The Impact of Vocational Rehabilitation on Mentally III Veterans: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for The Impact of Vocational Rehabilitation on Mentally III Veterans (NCT00333801), a randomized na study. The primary outcome is Obtain Competitive Employment.

2. Clinical Question

The trial addressed an employment outcome rather than a measure of PTSD symptoms as its primary endpoint. It compared two behavioral vocational approaches: a supported-employment model and the VA’s usual vocational rehabilitation program. Employment counted only if it met the registry’s definition of competitive work. Day labor—including babysitting, day-by-day manual labor, temporary work for family or friends, and similar work—did not qualify.

The registry record does not report additional eligibility criteria, the recruitment setting, or participant characteristics beyond the study population description. The record also does not establish whether the results apply to veterans outside the enrolled population or to other vocational services. Those questions require information beyond this registered comparison.

3. Trial Design

This was a randomized, parallel-group trial with no masking reported. The registry lists an enrollment of 100. For the primary outcome, the posted arm counts were 43 participants in the VA Vocational Rehabilitation Program group and 42 in the IPS supported-employment group. The primary outcome description states that all randomized participants were included in the analysis on an intent-to-treat basis.

The study began on 2006-04-01, had a primary completion date of 2010-04-24, and is recorded as completed. The sponsor was the VA Office of Research and Development. The phase is listed as not applicable. The registry record does not report why the primary-outcome analysis counts are smaller than the enrollment figure, or provide a participant-flow account explaining enrollment, assignment, follow-up, and analysis by arm.

VRP is described as treatment as usual and mostly involving a transitional work program, with a set-aside job for a limited period before pursuing competitive employment at discharge. IPS is described as using a specialist to help identify work preferences, begin a community-based job search, pursue competitive employment, and provide follow-along supports. The registry descriptions do not provide complete intervention details, including the full text of each description.

The record does not report the original sample-size justification, planned allocation ratio, or a detailed statistical analysis plan. It also does not report interim analyses, trial sites, or whether recruitment and follow-up procedures differed between groups. The design information therefore establishes the randomized parallel comparison and the primary analysis population, but does not answer all questions about trial conduct.

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: 100 Veterans were enrolled; 85 were randomized to IPS-SE (n=42) or VRP (n=43) and followed 12-months. 15 participants were excluded: 2 did not have PTSD, 2 lived \>100 miles away, 3 withdrew consent, 4 had exclusionary diagnosis, and 4 were not eligible to be enrolled in the VRP. Participants were allowed to continue all other treatments for PTSD.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All randomized participants were included in the analysis (intent-to-treat)

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: Veterans with PTSD were randomly assigned to a treatment goup.

6. Primary Endpoint

The registry lists the primary outcome as Obtain Competitive Employment, measured over 1 calendar year.

The primary endpoint was obtaining competitive employment, recorded as a yes-or-no outcome over one calendar year. The registry defined competitive employment as a job for regular wages in a setting that was not sheltered or set aside, and one that could be held by people without a mental illness or disability. Day labor, including babysitting, manual labor by the day, temporary work, or work for family or friends, did not meet the definition.

This endpoint focuses on whether a participant entered competitive work at least once during the specified period. It does not, by itself, describe how many weeks, days, or hours the participant worked, the amount earned, or whether employment was sustained. The registry lists separate secondary employment outcomes addressing some duration and income measures, but those are distinct from the primary yes-or-no endpoint.

As a binary endpoint, the primary outcome is summarized by the number of participants meeting the definition in each arm, relative to the number analyzed. The registry states that the primary analysis included all randomized participants under an intent-to-treat approach. In general, an intent-to-treat analysis compares groups according to their randomized assignment, even when participants’ experiences after assignment vary. The record does not explain how missing primary-outcome information, if any, was handled.

The endpoint was used to compare vocational approaches on a defined employment milestone. It should not be interpreted as a direct measure of symptoms, quality of life, job stability, or overall functioning.

7. Results Posted in the Registry

• Individual Placement and Support (IPS) Supported Employment: 76.2% (32/42) versus 27.9% (12/43) on control; risk difference 48.3 percentage points (95% CI 29.7 to 66.9), risk ratio 2.73 (1.64–4.54); unadjusted two-proportion 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.

Employment Outcomes (Weeks Competitively Employed)

• 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:
  • Individual Placement and Support (IPS) Supported Employment: mean 21.60 versus 6.80 on control; difference 14.80 (95% CI 8.04 to 21.56), standardized effect d = 0.93; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Individual Placement and Support (IPS) Supported Employment than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Employment Outcomes (Days Competitively Employed

• 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:
  • Individual Placement and Support (IPS) Supported Employment: mean 83.80 versus 29.30 on control; difference 54.50 (95% CI 23.90 to 85.10), standardized effect d = 0.76; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Individual Placement and Support (IPS) Supported Employment than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Employment Outcomes (Hours Competitively Employed)

• 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:
  • Individual Placement and Support (IPS) Supported Employment: mean 656.00 versus 236.00 on control; difference 420.00 (95% CI 171.48 to 668.52), standardized effect d = 0.72; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Individual Placement and Support (IPS) Supported Employment than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The posted secondary employment summaries have a consistent overall direction: the IPS group has higher reported mean weeks, days, and hours competitively employed over one year than the VRP group. These measures describe the amount of competitive employment, extending the primary yes-or-no outcome by considering duration. They are related but distinct: a participant can meet the primary definition without having a high total number of weeks, days, or hours.

The record lists broader secondary outcome topics, including PTSD, depression, disability, and employment outcomes such as duration and income. It reports six posted secondary outcomes in total, but the registry record does not report detailed results for every listed topic in the information available for this analysis. Results for PTSD, depression, disability, and the other income outcomes should not be inferred from the employment-duration summaries.

The registry reports a Wilcoxon (Mann-Whitney) analysis for weeks competitively employed. The recalculated unadjusted mean comparisons for weeks, days, and hours each have p < 0.001. The registry does not list sponsor-reported analyses for days and hours. These statistical results support describing the posted summaries as showing a consistent pattern, but they do not establish that all secondary outcomes had the same direction or magnitude.

Secondary outcomes generally carry less weight than a prespecified primary outcome because multiple outcomes and tests create more opportunities for apparently notable findings. In general, multiplicity procedures can control the chance of false-positive findings across a family of tests, but the registry record does not report whether any such procedure was planned or applied. The secondary comparisons should therefore be interpreted as supportive and exploratory rather than as independent confirmation of the primary result.

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

• PTSD, Depression, Disability Outcomes
• Employment Outcomes (Weeks Competitively Employed)
• Employment Outcomes (Days Competitively Employed
• Employment Outcomes (Hours Competitively Employed)
• Employment Outcomes (Gross Income Competitive)
• Employment Outcomes (Total Gross Income From All Sources

9. Statistical Interpretation

10. Statistical Methods Explained

The registry reports a chi-squared test for the primary binary outcome. The unadjusted recalculation uses the posted counts in each group to compare the proportions who obtained competitive employment. A binary endpoint has two possible states—in this case, obtaining qualifying competitive employment or not obtaining it—so the group summaries are counts of participants with the event and the numbers analyzed. The recalculated effect is expressed as a risk difference, an absolute comparison of the two group proportions. No adjustment for other participant characteristics is described in that recalculation.

What does a p-value tell the reader?

In general, a p-value measures how unusual data at least as different as those observed would be under a specified null model and the assumptions of the statistical test. It does not measure the probability that a treatment is effective, the chance that the result will replicate, or the size of the difference. The registry reports p < 0.001 for the primary chi-squared analysis. The recalculated unadjusted result is also p < 0.001. A small p-value should be read alongside the effect estimate and its interval, since a test alone does not convey the range of plausible effect sizes.

Why report an effect estimate and confidence interval?

In general, an effect estimate describes the direction and magnitude of a comparison, while a confidence interval expresses uncertainty around that estimate under the method used. For the primary outcome, the recalculated risk difference compares the groups’ event proportions. The interval is calculated from the posted summary counts and is unadjusted. It is not a confidence interval reported by the registry’s own chi-squared analysis. The interval’s position relative to zero conveys whether the estimated direction is positive or negative under that calculation; its width conveys precision.

How are the continuous secondary outcomes compared?

The posted summaries for weeks, days, and hours competitively employed give a mean and standard deviation for each arm. The recalculation compares the arm means and reports a mean difference with a confidence interval. In general, the standard deviation describes variation among participant values; it is not the uncertainty of the mean itself. The standard error is related to the precision of a mean estimate and depends on both the spread of observations and the sample size. These outcomes can be skewed or contain many low or zero values, but the registry record does not provide participant-level distributions or enough information to assess their shape.

The registry reports a Wilcoxon (Mann-Whitney) analysis for weeks competitively employed, with Cohen’s d reported as .93 and p < 0.001. The recalculated summary-statistic comparison is a mean difference; it is not the same procedure as the reported rank-based test. A rank-based test compares the ordering of observations between groups, whereas a mean difference describes the difference in arithmetic averages. Their results need not answer identical questions. For days and hours, the registry record does not list a sponsor-reported analysis, although the recalculated unadjusted comparisons are available from the posted summary statistics.

What does over-dispersion mean in general?

For count outcomes, over-dispersion means that observed variability is greater than a simple model assumes. It can make standard errors and p-values from an overly restrictive model misleading. The listed primary endpoint is binary rather than a count of repeated events, and the record does not report a count model for it. For the continuous employment-duration outcomes, the posted standard deviations are substantial relative to their means, but that alone does not establish a particular distribution or prove over-dispersion. Participant-level data and model details would be needed to assess distributional assumptions more fully.

What is the role of the standardized effect?

In general, Cohen’s d expresses a difference in means relative to a pooled standard deviation. It puts a continuous-outcome difference on a scale that can be compared across outcomes measured in different units, but it does not replace the original-unit difference. The registry reports Cohen’s d as .93 for weeks competitively employed. The recalculation also reports standardized effect summaries for the continuous outcomes, while the posted mean differences retain the practical units of weeks, days, or hours. Standardized effects depend on the variability in the sample and should be interpreted together with the original outcome units.

All recalculated comparisons are unadjusted. The registry record does not report a comprehensive analysis plan describing covariate adjustment, handling of multiple outcomes, missing-data methods, or sensitivity analyses. Statistical methods can clarify what the posted comparisons represent, but they cannot fill in those unreported design and analysis details.

11. Confidence Intervals and Precision

The recalculated primary risk-difference interval is based on the posted event counts and group sizes, and its lower and upper limits are both on the positive side of zero. This indicates that, under the unadjusted calculation, the estimated difference is not close to zero within the interval. The interval nevertheless spans a range of possible effect sizes, so the observed estimate should not be treated as an exact measure of the difference that would occur in another sample. The registry’s reported chi-squared result does not include a confidence interval or effect estimate.

For weeks competitively employed, the recalculated mean difference is 14.8 weeks, with a 95% confidence interval from 8.042 to 21.56. For days, the recalculated mean difference is 54.5 days, with a 95% confidence interval from 23.9 to 85.1. For hours, the recalculated mean difference is 420.0 hours, with a 95% confidence interval from 171.5 to 668.5. These intervals are unadjusted comparisons calculated from posted summary statistics.

In general, a confidence interval is a range generated by a statistical procedure that, over repeated samples under its assumptions, would cover the true parameter at the stated rate. It is not a range containing a fixed proportion of individual participants’ outcomes, nor does it mean that every value in the interval is equally plausible. Precision depends on the sample size, outcome variability, and calculation method. The registry record does not provide participant-level data, so the recalculated intervals cannot address distributional features or alternative analysis choices beyond those reflected in the summary statistics.

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 for 80% power and participants per arm for 90% power. They are not total sample sizes and are not total event counts. The figures use the observed effect or a stated fraction of that observed effect as the effect to detect. Planning for a smaller fraction of the observed effect requires more participants per arm in the listed illustrations.

The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The figures are calculated from the observed effect; they are not the original trial’s design assumptions. The original sample-size justification and assumptions are not reported in the registry record. The reported illustration method specifies total events for time-to-event outcomes, but the outcomes summarized for these comparisons are not time-to-event outcomes, so the relevant figures are participants per arm.

In general, 80% power means that a study designed under the stated assumptions would have an 80% chance of detecting an effect of the specified size at the chosen significance level, if that effect were the true one and the model assumptions held. The 90% column describes the corresponding planning target with a higher probability of detection, which generally requires more participants for the same target effect. The 80% and 90% figures are therefore planning targets, not descriptions of how likely the completed trial was to find a result.

These values are planning illustrations based on observed results, not observed power. Observed power calculated from an effect estimate after a study is complete does not provide a useful substitute for the effect estimate and confidence interval. The figures should be read as sensitivity to different assumed effect sizes: smaller effects need more participants to detect under otherwise unchanged planning assumptions.

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, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.279; treatment proportion p₁ = 0.762; average p̄ = 0.520.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.7065 = 1.3847.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6185 = 0.5206.
4. n = (1.3847 + 0.5206)² / (0.4828)² = 15.6, rounded up to 16 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
p0 <- 0.27906976744186                          # control proportion
p1 <- 0.761904761904762                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^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 (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(333801)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 16 (80% power), 21 (90% power)
 75% of observed effect: n per arm = 29 (80% power), 38 (90% power)
 50% of observed effect: n per arm = 64 (80% power), 85 (90% power)
power.prop.test: 15.6 
Simulated power at n = 16 per arm: 0.819 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-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 registry’s own primary-outcome analysis is a chi-squared test comparing the two groups, with p < 0.001. The recalculation uses the posted counts to produce an unadjusted risk-difference estimate and confidence interval, and its p-value is also p < 0.001. The record does not flag a discrepancy between these results.

The sponsor analysis and the recalculation do not report precisely the same statistical quantities. The registry’s primary analysis reports the test method and p-value, but not an effect estimate or confidence interval. The recalculation adds an unadjusted effect estimate and interval based on the posted summary counts. It is useful for describing the size and precision of the observed comparison, but it should not be mistaken for an analysis reported by the sponsor or for an adjusted model.

For weeks competitively employed, the registry reports a Wilcoxon (Mann-Whitney) test, Cohen’s d of .93, and p < 0.001. The recalculated comparison of means is a different summary-statistic analysis. No sponsor-reported analyses are listed for the days and hours outcomes. Where a sponsor analysis is reported, it remains the primary account of the registry’s own analysis; recalculations add context rather than replace it.

15. Multiplicity, Missing Data and Analysis Populations

The primary endpoint is identified as obtaining competitive employment, and the analysis population is described as all randomized participants under an intent-to-treat approach. The registry record reports a chi-squared analysis for this endpoint. It does not report a full statistical analysis plan explaining whether the primary outcome was prespecified in a separate protocol, what other analyses were planned, or whether models adjusted for baseline characteristics.

The registry lists several secondary outcome areas and multiple employment measures. It does not report a multiplicity strategy, such as a testing hierarchy or adjustment across outcomes. In general, testing many outcomes without accounting for multiplicity can increase the chance that at least one result appears statistically significant by chance. That general concern does not determine whether any particular reported result is false; it means secondary p-values should be interpreted in the context of the number of outcomes and the absence of a reported correction plan.

Missing-data handling is not reported in the registry record. Although the primary analysis is described as intent-to-treat, the record does not explain whether every randomized participant had an observed primary outcome, how missing employment information was treated, or whether imputation or sensitivity analyses were used. The primary arm counts are 43 and 42, while enrollment is listed as 100; the reason for that difference is not reported in the registry record. No inference about the reasons for the counts or the effect of missing data can be made from those figures alone.

The record does not report interim analyses, stopping rules, or sensitivity analyses. It also does not provide enough information to evaluate whether outcome definitions and analysis choices were finalized before results were examined. These omissions limit assessment of the risk that analysis decisions, incomplete follow-up, or multiple testing influenced the reported estimates.

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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: suicidal ideation (11.6%), surgical procedure (2.4%), severe staph infection (2.3%), pneumonia (2.3%), motor vehicle accident complications (2.3%).

The registry lists serious adverse events affecting 11 of 43 participants in the VA Vocational Rehabilitation Program group and 5 of 42 participants in the IPS group. It lists other adverse events affecting 7 of 43 participants in the VA group and 2 of 42 participants in the IPS group. These are counts of affected participants as recorded, not measures of the severity, duration, or causal relationship of every event.

The listed serious-event terms include suicidal ideation at 11.6%, surgical procedure at 2.4%, and severe staph infection, pneumonia, motor vehicle accident complications, and medical stabilization at 2.3% each. The registry summary does not connect these listed percentages to particular arms in the information, so they should not be assigned to one group. Deaths are not reported in the registry record.

Adverse-event counts in a trial provide descriptive safety information, but the comparison alone does not establish that an intervention caused or prevented an event. The record does not provide event-level timing, duration, relatedness assessments, or a detailed account of how events were collected. Differences in the counts should therefore be read as a description of what was recorded, not as a causal safety conclusion.

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

The posted primary result describes a difference between randomized groups in the proportion of participants recorded as obtaining competitive employment during one calendar year. More participants in the IPS group than in the VRP group met the registry’s definition in the posted counts. The reported chi-squared p-value is p < 0.001, and the unadjusted recalculation yields a confidence interval for the risk difference that is above zero.

This finding does not mean that every participant assigned to IPS obtained work, or that a participant assigned to VRP could not obtain work. The endpoint is a group-level comparison, not a prediction for an individual. Nor does obtaining any qualifying competitive job establish how long the job lasted, how many hours were worked, what income was earned, or how the participant experienced the work. The secondary employment measures address some of those other dimensions, but they remain distinct outcomes.

The result does not directly answer whether PTSD symptoms, depression, disability, or broader health outcomes changed. Those topics appear among the registry’s secondary outcome areas, but complete corresponding results are not reported in the information available for this analysis. The employment endpoint should not be treated as a substitute for those outcomes.

In general, randomization supports a comparison between assigned groups, but interpretation still depends on trial conduct, outcome measurement, follow-up, and analysis. The registry record does not report enough detail about missing data, sites, participant characteristics, or intervention delivery to resolve all questions about the setting and applicability of the results. The posted findings describe this trial’s comparison; they do not establish what will happen in another program or population and do not constitute medical or vocational advice.

18. Why This Trial Matters Statistically

This trial illustrates how a binary primary outcome can be paired with measures of employment duration. The primary endpoint asks whether a participant obtained qualifying competitive employment during the specified period. Weeks, days, and hours employed ask related questions about the amount of work. Keeping those outcome definitions distinct is important: a difference on one measure does not automatically describe all aspects of employment.

The registry’s reported primary analysis gives a chi-squared p-value, while the recalculation adds an unadjusted risk difference and confidence interval from the posted counts. That contrast demonstrates why statistical significance and effect size answer different questions. The p-value describes evidence against a no-difference model under the test assumptions; the effect estimate and interval describe the direction, scale, and uncertainty of the group comparison.

The trial also illustrates the importance of distinguishing the sponsor’s reported analysis from a recalculation. The registry reports a rank-based analysis for weeks employed, while the recalculation compares means using summary statistics. Those approaches do not necessarily estimate or test the same feature of the data. Clear labeling helps prevent one result from being mistaken for the other.

Finally, the registry’s secondary employment results sit alongside several other outcome areas, and no multiplicity plan is reported. This makes the primary-versus-secondary distinction statistically relevant. The evidence is best read with attention to the endpoint, method, uncertainty, and what the registry does not document, rather than relying on a p-value alone.

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.

The primary endpoint is based on a specific definition of competitive employment. It excludes day labor and sheltered or set-aside jobs, so the result does not describe every form of work or vocational participation. It records whether qualifying employment was obtained during one calendar year, not job retention, job quality, hours worked by each participant, or income.

The registry lists enrollment of 100 but gives primary-outcome analysis counts of 43 in the VRP group and 42 in the IPS group. The reason for this difference is not reported in the registry record. Although the primary outcome is described as intent-to-treat, the record does not explain missing-outcome handling or provide participant flow by arm. These gaps limit assessment of how complete the analysis was and whether missing information could affect interpretation.

Masking is listed as none. The registry record does not report how the outcome was assessed or whether employment information was verified independently. It also does not provide detailed baseline characteristics, recruitment sites, or enough information to determine how closely the enrolled participants represent other veteran populations with PTSD.

The intervention descriptions do not contain complete text, and the record does not provide detailed information about implementation or adherence. The original sample-size justification, full statistical analysis plan, multiplicity procedures, and interim analyses are not reported. These limitations matter when considering precision, secondary outcomes, and the extent to which the posted analyses capture the trial’s original plan.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: The Impact of Vocational Rehabilitation on Mentally III Veterans — NCT00333801.
• Registry record last updated 2024-02-20; status last verified 2024-01.
• Results first posted 2015-03-19.
• 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 trial sponsored by the VA Office of Research and Development, comparing IPS supported employment with the VA Vocational Rehabilitation Program among veterans with PTSD. The primary endpoint was obtaining competitive employment during one calendar year, analyzed as an intent-to-treat outcome.

The primary counts were 12 of 43 participants in the VRP group and 32 of 42 in the IPS group. The registry reports a chi-squared p-value of p < 0.001. Recalculated unadjusted summaries for weeks, days, and hours competitively employed also favor higher reported means in the IPS group, while the registry reports a Wilcoxon analysis for weeks.

The findings are limited by details not reported in the registry record, including missing-data handling, the original sample-size justification, and a multiplicity plan. The posted results describe employment outcomes in this trial and should be interpreted in light of those reporting limits.