Metformin Versus Insulin in Pregnant Women With Type 2 Diabetes: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Metformin Versus Insulin in Pregnant Women With Type 2 Diabetes (NCT00678080), a randomized na study. The primary outcome is The Number of Participants Who Achieved a Hemoglobin A1C <7%.

2. Clinical Question

The study description frames the question around blood-sugar control during pregnancy. Hemoglobin A1c reflects average blood sugar over the preceding 2 to 3 months, according to the registry record. The primary endpoint therefore records whether participants reached the stated threshold during the third trimester, rather than reporting each participant’s measured A1c value or the average value within a group.

The registry record describes a randomized, controlled pilot study. Its stated hypothesis was that metformin would achieve similar levels of blood-sugar control compared with insulin. That stated hypothesis is not itself a result: the posted endpoint counts and the absence of an estimable two-arm comparison do not establish whether the groups had similar outcomes. The number of participants with third-trimester A1c measurements was limited to 5 in the metformin arm and 12 in the insulin arm. These are the populations reported for the primary endpoint, not the full enrollment count.

3. Trial Design

This was a randomized, controlled, parallel-group trial comparing metformin with insulin during pregnancy in women with type 2 diabetes mellitus. The registry lists 25 participants as the enrollment. The record characterizes the study as a pilot and lists its status as completed. The study phase is listed as NA.

Participants were allocated to one of two drug interventions: metformin or insulin (NPH and Regular). The arm descriptions state that treatment was prescribed by participants’ health care providers. The registry lists no masking. The allocation method beyond being randomized, the randomization procedure, and any details about treatment adjustment are not reported in the registry record.

The study began in 2008-09, with primary completion in 2011-08. The primary endpoint was the number of participants who achieved hemoglobin A1c below 7% during the third trimester. The record says that third-trimester A1c levels were obtained for only 5 participants in the metformin arm and 12 in the insulin arm. The primary outcome counts are 5 and 12, respectively. Those counts describe the posted endpoint; they do not supply a model-based or otherwise estimable comparison between groups.

The record lists 8 secondary outcomes, including body mass index, hypoglycemia, failure of metformin therapy, cesarean section, macrosomia, shoulder dystocia, newborn respiratory distress syndrome, and newborns needing neonatal dextrose. Statistical analyses for these outcomes are not reported in the registry record. Further details of the protocol, including eligibility criteria, study visits, treatment adherence, and analysis populations, 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.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: During the third trimester, hemoglobin A1c levels were obtained for only 5 in the metformin arm and 12 in the insulin arm.

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.

6. Primary Endpoint

The registry lists the primary outcome as The Number of Participants Who Achieved a Hemoglobin A1C <7%, measured over during third trimester.

The primary endpoint is the number of participants who achieved a hemoglobin A1c below 7% during the third trimester. It is a binary outcome at the participant level—each participant either met the stated threshold or did not—with the reported group result expressed as a count of participants meeting it. The registry description calls hemoglobin A1c an indicator of glycemic control and says it reflects average blood sugar over the preceding 2 to 3 months. It also notes that levels of 6.5% and higher indicate diabetes.

A threshold-based endpoint can make a clinical outcome easier to interpret as the number who reached a defined target. However, it does not show how close other participants were to the threshold, how far values were from it, or how measurements changed over time. Those details are not reported in the registry record.

The posted counts are 5 in the metformin arm and 12 in the insulin arm, while the endpoint population description states that measurements were available for 5 and 12 participants, respectively. The record does not provide an estimable two-arm comparison for this binary endpoint. It also does not report a confidence interval, p-value, or adjusted analysis for the primary endpoint. Consequently, the counts are descriptive and should not be read as a quantified treatment contrast.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Randomization assigns participants to treatment groups using a chance-based process. In general, this design can reduce systematic differences between groups at the start of a study, although it does not guarantee that every characteristic will be balanced, particularly in a small study. The registry identifies this trial as randomized and parallel-group, but details of the randomization sequence, concealment, and implementation are not reported in the registry record. No masking is listed.

What does a count outcome represent?

The primary outcome counts participants who achieved hemoglobin A1c below 7% during the third trimester. Each person contributes a yes-or-no outcome under this definition, and the result is reported as a count. A comparison of two such groups would ordinarily require a defined statistical method and an estimate of the contrast, with uncertainty expressed where appropriate. The registry record does not report an estimable two-arm comparison for this endpoint. It gives counts but no confidence interval, p-value, or adjusted estimate.

Why does uncertainty matter?

In general, a confidence interval describes a range of values compatible with the data and the statistical method used. Its width reflects precision: a wider interval indicates greater uncertainty about the underlying effect. A p-value, in general, describes how compatible the observed data are with a specified null hypothesis under a particular model; it does not measure the probability that a treatment works or that the null hypothesis is true. Neither a confidence interval nor a p-value for the primary comparison is reported in the registry record. With limited endpoint populations, uncertainty is especially important, but it cannot be quantified from the posted summary counts alone.

What can be learned from the posted counts?

The counts describe the recorded primary outcomes in the participants with third-trimester measurements. They do not, by themselves, account for missing measurements, differences in follow-up, or other participant characteristics. Nor do they reveal individual A1c values or changes over time. A statistical model could sometimes account for selected design features or covariates, but the model and any adjusted results are not reported in the registry record. No model should be inferred from the count presentation.

How should planning calculations be read?

The registry record describes planning illustrations based on the observed effect, with a two-sided significance level of 0.05, equal allocation, and a normal approximation; the effect to detect is expressed as a stated fraction of the observed effect. These are hypothetical planning assumptions, not evidence about the trial’s actual design justification. The registry record does not report the trial’s own sample-size justification. Planning illustrations are not observed power and do not show the probability that this completed study detected an effect. No planning sample-size values are reported in the registry record.

In general, power is a design-stage probability tied to assumptions about an effect, variability, sample size, and analysis method. It is not a measure of how likely the observed result is to be true. Calculations based on a study’s observed effect can be particularly unstable when the study is small, because the observed effect itself may be imprecise. Thus, a planning illustration that uses an observed effect should be interpreted as a conditional scenario, not as a property established by the trial.

What is the role of analysis populations?

In general, an analysis population specifies which randomized participants contribute to a particular estimate and how participants are handled if they do not follow the assigned treatment or have missing data. The registry record reports the number with primary endpoint measurements in each arm, but does not name an analysis population or explain the handling of missing outcomes. It also does not report an intention-to-treat analysis, a per-protocol analysis, or another population definition. Therefore, the reported endpoint counts should be described as the recorded counts in the stated measurement groups, not assumed to represent a particular analysis strategy.

Over-dispersion is a general concern when data show more variability than a statistical model assumes. It is more directly relevant to some count models than to a simple participant-level binary endpoint. The registry record does not report a model for the primary outcome, so there is no basis to assess whether over-dispersion was considered or relevant to the actual analysis. More broadly, statistical methods, covariate adjustments, interim analyses, and multiplicity procedures are not reported in the registry record.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("Metformin", "Insulin")
n      <- c(5, 12)
events <- c(5, 12)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  Metformin  5  5  100.0  47.8  100.0  56.6  100.0  54.1  99.6
ARM  Insulin  12  12  100.0  73.5  100.0  75.8  100.0  75.3  99.8

11. Multiplicity, Missing Data and Analysis Populations

The registry record lists 8 secondary outcomes in addition to the primary endpoint. They cover body mass index, hypoglycemia, failure of metformin therapy, cesarean section, macrosomia, shoulder dystocia, newborn respiratory distress syndrome, and newborns needing neonatal dextrose. The record does not report statistical analyses for these outcomes. It also does not report whether the outcomes were tested, which outcomes were designated as exploratory, or whether a multiplicity adjustment was planned.

In general, multiplicity arises when several outcomes or comparisons are examined. The more questions tested, the greater the chance of observing an apparently notable result by chance if each is treated as a separate test. A prespecified hierarchy or adjustment can help define how to interpret multiple tests. No such procedure is reported in the registry record, so no claim can be made about how multiplicity was handled.

Missing primary outcome data are a central limitation of the posted record: third-trimester A1c measurements were obtained for 5 participants in the metformin arm and 12 in the insulin arm, while enrollment was 25. The record does not explain why the other participants lack reported measurements, how missing values were handled, or whether the analysis was based on all randomized participants. It does not report sensitivity analyses for missing outcomes or details of follow-up completeness.

In general, missingness can affect both precision and interpretation, especially if the likelihood of having a measurement is related to treatment or outcome. The available record does not establish whether that was the case in this trial. Consequently, the reported counts should be read as observations among participants with recorded third-trimester measurements, without assuming that they fully represent all randomized participants.

12. 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 0% reporting threshold.

The posted serious adverse event figures are 0 affected participants among 8 at risk in the metformin group and 0 among 13 at risk in the insulin group. These safety populations differ from the primary endpoint populations, which were 5 and 12 participants with third-trimester hemoglobin A1c measurements. The counts should therefore be interpreted using their stated safety denominators, not as if they came from the primary endpoint analysis groups.

The record also reports 0 other adverse events among 8 at risk in the metformin group and 0 among 13 at risk in the insulin group. These are counts in the registry entry, not estimates of comparative safety. The record does not report the observation period, detailed adverse event definitions, event collection procedures, or a statistical comparison. Death counts are not reported in the registry record.

In general, a zero count means that no event of the specified category was recorded in the stated group and observation context. It does not establish that an event could never occur, nor does it establish equivalent safety between groups. With limited safety populations and no reported uncertainty estimates, the registry counts give only a narrow description of the events recorded.

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

The posted primary outcome documents 5 participants achieving hemoglobin A1c below 7% in the metformin arm and 12 in the insulin arm. The record says measurements were obtained for 5 and 12 participants, respectively. These figures describe the recorded third-trimester endpoint in those groups. They do not provide an estimable statistical comparison or a measure of uncertainty around a difference.

Accordingly, the registry information does not establish that metformin and insulin produced similar blood-sugar control, nor does it establish that one group had better control. The study’s stated hypothesis of similar control should be understood as the question motivating the trial, not as a conclusion demonstrated by the posted count data. A result that is not statistically estimable in the record is not evidence of equivalence or of no difference.

The endpoint itself has a defined threshold, but it does not convey the distribution of measured A1c values or how outcomes varied among participants. The registry record does not report individual measurements, group averages, changes from baseline, confidence intervals, or p-values. It also does not explain the absence of measurements for other enrolled participants or describe an analysis population that would show how missing data were handled.

The safety figures record zero serious adverse events and zero other adverse events in each stated safety group. They do not establish that the two interventions have the same safety profile, particularly because the record does not report observation time, event definitions, or comparative estimates. Secondary outcomes are listed, but their statistical analyses are not reported in the registry record.

In general, results from a randomized pilot trial can contribute descriptive information and help define questions for further study. Here, the registry record supports a limited account of the primary endpoint counts and adverse event counts, but not a precise estimate of comparative effects. No medical conclusion about treatment choice follows from these posted data alone.

14. Why This Trial Matters Statistically

This trial illustrates why the denominator and the analysis population matter as much as a reported outcome count. The total enrollment is 25, while the registry says third-trimester A1c measurements were obtained for 5 participants in the metformin arm and 12 in the insulin arm. Those measurement counts define the observed primary endpoint populations; the record does not explain the remaining gap or how it was addressed analytically.

It also shows the difference between a study hypothesis and an inferential result. The study proposed that metformin would achieve similar blood-sugar control compared with insulin. The posted endpoint counts do not provide an estimable comparison, so they cannot confirm similarity or rule out a meaningful difference. Without an uncertainty estimate, readers cannot judge the precision of a comparative effect.

The example is also relevant to interpreting pilot studies. A pilot designation describes the study’s role, but it does not by itself establish that a sample size is adequate for a definitive efficacy or safety conclusion. The registry record does not report the trial’s own sample-size justification or its statistical analysis plan. Planning illustrations based on the observed effect are hypothetical and should not be mistaken for observed power or proof that the study could resolve the stated question.

Finally, a list of outcomes is not equivalent to a set of interpretable results. The record names secondary outcomes but does not report analyses for them, and it does not state how multiplicity was handled. Together, these features demonstrate why clear reporting of denominators, missing-data methods, analysis populations, uncertainty, and outcome-specific results is essential for evaluating what a clinical trial can 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 number of participants contributing to the primary endpoint is limited: measurements were obtained for 5 in the metformin arm and 12 in the insulin arm. Although overall enrollment is 25, the registry record does not explain why third-trimester measurements were unavailable for other participants. It does not describe loss to follow-up, reasons for missing measurements, or methods used to address missing data.

The registry entry reports counts for the primary outcome but no estimable two-arm comparison, confidence interval, p-value, or adjusted result. As a result, the comparative effect and its precision are not available. The record also does not report participant-level A1c values or other summaries that would show the distribution of measurements or changes over time.

Important design and analysis details are not reported in the registry record. These include eligibility criteria, baseline characteristics, the randomization procedure, treatment adherence, the planned analysis population, statistical methods, and the sample-size justification. No interim analysis plan or multiplicity procedure is reported. The record lists secondary outcomes but does not report statistical analyses for them.

The safety information is also limited. Serious and other adverse events are reported as counts in groups of 8 and 13 at risk, but event definitions, surveillance duration, and comparative estimates are not reported. Death counts are not reported in the registry record. These limitations mean the posted information is primarily descriptive and does not support a precise assessment of comparative efficacy or safety.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Metformin Versus Insulin in Pregnant Women With Type 2 Diabetes — NCT00678080.
• Registry record last updated 2018-11-28; status last verified 2018-11.
• Results first posted 2018-11-28.
• 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

NCT00678080 is a completed randomized, controlled, parallel-group pilot trial of metformin versus insulin (NPH and Regular) in pregnant women with type 2 diabetes mellitus. Enrollment is listed as 25. The study began in 2008-09, and primary completion was in 2011-08. No masking is listed.

The primary outcome was the number of participants achieving hemoglobin A1c below 7% during the third trimester. The registry reports counts of 5 for metformin and 12 for insulin, with measurements obtained for 5 and 12 participants, respectively. It does not report an estimable two-arm comparison. The record lists 8 secondary outcomes but no statistical analyses for them. Serious adverse events and other adverse events are reported as 0 in each safety group, with 8 and 13 participants at risk, respectively. Details of missing outcome data, analysis methods, and the trial’s sample-size justification are not reported in the registry record.