Treatment Strategy to Prevent Mood Disorders Following Traumatic Brain Injury: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Treatment Strategy to Prevent Mood Disorders Following Traumatic Brain Injury (NCT00704379), a randomized phase 2, phase 3 study. The primary outcome is Time to Onset of Diagnostic and Statistical Manual (DSM) IV Defined Mood and Anxiety Disorders Associated With Traumatic Brain Injury (TBI).

2. Clinical Question

3. Trial Design

This completed study is listed as phase 2 and phase 3, with randomized allocation and a parallel-group model. The registry record lists 94 participants enrolled: 46 in the placebo arm and 48 in the sertraline arm. The sponsor is Baylor College of Medicine. The record gives a start date of 2008-06 and a primary completion date of 2014-04.

The study compared once-daily sertraline with once-daily placebo. The registry describes the intervention arms as double blind. Its overall masking field says triple masking, but the record does not specify which three groups or roles were masked. The allocation process, concealment procedures, and details of how masking was maintained are not reported in the registry record.

The study purpose was to examine sertraline for prevention of mood and anxiety disorders during the first six months after TBI. The primary outcome is titled time to onset of DSM-IV-defined mood and anxiety disorders associated with TBI, with a timeframe of six months after TBI. Several secondary outcomes concern community integration, decision-making task performance, memory, social functioning, and neuroimaging. Their presence broadens the outcomes being studied beyond the primary time-to-onset question.

The registry record does not report the original sample-size justification, the trial’s full statistical analysis plan, interim analyses, or the methods used to handle missing outcome data. These details cannot be inferred from the sample-size illustrations or the posted outcome summaries.

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 Time to Onset of Diagnostic and Statistical Manual (DSM) IV Defined Mood and Anxiety Disorders Associated With Traumatic Brain Injury (TBI), measured over 6 months after TBI.

The primary outcome concerns the time until a participant first meets the registry’s definition of a mood or anxiety disorder associated with TBI. It is expressed in weeks, with a timeframe of six months after TBI. Time-to-onset is useful when the timing of an event matters, rather than only whether an event occurred by a single endpoint assessment. General statistical context: time-to-event methods can use information about when events occur and can account for participants whose event time is not observed within the follow-up period, subject to the method’s assumptions and the way censoring is handled.

The outcome description uses DSM-IV categories for mood disorders associated with TBI and describes a related framework for anxiety disorders. It notes that DSM-IV has since been updated by DSM-5; the registry does not state that DSM-5 criteria were used in the trial. The record’s outcome title refers to diagnostic disorders, but the diagnostic assessment procedures and adjudication process are not reported in the registry record.

There is an important distinction between the outcome’s stated unit and the registry’s reported analysis. The primary outcome is listed as a mean in weeks, while the sponsor’s posted analysis is a log-rank analysis reported as a hazard ratio. The record does not explain the relationship between those two presentations. The mean summaries and their recalculated difference therefore describe a different statistical summary from the sponsor’s reported time-to-event result.

7. Results Posted in the Registry

• Sertraline: mean 15.78 versus 21.43 on control; difference -5.64 (95% CI -7.64 to -3.64), standardized effect d = -1.16; 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.

Total Community Integration Questionnaire Scores

• 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 deviations.
• Result:
  • Sertraline: mean 17.67 versus 16.77 on control; difference 0.90 (95% CI -1.02 to 2.82), standardized effect d = 0.22; z-test p = 0.357.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Sertraline and the control. The interval gives the range of effects that remain compatible with the data.

Iowa Gambling Task 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 deviations.
• Result:
  • Sertraline: mean -0.77 versus 16.67 on control; difference -17.44 (95% CI -31.41 to -3.47), standardized effect d = -0.61; z-test p = 0.014.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Sertraline than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Memory Function Composite

• 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 deviations.
• Result:
  • Sertraline: mean 0.07 versus -0.08 on control; difference 0.15 (95% CI -0.26 to 0.56), standardized effect d = 0.18; z-test p = 0.473.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Sertraline and the control. The interval gives the range of effects that remain compatible with the data.

The posted secondary outcomes cover community integration, gambling-task performance, memory, social functioning, and neuroimaging. Summary statistics and recalculated comparisons are available for community integration, gambling-task score, and memory function. Detailed result statistics for the social-functioning and neuroimaging outcomes are not reported in the registry record.

Across the outcomes with posted recalculations, estimates do not all point in the same direction, and the confidence intervals vary in their precision. This pattern does not create a consistent secondary-outcome signal that resolves the uncertainty around the primary presentation. Secondary measures also address different constructs from time to onset of mood and anxiety disorders, so they are not interchangeable evidence about the primary question.

The registry record does not report a multiplicity adjustment for the secondary analyses. General statistical context: when many outcomes are tested, the chance of at least one apparently notable result can increase, even if there is no corresponding underlying effect. Unadjusted secondary p-values should therefore carry less weight than a prespecified primary analysis, particularly when outcome definitions, analysis populations, and the analysis plan are not fully described. The secondary findings are best read as supportive or exploratory summaries, not as independent confirmation of the primary result.

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

• Total Community Integration Questionnaire Scores
• Iowa Gambling Task Score
• Memory Function Composite
• Social Functioning Examination Total Score
• Neuroimaging Variables (i.e., Fractional Anisotropy [FA] of Frontal White Matter Such as the Cingulate Gyrus)

9. Statistical Interpretation

10. Statistical Methods Explained

General statistical context: a mean describes the average observed value in a group. For a continuous outcome, a difference between group means expresses how far apart the group averages are, in the outcome’s original units. It does not describe how values are distributed among individuals, and it may not be an appropriate summary when the analysis is based on event timing or when the outcome has a substantially different structure.

The primary outcome is recorded as a mean in weeks with standard error. A standard error describes the uncertainty in an estimated mean; it is not the spread of individual participants’ measurements. The registry also reports an analysis using the log-rank method and a hazard ratio. General statistical context: the log-rank method compares event-time patterns between groups, while a hazard ratio compares event rates over time under a model or interpretation that depends on event coding and the reference group. The registry does not report those conventions, and its mean and log-rank presentations are not interchangeable summaries.

The recalculated primary mean difference is an unadjusted comparison based on posted arm-level summary statistics. Its p-value and confidence interval are recalculated rather than the sponsor’s own analysis. The record identifies a discrepancy, so the sponsor’s reported log-rank analysis takes precedence. No adjustment variables or covariates are specified for the sponsor’s analysis in the registry record.

The posted secondary outcomes are continuous measures summarized by group means and standard deviations. General statistical context: a standard deviation describes variation among individual observations, unlike a standard error, which describes uncertainty in an estimated mean. The recalculated mean differences compare the posted arm means. Their confidence intervals express uncertainty around those differences under the assumptions of the unadjusted calculation; they are not evidence that the sponsor used the same calculation or model.

General statistical context: a confidence interval gives a range of values compatible with the estimate and the method’s assumptions. A broad interval indicates greater uncertainty about the size and potentially the direction of an effect. An interval that crosses zero in a mean-difference analysis includes no difference as a value compatible with that calculation. That interpretation concerns the recalculated comparison, not a replacement for a sponsor analysis using a different method.

Over-dispersion is a general concern when outcome variability is greater than a statistical model assumes. If it is not accounted for, standard errors may be too small and confidence intervals may be too narrow. The registry record does not report whether over-dispersion was assessed or relevant to the primary analysis. The secondary outcomes are reported with standard deviations, but the record does not give enough distributional detail to assess model assumptions or the shape of the individual-level data.

The registry reports a log-rank result but does not describe censoring rules, event adjudication, proportional-hazards assessment, or the handling of participants without an observed event. General statistical context: a hazard ratio is often interpreted under assumptions about how the relative event rate behaves over follow-up; when those assumptions are not met, a single ratio may not capture changes in the comparison over time. Whether those assumptions were assessed is not reported in the registry record.

The comparison calculations use the posted group summaries and are unadjusted. They do not recover participant-level information, account for unreported covariates, or establish how missing data were handled. The registry record does not report a full statistical analysis plan, analysis populations, or interim analyses. Those omissions limit the ability to reproduce the sponsor’s analysis or determine whether the posted summaries cover all randomized participants.

11. Confidence Intervals and Precision

The sponsor’s reported hazard ratio is 3.6, with a 95% confidence interval from 1.1 to 16.2. The interval spans a broad range of estimates. Its lower and upper bounds are both above 1, but the registry does not report the reference group or event coding needed to identify the direction of the comparison. The interval therefore indicates uncertainty about the magnitude without supporting a clear direction-specific account based on the registry record alone.

The recalculated unadjusted mean difference is -5.643 weeks, with a 95% confidence interval from -7.642 to -3.644. This interval does not include zero under the recalculation, but it belongs to the mean-difference analysis, not the sponsor’s log-rank analysis. Since the registry identifies a discrepancy, the recalculation should be read as a separate descriptive check; the sponsor’s analysis takes precedence.

Standard errors are reported for the primary arm means, while standard deviations are reported for the secondary arm means. These quantities describe different kinds of variability and should not be compared as if they meant the same thing. The registry does not provide participant-level distributions, so the precision of the group summaries cannot be evaluated beyond their reported uncertainty measures and intervals.

General statistical context: precision concerns how narrowly a result is estimated, not whether the result is important or whether a particular interpretation is correct. The wide reported hazard-ratio interval and the mismatch between the outcome’s mean presentation and the sponsor’s time-to-event analysis make careful interpretation especially important.

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 planning entries are illustrations based on the observed effect, not the trial’s original design calculation. For continuous outcomes, the figures are participants per arm. The stated assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation; the effect to detect is the stated fraction of the observed effect. For time-to-event tables, the figures refer to total events, using the Schoenfeld method. The registry record does not report the trial’s own original sample-size assumptions or justification.

The 80% power column gives the planned participants per arm, or total events for a time-to-event illustration, under which the assumed effect would be detected in 80% of repetitions of the study under the planning assumptions. The 90% power column gives the corresponding planning quantity for 90% of repetitions. These are design targets, not guarantees about the result of a particular study.

Rows that plan for a smaller fraction of the observed effect require more participants per arm in the continuous-outcome illustrations. This is because smaller differences are harder to distinguish from sampling variation under otherwise stated assumptions. The values are illustrative and depend on using the observed effect as their starting point; they do not show that the completed trial had a particular probability of detecting an effect. This is 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 = -5.643; pooled standard deviation σ = 4.847.
2. n = 2 × (1.96 + 0.842)² × 4.847² / -5.643² = 11.6, rounded up to 12 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 <- -5.6431                      # observed difference in means
sigma <- 4.8467754477456                     # 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(704379)
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 = 12 (80% power), 16 (90% power)
 75% of observed effect: n per arm = 21 (80% power), 28 (90% power)
 50% of observed effect: n per arm = 47 (80% power), 63 (90% power)
power.t.test: 12.6 
Simulated power at n = 12 per arm: 0.779 (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 posted primary analysis is a log-rank result: hazard ratio 3.6, 95% confidence interval 1.1 to 16.2, and p < 0.05. A separate unadjusted recalculation from the arm means gives a mean difference of -5.643 weeks, with a 95% confidence interval from -7.642 to -3.644 and p < 0.001. The registry flags a discrepancy between these presentations.

The recalculation is based on group-level means and is not the same analysis as the reported log-rank comparison. It does not explain the sponsor’s event coding, reference group, censoring methods, or any participant-level analysis choices. The registry record does not state why the presentations differ. In line with the discrepancy flag, the sponsor’s own analysis takes precedence; the recalculated result is a secondary check and should not be used to substitute a direction or interpretation for the reported hazard ratio.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists several secondary outcomes in addition to the primary time-to-onset outcome. It does not report whether the primary endpoint was selected from a prespecified hierarchy, whether secondary outcomes were ordered, or whether p-values or confidence intervals were adjusted for multiple comparisons. Without those details, the secondary p-values should be interpreted as unadjusted and considered exploratory rather than as a collection of confirmatory tests.

The available record also does not report the analysis populations used for the primary or secondary outcomes. Although the enrolled total and primary arm counts are listed, the secondary outcome summaries have different group counts. The registry does not explain why those counts differ, whether they reflect missing measurements, eligibility for a particular assessment, or another reason. The handling of missing observations, withdrawals, and incomplete follow-up is not reported in the registry record.

For the primary endpoint, the record does not describe event ascertainment, censoring, or how participants without a recorded onset were handled. The sponsor’s log-rank analysis is reported, but its detailed model and analysis plan are not. The registry does not report interim analyses, stopping rules, or whether any analyses were changed after looking at the data.

These omissions matter because analysis choices affect the interpretation of uncertainty and of comparisons across many outcomes. They do not establish that the analyses were inappropriate; they mean the registry record does not provide enough information to evaluate those aspects of the study. The posted results should therefore be interpreted in light of both the reported discrepancy and the limits of the available methods documentation.

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: hip fracture (2.2%).

The registry lists serious adverse events in 1 of 46 participants in the placebo arm and 1 of 48 participants in the sertraline arm. It also lists other adverse events in 31 of 46 participants in the placebo arm and 35 of 48 participants in the sertraline arm. Hip fracture is the leading serious-event term listed, at 2.2%.

These figures describe the counts recorded in the registry and do not by themselves establish that an event was caused by a study intervention. The record does not provide event-level descriptions, timing, severity grading beyond the serious-event classification, or an attribution assessment. Deaths are not reported in the registry record. The adverse-event summaries should be interpreted descriptively, especially because the listed group counts do not provide enough detail to assess follow-up completeness or the methods used to collect events.

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

The registry record describes a randomized comparison of sertraline and placebo after TBI, with the aim of examining the onset of mood and anxiety disorders during the first six months. It reports a statistically significant log-rank analysis, but the hazard ratio’s reference group and event coding are not stated. The reported estimate therefore cannot be assigned a clear direction from the registry record alone.

A recalculation from the posted means yields a negative unadjusted difference in weeks, but this is not the sponsor’s reported analysis. The record flags a discrepancy, and the sponsor’s analysis takes precedence. The difference in statistical summaries, together with the lack of detail about event definitions and analysis methods, means that the recalculation cannot settle the interpretation of the primary result.

The secondary outcomes describe a range of functional and cognitive measures, but not all listed secondary outcomes have result statistics posted. The available recalculated comparisons are not a substitute for the primary endpoint, and unadjusted secondary analyses carry less weight when multiplicity procedures and the analysis plan are not reported.

These results do not establish a clinical recommendation. They describe registry-recorded analyses and their precision, not a complete account of every methodological choice or participant-level result. A more definitive assessment of the primary comparison would require clarification of the hazard-ratio direction, the relation between the reported time-to-event analysis and the mean summaries, and details of the analysis populations and missing-data handling.

18. Why This Trial Matters Statistically

This trial is statistically instructive because its primary outcome is described as a mean time in weeks, while the sponsor’s reported analysis is a log-rank hazard ratio. Those are distinct summaries of event-time data. The registry also flags a discrepancy between the sponsor’s analysis and a simple recalculation from the posted means. Keeping those results separate is essential: a calculation that can be reproduced from summary statistics is not automatically the analysis that should govern interpretation.

The confidence interval for the reported hazard ratio is wide, which highlights that statistical significance and precision are different concepts. A result can meet a conventional significance threshold while leaving substantial uncertainty about its magnitude. The missing reference-group and event-coding details further show why a numerical estimate must be interpreted in context, rather than by its value alone.

The trial also illustrates why secondary outcomes need careful framing. Multiple outcomes describe different aspects of functioning, and the registry does not report whether multiplicity was addressed. Unadjusted findings across those measures should not be treated as equal in evidential weight to a clearly specified primary analysis. Finally, the sample-size illustrations are planning examples based on observed effects, not evidence about the completed trial’s original design or its probability of detecting an effect.

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 details needed to interpret the primary result are not reported in the registry record. The hazard-ratio reference group and event coding are unspecified, and the relationship between the primary outcome’s mean-in-weeks presentation and the sponsor’s log-rank analysis is not explained. The registry flags a discrepancy with the recalculation, but does not describe its source.

The record does not give diagnostic assessment procedures, event adjudication details, censoring rules, or the handling of participants without an observed onset. Nor does it report the analysis population, missing-data methods, or why the secondary-outcome group counts differ from the primary arm counts. These gaps limit assessment of how complete and comparable the outcome data are.

The record lists secondary outcomes for social functioning and neuroimaging, but their result statistics are not reported in the registry record. It also does not report multiplicity procedures, the original sample-size justification, interim analyses, or a full statistical analysis plan. The sample-size illustrations therefore cannot be read as evidence about the trial’s original design. Finally, the record’s overall triple-masking label does not identify which roles were masked, while the intervention descriptions refer to double-blind administration. These reporting limits restrict methodological appraisal without establishing that any particular procedure was or was not used.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Treatment Strategy to Prevent Mood Disorders Following Traumatic Brain Injury — NCT00704379.
• Registry record last updated 2015-08-25; status last verified 2015-08.
• Results first posted 2015-05-12.
• 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, parallel-group phase 2 and phase 3 study sponsored by Baylor College of Medicine, with 94 participants enrolled. It compares once-daily sertraline with placebo after TBI and focuses on time to onset of DSM-IV-defined mood and anxiety disorders during the six months after TBI.

The sponsor reported a log-rank hazard ratio of 3.6, with a 95% confidence interval from 1.1 to 16.2 and p < 0.05. A recalculation from the posted means is separately available, but the registry flags a discrepancy and the sponsor’s analysis takes precedence. The record does not report enough information to determine the hazard ratio’s direction or reconcile the two presentations. Secondary outcomes and adverse-event counts are posted with notable gaps in methods and outcome reporting.