Treating Nonalcoholic Steatohepatitis With Pioglitazone: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Treating Nonalcoholic Steatohepatitis With Pioglitazone (NCT00062764), a na phase 2 study. The primary outcome is Number of Patients With Improvement in Liver Histology.

2. Clinical Question

3. Trial Design

This was a phase 2, single-group study of pioglitazone, registered under the title “Treating Nonalcoholic Steatohepatitis With Pioglitazone.” The listed sponsor was the National Institute of Diabetes and Digestive and Kidney Diseases. Enrollment was 18. The study is marked completed, with a start date of 2003-06 and a primary completion date of 2009-02.

The registry lists allocation as NA and masking as none. With a single-group model, the design does not include a randomized or concurrent control arm. Thus, the primary outcome is a count among treated participants rather than a treatment contrast between groups. The registry record does not report a protocol-level statistical plan or explain why a single-group design was selected.

The summary describes offering retreatment to people who had completed a pioglitazone pilot and been off therapy for 48 weeks, as well as treating people whose response to metformin was unsatisfactory. It also describes repeat medical and metabolic evaluation and liver biopsy before restarting treatment. The intended eligibility description specifies moderate-to-severe NASH, with an activity score greater than or equal to 4. The record does not report the enrolled group’s baseline characteristics, how eligibility was assessed in practice, or how many participants belonged to each of the described groups.

Pioglitazone was to be restarted at 15 mg daily. The summary describes a possible increase to 30 mg daily in relation to ALT response, but the precise timing and operational details of that dose change are not clear in the registry description. The planned treatment duration was 3 years. The primary endpoint, however, is listed at 48 weeks. Whether later planned evaluations were completed and what their results were 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 Number of Patients With Improvement in Liver Histology, measured over 48 weeks.

The primary endpoint is the number of participants with improvement in liver histology at 48 weeks. The registry defines a histological response as a reduction in the NASH activity index by 3 points or more, together with improvement of at least 1 point each in steatosis, parenchymal inflammation, and hepatocellular injury. This definition requires improvement across all listed components rather than relying on change in only one component.

Histology is used in the endpoint because the study concerns changes in liver tissue. As a general statistical and clinical-trial concept, a prespecified endpoint definition makes the outcome more reproducible: it clarifies which observations count as a response and when that response is assessed. A count endpoint also communicates how many participants met the criterion, although interpretation depends on knowing the number assessed and how missing assessments were handled.

The registry lists the endpoint unit as participants and the timeframe as 48 weeks. It does not report the number who met the response definition, the number with an evaluable biopsy, or the analysis population. The registry record therefore supplies the response rule but not the observed endpoint result. No claim about the frequency of histological improvement can be drawn from the endpoint definition alone.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry record identifies the study model and endpoint, but it does not report a detailed statistical analysis plan. The explanations below describe general statistical concepts; they should not be read as methods confirmed to have been used in this study.

What does a single-group design estimate?

A single-group study can describe outcomes among participants who receive the intervention. For a binary endpoint such as meeting or not meeting a response definition, a basic summary may be the count meeting the definition and the number assessed. That description is not the same as a comparative treatment effect. A control group helps estimate what might have happened over the same period without the intervention, subject to the design’s assumptions. Without a concurrent comparator, the registry’s design cannot directly estimate a between-group difference.

For this study, the primary endpoint is recorded as the number of participants with histological improvement at 48 weeks. The registry gives the response definition but does not report the count of responders or the number assessed. It also does not report a percentage, confidence interval, or test result for this endpoint. No recalculation is possible from the registry record because the relevant endpoint summary statistics are absent.

What does the response definition contribute?

A clear outcome definition reduces ambiguity about classification. Here, response requires a reduction in the NASH activity index by 3 points or more and improvements of at least 1 point each in three specified histological components. This is a composite rule: all listed conditions must be satisfied for a participant to meet the stated definition. A reader should distinguish this operational definition from the observed outcome. The existence of a rule does not indicate how many participants satisfied it.

In general, a binary outcome discards information about the degree of change beyond the cutoff: participants on either side of the threshold are classified differently, even if their underlying measurements are close. Continuous measurements can preserve more detail, but they require their own analysis choices and interpretation. The registry also lists secondary outcomes concerning impaired glucose tolerance, insulin sensitivity, weight, and BMI. Their numerical results and statistical methods are not reported in the registry record.

How should uncertainty be understood?

In general, a confidence interval expresses uncertainty around an estimate under a specified statistical model. A narrower interval indicates greater precision under that model; a wider interval indicates more uncertainty. An interval is not a guarantee that the true value lies within its endpoints, nor does it correct for bias or limitations in design. To interpret an interval, the estimate, interval limits, analysis population, and method are needed. None of those primary-outcome estimates are reported in the registry record.

For a count or proportion, the precision depends in part on how many observations contribute and how the outcome is distributed. The number enrolled alone does not establish how many had an evaluable assessment. Missing biopsies, withdrawals, or other unavailable outcomes could affect the analysis, but these details are not reported in the registry record. No assumptions about their frequency or handling can be made.

What do the listed analysis and planning assumptions tell us?

The registry lists 1 analysis, but does not explain which outcome it applies to, whether it is a primary or secondary analysis, or what model or test was used. The entry does not establish that only one endpoint was examined or that no other analyses were performed. Statistical methods, covariates, subgroup analyses, and any interim analyses are not reported in the registry record.

The planning illustration assumptions describe a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. They also state that the effect to detect is a stated fraction of the observed effect. This is a hypothetical two-arm planning framework, whereas the registered study model is single group. The actual illustration values and the observed effect used are not reported in the registry record. Such planning values are not a retrospective measure of power and should not be interpreted as observed study findings.

As a general concept, power is a design quantity based on assumptions about an effect, variability, significance level, and sample size. Calculating power after observing results does not resolve uncertainty in the estimate or compensate for a lack of control group. A sample-size illustration is useful only within the assumptions that generated it; it cannot establish that the observed study was adequately sized for its actual design.

Why can over-dispersion matter?

For count outcomes, a statistical model may assume a particular relationship between the mean and variability. Over-dispersion means the observed variability exceeds what that model expects. If it is ignored, standard errors may be too small and interval estimates or tests may appear more precise than warranted. This is a general consideration for count data, not a reported feature of this study. The primary endpoint is described as a participant count, but the registry does not report a model or information that would allow an assessment of over-dispersion.

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("Pioglitazone")
n      <- c(18)
events <- c(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  Pioglitazone  18  12  66.7  41.0  86.7  43.7  83.7  43.4  83.7

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 5 outcome measures, including the primary histological response outcome and secondary outcomes related to impaired glucose tolerance, insulin sensitivity, weight, and BMI. It records 1 analysis, but does not identify the outcome or outcomes covered by that analysis. A prespecified multiplicity plan—such as how to interpret several outcome tests together—is not reported in the registry record. It is therefore not possible to determine whether the listed outcomes were treated as separate descriptive measures, tested inferentially, or addressed through an adjustment strategy.

Missing-data methods are also not reported in the registry record. In particular, the record does not state how participants without a 48-week biopsy or other outcome assessment were handled, whether the analysis used all enrolled participants or only those with evaluable data, or whether any imputation method was used. The number assessed for the primary outcome is not reported. Enrollment of 18 should not be taken to mean that all 18 contributed to every endpoint.

The analysis population is not reported in the registry record. There is no stated intention-to-treat, per-protocol, or other population definition. The registry also does not report exclusions after enrollment, protocol deviations, or participant flow through follow-up. Without these details, the relationship between the enrolled group and the endpoint analysis cannot be evaluated.

Interim analyses, stopping rules, and the statistical analysis plan are not reported in the registry record. These points are unknown rather than evidence that such procedures were absent. The available information does not support assumptions about the handling of multiple outcomes, missing observations, or analysis populations.

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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: worsened serum alt levels (0.0%).

The registry lists 18 participants at risk for serious adverse events and 0 participants affected. It also lists “Worsened serum ALT levels” as a top serious adverse-event term at 0.0%. These are registry-reported safety figures. The record does not report deaths; a missing death value should not be read as a confirmed zero.

The listing separately records 14 participants affected in the other adverse-event category, with 18 at risk. The registry information does not describe the events in that category in enough detail to characterize their types, timing, severity, or relation to treatment. The number affected in the other category should not be treated as a serious-event count.

As a general statistical point, observing no serious events in a study group does not prove that the underlying risk is zero. A finite study may fail to observe uncommon events, and the precision of a safety estimate depends on the number followed and the completeness of event ascertainment. No confidence interval or detailed follow-up information for safety is reported in the registry record. These figures also do not compare risk with an untreated or alternative-treatment group, because the design is single group.

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

The registry describes a phase 2 study in which 18 participants were enrolled in a single-group pioglitazone study for NASH. Its primary endpoint is clearly defined as the number of participants with histological improvement at 48 weeks, using a threshold involving the NASH activity index and specified component improvements. Yet the endpoint result itself is not reported in the registry record. The response definition cannot stand in for an observed count, and no conclusion about the number or proportion responding is possible from the record.

Even if the primary count were available, a single-group result would describe outcomes in the treated participants rather than establish what would have occurred without treatment. As a general principle, a concurrent control group supports estimation of a treatment contrast; this study has no listed comparator. The registry record therefore does not provide a randomized comparison or a like-for-like estimate of treatment effect.

The study summary describes an intended treatment duration of 3 years and a primary endpoint timeframe of 48 weeks. It gives a planned starting dose of 15 mg daily and describes a possible increase to 30 mg daily in relation to ALT response. The registry record does not provide detailed implementation information or the primary endpoint count. These details describe the stated study plan, not proof that every participant followed the same treatment course.

The safety listing reports 0 serious adverse-event participants among 18 at risk and 14 participants affected in the other adverse-event category among 18 at risk. These are descriptive figures without a control comparison, and the registry does not report a detailed event breakdown or death count. They do not establish comparative safety or characterize all possible risks.

Overall, the record identifies the intervention, design, endpoint definition, and some safety counts, but leaves important outcome and analysis information unavailable. It does not support a numerical interpretation of the primary histological result, a comparative treatment-effect estimate, or a determination of how missing endpoint data were handled.

14. Why This Trial Matters Statistically

This study illustrates why design information and outcome results must be read together. The registry states a specific histological response rule and a 48-week assessment timeframe, but does not report the number meeting that rule. Without the observed endpoint count, a reader cannot quantify the primary result. A detailed endpoint definition improves clarity about what would count as response; it does not substitute for reporting the result.

The single-group model also defines what the data can address. It permits description of outcomes among participants receiving pioglitazone, but it does not include a concurrent comparator for estimating a treatment contrast. This distinction matters because within-group change and comparative effect answer different questions. The registry does not report the primary outcome estimate or uncertainty interval, so neither magnitude nor precision can be evaluated.

The record also demonstrates the importance of documenting analysis populations and missing outcomes. Enrollment is reported as 18, but the number with an evaluable 48-week histology assessment and the handling of unavailable assessments are not reported in the registry record. Without those details, the denominator behind the primary result would be unclear even if a count were available.

Finally, the safety figures require context. The record reports no participants affected by serious adverse events among 18 at risk, while separately recording 14 affected in the other adverse-event category. These numbers describe the registry’s categories, not comparative risks. Reporting denominators, event definitions, follow-up, and uncertainty helps readers distinguish absence of observed events from evidence that an event cannot occur. These are general statistical lessons; the missing details remain unreported for this study.

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 primary limitation for interpreting the trial is that the registry record does not report the primary outcome count, despite defining histological response and its 48-week timeframe. It also does not report an estimate of uncertainty for that outcome. As a result, the frequency and precision of the primary result cannot be assessed from the record.

The single-group design has no concurrent comparator. Any observed within-group change, if reported elsewhere, would not by itself separate changes associated with the treatment from other possible explanations. The registry record does not provide a between-group comparison or a randomized treatment contrast.

Details about the analyzed population are not reported in the registry record. The number of participants with an evaluable biopsy, withdrawals, exclusions, protocol deviations, and the handling of missing observations are unknown. Although enrollment is listed as 18, that does not establish that all enrolled participants contributed to the primary endpoint or each secondary outcome.

The description of the intervention leaves operational details uncertain. It states a starting dose of 15 mg daily and refers to an increase to 30 mg daily in relation to ALT response, but the timing and implementation of the dose change are not clear in the registry summary. The actual treatment exposure and adherence are not reported in the registry record.

The statistical analysis plan, endpoint-specific analysis methods, multiplicity approach, and interim analyses are not reported in the registry record. The safety information is also limited: detailed descriptions for the other adverse events, follow-up details, and deaths are not reported. These gaps prevent a fuller assessment of outcome ascertainment, analysis choices, and safety precision.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Treating Nonalcoholic Steatohepatitis With Pioglitazone — NCT00062764.
• Registry record last updated 2012-12-19; status last verified 2012-12.
• Results first posted 2011-04-05.
• 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 completed phase 2 study enrolled 18 participants and evaluated pioglitazone in a single-group design with no masking. The registry lists allocation as NA and identifies the primary endpoint as the number of participants with histological improvement at 48 weeks. Response required a reduction in the NASH activity index by 3 points or more and improvement of at least 1 point in each of steatosis, parenchymal inflammation, and hepatocellular injury.

The registry summary describes a planned 3-year treatment course, starting at 15 mg daily with a possible increase to 30 mg daily in relation to ALT response. It lists 5 outcome measures and 1 analysis, but does not report the primary outcome count, endpoint analysis methods, or missing-data approach. Safety entries record 0 serious adverse-event participants among 18 at risk and 14 participants affected in the other adverse-event category among 18 at risk. The registry record does not report deaths.