A Study of Duloxetine in Elderly Generalized Anxiety Disorder: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of Duloxetine in Elderly Generalized Anxiety Disorder (NCT01118780), a randomized phase 4 study. The primary outcome is Change From Baseline to Week 10 in Hamilton Anxiety Rating Scale (HAMA) Total Score.

2. Clinical Question

The scale assesses anxiety symptom severity. Lower scores indicate less severe symptoms, so a more negative change represents a larger reduction from baseline. The registry record reports adjusted group estimates and a sponsor analysis of the between-group difference. Recalculation from the posted group summaries gives an unadjusted estimate, which is a different analysis and should be read as a supplementary statistical check rather than a replacement for the sponsor’s model.

3. Trial Design

This was a randomized, double-masked, parallel-group phase 4 study comparing duloxetine with placebo. The registry record lists enrollment as 291 and the study status as completed. The recorded study dates are 2010-10 for start and 2012-07 for primary completion. The stated purpose was to test the safety and efficacy of duloxetine versus placebo in elderly patients with generalized anxiety disorder.

For the primary outcome, the registry identifies randomized participants who had a baseline and at least 1 post-baseline Hamilton Anxiety Rating Scale total score as the analysis population. The posted endpoint summaries show 143 participants for duloxetine and 131 for placebo. These are endpoint-specific counts; the registry record does not explain all differences between enrollment, randomized groups, and the analysis population.

The primary measure was assessed from baseline to Week 10. The registry also lists secondary assessments of functioning, anxiety symptoms, depression and anxiety scales, global impressions, pain, quality of life, and suicide-related ideation and behavior. It records 21 outcome measures and 38 analyses overall, with 19 secondary outcomes posted. The registry record does not report the original sample-size justification, planned enrollment by arm, or detailed rules governing which secondary analyses would be emphasized.

Randomization is intended to reduce systematic differences between treatment groups at the start of a trial. Double masking is intended to limit the influence of treatment knowledge on participants’ reports and study assessments. These design features support a comparative interpretation, but the details of allocation concealment and masking procedures 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.

Before assignment: Study had 3 periods: a screening period (3 to 30 days prior to randomization, no study drug administered), a treatment period (10 weeks), and a taper period (2 weeks). Participants who completed the treatment period were considered to have completed the study. Participant Flow and results, unless specified otherwise, are during treatment period.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Randomized participants with a baseline and at least 1 post-baseline HAMA Total Score.

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.

Population: All randomized participants.

6. Primary Endpoint

The registry lists the primary outcome as Change From Baseline to Week 10 in Hamilton Anxiety Rating Scale (HAMA) Total Score, measured over Baseline, Week 10.

The primary endpoint was change from baseline to Week 10 in the Hamilton Anxiety Rating Scale (HAMA) total score. The Structured Interview Guide for the Hamilton Anxiety Rating Scale was used to collect the measure. Its 14 items assess anxiety severity, with each item scored from 0, meaning not present, to 4, meaning very severe. The total score can range from 0 to 56, and higher scores indicate greater symptom severity.

Change from baseline is used to describe how a measured outcome shifts during the study relative to each participant’s starting value. Comparing change between randomized groups can help distinguish a difference associated with assignment from change occurring in both groups over time. In this trial, the endpoint is a continuous score, so the group results are expressed as means and a between-group mean difference.

The primary analysis used least-squares means from a mixed-model repeated-measures analysis. Such adjusted estimates account for specified factors in the model and use repeated outcome measurements over time. The registry record reports adjustment for treatment, pooled investigator, age category, visit, treatment-by-visit, baseline score, and baseline-by-visit. A threshold for a clinically important change is not reported in the registry record, so the numerical difference alone does not establish how important it is to patients.

7. Results Posted in the Registry

• Duloxetine: mean -15.86 versus -11.69 on control; difference -4.17 (95% CI -5.97 to -2.37), standardized effect d = -0.55; 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 to Week 10 in Sheehan Disability Scale (SDS) Global Functional Impairment Score

• 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:
  • Duloxetine: mean -8.60 versus -5.37 on control; difference -3.23 (95% CI -4.95 to -1.51), standardized effect d = -0.45; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Duloxetine than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Clinical Global Impressions of Improvement Scale (CGI-Improvement) at Week 10

• 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:
  • Duloxetine: mean 2.10 versus 2.63 on control; difference -0.53 (95% CI -0.81 to -0.25), standardized effect d = -0.45; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Duloxetine than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Patient's Global Impressions of Improvement Scale (PGI-Improvement) at Week 10

• 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:
  • Duloxetine: mean 2.35 versus 2.97 on control; difference -0.62 (95% CI -0.94 to -0.30), standardized effect d = -0.46; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Duloxetine than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The posted secondary results cover several kinds of outcomes, including functioning, anxiety and mood symptoms, clinician and patient global impressions, pain, quality of life, and suicide-related ideation and behavior. Among the continuous secondary comparisons reported in detail, the posted estimates generally favor a larger improvement or lower score in the duloxetine group than in the placebo group. These findings are directionally consistent with the primary HAMA result, but they do not replace it: each measure captures a different aspect of symptoms or functioning.

The registry lists 21 outcome measures, 19 secondary outcomes posted, and 38 analyses overall. A reader should distinguish an overall pattern from a claim that every outcome was positive or that every difference is reliable. The registry record does not provide a complete account of all secondary estimates in the available summary or a prespecified hierarchy for interpreting them.

In general, testing many outcomes increases the chance that at least one result will appear statistically notable by chance. Unless an analysis plan accounts for multiple comparisons, unadjusted secondary p-values carry less confirmatory weight than a prespecified primary result. The registry record does not report a multiplicity adjustment or other strategy for controlling the chance of false-positive findings across secondary outcomes. Secondary results are therefore best read as supportive or exploratory evidence, with attention to effect estimates, uncertainty, outcome relevance, and the missing multiplicity details.

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

• Change From Baseline to Week 10 in Sheehan Disability Scale (SDS) Global Functional Impairment Score
• Change From Baseline to Week 10 in Hamilton Anxiety Rating Scale (HAMA) (Psychic Anxiety Factor Score, Somatic Anxiety Factor Score, and Individual Item Scores: Anxious Mood Item and Tension Item)
• Change From Baseline to Week 10 Endpoint in Hospital Anxiety Depression Scale (HADS) Subscale Scores
• Clinical Global Impressions of Improvement Scale (CGI-Improvement) at Week 10
• Patient's Global Impressions of Improvement Scale (PGI-Improvement) at Week 10
• Change From Baseline to Week 10 in Brief Pain Inventory-Modified Short Form (BPI-SF) Pain Severity and Interference Subscales
• Change From Baseline to Week 10 in Quality of Life Enjoyment and Satisfaction Questionnaire - Short Form (Q-LES-Q-SF) Total Score
• Number of Participants With Treatment-Emergent Suicide-Related Ideation and Behavior Based on the Columbia Suicide Severity Rating Scale (C-SSRS)

9. Statistical Interpretation

10. Statistical Methods Explained

What does the mixed-model analysis do?

The primary analysis used mixed-model repeated measures (MMRM) to estimate least-squares mean changes and compare treatment groups over repeated visits. In general, a repeated-measures model can use information from more than one measurement time while accounting for the fact that observations from the same participant are related. It estimates group results after adjusting for factors specified in the model, rather than simply comparing raw group averages.

The registry says the primary model adjusted for treatment, pooled investigator, age category, visit, treatment-by-visit, baseline score, and baseline-by-visit. A treatment-by-visit term allows the modeled treatment contrast to vary across visits. Baseline score and its interaction with visit allow the relationship between starting score and later measurements to be represented in the model. Details such as the covariance structure, estimation method, and handling of incomplete repeated measurements are not reported in the registry record.

What are least-squares means?

Least-squares means are model-based group estimates adjusted for the factors included in the analysis. They are not necessarily identical to unadjusted averages calculated directly from the observed scores. The sponsor reported a least-squares mean difference of 4.17 for the primary endpoint, with a 95% confidence interval from 2.47 to 5.88 and p < 0.001. The group changes were -15.86 for duloxetine and -11.69 for placebo.

What does the recalculation add?

The recalculated primary comparison uses the posted group summaries to produce an unadjusted mean difference of -4.17 and a 95% confidence interval from -5.973 to -2.367. Its p-value is <0.001. This is a simple check based on summary statistics, not a reproduction of the sponsor’s MMRM. It does not adjust for the model factors or repeated-measures structure. The direction is written as duloxetine minus placebo, which accounts for its negative sign relative to the sponsor’s positive difference. Because the registry discrepancy flag is false, no discrepancy between the analyses is identified; the model-based sponsor result remains the primary reported analysis.

Why can variation matter?

For a continuous outcome, a mean difference gives the contrast in average scores, while a measure of spread describes how variable scores are within groups. Greater variability makes a given difference harder to estimate precisely. In general, over-dispersion means the data vary more than a model’s assumed distribution or variance structure allows; failing to account for it can make uncertainty estimates too small. The registry describes the primary dispersion as a standard error and reports least-squares means. The recalculation uses an approximate pooled standard deviation derived from the posted standard errors and group sizes, so its uncertainty calculation is an approximation rather than the sponsor’s model-based interval.

What do the p-value and interval convey?

In general, a p-value measures how unusual data at least as incompatible with a null value would be if that null model and its assumptions held. A small p-value does not quantify clinical importance or the probability that the treatment effect is real. A confidence interval gives a range of estimates compatible with the data under the analysis assumptions and conveys precision as well as direction. The reported and recalculated intervals use different analyses, so their values should not be treated as interchangeable. The registry reports a 95% confidence interval for the sponsor analysis and an unadjusted recalculated interval.

What is not described?

The registry record does not report the original sample-size justification, the trial’s planned power, interim analyses, or a detailed statistical analysis plan. The MMRM adjustment variables and primary analysis population are stated, but the record does not give enough detail to reconstruct every modeling choice or assess how incomplete data were handled. The registry also does not report a prespecified multiplicity strategy for the many outcomes and analyses. These omissions limit independent assessment of analysis choices, especially for outcomes beyond the primary endpoint.

11. Confidence Intervals and Precision

The sponsor’s primary estimate is a least-squares mean difference of 4.17, with a 95% confidence interval from 2.47 to 5.88. The interval lies on the positive side for the sponsor’s reported comparison direction. The recalculated unadjusted estimate is -4.17, with a 95% confidence interval from -5.973 to -2.367, because that contrast is expressed in the opposite direction. Both intervals exclude zero, but they arise from different analyses and have different limits.

In general, a confidence interval describes the range of values reasonably compatible with the observed data under the model and its assumptions. A narrower interval indicates greater statistical precision; a wider interval indicates more uncertainty about the size of the effect. An interval does not capture uncertainty from all sources, including possible bias, model misspecification, or selective reporting.

The sponsor’s interval is based on the reported mixed-model analysis, which adjusts for the listed covariates and visit structure. The recalculated interval is a simple unadjusted estimate based on posted group summaries. The recalculation therefore provides a check on the direction and rough scale of the contrast, not a substitute for the adjusted analysis. A confidence interval that excludes zero does not itself establish clinical importance. The registry record does not report a threshold for a clinically meaningful HAMA difference, nor does it provide enough detail to evaluate every assumption behind the model-based interval.

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 illustrations are planning calculations based on the observed primary effect, not the trial’s original design. They assume a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each illustrated target effect is the stated fraction of the observed effect. The registry record does not report the original sample-size justification or design assumptions.

For the primary endpoint, the figures are participants per arm, not a combined total. Planning for 100% of the observed effect gives 53 participants per arm for 80% power and 70 participants per arm for 90% power. Planning for 75% of the observed effect gives 93 participants per arm for 80% power and 125 participants per arm for 90% power. Planning for 50% of the observed effect gives 209 participants per arm for 80% power and 280 participants per arm for 90% power.

In general, power is the chance that a study would detect an effect of a specified size at the chosen significance level if that effect were present and the assumptions held. The 80% and 90% columns give different planning targets: the 90% target aims for a higher chance of detection and therefore calls for more participants at the same effect size. Smaller effects are harder to detect and require more participants. These figures are planning illustrations, not observed power.

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 = -4.170; pooled standard deviation σ = 7.598.
2. n = 2 × (1.96 + 0.842)² × 7.598² / -4.170² = 52.1, rounded up to 53 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 <- -4.17                      # observed difference in means
sigma <- 7.59842727535268                     # 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(1118780)
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 = 53 (80% power), 70 (90% power)
 75% of observed effect: n per arm = 93 (80% power), 125 (90% power)
 50% of observed effect: n per arm = 209 (80% power), 280 (90% power)
power.t.test: 53.1 
Simulated power at n = 53 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 primary analysis used mixed models and reported a least-squares mean difference of 4.17, a 95% confidence interval from 2.47 to 5.88, and p < 0.001. The reported contrast uses the opposite direction from the recalculated duloxetine-minus-placebo contrast, which is -4.17 with a 95% confidence interval from -5.973 to -2.367. The sign difference reflects comparison order, not a difference in which group had the larger reduction.

The recalculation is unadjusted and uses posted summary statistics; it does not reproduce the sponsor’s MMRM adjustment for treatment, pooled investigator, age category, visit, treatment-by-visit, baseline score, and baseline-by-visit. Its p-value is also unadjusted. The registry record does not flag a discrepancy between the analyses. The sponsor’s model-based result is the reported primary analysis, while the recalculation offers a simpler statistical check. Their estimates should be interpreted with their respective methods and comparison directions in mind.

15. Multiplicity, Missing Data and Analysis Populations

The registry record identifies a primary endpoint and lists numerous secondary outcomes and analyses, but a multiplicity strategy is not reported in the registry record. Multiplicity arises when several hypotheses are tested: even when no real effects exist, the chance of at least one apparently unusual result can rise as more tests are conducted. In general, prespecifying a hierarchy or adjusting the analysis can help control that risk. Without the stated plan, the secondary p-values should not be treated as if each were the sole confirmatory test.

The primary analysis population is described as randomized participants with a baseline and at least 1 post-baseline HAMA total score. That definition clarifies who contributed to the primary endpoint analysis, but the registry record does not explain how participants without post-baseline observations were handled, how intermittent missing visits were treated, or whether sensitivity analyses assessed departures from the model’s missing-data assumptions. MMRM can use available repeated measurements under its assumptions; the registry’s description is not sufficient to determine whether those assumptions were plausible or how robust the result was to missing data.

The registry record reports adjustment variables for the primary MMRM but does not report the full analysis plan, interim analyses, the original sample-size justification, or a detailed method for handling all analysis populations. It also does not describe allocation concealment or the procedures used to maintain masking. These are reporting limitations, not proof that the procedures were absent. They matter because interpretation depends not only on the estimated difference but also on prespecification, missingness, and how many analyses were considered.

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 3% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: angina pectoris (0.7%), large intestinal obstruction (0.7%), hypertensive crisis (0.7%).

The registry reports serious adverse events in 3 of 151 participants at risk in the duloxetine group and 0 of 140 at risk in the placebo group. Other adverse events are reported for 89 of 151 and 71 of 140, respectively. These counts describe events recorded in this trial; they do not by themselves establish that an event was caused by the assigned treatment.

The serious-event terms listed include angina pectoris, large intestinal obstruction, and hypertensive crisis, each reported as 0.7%. The registry record does not report death counts. Serious adverse events and other adverse events are distinct categories in the record, and the counts should not be combined. In general, adverse-event comparisons need to be interpreted in light of event definitions, observation time, severity, and uncertainty. Those details are not fully reported in the registry record, and the event counts alone do not support a broad safety conclusion.

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

The primary result describes a difference in average change in HAMA total score from baseline to Week 10 between randomized groups. The reported group changes were -15.86 for duloxetine and -11.69 for placebo, with a sponsor-reported adjusted difference of 4.17 in the sponsor’s comparison direction. The sponsor reported a 95% confidence interval from 2.47 to 5.88 and p < 0.001. The recalculated unadjusted contrast uses the reverse direction and has an estimate of -4.17.

These data support a statistical distinction between the group estimates under the analyses reported. They do not show that all participants improved, that the same change occurred for each participant, or that an individual response can be predicted from the group average. A statistical difference also does not by itself establish clinical importance. The registry record does not report a threshold for a clinically meaningful change on HAMA.

The result applies to the study’s population, outcome, follow-up period, and analysis approach. It does not establish effects beyond the Week 10 endpoint or answer every question about functioning, quality of life, or safety. Secondary results add context but have less confirmatory weight when the multiplicity plan is not reported. Missing-data handling and several other design and analysis details are also not fully described in the registry record.

In general, trial estimates are most informative when considered alongside their confidence intervals, design, analysis assumptions, and completeness of reporting. The interval indicates statistical uncertainty under the model, not certainty about the true effect or protection from possible biases. The findings should therefore be understood as the comparative results recorded for this trial, rather than as a guarantee of an outcome for any individual.

18. Why This Trial Matters Statistically

This trial illustrates why the estimated treatment contrast, its direction, and the analysis method all matter. The sponsor’s mixed-model result is adjusted for specified baseline and study factors, while the recalculation uses summary statistics without those adjustments. Their opposite signs reflect different comparison directions, so interpretation requires identifying which group is subtracted from which rather than reading the sign in isolation.

It also shows the distinction between statistical evidence and clinical importance. The primary result has a confidence interval excluding zero and a reported p-value of <0.001, but the registry does not provide a clinically meaningful threshold for the HAMA scale. Precision and statistical significance cannot answer that separate question on their own.

The range of secondary outcomes highlights the importance of multiplicity. A broad set of measures can offer useful perspectives on functioning, symptoms, and patient or clinician assessments, but unadjusted findings are more vulnerable to chance patterns. Because a multiplicity strategy is not reported, the primary endpoint should remain central and secondary findings should be interpreted cautiously.

Finally, the record provides a basis for examining model-adjusted results, unadjusted checks, and planning illustrations, while leaving some protocol and analysis details unreported. The distinction between observed trial results and planning calculations is important: the sample-size illustrations describe hypothetical future designs, not the power achieved by this completed trial.

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 trial-specific limits follow from the registry record. The primary analysis includes randomized participants with a baseline and at least 1 post-baseline HAMA total score, but the record does not explain how missing observations or participants outside that analysis population were handled. That limits assessment of possible bias from incomplete follow-up.

The primary estimate comes from an MMRM adjusted for listed factors, but key modeling details, including covariance structure and sensitivity analyses, are not reported in the registry record. The unadjusted recalculation is based on posted summary statistics and uses an approximate standard deviation; it cannot verify all features of the sponsor’s model. The original sample-size justification, interim analyses, and a detailed statistical analysis plan are also not reported.

The registry lists many outcome measures and analyses but does not report a multiplicity strategy. This makes secondary statistical findings less conclusive than a single prespecified primary comparison. The record also does not state a clinically meaningful threshold for the HAMA change, limiting interpretation of the practical importance of the estimated difference.

Safety reporting includes event counts and selected serious-event terms, but death counts are not reported in the registry record, and the summary does not provide all details needed to assess event timing or attribution. Finally, masking and randomization are recorded, but the procedures for allocation concealment and maintaining masking are not described. These reporting gaps limit independent evaluation; they do not establish that the unreported procedures were not conducted.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Study of Duloxetine in Elderly Generalized Anxiety Disorder — NCT01118780.
• Registry record last updated 2013-09-02; status last verified 2013-06.
• Results first posted 2013-09-02.
• 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 record describes a completed, randomized, double-masked, parallel-group phase 4 study of duloxetine versus placebo in elderly patients with generalized anxiety disorder. Enrollment was 291. The primary endpoint was change from baseline to Week 10 in HAMA total score, analyzed using mixed-model repeated measures in randomized participants with baseline and at least 1 post-baseline score.

The sponsor reported an adjusted least-squares mean difference of 4.17, with a 95% confidence interval from 2.47 to 5.88 and p < 0.001. An unadjusted recalculation expresses the contrast in the opposite direction as -4.17. The registry also reports secondary outcomes and adverse-event counts, but multiplicity procedures, the original sample-size justification, and several missing-data and modeling details are not reported in the registry record.