Glargine Dosing in Hospitalized Patients With Type 2 Diabetes and Renal Insufficiency: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Glargine Dosing in Hospitalized Patients With Type 2 Diabetes and Renal Insufficiency (NCT00911625), a randomized phase 4 study. The primary outcome is Average Blood Glucose Over 6 Days.

2. Clinical Question

The study addressed insulin dosing in the context of renal insufficiency, a setting in which the study description identifies dose adjustment as an issue of interest. Its brief summary also describes a dosing-reduction approach tied to estimated GFR by MDRD, including a stated reduction for participants below the specified threshold or on dialysis. The arm descriptions, however, identify the randomized regimens by daily weight-based dose. The registry record does not provide further detail about how the dosing description was implemented for individual participants.

3. Trial Design

The registry record describes a completed phase 4, randomized, parallel-group trial with no masking. Loyola University is listed as the sponsor. Enrollment is listed as 114. The study began on 2009-01-21, and primary completion is listed as 2011-06-30. These record details describe the registered study; they do not provide a fuller account of how the trial was conducted.

Participants were assigned to one of two insulin regimens. The 0.5 units/kg group received daily insulin divided equally between glargine and glulisine. The 0.25 units/kg group received the same two insulins in equal portions of the lower daily dose. The study description presents the comparison as standard-dose versus reduced-dose insulin in hospitalized patients with type 2 diabetes and renal insufficiency.

The primary outcome was average blood glucose over 6 days. The stated primary analysis population comprises randomized participants who completed all study activities. The registry record gives group summaries for 50 participants in the 0.5 units/kg arm and 57 participants in the 0.25 units/kg arm. It does not explain the difference between those analysis counts and the enrollment figure, or report how many randomized participants did not complete all study activities.

The registry record lists one secondary outcome: the number of participants who experienced at least one blood glucose level below 70 milligrams per deciliter. It also reports analyses of the primary outcome and safety counts. Detailed eligibility criteria, site information, concealment procedures, treatment adherence, co-interventions, and the trial's original sample-size justification are not reported in the registry record. Randomization and lack of masking are recorded, but the record does not give enough detail to assess all aspects of trial conduct.

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.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The analysis population for the primary outcome comprises all participants who were randomized and completed all study activities

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: The baseline analysis population comprises all patients who were randomized and completed all study activities

6. Primary Endpoint

The registry lists the primary outcome as Average Blood Glucose Over 6 Days, measured over 6 Days.

The primary endpoint is average blood glucose over 6 days, measured in milligrams per deciliter. Participants had blood glucose measured daily for 6 days, and the average across those days was compared between the two treatment groups. The registry record identifies this as a mean of a continuous measure and describes the reported dispersion as a standard deviation.

A group mean gives a summary of the central level of the measured outcome. Comparing means is a way to express how the average measured blood glucose differed between the randomized regimens. The standard deviation describes variation among individual participant values around each group's mean; it is not the uncertainty of the estimated difference between groups.

As a general statistical point, an average across repeated daily measurements can provide a compact summary of a participant's measured levels during the stated period. It does not, by itself, describe every daily value, the pattern of change from day to day, or the frequency of unusually high or low readings. The registry record does not report those details for the primary endpoint. It also does not describe how missing daily measurements, if any, were handled.

This outcome is used to compare measured blood glucose between the two assigned regimens during the stated 6-day period. It is not the same as the listed secondary outcome, which counts participants with at least one reading below 70 milligrams per deciliter. Neither measure alone provides a complete account of all clinical outcomes or experiences during hospitalization.

7. Results Posted in the Registry

• 0.25 Units/kg: mean 174.50 versus 174.00 on control; difference 0.50 (95% CI -18.28 to 19.28), standardized effect d = 0.01; z-test p = 0.958.

8. Statistical Interpretation

9. Statistical Methods Explained

The registry reports a two-sided t-test comparing the primary-outcome means. The posted results also permit an unadjusted re-calculation from the group summaries. These are related summaries of the same comparison, but the registry analysis is the sponsor's reported inferential analysis. The re-calculation should be read as a check based on posted summary statistics, not as a replacement for the trial's own analysis.

What does a mean difference represent?

A mean difference subtracts one group's average from the other's. Its sign depends on which group is subtracted from which. In the re-calculation, the 0.25 units/kg mean is compared with the 0.5 units/kg mean, giving an estimated difference of 0.5 milligrams per deciliter. The registry reports -0.50, consistent with reversing the direction of subtraction. Reversing the order changes the sign of the estimate and the corresponding interval endpoints, but not the underlying comparison or its p-value. The two descriptions therefore point to the same near-zero observed difference, with the same uncertainty about its direction.

Why use a t-test?

A t-test is a method for comparing means while accounting for sampling variability. In general, it asks whether the observed difference is large relative to the uncertainty expected under a null model of no mean difference. The registry lists a two-sided test, which considers departures in either direction rather than testing only whether one specified group has a larger mean. Its reported p-value is .958. The record does not report additional details about the test implementation, such as assumptions checks or any adjustment for covariates.

The standard deviations reported for the groups describe spread in participant-level average blood glucose: 52.3 in the 0.5 units/kg group and 46.0 in the 0.25 units/kg group. They are not standard errors or confidence intervals. A standard error describes the sampling uncertainty of an estimate, such as a group mean or mean difference. The re-calculation lists standard errors of 7.396 and 6.093 for the two group means, respectively. These quantify uncertainty in the estimated group means rather than variation among individual measurements.

How should the confidence interval be read?

A confidence interval describes a range of values compatible with the data under the method's assumptions. In general, a 95% confidence procedure is designed to cover the true parameter in 95% of repeated applications under those assumptions. It is not a statement that there is a 95% probability that this particular fixed parameter lies within the observed interval. For the re-calculated comparison, the interval is -18.28 to 19.28 milligrams per deciliter. The registry's reported interval is -19.341 to 18.341 for its oppositely directed estimate. Both include zero and allow a range of possible differences in either direction.

The interval's width reflects limited precision in the estimate. The point estimate is close to zero, but the confidence interval does not tightly constrain the size or direction of the underlying mean difference. A result can therefore have a small observed estimate and still leave meaningful uncertainty about other values compatible with the data. Statistical significance and precision are related but distinct: a p-value summarizes evidence against a null hypothesis under a model, while an interval conveys a range of estimates consistent with the data and method.

What does over-dispersion mean in general?

In general, over-dispersion means that observed variability is greater than a statistical model assumes. When a model understates variability, standard errors can be too small and confidence intervals too narrow, making results appear more precise than they are. This concept is often discussed for count outcomes. The primary endpoint is a continuous mean, and the registry record does not report an over-dispersion analysis or identify over-dispersion as a concern. The general concept should not be taken as a finding about this trial.

The posted group summaries identify participant counts, means, and standard deviations, and the analysis population is described as randomized participants who completed all study activities. The registry does not report individual-level data, the handling of missing daily measurements, adjusted analyses, or model diagnostics. Those omissions limit what can be independently assessed from the posted summary statistics. The re-calculation is explicitly unadjusted and should not be confused with a fuller analysis of participant-level data.

10. Confidence Intervals and Precision

Precision concerns how narrowly the data estimate the difference between group means. The re-calculated mean difference is 0.5 milligrams per deciliter, with a confidence interval from -18.28 to 19.28. The interval spans zero, so the data are compatible with a small difference in either direction and with larger differences in either direction. The point estimate alone does not convey this uncertainty.

The sponsor's reported estimate is -0.50, with a 95% confidence interval from -19.341 to 18.341. The sign and interval direction reflect the opposite subtraction order. Like the re-calculated interval, the sponsor's interval includes zero. The registry reports p = .958 for its two-sided t-test, matching the re-calculated p-value of 0.958. These summaries do not indicate a precise directional difference between the groups.

In general, a confidence interval reflects both the estimate and its sampling uncertainty under the method used. A narrow interval would constrain the range of compatible differences more tightly; a wide interval leaves more uncertainty. The interval in this comparison is wide relative to the near-zero point estimate. That does not prove that the true difference is large, but it means that the posted results do not pin down the size of any difference very precisely.

Precision is not the same as clinical importance. The registry record does not report a threshold for a clinically important difference, so the interval cannot be classified against such a threshold from this record alone. Nor does an interval containing zero demonstrate equivalence. Equivalence requires a defined margin and an analysis designed to assess whether the plausible difference lies within that margin; such a margin is not reported in the registry record.

11. 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 displayed planning figures are participants per arm, not total event counts. They are illustrations based on the observed effect, not a report of the trial's own design calculation. The stated assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. For each scenario, the effect to detect is the stated fraction of the observed effect.

The 80% power column gives the planned number of participants per arm for a design that would have an 80% chance of detecting the specified effect under the planning assumptions, if that effect were the true effect. The 90% power column gives the corresponding planning number for a 90% chance. These are design probabilities under assumptions, not guarantees about any future study's result. The figures are based on the very small observed difference, so the resulting planning counts are extremely large.

The rows describe planning for 100%, 75%, and 50% of the observed effect. A smaller effect is harder to distinguish from random variation, so detecting smaller effects requires more participants under the same assumptions. The posted illustrations show this pattern: the required participants per arm rise as the target effect becomes a smaller fraction of the observed effect. Because these scenarios are anchored to the observed estimate, they should not be read as a recommended design or as a substitute for a protocol-based sample-size justification.

The trial's original sample-size assumptions and justification are not reported in the registry record. The displayed planning values are not observed power. Observed power is not a useful substitute for interpreting the estimate and its confidence interval; the interval directly conveys the uncertainty in the posted comparison.

12. 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 = 0.500; pooled standard deviation σ = 49.041.
2. n = 2 × (1.96 + 0.842)² × 49.041² / 0.500² = 151,012.6, rounded up to 151,013 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 <- 0.5                      # observed difference in means
sigma <- 49.0408197321374                     # 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(911625)
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 = 151,013 (80% power), 202,163 (90% power)
 75% of observed effect: n per arm = 268,467 (80% power), 359,401 (90% power)
 50% of observed effect: n per arm = 604,051 (80% power), 808,652 (90% power)
power.t.test: 151013.5 
Simulated power at n = 151,013 per arm: 0.802 (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.

13. Analyses Reported by the Sponsor

The registry's reported analysis is a two-sided t-test of the final values. It gives a mean difference of -0.50, a 95% confidence interval from -19.341 to 18.341, and p = .958. The re-calculation uses the posted group means, standard deviations, and analysis counts, and expresses the difference in the reverse direction: 0.5, with an interval from -18.28 to 19.28 and p = 0.958.

The opposite signs reflect the order of subtraction, rather than a different direction of evidence. Changing the comparison order reverses the sign of a mean difference and swaps the interval endpoints. Both analyses place the estimated difference close to zero, both confidence intervals include zero, and both give the same p-value. The registry record does not flag a discrepancy between the sponsor's result and the re-calculation.

The sponsor's result remains the trial's own reported analysis. The re-calculation is an unadjusted reconstruction from posted summaries and may not reproduce every detail of an analysis using participant-level data. The registry record does not report additional adjustments or further analysis details that would allow a fuller comparison of procedures. The two-sided test addresses evidence for a difference in either direction; it does not establish equivalence or show that the regimens have identical effects.

14. Multiplicity, Missing Data and Analysis Populations

The registry record lists a primary outcome and one secondary outcome. The primary outcome is average blood glucose over 6 days. The secondary outcome is the number of participants who experienced at least one blood glucose level below 70 milligrams per deciliter. The record lists two outcome measures and two analyses, but it does not report a multiplicity plan or explain how inference across outcomes was handled.

As a general statistical concept, multiplicity arises when several outcomes, time points, groups, or analyses are tested. Conducting many tests can increase the chance of observing at least one apparently notable result by chance. Methods such as prespecifying a primary comparison or adjusting for multiple tests can address particular multiplicity concerns, but the registry record does not report whether such methods were planned or used. No correction should be assumed from the information recorded.

The primary-outcome population is described as all randomized participants who completed all study activities. This is a completion-based analysis population, rather than an explicitly stated analysis of every randomized participant regardless of completion. The record does not report the number who failed to complete all activities, reasons for non-completion, or how missing outcome measurements were handled. It also does not say whether participants were analyzed according to assigned group if treatment was not received as assigned.

These omissions matter because missing outcomes and analysis-population choices can affect the interpretation of a randomized comparison. In general, excluding participants with incomplete data can compromise the protection randomization is intended to provide if completion is related to treatment or outcome. That is a general concern, not evidence that such bias occurred in this trial. The registry does not report the information needed to evaluate the relationship between completion, treatment assignment, and outcome.

The record also does not report an interim-analysis plan, stopping rules, a prespecified statistical analysis plan, or a detailed approach to missing data. Their absence from the registry record does not establish that they did not exist; it means the details are not reported in the registry record. The available results should therefore be interpreted as the posted analysis and summaries, with limited information about broader analysis planning.

15. 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.

The serious adverse event figures list 57 participants at risk in each group and zero participants affected in each group. The figures for other adverse events also list 57 participants at risk in each group and zero affected in each group. Deaths are listed as zero in both groups. The registry record lists no serious adverse event terms.

These counts describe what is recorded for the groups and should not be read as proof that no adverse events of any kind occurred. The record reports zero affected participants for the specified serious and other adverse event groupings, but it does not provide a narrative explaining how events were collected, classified, or followed up. The risk counts used in the safety record also differ from the primary-outcome analysis counts; the registry does not explain the distinction.

In general, safety counts depend on the definitions, observation period, and completeness of event reporting. The registry record does not provide additional detail about those aspects. The posted counts are therefore best read narrowly as the recorded participant counts for the listed categories, not as a comprehensive account of every possible harm or as a comparison establishing a safety advantage.

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

The posted primary-outcome results show similar observed average blood-glucose values in the two groups over 6 days: 174.0 milligrams per deciliter for the 0.5 units/kg regimen and 174.5 milligrams per deciliter for the 0.25 units/kg regimen. The estimated mean difference is close to zero, and the reported p-value is .958. The confidence intervals include zero and allow differences in either direction, so the estimate is not precise enough to establish the direction or size of a true difference.

This result does not show that the regimens are equivalent. A non-significant test is not proof of no effect, and equivalence requires a prespecified margin and an analysis that evaluates that margin. The registry record does not report an equivalence or non-inferiority design, or a threshold for a clinically important difference. The interval's inclusion of zero therefore should not be translated into a conclusion that the regimens produce identical outcomes.

The result also does not answer every question about blood glucose. An average over the stated period does not reveal the distribution of individual values, daily patterns, or all episodes of low blood glucose. The registry lists a secondary outcome concerning at least one value below 70 milligrams per deciliter, but the posted primary-outcome comparison is not a substitute for that outcome. Nor does the primary endpoint alone address other clinical outcomes or overall experience during hospitalization.

The analysis population comprises randomized participants who completed all study activities, according to the outcome description. The record does not report how missing data were handled or provide enough detail to assess the impact of completion-based analysis. The findings should be interpreted in that recorded population and within the stated short measurement period. They do not, by themselves, establish how the results would apply to other patients or settings.

The statistics describe the posted comparison; they do not give medical advice and do not determine which regimen should be used. The study's stated purpose was to compare dosing approaches in hospitalized patients with type 2 diabetes and renal insufficiency. The available primary result leaves substantial uncertainty about the size and direction of the difference in average blood glucose.

17. Why This Trial Matters Statistically

This trial provides an example of why a point estimate and a p-value are not enough to describe a result. The observed group means are close, and the p-value is high, but the confidence interval is wide enough to include differences in either direction. Reading the estimate together with its interval makes the uncertainty visible rather than treating the observed near-zero difference as proof of no difference.

It also illustrates the importance of keeping statistical questions distinct. A superiority test asks whether the data provide evidence of a difference under a null model. It does not establish equivalence when the test does not reject that null. Equivalence requires a defined margin and an appropriate analysis; neither is reported in the registry record. The distinction is important when readers are tempted to interpret a non-significant result as showing that two regimens are interchangeable.

The comparison also highlights how a mean difference depends on group order. The registry reports -0.50, while the re-calculation reports 0.5 because it uses the opposite subtraction direction. The p-value and substantive interpretation do not change simply because the contrast is written in reverse. Understanding that convention helps prevent a sign change from being mistaken for a conflict in results.

Finally, planning figures derived from a small observed effect can become very large, and they remain hypothetical design illustrations rather than evidence about the original design or observed power. The registry record does not report the trial's original sample-size justification. The trial therefore offers a useful case for separating the posted result, its precision, and hypothetical planning calculations, while recognizing the limits of the recorded analysis details.

18. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The registry record does not report detailed eligibility criteria, participant characteristics, or enough information to judge how representative the analyzed participants were. Although the intended population is described as hospitalized patients with type 2 diabetes and renal insufficiency, the record does not provide further clinical detail that would support broader conclusions about other populations or care settings.

The primary analysis population is limited to participants who were randomized and completed all study activities. The record does not explain why the primary-outcome counts differ from the enrollment figure, how many participants did not complete activities, or how missing daily blood-glucose measurements were addressed. Without those details, the potential effect of incomplete outcome data cannot be assessed from the registry record.

The primary endpoint summarizes average blood glucose over 6 days. The registry does not report daily trajectories or a fuller description of the distribution of individual values. The posted comparison therefore cannot show whether group patterns differed across days or whether particular values were concentrated in a subset of participants. The record lists a secondary low-blood-glucose outcome, but it does not provide further detail in the information described for this analysis.

The study was unmasked, according to the registry record. The record does not explain whether outcome measurement or other procedures could have been influenced by awareness of assigned treatment. It also does not report allocation-concealment procedures, adherence, or co-interventions. These are limitations of the available record and do not establish that a particular bias occurred.

Finally, the registry does not report a clinical-importance threshold, an equivalence margin, an original sample-size justification, or a detailed multiplicity and missing-data plan. Those omissions limit assessment of how the study was designed to distinguish important effects and how the posted results should be interpreted across outcomes. The confidence interval itself indicates that the estimated difference is imprecise.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Glargine Dosing in Hospitalized Patients With Type 2 Diabetes and Renal Insufficiency — NCT00911625.
• Registry record last updated 2017-04-06; status last verified 2017-03.
• Results first posted 2017-03-10.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

22. Record Summary

The registry record describes a completed phase 4 randomized, unmasked, parallel-group trial sponsored by Loyola University, with enrollment listed as 114. It compared daily insulin regimens of 0.5 units/kg and 0.25 units/kg, each split equally between glargine and glulisine, in hospitalized patients with type 2 diabetes and renal insufficiency.

The primary outcome was average blood glucose over 6 days. The posted group means were 174.0 and 174.5 milligrams per deciliter, respectively. The registry's two-sided t-test reported a mean difference of -0.50, a 95% confidence interval from -19.341 to 18.341, and p = .958. The unadjusted re-calculation expresses the comparison in the reverse direction and gives a mean difference of 0.5, with a confidence interval from -18.28 to 19.28 and p = 0.958.

The interval includes zero and leaves substantial uncertainty about the direction and size of the difference. The registry record does not report an equivalence margin, the original sample-size justification, or detailed methods for missing data and multiplicity.