Hydroxychloroquine Versus Pioglitazone in Combination Treatment for Type 2 Diabetes Mellitus: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Hydroxychloroquine Versus Pioglitazone in Combination Treatment for Type 2 Diabetes Mellitus (NCT02303405), a randomized phase 2 study. The primary outcome is Hemoglobin A1c.

2. Clinical Question

The registry describes a prospective, randomized, open-label comparison. Its primary outcome is hemoglobin A1c, reported as a percentage of hemoglobin, with the measurement scheduled at 4 months. The clinical question is therefore a comparison of the groups’ average measured values at that time point, rather than a comparison of how many participants crossed a specified threshold.

The record also lists secondary outcomes including fasting plasma glucose, the percentage of participants achieving HbA1c < 7.5%, weight, body mass index, and measures related to insulin resistance and inflammation. These are distinct outcomes and should not be substituted for the primary endpoint when interpreting the central comparison. The registry record does not report a two-group statistical test or estimate for the primary endpoint, so the posted group summaries alone do not answer whether the groups differed beyond the observed descriptive values.

3. Trial Design

This was a Phase 2, randomized, parallel-group trial with no masking. The registry describes it as a prospective, open-label comparison lasting 4 months. Participants were assigned to hydroxychloroquine or pioglitazone, with both treatments added in the context of inadequate control on maximally tolerated metformin plus a sulfonylurea. The posted arm descriptions specify hydroxychloroquine 400 mg by mouth once daily or pioglitazone 45 mg by mouth once daily, each for 4 months.

The registry record reports enrollment of 22 participants. The primary-outcome summaries include 15 participants in the hydroxychloroquine arm and 7 in the pioglitazone arm. These are the counts attached to the posted endpoint summaries; the record does not explain the difference in arm counts or describe participant flow, withdrawals, or the number assessed at each visit. It also does not report whether these endpoint counts represent all randomized participants.

The listed study start was 2014-11 and the primary completion was 2017-11. The record status is terminated. It does not state why the trial was terminated, how termination affected recruitment or follow-up, or whether the planned duration or procedures changed. Those reasons should not be inferred from the status alone.

The registry identifies a randomized design but does not report the randomization sequence method, allocation concealment, stratification, or any other implementation details. Because masking was none, the trial was open-label; the record does not describe procedures to reduce potential bias in outcome assessment or participant reporting. Further protocol details, including the intended sample-size justification and the statistical 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.

Overall Study

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 Hemoglobin A1c, measured over 4 months.

The primary endpoint is hemoglobin A1c, abbreviated HbA1c, measured as a percentage of hemoglobin at 4 months. It is described in the registry as an indicator of glycemic control. HbA1c reflects glucose exposure over a period of time, so it is used in clinical research to summarize longer-term blood glucose control rather than a single glucose reading. This is a general description of the measure.

The registry reports the endpoint as a mean with standard deviation. A mean describes the group’s average measured value, while a standard deviation describes how spread out the participant values are around that average. They answer different questions: the mean locates the group average, and the standard deviation describes variation among observations. Neither statistic alone quantifies the uncertainty in a comparison between groups.

At 4 months, the posted mean was 7.4% with standard deviation 0.9 in the hydroxychloroquine group, based on 15 participants. In the pioglitazone group, the posted mean was 6.3% with standard deviation 0.2, based on 7 participants. The registry record does not post a confidence interval, p-value, or estimable two-arm comparison for this continuous endpoint. Accordingly, these values describe the observed group summaries but do not establish the precision or statistical significance of the difference between groups.

A separate secondary outcome is the percentage of participants achieving HbA1c < 7.5%. That threshold-based outcome is not the same as the primary endpoint, which uses the continuous HbA1c measurement.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

For a continuous outcome such as HbA1c, the registry can report each group’s mean and standard deviation. The mean is the arithmetic average of the observed measurements in a group. The standard deviation describes the spread of individual measurements around that group mean. In this trial, the posted values are 7.4% with standard deviation 0.9 for hydroxychloroquine and 6.3% with standard deviation 0.2 for pioglitazone. The associated counts are 15 and 7, respectively.

Those summaries are not the same as a treatment-effect estimate with uncertainty. A two-group analysis would ordinarily express the contrast between groups and quantify how uncertain that estimate is, often with a confidence interval and, depending on the analysis, a p-value. A confidence interval is a range of values compatible with the data and the statistical model under repeated sampling; as a general concept, it communicates precision as well as the direction and possible size of an effect. No confidence interval or p-value for the primary comparison is reported in the registry record.

The posted results flag the continuous two-arm comparison as having values missing or not estimable. That distinction matters: the means and standard deviations remain descriptive summaries, but they do not become a formal comparison merely because both arms have results. The registry record does not state whether the problem arose from the analysis method, the available data, or another reason. It does not provide enough detail to reconstruct or verify a treatment-effect estimate.

Randomization is intended to create groups that are comparable on average with respect to factors measured and unmeasured at baseline. It does not guarantee exact balance in a particular small trial. The record identifies the study as randomized but does not report the sequence-generation procedure, allocation concealment, or baseline comparability results. Without those details, the statistical interpretation should stay focused on the posted design description and the available outcome summaries.

Masking was none, so treatment assignment was not blinded. In general, open-label designs can influence behaviors, reporting, and assessment, depending on the outcome and study procedures. HbA1c is a laboratory measure, but the registry record does not describe laboratory masking, measurement procedures, or safeguards for outcome assessment. The lack of masking is a design feature, not proof that bias occurred.

Standard deviations describe variability among participant measurements, whereas standard errors and confidence intervals describe uncertainty in an estimated mean or contrast. They are not interchangeable. A small standard deviation does not automatically imply that a group comparison is precise, especially when the sample count is limited or the groups have unequal counts. A comparison’s precision depends on the counts, the variability, the analysis model, and other design and data features. The registry record does not report a standard error or confidence interval for the between-group contrast.

Many analysis choices can be relevant to a continuous endpoint, including whether to compare final values or changes from baseline, whether to adjust for baseline HbA1c, and how to handle incomplete observations. These are general examples of possible methods, not claims about this trial. The statistical analysis plan, baseline-adjustment strategy, and outcome model are not reported in the registry record. No method should be attributed to the investigators without a record of it.

Sample-size planning is also distinct from analysis of observed results. A planning calculation uses assumptions about an effect, variability, significance level, allocation, and desired power to illustrate the number of participants that might be needed for a future or planned study. Such a calculation is not observed power and cannot repair a missing inferential comparison. The record describes planning illustrations based on fractions of an observed effect, with a two-sided significance level of 0.05, equal allocation, and a normal approximation. Those illustrations are not the trial’s own sample-size justification; that justification is not reported in the registry record. No planning sample-size values are.

For outcomes with over-dispersion, a model that assumes variance equals the mean may describe the data poorly; in general, over-dispersion means that observed variation exceeds what that model expects. This is a general statistical concept, not a reported feature of this trial’s HbA1c outcome. The primary endpoint is continuous, and the registry record does not report an over-dispersion assessment or a model for it. The relevant issue for this endpoint is that the posted continuous two-arm comparison was not estimable or its values are missing.

The record lists 11 outcome measures and 10 secondary outcomes, but reports an analysis count of 0. This cautions against interpreting the presence of an outcome label as evidence that a comparison was successfully analyzed or posted. The registry record does not report which analyses, if any, were conducted outside the registry, or how the primary and secondary outcomes were prioritized in the analysis.

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("Hydroxychloroquine", "Pioglitazone")
n   <- c(15, 7)
m   <- c(7.4, 6.3)
sp  <- c(0.9, 0.2)
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  Hydroxychloroquine  15  7.400  0.900  6.902  7.898
CARM  Pioglitazone  7  6.300  0.200  6.115  6.485

11. Multiplicity, Missing Data and Analysis Populations

The registry record lists 10 secondary outcomes in addition to the primary HbA1c endpoint. When a study examines multiple outcomes, the chance of at least one apparently unusual result can increase if each is tested separately without a plan for multiplicity. This is a general statistical concern; the record does not report a multiplicity adjustment, a testing hierarchy, or another strategy for handling multiple comparisons. The analysis count is 0, so the posted registry results do not establish that inferential tests were reported for those secondary outcomes.

The registry lists fasting plasma glucose, the percentage of participants achieving HbA1c < 7.5%, weight, body mass index, HOMA-IR, QUICKI, Hs-CRP, and leucocyte count among the secondary outcomes. The record does not provide analysis details for these outcomes in the registry record used for this explanation. Their presence in the outcome list should not be taken as evidence of a reported treatment comparison.

The number of participants with primary endpoint summaries is 15 for hydroxychloroquine and 7 for pioglitazone. The registry does not describe losses to follow-up, missing measurements, withdrawals, exclusions, or reasons for any missing data. It also does not report whether the posted arm counts correspond to all randomized participants or only participants with an available endpoint measurement. No imputation method, sensitivity analysis, or missing-data assumption is reported in the registry record.

The analysis population is not reported in the registry record. In particular, the record does not specify whether the primary analysis followed an intention-to-treat principle, used a per-protocol population, or applied another rule. The statistical analysis plan and the approach to multiple outcomes, missing data, and analysis populations are not reported. These omissions limit assessment of how the posted summaries relate to the randomized groups and how robust any intended comparison might have been.

12. Safety Results

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

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: transient ischemic attack (6.7%).

The registry reports serious adverse event data separately by treatment group. In the hydroxychloroquine group, 15 participants were at risk and 1 was affected by a serious adverse event. In the pioglitazone group, 7 participants were at risk and 0 were affected. Deaths are reported as 0 in both groups. The listed serious adverse event term is Transient Ischemic Attack, with 6.7% shown in the registry record.

These are event counts and a listed term, not a comparative safety analysis. The registry record does not provide a confidence interval, statistical test, exposure-adjusted rate, or explanation of how the term relates to the group-level serious-event count. With these small arm counts and few reported serious events, the summaries alone do not establish whether the groups differ in safety. They also do not establish that a treatment caused or prevented an event.

The record separately reports other adverse events: 0 affected participants among 15 at risk in the hydroxychloroquine group and 1 among 7 at risk in the pioglitazone group. Serious and other adverse events are separate categories in the posted results and should not be combined. The registry record does not report further detail about event timing, severity beyond the serious classification, recurrence, or relatedness to study treatment.

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

The posted primary results are group-level descriptive summaries at 4 months: mean HbA1c was 7.4% with standard deviation 0.9 among 15 participants assigned to hydroxychloroquine, and 6.3% with standard deviation 0.2 among 7 assigned to pioglitazone. The registry record does not report an estimable two-arm comparison for this continuous endpoint. Consequently, the summaries do not quantify the uncertainty around a between-group effect.

These results do not establish that one treatment is better, worse, equivalent, or noninferior to the other. No formal comparison, confidence interval, or p-value for the primary endpoint is reported. Equivalence and noninferiority require specific designs and analyses; the registry record does not report either. A reader should not infer those conclusions from the difference between the group means alone.

The trial was randomized, which is relevant to how treatment groups are compared, but randomization does not eliminate uncertainty or compensate for an unreported analysis. The record reports enrollment of 22 and endpoint counts of 15 and 7, but does not explain participant flow or identify the analysis population. The record status is terminated, but no reason for termination is reported. These design and reporting details constrain interpretation without indicating why the result is incomplete.

The outcome is a measure of glycemic control, but a posted summary at one follow-up time does not describe all aspects of treatment response, durability, or safety. The record lists additional outcomes, yet it reports an analysis count of 0. It therefore does not support broad conclusions across the listed outcomes. The serious adverse event summaries likewise describe a small number of recorded events and do not establish causal relationships or comparative safety.

The most defensible reading is narrow: the registry presents the observed HbA1c group summaries and selected adverse-event counts, while leaving the primary inferential comparison and important analysis details unreported. That reporting gap limits how precisely the relative outcomes can be characterized. It is not evidence that the treatments have the same effect, and it is not evidence of a particular effect beyond the observed summaries.

14. Why This Trial Matters Statistically

This trial illustrates a central distinction in clinical statistics: a group mean is a description, while a treatment comparison requires an estimate of the contrast and its uncertainty. The registry posts means and standard deviations for both arms, but indicates that values are missing or not estimable for the two-arm comparison of the continuous primary endpoint. A reader can describe the observed group summaries, but cannot assess the precision of a formal comparison from those numbers alone.

The trial also shows why participant counts and analysis populations matter. The primary summaries are based on 15 participants in one arm and 7 in the other, while the overall enrollment is 22. The registry does not describe participant flow or explain these counts. Without knowing how the analysis population was defined and whether outcome data were missing, it is difficult to understand how the summaries correspond to the original randomized groups.

Open-label status, small arm counts, multiple listed outcomes, and the terminated status are relevant context, but none explains the missing comparison on its own. Statistical interpretation depends on the analysis actually reported, not on assumptions about what investigators might have planned or performed. Here, the statistical analysis plan, multiplicity strategy, and missing-data approach are not reported in the registry record.

The record is therefore useful as an example of the limits of registry summaries. Numerical results can be informative without being sufficient for an inferential conclusion. Transparent reporting of effect estimates, uncertainty, participant flow, analysis populations, and analysis methods is essential for readers to distinguish observed values from evidence about comparative effects.

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.

Several limitations specific to this trial affect interpretation. The registry reports enrollment of 22 and endpoint summaries for 15 participants in the hydroxychloroquine group and 7 in the pioglitazone group, but does not explain participant flow, missing measurements, withdrawals, or exclusions. It is therefore unclear how the endpoint counts relate to the randomized population.

The primary outcome is reported descriptively, but the continuous two-arm comparison is marked as having values missing or not estimable. No confidence interval, p-value, or treatment-effect estimate is reported for the primary endpoint. The analysis method and analysis population are also not reported, so the comparison cannot be reconstructed from the registry summaries.

The registry identifies an open-label randomized design but gives no further information about sequence generation, allocation concealment, baseline balance, laboratory procedures, or outcome-assessment safeguards. The status is terminated, but the reason and its implications for recruitment, follow-up, or analysis are not reported. The statistical analysis plan, sample-size justification, missing-data approach, and multiplicity strategy are not reported in the registry record.

The safety results are sparse, with few affected participants, and do not include a comparative analysis or detailed event context. The listed secondary outcomes are not accompanied by reported analyses in the record used for this interpretation. These limitations mean that the posted results offer only a restricted description of outcomes and do not support a precise assessment of comparative effects or safety.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Hydroxychloroquine Versus Pioglitazone in Combination Treatment for Type 2 Diabetes Mellitus — NCT02303405.
• Registry record last updated 2022-10-20; status last verified 2022-09.
• Results first posted 2022-10-20.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

The registry describes a Phase 2, randomized, parallel, open-label comparison of hydroxychloroquine and pioglitazone in people with type 2 diabetes inadequately controlled on metformin plus a sulfonylurea. The planned treatment period was 4 months. The record reports enrollment of 22 and a terminated status.

The primary endpoint was mean HbA1c at 4 months. The posted summaries are 7.4% with standard deviation 0.9 for hydroxychloroquine, based on 15 participants, and 6.3% with standard deviation 0.2 for pioglitazone, based on 7 participants. The record does not report an estimable two-arm comparison for this continuous outcome.

Serious adverse events were reported for 1 participant in the hydroxychloroquine group and 0 in the pioglitazone group; deaths were 0 in both groups. The record does not report the analysis population, participant flow, reason for termination, or key statistical methods for the primary comparison.