Dutasteride Versus Placebo and Finasteride in Men With Androgenetic Alopecia: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Dutasteride Versus Placebo and Finasteride in Men With Androgenetic Alopecia (NCT01231607), a randomized phase 3 study. The primary outcome is Change From Baseline (BL) in Target Area Hair Count (HC) Within a 2.54 Centimeter (cm) (1 Inch) Diameter Circle at the Vertex at Week 24, as Assessed by Macrophotographic Technique (MT).

2. Clinical Question

The registry describes the study purpose as evaluating safety and efficacy over six months, including whether dutasteride was more effective than placebo and at least as safe and effective as finasteride. The posted primary results are changes in hair count, not a complete account of all clinical benefits or risks. The results can therefore be read as evidence about this particular measured hair-count outcome and these study groups. They do not, by themselves, resolve every comparison named in the study purpose, especially safety or any broader assessment of treatment effects.

3. Trial Design

This was a Phase 3, randomized, parallel-group study with quadruple masking. The registry lists enrollment of 917 and describes the study as completed. The listed study dates are 2010-10-28 for start and 2012-02-01 for primary completion. The sponsor was GlaxoSmithKline.

Five treatment groups appear in the primary outcome results: placebo, three dutasteride doses, and finasteride 1 mg. The arm descriptions specify once-daily oral treatment and corresponding placebo regimens. The registry's intervention list includes active drugs and placebos, while the results distinguish the five treatment groups. The posted primary outcome analysis is based on the Intent-to-Treat population, defined in the record as randomized participants regardless of whether treatment was administered. Missing primary-outcome data were handled using last observation carried forward (LOCF), according to the registry description.

The primary endpoint was change from baseline to Week 24 in target-area hair count, measured within a 2.54 cm diameter circle at the vertex. The registry reports least-squares means and standard errors for group results and general linear model comparisons. Week 12 and smaller-area hair-count outcomes, hair width, terminal hair counts, and global assessments are among the secondary outcomes listed. The registry record does not report the original sample-size justification, the full statistical analysis plan, or details of the masking procedures. It also does not report whether interim analyses were planned or conducted. Those details should not be inferred from the randomized, masked design or from the posted results alone.

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: Intent-to-Treat (ITT) Population: all randomized participants regardless of whether or not treatment was administered. Calculation was based on the last observation carried forward (LOCF) imputation method for missing data. All participants for whom data were collected at both Baseline and Week 24 were assessed.

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 (BL) in Target Area Hair Count (HC) Within a 2.54 Centimeter (cm) (1 Inch) Diameter Circle at the Vertex at Week 24, as Assessed by Macrophotographic Technique (MT), measured over Baseline and Week 24.

The primary endpoint is change from baseline in nonvellus hair count within a target circle at the vertex, or crown, at Week 24. The circle had a diameter of 2.54 cm. The registry defines nonvellus hairs as at least 30 micrometers in width and describes them as thick and noticeable. Hair was clipped before photographs were taken, and the macrophotographic technique was used to count follicles in those photographs. A tattooed cosmetic ink dot marked the target location so the same area could be identified at baseline and after baseline; the mark was refreshed if it faded. Change was defined as the Week 24 value minus the baseline value.

A localized, consistently identified area allows the same measure to be compared over time within a participant. Hair count is a continuous outcome, so group results can be expressed as mean changes and mean differences. A positive change indicates a higher count at Week 24 than at baseline; a negative change indicates a lower count. This is an explanation of the measure's statistical meaning, not a claim about the clinical importance of any particular change. The registry reports least-squares means with standard errors, and specifies an Intent-to-Treat population with LOCF imputation for missing primary data. The registry record does not report a threshold for a clinically meaningful change.

7. Results Posted in the Registry

• Dutasteride 0.02 mg: mean 17.10 versus -4.90 on control; difference 22.00 (95% CI 0.34 to 43.66), standardized effect d = 0.23; z-test p = 0.047.
• Dutasteride 0.1 mg: mean 63.00 versus -4.90 on control; difference 67.90 (95% CI 46.33 to 89.47), standardized effect d = 0.71; z-test p = <0.001.
• Dutasteride 0.5 mg: mean 89.60 versus -4.90 on control; difference 94.50 (95% CI 72.66 to 116.34), standardized effect d = 0.98; z-test p = <0.001.
• Finasteride 1 mg: mean 56.50 versus -4.90 on control; difference 61.40 (95% CI 39.21 to 83.59), standardized effect d = 0.64; z-test p = <0.001.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Change From Baseline in Target Area Hair Count Within a 1.13 cm (0.44 Inch) Diameter Circle at the Vertex at Week 24, as Assessed by MT

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard errors.
• Result:
  • Dutasteride 0.02 mg: mean 4.20 versus -0.30 on control; difference 4.50 (95% CI -0.28 to 9.28), standardized effect d = 0.21; z-test p = 0.065.
  • Dutasteride 0.1 mg: mean 12.40 versus -0.30 on control; difference 12.70 (95% CI 7.93 to 17.47), standardized effect d = 0.59; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Dutasteride 0.02 mg and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Dutasteride 0.1 mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Change From Baseline in Target Area Hair Count Within a 2.54 cm (1 Inch) Diameter Circle at the Vertex at Week 12 as Assessed by MT

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard errors.
• Result:
  • Dutasteride 0.02 mg: mean 22.90 versus -4.00 on control; difference 26.90 (95% CI 6.75 to 47.05), standardized effect d = 0.31; z-test p = 0.009.
  • Dutasteride 0.1 mg: mean 59.60 versus -4.00 on control; difference 63.60 (95% CI 43.68 to 83.52), standardized effect d = 0.73; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Dutasteride 0.02 mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Dutasteride 0.1 mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Change From Baseline in Target Area Hair Count Within a 1.13 cm (0.44 Inch) Diameter Circle at the Vertex, as Assessed by MT at Week 12

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard errors.
• Result:
  • Dutasteride 0.02 mg: mean 5.10 versus -0.40 on control; difference 5.50 (95% CI 0.98 to 10.02), standardized effect d = 0.27; z-test p = 0.017.
  • Dutasteride 0.1 mg: mean 12.80 versus -0.40 on control; difference 13.20 (95% CI 8.70 to 17.70), standardized effect d = 0.66; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Dutasteride 0.02 mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Dutasteride 0.1 mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The secondary outcomes extend the primary hair-count assessment across other time points and measurement areas. The listed outcomes include Week 12 hair-count changes, counts in a smaller vertex circle, hair width, terminal hair counts, and global assessments of improvement. The posted numerical secondary results include some Week 12 and smaller-area hair-count comparisons. Their broad pattern is consistent with the primary endpoint in that the estimated mean changes generally differ across active treatment groups and placebo, with larger estimates appearing for higher dutasteride doses in the posted hair-count summaries. This is a descriptive pattern across related measures, not a substitute for the sponsor's primary analysis.

Secondary outcomes can help assess whether the primary result is accompanied by similar patterns in related measurements and assessment times. However, they are not independent confirmations simply because they use different time points or measurement circles. Outcomes measured in the same participants may be correlated, and many statistical comparisons increase the chance of at least one apparently notable result occurring by chance. The registry lists 16 secondary outcomes posted and reports 10 analyses, but a multiplicity adjustment for secondary comparisons is not reported in the registry record. The statistical analysis plan, including any prespecified hierarchy among secondary outcomes, is also not reported.

For these reasons, unadjusted secondary findings carry less inferential weight than the prespecified primary outcome and should be read as supportive or exploratory evidence. Their p-values and intervals do not establish a comprehensive pattern of efficacy, and the posted values cannot determine whether any secondary finding was selected after examining the data. Results for hair width, terminal hair count, and global assessment should be interpreted in light of their definitions and analysis details, which are not all included in the posted numerical summaries. The secondary results provide context for the primary hair-count finding, but they do not remove uncertainty about multiplicity, missing data, or clinical importance.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Change From Baseline in Target Area Hair Count Within a 1.13 cm (0.44 Inch) Diameter Circle at the Vertex at Week 24, as Assessed by MT
• Change From Baseline in Target Area Hair Count Within a 2.54 cm (1 Inch) Diameter Circle at the Vertex at Week 12 as Assessed by MT
• Change From Baseline in Target Area Hair Count Within a 1.13 cm (0.44 Inch) Diameter Circle at the Vertex, as Assessed by MT at Week 12
• Change From Baseline in Target Area Hair Width Within a 2.54 cm (1 Inch) Diameter Circle at the Vertex at Week 12 and Week 24, as Assessed by MT
• Change From Baseline in Target Area Hair Width Within a 1.13 cm (0.44 Inch) Diameter Circle at the Vertex at Week 12 and Week 24, as Assessed by MT
• Change From Baseline in Terminal Hair Count (THC) Within a 2.54 cm (1 Inch) Diameter Circle at the Vertex at Week 12 and Week 24, as Assessed by MT
• Change From Baseline in Terminal Hair Count Within a 1.13 cm (0.44 Inch) Diameter Circle at the Vertex at Week 12 and Week 24, as Assessed by MT
• Global Assessment of Improvement From Baseline to Week 24 Assessed for Vertex and Frontal Views Separately

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is continuous: it is a numerical hair count, and the analysis concerns change in that measure between baseline and Week 24. The registry presents least-squares means for each group. In general, a least-squares mean is a model-based group estimate, which can account for terms included in the model; the registry identifies the sponsor's comparison method as a general linear model but does not report the complete model specification or its covariates. The registry also reports standard errors for the group means. A standard error describes the statistical uncertainty in an estimated mean, rather than the spread of individual participants' measurements.

What does a mean difference represent?

A mean difference compares the estimated average change in one group with the estimated average change in another. The sign depends on the order of the groups: the reported comparisons with placebo use the active treatment minus placebo interpretation, while comparisons between dutasteride and finasteride are presented in that order. A difference above zero indicates a higher estimated change for the first named treatment in that comparison; a difference below zero indicates a lower estimated change. This describes the direction of the measured outcome, not whether the difference is clinically important.

The recalculated comparisons use unadjusted calculations from posted summary statistics. They are not the sponsor's general linear model estimates. The recalculation uses approximate pooled standard deviations, derived from the reported standard errors, and therefore relies on an approximation. In general, a pooled standard deviation represents a combined estimate of outcome variability across two groups. The sponsor's model may incorporate analysis features that cannot be recovered from the posted group summaries. Since the record flags a discrepancy, the sponsor's reported analysis takes precedence.

Why do confidence intervals matter?

A confidence interval gives a range of values compatible with the data and the statistical procedure, under its assumptions. It communicates precision as well as direction: a narrower interval indicates greater precision than a wider interval. For a mean difference, an interval that includes zero is compatible with no difference under that interval procedure; an interval entirely on one side of zero is less compatible with zero. Neither pattern alone establishes clinical importance. The confidence levels differ across the sponsor's listed comparisons: 98.33% for comparisons with placebo and 99.165% for the listed dutasteride-versus-finasteride comparisons classified as non-inferiority or equivalence. The registry record does not explain the rationale for those confidence levels.

What does the p-value contribute?

A p-value describes how unusual data at least as incompatible with a specified null hypothesis would be under that hypothesis and the model assumptions. It is not the probability that the null hypothesis is true, nor a measure of the size or importance of a difference. The posted sponsor p-values belong to the sponsor's model-based analyses. Recalculated unadjusted p-values may differ because they are based on summary-statistic approximations rather than the sponsor's model. The statistical analysis plan and full model details are not reported in the registry record, so the analyses should not be reverse-engineered beyond what is posted.

Over-dispersion is a general issue when outcome variability is greater than assumed by a statistical model: standard errors can be too small if the model understates variability, making intervals too narrow and tests too optimistic. It is not reported as a concern for this analysis, and the record does not provide enough model detail to assess it. Likewise, the registry record does not report whether model diagnostics, transformations, or robust variance methods were used. These are general considerations, not claims that a problem occurred in this trial.

For missing primary endpoint data, the record specifies LOCF. In general, LOCF carries a participant's last observed value forward to a later assessment where data are missing. This creates an analyzed value but depends on assumptions about what would have happened after the last observation. The registry record does not report sensitivity analyses that test the effect of alternative missing-data assumptions. The stated Intent-to-Treat definition includes randomized participants regardless of treatment administration, but the record's primary analysis description also refers to participants with baseline and Week 24 data; the remainder of that description is truncated. The exact handling of all analysis cases beyond the stated LOCF method is therefore not fully reported.

11. Confidence Intervals and Precision

The sponsor's placebo comparisons are reported with 98.33% confidence intervals. For dutasteride 0.02 mg, the estimated difference is 22 with an interval from -4.4 to 48.4. The interval includes zero, so the posted estimate remains compatible with no difference under that interval procedure as well as with positive differences. For dutasteride 0.1 mg, the difference is 67.9 and the interval runs from 41.6 to 94.2. For dutasteride 0.5 mg, the difference is 94.4 and the interval runs from 67.8 to 121.0. Finasteride 1 mg has a reported difference of 61.4, with an interval from 34.4 to 88.4. These intervals are entirely above zero.

The comparisons with finasteride that are classified as non-inferiority or equivalence use 99.165% confidence intervals. The dutasteride 0.02 mg difference is -39.4, with an interval from -66.1 to -12.7; dutasteride 0.1 mg is 6.5, with an interval from -20.1 to 33.1; and dutasteride 0.5 mg is 33, with an interval from 6.1 to 60. The intervals indicate the range of uncertainty around the estimated differences under the sponsor's analysis. An interval that crosses zero does not establish equivalence, and a non-inferiority interpretation requires a prespecified margin; that margin is not reported in the registry record.

Precision is not the same as clinical importance. The registry provides no threshold for a meaningful hair-count difference, and the intervals do not describe the full range of individual responses. The recalculated intervals are unadjusted approximations based on summary statistics and differ from some sponsor results. With a flagged discrepancy, the sponsor's model-based results should be used as the primary inferential results.

12. 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 sample-size figures are planning illustrations expressed as participants per arm, not total events. They are based on the observed effect and the listed assumptions: a two-sided significance level of 0.05, equal allocation between the two arms, a normal approximation, and an effect to detect set at the stated fraction of the observed effect. The original trial's own design assumptions and sample-size justification are not reported in the registry record.

The 80% power column gives the estimated participants per arm for a future two-arm study to have an 80% chance of detecting the specified effect if the assumptions hold and that effect is real. The 90% power column gives the corresponding planning estimate for a 90% chance. These are prospective design quantities, not evidence that the current trial had a particular probability of detecting its findings. This is not observed power.

The rows for planning for 100%, 75%, and 50% of the observed effect illustrate how required enrollment changes when the target difference is smaller. Smaller effects need more participants to detect under otherwise unchanged assumptions. For the same outcome, the estimates also vary by comparison because the observed effect differs by treatment contrast. The calculated figures do not account for details absent from the stated assumptions, such as attrition or a more complex analysis. They should be used as rough planning illustrations, not as a substitute for a prospective sample-size calculation based on a clinically meaningful target effect and an appropriate analysis plan. The registry does not say that these planning values were used to design this trial.

13. 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 = 22.000; pooled standard deviation σ = 96.179.
2. n = 2 × (1.96 + 0.842)² × 96.179² / 22.000² = 300.0, rounded up to 301 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 <- 22                      # observed difference in means
sigma <- 96.1786664446423                     # 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(1231607)
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 = 301 (80% power), 402 (90% power)
 75% of observed effect: n per arm = 534 (80% power), 715 (90% power)
 50% of observed effect: n per arm = 1,201 (80% power), 1,607 (90% power)
power.t.test: 301 
Simulated power at n = 301 per arm: 0.798 (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.

14. Analyses Reported by the Sponsor

The sponsor's reported primary comparisons use a general linear model and list least-squares mean differences, confidence intervals, and p-values. The recalculated comparisons are simple, unadjusted checks based on posted group summary statistics. They do not reproduce the sponsor's model, and some estimates, intervals, and p-values differ. For example, the recalculated dutasteride 0.5 mg versus placebo difference is 94.5, while the sponsor reports 94.4; the recalculated interval and p-value also differ from the sponsor's reported interval and p-value.

The record flags a discrepancy, so the sponsor's own analysis takes precedence for interpretation of the trial's inferential results. The recalculations can help readers see how a basic comparison relates to the group summaries, but they should not be treated as a correction to the sponsor analysis or as an independent replication. The full model specification and statistical analysis plan are not reported in the registry record, so the source of every difference cannot be determined from the posted information. The sponsor's analyses should be quoted with their own stated confidence levels and comparison categories, rather than mixed with recalculated values.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists one primary outcome and multiple secondary outcomes, with 17 outcome measures and 10 reported analyses. The primary outcome was change in target-area hair count at Week 24. Multiple comparisons arise because results are reported across treatment groups and because additional endpoints and time points are assessed. In general, conducting many tests can increase the chance of obtaining one small p-value even when there is no underlying effect for the comparisons being tested. Methods such as an adjustment procedure or a prespecified testing hierarchy can address this issue, but whether such methods were used is not reported in the registry record.

The sponsor reports different confidence levels for comparison categories: 98.33% for the listed placebo comparisons and 99.165% for comparisons classified as non-inferiority or equivalence. These levels may be relevant to the analysis strategy, but the registry record does not state the reasoning behind them or how they relate to multiplicity control. The meaning of the non-inferiority or equivalence category also depends on a prespecified margin. That margin and the detailed decision criteria are not reported in the registry record, so a confidence interval crossing zero cannot be taken as proof of equivalence or non-inferiority.

For the primary outcome, the record identifies the Intent-to-Treat population as all randomized participants regardless of whether treatment was administered and states that LOCF was used for missing data. In general, LOCF assumes the last recorded measurement can stand in for a later unobserved measurement. If the outcome would have changed after the last observation, that substitution can affect group estimates and uncertainty. The registry record does not report the amount or pattern of missing primary outcome data, sensitivity analyses, or alternative imputation methods. The primary population description includes a truncated statement about participants with baseline and Week 24 data, so the full operational analysis details are not available.

The registry record does not report a complete statistical analysis plan, the trial's original sample-size justification, or whether interim analyses took place. These omissions limit assessment of which analyses were prespecified, how multiplicity was handled, and how missing data influenced the reported results. They do not prove that such plans or procedures were absent; they mean that those details are not reported in the registry record.

16. 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: fractured sacrum (0.6%), laryngitis (0.6%), lower limb fracture (0.6%), pharyngeal abscess (0.6%), nephrolithiasis (0.6%).

The registry's safety summary lists serious adverse events and other adverse events separately by treatment group. Serious adverse events affected 0 participants in the dutasteride 0.02 mg group, 3 in the dutasteride 0.1 mg group, 1 in the dutasteride 0.5 mg group, 2 in the finasteride 1 mg group, and 2 in the placebo group. The corresponding numbers at risk are listed as 185, 188, 184, 179, and 181. The registry lists other adverse events affecting 34, 34, 41, 32, and 36 participants in those same groups, respectively.

The serious-event terms listed at 0.6% each are fractured sacrum, laryngitis, lower limb fracture, pharyngeal abscess, nephrolithiasis, and syncope. These entries describe the events recorded, not a comparative safety conclusion. Event counts and the registry's at-risk figures should be read together, and small counts do not provide precise estimates of between-group differences. The registry record does not report death counts in the posted safety summary. It also does not provide enough detail to assess event timing, severity beyond the serious classification, relatedness to treatment, or discontinuations due to adverse events. Safety should therefore be interpreted separately from the hair-count efficacy outcome.

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

The results describe average changes in a specific measured outcome: nonvellus hair count in a marked vertex area from baseline to Week 24. The sponsor's analysis reports larger estimated changes for the higher dutasteride doses than for placebo, while the lowest dose comparison has a confidence interval that includes zero. Finasteride also has a reported difference from placebo in the sponsor analysis. These statements concern the posted model results for this endpoint and follow-up period.

The findings do not establish how much any individual participant's hair count changed, because group averages do not show the full distribution of individual outcomes. The registry record does not report a threshold for a clinically meaningful difference, so statistical evidence of a difference should not automatically be equated with a meaningful change for participants. Nor does an outcome measured at the vertex describe every aspect of hair loss or every patient-important outcome.

The results do not by themselves establish that one treatment is safer than another. Safety requires consideration of the adverse-event data, their seriousness, duration, attribution, and uncertainty, as well as the efficacy outcome. The posted safety summary lists event counts but does not provide all details needed for a comprehensive benefit-risk assessment. Similarly, a comparison with finasteride that has an interval crossing zero does not prove that the treatments are equivalent. A formal non-inferiority or equivalence claim depends on a margin and analysis criteria that are not reported in the registry record.

The study was randomized and quadruple-masked according to the registry, which describes design features intended to support fair comparison, but the posted information does not resolve limitations from missing data, multiplicity, or incomplete reporting of model details. The evidence should be understood as applying to the studied population, regimens, outcome definition, and Week 24 assessment, rather than extrapolated beyond those details.

18. Why This Trial Matters Statistically

This trial is statistically informative because it compares several active regimens with both placebo and an active comparator in a randomized, masked, parallel-group design. The multiple dutasteride doses make it possible to examine whether observed group estimates vary across dose levels, while the finasteride group provides an active-treatment reference. The primary outcome is defined with a consistent photographic measurement of a marked vertex area, and the registry describes a common Week 24 assessment.

The reported results also illustrate why estimates, intervals, and analysis provenance all matter. A mean difference alone gives a point estimate but not its precision. Confidence intervals show a range of uncertainty, and the comparison with the lowest dutasteride dose illustrates that a positive point estimate may coexist with an interval including zero. The sponsor's model-based results differ in some details from simple unadjusted recalculations, making it important to distinguish the trial's own analysis from checks based on summary statistics.

The design and posted analyses do not remove the need to examine multiplicity, missing data, and safety separately. The record lists several outcomes and comparisons but does not report a full multiplicity strategy or all analysis-plan details. It specifies LOCF for the primary outcome, a method whose assumptions can influence results when data are missing. Safety counts and the primary hair-count endpoint answer different questions and should not be conflated. Taken together, these features make the trial a useful example of how to read comparative continuous-outcome results while keeping statistical significance, precision, clinical relevance, and evidence limitations distinct.

19. 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 are specific to the information reported for this trial. The primary outcome is a hair count in a defined vertex circle, so it does not directly describe changes across the entire scalp or outcomes beyond the measured area. The assessment ends at Week 24; longer-term outcomes are not reported in the primary results summarized in the registry record. The registry's description of the study purpose refers to six months, but the posted primary endpoint is specifically measured at Week 24.

The sponsor reports a general linear model, but the complete model specification, covariates, and diagnostic information are not reported in the registry record. This restricts independent assessment of model assumptions and makes the posted summary statistics insufficient to reproduce the sponsor analysis. Recalculated estimates are approximate and unadjusted; because a discrepancy is flagged, the sponsor analysis takes precedence.

The record specifies LOCF for missing primary data but does not report the amount or pattern of missingness, the assumptions supporting that method, or sensitivity analyses. The full operational description of the primary analysis population is truncated. The record also does not report a complete multiplicity strategy, an equivalence or non-inferiority margin, the original sample-size justification, or whether interim analyses occurred. These details matter for interpreting the numerous treatment comparisons and the active-comparator claims.

Safety information is limited to the posted counts and selected serious-event terms. Death counts are not reported in the registry record, and the information does not fully describe event timing, relatedness, or participant-level exposure. Finally, no threshold for a clinically meaningful hair-count change is reported, limiting the translation of statistical differences into clinical importance.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Dutasteride Versus Placebo and Finasteride in Men With Androgenetic Alopecia — NCT01231607.
• Registry record last updated 2018-08-10; status last verified 2018-06.
• Results first posted 2012-10-04.
• 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.

23. Record Summary

The registry describes a completed Phase 3 randomized, quadruple-masked, parallel-group study sponsored by GlaxoSmithKline, with enrollment of 917. Men with androgenetic alopecia were assigned to placebo, finasteride 1 mg, or one of three dutasteride doses. The primary endpoint was change from baseline to Week 24 in nonvellus hair count within a 2.54 cm diameter circle at the vertex, measured by macrophotographic technique.

The sponsor's reported general linear model results show positive estimated differences from placebo for the listed active treatments, with the lowest-dose interval including zero. Comparisons with finasteride vary by dutasteride dose; the registry does not report the margin needed to establish non-inferiority or equivalence. The primary analysis specifies an Intent-to-Treat population and LOCF for missing data. The record also lists secondary hair-count, hair-width, terminal-hair-count, and global-assessment outcomes, along with adverse-event counts.

The sponsor's analysis takes precedence over unadjusted recalculations because the registry record flags a discrepancy. The full statistical analysis plan, multiplicity strategy, original sample-size justification, and several details needed to evaluate missingness and safety are not reported in the registry record.