at n = 252 per arm: 0.803 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-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 sponsor reports the primary comparison using ANCOVA, with a mean difference of 0.24 and a 95% confidence interval from 0.08 to 0.40. The registry classifies the analysis as non-inferiority or equivalence but does not report its p-value or the margin against which the comparison was judged. The reported result is therefore identifiable, but the decision criterion behind the classification cannot be checked from the registry record alone.

The unadjusted recalculation from the posted arm means expresses the contrast in the reverse direction: aspart plus glargine minus TI plus glargine. It gives -0.24 and p = 0.004. Reversing the contrast changes the sign, not the underlying ordering or magnitude. The sponsor's positive estimate is consistent with TI plus glargine minus aspart plus glargine. Thus the estimates are directionally aligned once contrast order is taken into account.

The sponsor's ANCOVA is an adjusted analysis, while the recalculation is a simple unadjusted comparison based on summary statistics. They need not match exactly because the adjustment model and its covariates are not reported. The registry does not flag a discrepancy. The sponsor analysis remains the trial's reported analysis; the recalculation is best read as a check on the broad scale and direction of the posted summary results, rather than as an alternative adjusted analysis.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists one primary outcome and seven secondary outcomes. Considering several outcomes increases the chance that at least one result may appear notable by chance, even when no underlying difference exists. Statistical procedures can control this multiplicity, for example by adjusting decision thresholds or using a prespecified hierarchy of tests. The multiplicity strategy for this trial is not reported in the registry record. The unadjusted secondary p-values should therefore be read as individual comparisons rather than as multiplicity-controlled evidence.

The primary population is identified as intention to treat with last observation carried forward. That description gives important information about the analysis, but the registry does not report the number or timing of missing primary measurements, the reasons for missingness, or whether sensitivity analyses were performed. It also does not describe how missing values were handled for each secondary outcome. Missingness can affect both the estimated group difference and its uncertainty, particularly if the likelihood of a missing measurement is related to participant outcomes.

Analysis counts vary across outcomes. The primary arm counts are 277 and 262; the continuous secondary analyses and the threshold outcome use other counts. The record does not explain all differences in analysis populations or state whether each participant had an outcome measurement at the relevant time. This limits direct comparisons across endpoints and makes it difficult to determine how exclusions, unavailable observations, or other analysis choices affected the estimates.

Further details of the prespecified statistical analysis plan, interim analyses, covariate selection, and rules for handling multiple testing are not reported in the registry record. Those omissions do not invalidate the posted estimates, but they limit independent assessment of whether analyses followed a plan established before outcomes were known. The reported results should consequently be interpreted with attention to both the stated analysis population and the documentation gaps.

16. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

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: hypoglycaemia (12.5%), loss of consciousness (2.7%), diabetic ketoacidosis (1.7%), hypoglycaemic seizure (1.5%), ketoacidosis (1.0%).

The serious adverse event figures describe participants affected, not the total number of events. In the TI plus insulin glargine group, 52 of 293 participants at risk were recorded as affected by a serious adverse event. In the insulin aspart plus insulin glargine group, 47 of 272 participants at risk were recorded as affected. These are descriptive arm-level counts; the registry does not provide an adjusted comparative safety analysis or a confidence interval for the difference.

The record separately lists other adverse events affecting 265 of 293 participants in the TI plus glargine group and 257 of 272 in the aspart plus glargine group. The listed serious-event terms include hypoglycaemia, loss of consciousness, diabetic ketoacidosis, hypoglycaemic seizure, ketoacidosis, and ankle fracture, with the corresponding percentages recorded in the adverse-event results. Those terms are not necessarily mutually exclusive, and the percentages should not be interpreted as a comparative risk estimate without the relevant category definitions and analysis.

Death counts are not reported in the registry record. The listed adverse-event data do not by themselves establish causation, comparative risk, or the overall balance of safety outcomes. The registry's listed hypoglycemia secondary outcomes also lack numerical endpoint results, so the adverse-event summary should not be taken as a substitute for those analyses.

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

The results describe a randomized comparison of two insulin regimens in participants with type 1 diabetes. For the primary endpoint, the posted mean HbA1c changes differ between arms, and both the sponsor's ANCOVA and the unadjusted recalculation report an estimate away from zero when expressed in their respective contrast directions. The confidence interval and p-value summarize statistical uncertainty under their analysis methods; they do not determine clinical importance on their own.

The registry classifies the sponsor's primary analysis as non-inferiority or equivalence, but does not report the non-inferiority margin or the full analysis plan. Without that information, the classification cannot be independently evaluated from the estimate and confidence interval alone. The trial's open-label design is also relevant context, although the primary endpoint is a laboratory measure and the registry does not describe how participants or investigators' knowledge of treatment assignment affected measurement or care.

The secondary findings address other measures, but they are not a substitute for the primary outcome and include analyses with different sample counts. Some listed secondary outcomes have no numerical results in the registry record. Multiple secondary tests also raise the possibility of chance findings when results are considered individually; a multiplicity plan is not reported.

These data do not establish why the groups' outcomes differed, whether a particular difference would matter to participants, or how results would generalize beyond the enrolled study population. The record does not provide detailed eligibility criteria, missing-data sensitivity analyses, or all model specifications. Nor do the reported adverse-event counts alone establish a causal comparison of safety. The most defensible reading is descriptive: the registry reports a primary mean-change contrast and selected secondary and adverse-event summaries, with important design and analysis details not reported.

18. Why This Trial Matters Statistically

This trial illustrates why the statistical question must be tied to the endpoint and contrast. The primary result is a difference in mean change, not a direct comparison of individual responses or a count of participants crossing a threshold. Writing the contrast direction explicitly clarifies why the recalculated estimate is negative while the sponsor's estimate is positive: the two estimates subtract the groups in opposite orders.

It also shows why adjusted and unadjusted results should not be treated as interchangeable. The sponsor reports ANCOVA, whereas the recalculation uses posted summary statistics. The two approaches can address closely related comparisons while relying on different information and assumptions. The sponsor's result is the trial's reported analysis; the unadjusted estimate provides a useful but limited check on the summaries.

Precision and statistical significance are distinct. The confidence interval describes a range of estimates compatible with the adjusted analysis, while the unadjusted p-value addresses a test under a separate calculation. Neither answers whether a difference crosses a clinically meaningful threshold, particularly when no such threshold or non-inferiority margin is reported.

Finally, this record emphasizes the importance of outcome-specific populations, missing-data handling, and multiplicity. Enrollment is not the same as the analysis count for every outcome, and secondary endpoints differ in scale and completeness. Reading those details together is necessary for a measured interpretation, rather than treating a single p-value or an endpoint label as a complete account of the trial.

19. 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. The design was open-label, with masking listed as none. Knowledge of assignment can influence participant behavior, co-interventions, reporting, or decisions about follow-up, although the registry does not describe whether these pathways affected the measured outcomes.

The primary analysis used last observation carried forward, but the registry does not report the amount or pattern of missing data, reasons for missingness, or sensitivity analyses under alternative assumptions. The primary endpoint analysis counts are 277 and 262, compared with total enrollment of 589. The record does not fully explain the difference or how the primary analysis population was formed. Counts also vary for the secondary outcomes, limiting direct comparison across endpoints.

The sponsor's primary ANCOVA is labeled as non-inferiority or equivalence, but the margin, covariates, and full model details are not reported. The registry also does not report the original sample-size justification, randomization procedure, allocation ratio, interim analysis plan, or multiplicity strategy. These gaps limit assessment of whether the reported analysis was prespecified and how its assumptions affected inference.

Numerical results are absent for the listed hypoglycemia incidence and event-rate outcomes. Death counts are also not reported in the registry record. Serious adverse event counts are descriptive and do not include a comparative confidence interval or an adjusted analysis. Finally, detailed eligibility criteria and results beyond the stated follow-up outcome are not reported, limiting assessment of generalizability and longer-term outcomes.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Efficacy & Safety of Inhaled Insulin in Type 1 Diabetes — NCT00308308.
• Registry record last updated 2014-10-16; status last verified 2014-10.
• Results first posted 2014-10-16.
• 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.

23. Record Summary

NCT00308308 was a completed Phase 3 randomized, open-label, parallel-group trial sponsored by Mannkind Corporation. It enrolled 589 participants with type 1 diabetes and compared Technosphere insulin plus insulin glargine with rapid-acting insulin aspart plus insulin glargine. The official title describes a 52-week treatment period and a 4-week follow-up.

The primary endpoint was mean HbA1c change from baseline to Week 52, analyzed in an intention-to-treat population using last observation carried forward. The sponsor reported an ANCOVA mean difference of 0.24 with a 95% confidence interval from 0.08 to 0.40. An unadjusted recalculation from the posted summaries gives a mean difference of -0.24 in the opposite contrast direction and p = 0.004.

Secondary numerical analyses are reported for weight change, fasting plasma glucose change, and achievement of the stated HbA1c threshold. Numerical results for the listed hypoglycemia outcomes and death counts are not reported in the registry record. The original design assumptions, multiplicity strategy, and several analysis details are also not reported.