An Open Study to Evaluate Whether Pack Size Affects Compliance of Metformin Treatment in Subjects With Type II Diabetes: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for An Open Study to Evaluate Whether Pack Size Affects Compliance of Metformin Treatment in Subjects With Type II Diabetes (NCT01817777), a randomized phase 4 study. The primary outcome is Change From Baseline in HbA1c Values at Week 28.

2. Clinical Question

The study addresses a practical question about medication packaging and treatment compliance. Its primary endpoint, however, was change in HbA1c rather than a direct measure of compliance. Compliance was included among the secondary outcomes, along with preferences, missed doses, diabetes-management modifications, and health-care visits. Results for those secondary outcomes are not reported in the registry record.

Participants were initially observed while using metformin as usual, then randomized to one of the pack-size arms. In the interventional phase, study medication was provided free of charge. That means the study’s setup includes both a change in pack size and a change in how medication was posted. The registry does not report a separate analysis that isolates the effect of pack size from other aspects of the study setting.

The record identifies 31 enrolled participants, but the primary endpoint summaries are based on smaller reported arm counts. The registry states that early termination and reduced sample size led to descriptive analyses only, rather than testing the planned statistical hypotheses. Consequently, the posted results describe the observed HbA1c changes in the analyzed groups; they do not establish a tested comparative effect of pack size.

3. Trial Design

This was a Phase 4 randomized, open-label, parallel-group study sponsored by GlaxoSmithKline. The registry lists the condition as type 2 diabetes and the interventions as metformin small pack and metformin large pack. Local pharmacies served as study sites, with pharmacists acting as principal investigators at the sites. The record lists a start date of 2013-05 and a primary completion date of 2014-02. Its status is terminated.

The study plan began with an 8-week observational phase. During that phase, participants visited the pharmacy and purchased metformin according to their usual routines; purchase behaviour and compliance with usual metformin use were to be observed and recorded. After the observational phase, participants were randomized to one of the two pack-size arms for a 20-week interventional phase. The study medication in that phase was provided free of charge. The registry does not report additional details about the randomization procedure, such as how the allocation sequence was generated or concealed.

HbA1c was scheduled for collection at enrollment (Week 0), baseline (Week 8), and the end of the study (Week 28). Week 8 is the baseline for the primary endpoint, so the primary change compares the Week 28 result with the Week 8 result. Pharmacists also asked participants about tablet compliance, satisfaction with the received pack size, and reasons for missed doses during the interventional phase.

The registry reports enrollment of 31, but the primary endpoint summaries contain 8 participants in the small-pack arm and 9 in the large-pack arm. These are the posted counts for that endpoint; the registry does not report a complete account of participant flow from enrollment through analysis. Because the trial was terminated early and the sample size was reduced, the planned statistical hypotheses were not tested. The final analyses were limited to descriptive statistics.

Masking was listed as none. Participants and study staff therefore were not blinded to pack size, according to the registry design description. Whether this could have affected particular measurements or reports cannot be determined from the record. No further operational details about deviations from the planned procedures are 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: Enrolled participants entered an 8-week Observational Phase (OP) prior to randomization to one of the treatment arms, followed by a 20-week Interventional Phase (IP). A total of 288 participants were planned; however, the study was prematurely discontinued when 36 participants had entered the OP. Of these 36, 31 had entered the IP.

Observational Phase (8 Weeks)

Reasons for not completing this period:

Interventional Phase (20 Weeks)

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Per Protocol Population: all participants in the Intent-to-Treat Population (randomly assigned to treatment and who received \>= 1 dose \[or any portion of a dose\] of study medication and had an HbA1c test at Week 8) who did not have a protocol violation

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 Change From Baseline in HbA1c Values at Week 28, measured over Baseline (Week 8) and Week 28.

The primary endpoint was change from baseline in HbA1c at Week 28. Baseline was defined as the Week 8 measurement, after the observational phase. HbA1c was measured as a percentage using a point-of-care device: a finger prick provided a blood sample, and the device gave an immediate result. The registry also scheduled HbA1c measurements at enrollment in Week 0, but the primary change was specified from Week 8 to Week 28.

HbA1c represents the proportion of hemoglobin that has become glycated, or chemically bound to glucose. As a continuous measurement, it can be summarized using a mean and a measure of spread. The registry reports mean change and standard deviation for each arm. A mean describes the average observed change among the participants included in that arm’s endpoint summary; the standard deviation describes how variable individual changes were around that mean.

The planned comparison was the difference in mean change between small-pack and large-pack metformin, expressed as small pack minus large pack. The registry states that a two-sided 95% confidence interval for this difference was intended. It also states that early termination and reduced sample size meant the protocol’s statistical hypotheses were not tested, and final analyses were limited to descriptive statistics. A comparative confidence interval or hypothesis-test result is not reported in the registry record.

HbA1c is distinct from the study’s direct compliance measures. Although compliance was a central motivation for examining pack size, the primary endpoint measures HbA1c change, not the number of doses taken or the proportion of prescribed doses taken. The compliance measures were listed among secondary outcomes, but their results are not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does a mean change describe?

For a continuous endpoint such as HbA1c, a change score expresses how a participant’s later measurement differs from the baseline measurement. The study defined baseline as Week 8 and the later measurement as Week 28. A negative mean change therefore describes a lower average Week 28 value than the Week 8 baseline value in the analyzed arm. It does not mean every participant had a negative change. The mean summarizes the group, while individual changes may be more or less varied.

The posted arm summaries are -0.026 percentage points for small-pack metformin and -0.109 percentage points for large-pack metformin. Their standard deviations are 0.551 and 0.863, respectively. Standard deviation is a measure of dispersion: it describes how spread out individual observations tend to be around the arm mean. It is not a confidence interval and does not by itself quantify the uncertainty in a comparison between arms.

Why is the planned comparison different from the arm summaries?

The planned treatment contrast was small-pack mean change minus large-pack mean change. This is a between-arm comparison of changes, whereas each posted mean is a within-arm descriptive summary. To draw an inferential conclusion about the contrast, an analysis would need to quantify its uncertainty, typically with a confidence interval and, if planned, a hypothesis test. The registry says that the two-sided 95% confidence interval was intended, but also says that the statistical hypotheses were not tested after early termination and reduction in sample size. A comparative estimate or interval is not reported in the registry record.

A confidence interval is a statistical method for expressing uncertainty around an estimated quantity. In general, a narrower interval indicates greater precision, while a wider interval indicates less precision. An interval around a between-arm contrast would be relevant to how precisely the difference in mean changes was estimated. Since no such interval is reported, the precision of the planned comparison cannot be assessed from the registry results.

What does descriptive analysis imply?

Descriptive statistics present observed data without claiming that a planned hypothesis has been tested. Means and standard deviations can show the direction and variability of the recorded outcomes, but they do not by themselves establish whether an observed between-arm pattern is compatible with chance variation or reflects a reproducible treatment difference. The registry explicitly characterizes the final analyses as descriptive because the study ended early and its sample size was reduced.

The absence of a reported comparative test should not be replaced by a test reconstructed from the arm summaries. A simple recalculation from posted summaries, if one were displayed, would be an unadjusted calculation and would not supersede a sponsor analysis. In this record, the discrepancy flag is false, but the registry’s stated reason for not testing the hypotheses remains central: the planned statistical hypotheses were not tested, and no comparative result is reported.

How should planning illustrations be understood?

Sample-size planning calculations are prospective illustrations, not observed outcomes and not evidence that a study had a particular probability of detecting its observed result. The planning assumptions described for this analysis use a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation; the effect to detect is expressed as a stated fraction of the observed effect. Such calculations can illustrate how planning targets depend on assumptions about the effect and analysis, but they do not recover the trial’s original design justification. That justification is not reported in the registry record.

Because the planning section is based on the observed effect, it should not be described as observed power. Observed power is not a substitute for interpreting the estimated effect and its uncertainty. The registry does not report a protocol sample-size justification, and the assumptions alone do not establish what enrollment the investigators originally intended or why.

Why can over-dispersion matter?

In general, over-dispersion means that outcome variability is greater than a statistical model expects. When a model assumes less variability than is present, its uncertainty estimates may be too small and inferential conclusions may appear more precise than warranted. This concept is relevant to statistical interpretation broadly, but the primary endpoint is continuous and is reported with standard deviations. The registry does not report an over-dispersion analysis or a model-based adjustment for this endpoint.

What can be inferred from the reported precision?

The standard deviations show that the observed participant changes were variable, but they are not enough to construct a confidence interval for the treatment contrast without additional assumptions and information. The registry gives no confidence limits for either arm mean and no interval for the planned contrast. Therefore, a reader can describe the posted means and dispersion, but cannot use these summaries alone to claim a precise or statistically significant comparative effect. The registry’s explicit account—that the planned hypotheses were not tested—should guide interpretation.

10. Confidence Intervals for Each Arm

No two-arm comparison is shown for this trial, but each arm's mean can be given with a t-based 95% confidence interval computed from the posted mean, spread and number of participants.

R code and output

arm <- c("Small Pack Metformin", "Large Pack Metformin")
n   <- c(8, 9)
m   <- c(-0.026, -0.109)
sp  <- c(0.5511, 0.8631)
sd  <- sp      # posted standard deviations
for (i in seq_along(n)) {
  se <- sd[i] / sqrt(n[i]); h <- qt(0.975, n[i] - 1) * se
  cat(sprintf("CARM|%s|%d|%.3f|%.3f|%.3f|%.3f\n", arm[i], n[i], m[i], sd[i], m[i] - h, m[i] + h))
}

CARM  Small Pack Metformin  8  -0.026  0.551  -0.487  0.435
CARM  Large Pack Metformin  9  -0.109  0.863  -0.772  0.554

11. Multiplicity, Missing Data and Analysis Populations

The registry lists eight secondary outcomes in addition to the primary endpoint. They cover compliance in the interventional and observational phases, days with zero metformin pills, diabetes-management modifications, non-routine health-care visits, treatment-regimen preference, reasons for missed doses, and withdrawals for specified reasons. Results for these outcomes are not reported in the registry record. The record does not describe a multiplicity-adjustment strategy or explain how interpretation across the primary and secondary outcomes was to be controlled.

Multiplicity is a general statistical concern when several outcomes or comparisons are examined: the chance of at least one apparently unusual result can increase as more tests are considered. A prespecified strategy can help address that issue, but the registry does not report such a strategy. Since the planned primary hypotheses were not tested and secondary results are not posted, no multiplicity-adjusted or unadjusted secondary-outcome conclusions can be drawn.

The primary endpoint population is described as per protocol: participants in the stated intent-to-treat population who received at least one dose, had a Week 8 HbA1c test, and did not have a protocol violation. The registry does not report a separate intention-to-treat primary endpoint analysis. The meaning and handling of missing Week 28 values, other missing measurements, and protocol deviations beyond the population definition are not reported in the registry record.

The record reports enrollment of 31 and primary endpoint counts of 8 in the small-pack arm and 9 in the large-pack arm. It does not provide a complete accounting of missing measurements, withdrawals, exclusions, or reasons that participants were not represented in the primary summaries. The secondary outcome list includes withdrawals due to withdrawal of informed consent or loss to follow-up, but corresponding results are not reported in the registry record.

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.

Most frequent non-serious adverse event terms

The registry reports serious adverse events affecting 0 participants in the small-pack group and 0 in the large-pack group. The corresponding numbers at risk are 16 and 14. These are the reported serious-event counts and group denominators for the safety summary; they are not the same counts as those in the primary HbA1c endpoint summaries.

The record also reports other adverse events affecting 5 participants in the small-pack group and 2 in the large-pack group, with 16 and 14 participants at risk, respectively. These counts describe participants affected, not necessarily the total number of event episodes. The registry does not provide event terms for serious adverse events in the listed top serious terms, and deaths are not reported in the registry record.

Zero reported serious adverse events should be read as a description of the events recorded in this study, not as proof that serious events cannot occur or that the groups have equivalent safety. The study was terminated early and the registry reports reduced sample size. The record does not supply a comparative safety analysis or uncertainty estimates for these counts. The safety figures also do not establish a causal difference between pack sizes.

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

The results support a limited descriptive statement: the registry reports mean HbA1c changes from Week 8 to Week 28 of -0.026 percentage points for small-pack metformin and -0.109 percentage points for large-pack metformin. The reported standard deviations are 0.551 and 0.863, and the endpoint summaries include 8 and 9 participants, respectively. These numbers describe the analyzed groups; they are not a completed statistical test of pack size.

The record says that the intended comparison was the difference in mean change between the arms, with a two-sided 95% confidence interval. Because the study was terminated early and the sample size was reduced, the protocol hypotheses were not tested. The registry therefore does not report a tested comparative effect, p-value, or confidence interval for that contrast. The posted means should not be interpreted as proof that one pack size performs better or worse.

Nor do the primary results answer whether pack size changed medication compliance. Compliance was among the secondary outcomes, but those results are not reported in the registry record. The study also planned to collect information about satisfaction, missed doses, and related measures; no conclusions about those outcomes can be drawn from the HbA1c summaries alone.

The findings are specific to the reported study setup: an open-label pharmacy-based study with an observational phase followed by free study medication in the interventional phase. The record does not provide enough detail to determine how the results would generalize beyond that setting. The appropriate reading is therefore narrow: these are descriptive HbA1c observations in the reported endpoint groups, with the planned inferential comparison left untested.

14. Why This Trial Matters Statistically

This trial is statistically informative because it makes a distinction between a planned comparative question and the results that were ultimately reported. The protocol description specifies a contrast between small-pack and large-pack changes in HbA1c, but the registry says early termination and reduced sample size prevented testing the statistical hypotheses. The available arm summaries are descriptive rather than a completed answer to that comparative question.

The case also illustrates why an outcome’s summary and its uncertainty need to be considered together. Means communicate the average observed change, and standard deviations describe variability among observations. Neither substitutes for an interval around the between-arm contrast. Without the comparative confidence interval or a reported test, the precision and inferential strength of the planned comparison cannot be assessed from the registry record.

Another important distinction is between the endpoint chosen and the study’s motivating behaviour. Pack size was being evaluated in relation to compliance, yet the primary endpoint was HbA1c change. Compliance was listed as a secondary outcome, but its results are not reported. A biomarker summary cannot stand in for an unreported behavioural outcome.

Finally, the record demonstrates why enrollment, endpoint-analysis counts, and safety denominators should not be treated as interchangeable. The registry reports enrollment of 31, smaller counts for the primary endpoint summaries, and separate numbers at risk for adverse-event reporting. Those figures represent different reporting contexts. The record does not provide the participant-flow detail needed to reconcile them fully, so interpretation should preserve those distinctions rather than infer missing information.

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.

Early termination is a central limitation recorded for this study. The registry states that the sample size was reduced and the protocol’s statistical hypotheses were not tested; final analyses were limited to descriptive statistics. As a result, the primary summaries do not provide the planned comparative inference, and the registry does not report the intended confidence interval for the difference in mean change.

The endpoint summaries include 8 participants in the small-pack arm and 9 in the large-pack arm, while overall enrollment is reported as 31. The per-protocol population definition excludes participants with protocol violations and requires treatment receipt and a Week 8 HbA1c measurement. However, the record does not report a full flow of participants, missing Week 28 values, or the reasons for each exclusion or absence from the endpoint summaries.

The study was open-label and used local pharmacies as sites, with pharmacists as site investigators. The registry does not report further details about allocation concealment, adherence measurement procedures, or the handling of possible deviations from the planned schedule. The primary endpoint is HbA1c change; although compliance was a prominent study concern, the results for the listed compliance outcomes are not reported in the registry record.

The interventional medication was provided free of charge, whereas participants purchased metformin during the observational phase according to usual routines. The registry does not report an analysis separating the effect of pack size from this change in medication-supply context. It also does not report the original sample-size justification, a multiplicity strategy, or an account of missing-data methods. These omissions limit what can be concluded about the design rationale, outcome selection, and possible influence of incomplete data.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: An Open Study to Evaluate Whether Pack Size Affects Compliance of Metformin Treatment in Subjects With Type II Diabetes — NCT01817777.
• Registry record last updated 2017-01-09; status last verified 2016-11.
• Results first posted 2014-09-30.
• 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

This Phase 4 study was randomized, open-label, and parallel-group, comparing small-pack with large-pack metformin in people with type 2 diabetes established on metformin. Local pharmacies served as sites. After an 8-week observational phase, participants were assigned to a 20-week interventional phase, during which study medication was provided free of charge. HbA1c was collected at Week 0, Week 8, and Week 28.

The primary endpoint was change in HbA1c from Week 8 baseline to Week 28. The registry reports descriptive mean changes of -0.026 percentage points for small-pack metformin and -0.109 percentage points for large-pack metformin, with standard deviations of 0.551 and 0.863. The endpoint summaries include 8 and 9 participants, respectively.

The study status is terminated. The registry states that early termination and reduced sample size meant the planned statistical hypotheses were not tested, and final analyses were limited to descriptive statistics. Results for the listed secondary outcomes are not reported in the registry record.