Efficacy and Safety of Temazepam in HIV Seropositive Patients With Insomnia: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Efficacy and Safety of Temazepam in HIV Seropositive Patients With Insomnia (NCT02153788), a randomized phase 4 study. The primary outcome is Mean Change in the Insomnia Severity Index.

2. Clinical Question

The trial record identifies insomnia as the condition and describes participants as HIV seropositive. The endpoint is a self-reported questionnaire score, not an objective sleep measurement. Because higher Insomnia Severity Index scores indicate greater insomnia severity, a negative change represents a reduction in score from the starting assessment. The reported contrast is between mean changes, not between final scores alone. The record does not provide baseline and final score distributions, details of adherence, or a fuller description of the people who contributed outcome data. Those absences constrain interpretation of the mean comparison.

3. Trial Design

The registry record describes a completed phase 4 randomized, triple-masked, parallel-group, placebo-controlled study sponsored by Duke University. The recorded start was 2009-09, and primary completion was 2010-09. Enrollment is listed as 44. The record does not report the trial's original sample-size justification or the assumptions used in its design. The planning illustration based on the observed effect is a separate calculation and should not be mistaken for the original design plan.

The arm descriptions include a screening period followed by a single-blind study-drug period. Participants meeting entry criteria were asked to complete a daily sleep diary and return one week later. The participants randomized to temazepam were then assigned temazepam 15 mg for 12 weeks; those randomized to placebo were assigned placebo for 12 weeks. The brief summary describes nightly dosing approximately 30 minutes before bedtime and an approximately 12-week treatment period. The primary endpoint timeframe is randomization to the final study visit, approximately 12 weeks.

Randomized allocation means assignment to the study arms was intended to be determined by a random process. A parallel-group design means participants were assigned to an arm rather than crossing between the two assigned treatments. Masking is recorded as triple, but the registry record does not identify which groups of people were masked or give details of how masking was maintained. It also does not report the randomization procedure, allocation concealment, treatment adherence, or handling of protocol deviations.

The registry outcome results list 23 participants for the temazepam arm and 18 for the placebo arm for the primary measure. These are the reported arm-level outcome counts; the record does not explain the difference between the enrollment figure and the number reported for each arm. The posted result therefore describes the available arm summaries, while the reasons that some enrolled participants may not appear in those summaries are not reported in the registry record.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Overall Study

Reasons for not completing this period:

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 Mean Change in the Insomnia Severity Index, measured over Randomization to final study visit, approximately 12 weeks.

The primary endpoint was mean change in the total Insomnia Severity Index score from randomization to the final study visit, approximately 12 weeks. The Insomnia Severity Index is a seven-item self-report questionnaire that covers sleep initiation, sleep maintenance, satisfaction or distress about sleep problems, and daytime dysfunction. Responses are summed to form a total score. The registry record states that the score ranges from 0 to 28 and that higher scores indicate greater insomnia severity.

A change score compares a participant's score at follow-up with that participant's score at the starting assessment. The posted arm means describe average changes within each assigned group; the comparison of interest is the difference between those average changes. A mean summarizes the group's average direction and magnitude, but it does not show how variable individual responses were or whether the same pattern occurred across participants. The standard deviations reported for each arm give information about spread around the mean.

This endpoint was selected to represent several dimensions of perceived insomnia and its daytime consequences in a single score. Because it is self-reported, it reflects participants' questionnaire responses rather than a direct measurement of sleep duration or sleep physiology. It is therefore important to interpret it as a change in reported insomnia severity, not as a measurement of every aspect of sleep. The registry record does not report a threshold for a clinically important change or another interpretation rule that would translate a given score difference into a categorical clinical judgment.

7. Results Posted in the Registry

• Temazepam: mean -4.01 versus 0.72 on control; difference -4.73 (95% CI -8.28 to -1.18), standardized effect d = -0.84; z-test p = 0.009.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary measure is continuous: each participant has a score, and the result is summarized using an average and a measure of spread. The registry record reports the mean change and standard deviation for each arm. A mean change is calculated from individual changes, while the mean difference compares the arm-level averages. In this record, the reported mean difference is -4.73 for temazepam relative to placebo. Because higher scores indicate greater insomnia severity, the sign of the contrast indicates the direction of the difference in average change.

What does a standard deviation describe?

A standard deviation describes how dispersed individual observations are around an average. It is not the uncertainty of the average itself. The posted standard deviations are 5.3 for temazepam and 6.1 for placebo, while the reported comparison also includes a pooled standard deviation of 5.663. Pooling is a way of combining information about spread across groups for some summaries. It can be useful for describing a standardized contrast, but its interpretation depends on the assumptions made about the groups and variability. The registry record does not report a detailed statistical analysis plan or state the specific model used for the comparison.

How should the standardized effect be read?

The comparison reports a standardized effect, d, of -0.835. A standardized effect expresses a difference relative to a measure of variability, so it places the mean contrast on a scale that is less dependent on the original questionnaire units. The negative sign retains the direction of the contrast. This standardized summary is not a separate clinical endpoint, and it does not show how many participants experienced a particular degree of change. Its magnitude is sensitive to the variability used in the standardization, which is why the original mean difference and its units remain important.

What does the p-value test?

The reported p-value is unadjusted and equals 0.009. In general, a p-value describes how often a result at least as incompatible with a null hypothesis as the observed result would arise if that null hypothesis and the analysis assumptions held. It is not an effect-size measure and does not quantify the importance of the score change. An unadjusted p-value is not corrected for considering multiple outcomes or analyses. The registry record lists secondary outcomes but does not report a multiplicity strategy or provide analyses of those outcomes.

The confidence interval gives a range of values compatible with the observed data under the method and assumptions used. Its reported limits are -8.284 and -1.176. In general, a confidence interval communicates uncertainty around an estimate; it is not a range that contains every possible participant response. The width reflects the information and variability used in the calculation, as well as the method. Since the registry record does not describe the inferential method, readers cannot verify the exact assumptions behind the interval from the registry entry alone.

Why does the analysis method matter?

Different methods can address different features of trial data. For a continuous outcome, a method may compare group means, adjust for starting values, or account for other prespecified variables. Repeated measurements may require a method that represents the relationship among measurements from the same participant. These are general possibilities; the registry record does not say that any particular adjustment, repeated-measures method, or covariate model was used. The posted summary alone does not reveal whether the reported comparison was prespecified or how it relates to a complete protocol analysis.

Change scores are understandable because they express movement from a participant's starting score. However, a comparison of change scores and a comparison of follow-up scores adjusted for baseline are not automatically identical. Their estimates may depend on starting values, variability, and the analysis model. Baseline arm summaries and the model specification are not reported in the registry record. It is therefore most precise to describe the posted result as the reported comparison of mean changes, without assigning it an undocumented model or adjustment.

The outcome summaries include arm counts and standard deviations, but the record does not identify whether the analysis used all randomized participants, participants with observed follow-up, or another defined analysis population. In general, missing outcome data can affect a group comparison if the availability of measurements is related to participant characteristics or outcome values. Statistical methods for missing data rely on assumptions; those assumptions and any sensitivity analyses are not reported in the registry record. The reasons for missing or unreported outcomes are also not reported.

The registry lists six secondary outcomes, including sleep time, fatigue, anxiety, distress, and depression measures. Considering several outcomes can raise the chance that at least one comparison appears notable by chance if each is tested separately. In general, multiplicity methods can adjust interpretation or testing when multiple comparisons are planned. No such method, testing hierarchy, or analysis plan is reported in the registry record. The primary endpoint result can be described as posted, but the available entry does not fully document the wider inferential framework.

Finally, the record's reported comparison should not be mistaken for a complete reproducible analysis. The arm means, standard deviations, counts, mean difference, confidence interval, and unadjusted p-value provide useful summary information. The calculation procedure, data-handling decisions, and analysis populations are not reported in the registry record. General statistical interpretation can explain what these summaries mean, but cannot fill in those trial-specific details.

10. Confidence Intervals and Precision

The reported mean difference is -4.73, with a 95% confidence interval from -8.284 to -1.176. The interval describes uncertainty around the estimated contrast in mean change. Every value in the reported interval is below zero, which aligns with a lower average change in the temazepam arm relative to placebo under the posted comparison. The interval does not mean that individual participants' changes fall within those limits; it concerns the estimated group contrast.

In general, a confidence interval is a range generated by a method that, across repeated use under its assumptions, has a stated coverage property. It communicates both the estimate and its precision. A narrower interval indicates less statistical uncertainty under the method, while a wider interval indicates more. The width in this trial must be interpreted in light of the observed variability and the outcome counts. The registry record does not describe the method used to construct the interval, so the assumptions that determine its exact coverage cannot be checked from the entry.

The unadjusted p-value of 0.009 is another summary of uncertainty under a null hypothesis, but it does not replace the interval. The interval shows a range and direction for the estimated mean contrast, whereas the p-value gives a measure of compatibility with a null hypothesis under the analysis assumptions. Neither statistic describes the distribution of participant-level responses, establishes clinical importance, or accounts automatically for unreported missing-data and multiplicity decisions.

Precision also depends on what question is being asked. These statistics concern mean change in the Insomnia Severity Index over the trial's stated follow-up, not longer-term change or other outcomes. The registry record does not report a threshold for a clinically important difference, so the interval cannot be classified against such a threshold using this record. It is best understood as uncertainty around the posted average contrast, with the caveat that the underlying analysis details are not fully reported.

11. 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 based on fractions of the observed effect, not the trial's original design assumptions. For this continuous outcome, the figures are participants per arm, not total events. The calculation assumptions listed for the illustration 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. The original trial's sample-size justification and design assumptions are not reported in the registry record.

The 80% power column gives the illustrative number of participants per arm for a design with an 80% chance of detecting the assumed effect under the calculation assumptions if that effect is the true one. The 90% power column uses a higher target chance of detection under the same general framework. These are prospective planning quantities: they describe what a hypothetical study might require under specified assumptions, not what happened in this completed trial.

The rows that plan for 100% of the observed effect, 75% of the observed effect, and 50% of the observed effect illustrate how the assumed effect changes the planning requirement. Smaller effects are harder to distinguish from random variation, so they need more participants to reach the same target power under otherwise comparable assumptions. The listed figures should therefore be read as scenario-based illustrations rather than a recommendation for a future trial; design decisions would also depend on the intended analysis and other assumptions.

This is not observed power. It is a planning illustration derived from the observed effect, and it does not measure the probability that the completed study correctly detected an effect. The accompanying curve should be read in the same way: it reflects planning under the stated assumptions, rather than a new result from the trial.

12. 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 = -4.730; pooled standard deviation σ = 5.663.
2. n = 2 × (1.96 + 0.842)² × 5.663² / -4.730² = 22.5, rounded up to 23 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 <- -4.73                      # observed difference in means
sigma <- 5.66263053848515                     # 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(2153788)
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 = 23 (80% power), 31 (90% power)
 75% of observed effect: n per arm = 40 (80% power), 54 (90% power)
 50% of observed effect: n per arm = 90 (80% power), 121 (90% power)
power.t.test: 23.5 
Simulated power at n = 23 per arm: 0.789 (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.

13. Multiplicity, Missing Data and Analysis Populations

The registry record lists six secondary outcomes in addition to the primary Insomnia Severity Index endpoint. These include mean change in self-reported total sleep time, Piper Fatigue Score, anxiety and depression measures from the Hospital Anxiety and Depression Scale, and the Distress Thermometer. Mean change in self-reported total sleep time appears twice in the listed outcome names. The record counts 7 outcome measures and 0 analyses, but it does not provide a set of posted secondary analysis results.

When multiple outcomes or comparisons are considered, the chance of at least one apparently unusual finding can increase if each is tested separately. In general, an analysis plan may specify a primary outcome, a testing order, or an adjustment for multiple comparisons. The registry record does not report whether such a plan was used, whether any multiplicity adjustment was applied, or whether the listed secondary measures were analyzed. The primary comparison's p-value is explicitly unadjusted, so it should not be read as correcting for consideration of other endpoints.

The record also does not report how missing observations were handled. It does not say whether participants with incomplete endpoint data were included in a particular analysis, whether values were imputed, or whether sensitivity analyses were performed. The primary result has arm-level counts of 23 and 18, but the registry does not explain how those counts relate to the enrollment figure or identify the reasons for incomplete or unreported data.

The analysis populations are not reported in the registry record. In general, trials may define populations according to randomization, treatment received, or availability of outcome measurements, and those definitions can affect the interpretation of estimates. No such definition should be inferred from the arm summaries alone. The registry record also does not report an analysis plan, protocol deviations, or a description of interim analyses; whether interim analyses occurred is not reported in the registry record. These gaps limit assessment of how outcomes were selected, how incomplete data affected estimates, and how the posted primary comparison fits into the full analysis.

14. 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 non-serious adverse event terms

The serious adverse event summary reports 0 participants affected in the temazepam arm among 23 at risk and 0 affected in the placebo arm among 18 at risk. These are the registry's reported arm-level figures for serious adverse events. The record lists no serious adverse event terms. A zero count in this summary means no serious adverse events are recorded as affecting participants in the posted data; it does not establish that the risk of serious events is absent.

For other adverse events, the summary lists 4 affected in the temazepam arm among 23 at risk and 2 affected in the placebo arm among 18 at risk. These counts are descriptive and should not be treated as a complete comparison of safety without information about event definitions, severity, timing, duration, relatedness, and completeness of follow-up. The registry record does not provide those details in the safety summary.

Deaths are not reported in the registry record. The number of participants at risk is given separately for each arm, and the figures should not be combined or converted into rates not reported in the record. Safety counts also need to be interpreted in the context of the study's limited reported arm sizes and the absence of detailed event listings. They describe what is recorded in this registry entry, not a definitive estimate of uncommon or delayed risks.

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

The result describes a difference in average change in self-reported insomnia severity between the randomized temazepam and placebo arms over approximately 12 weeks. The posted mean changes were -4.01 and 0.72, respectively, and the reported mean difference was -4.73. The confidence interval, from -8.284 to -1.176, and the unadjusted p-value of 0.009 describe statistical uncertainty under the analysis represented by the registry comparison. They do not show what happened to every participant or establish the importance of the contrast for an individual.

The result does not establish that all participants experienced a reduction in insomnia severity, nor does it describe the distribution of individual changes. A group mean can coexist with varied responses, but participant-level results are not reported in the registry record. The endpoint is a questionnaire score, so the finding concerns reported insomnia severity rather than every possible measure of sleep. The registry entry does not provide an interpretation threshold for clinically important change.

The finding also does not establish persistence beyond the recorded study period. The primary endpoint timeframe is randomization to the final study visit, approximately 12 weeks; follow-up beyond that period is not reported in the registry record. Nor does the result establish findings for the six listed secondary outcomes, because their analysis results are not posted in this record.

Statistical interpretation is constrained by missing documentation. The registry record does not report the analysis population, the handling of missing outcome data, the inferential method, or any multiplicity plan. These details matter when judging how robustly the posted comparison represents the randomized groups and how the primary finding relates to other measured outcomes. The result is evidence about the reported comparison in this trial, not a complete account of all possible outcomes or settings.

In general, a randomized placebo-controlled design can support a comparison between assigned groups when allocation and outcome follow-up are carried out as intended. The registry record does not supply enough detail to evaluate all such conditions. The statistics should therefore be stated as posted and interpreted with their uncertainty and reporting limits, without extending them into a treatment recommendation or a claim about broader populations.

16. Why This Trial Matters Statistically

This trial is statistically informative because it reports arm-level mean changes, standard deviations, group counts, a mean difference, a confidence interval, and an unadjusted p-value for a patient-reported continuous outcome. Reading those measures together helps distinguish the estimated size and direction of a group contrast from the uncertainty around it. The reported interval stays below zero, while also showing a range of plausible values under the method used.

The comparison also illustrates why a p-value should not be read in isolation. The p-value of 0.009 indicates compatibility with a null hypothesis under specified assumptions, but it does not quantify the magnitude or practical meaning of the score change. The mean difference of -4.73 gives the contrast in the questionnaire's units, and the confidence interval communicates uncertainty around that estimate. The standard deviations show that individual scores vary around the arm means, though participant-level patterns are not available.

The record highlights the importance of registry completeness. The endpoint and outcome summaries are clear enough to describe the reported comparison, yet important interpretive details are not reported: the analysis population, missing-data handling, inferential method, and multiplicity strategy. The six listed secondary outcomes have no posted analysis results in the record. These omissions make it difficult to assess the full outcome picture or reproduce the analysis.

Finally, the planning illustrations demonstrate a general relationship between effect assumptions and sample size. Planning for a smaller fraction of the observed effect requires more participants per arm under the stated assumptions. That is useful for understanding statistical design, but it is separate from the original study's design rationale, which is not reported. Distinguishing planning calculations from trial findings prevents the illustrative figures from being mistaken for evidence about the completed study's power.

17. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

Several limitations specific to this registry record affect interpretation. The reported primary outcome counts are 23 for temazepam and 18 for placebo, while enrollment is listed as 44. The record does not explain the relationship between those figures, give a participant flow account, or state why outcome data are not reported for every enrolled participant. It is not possible from the entry to determine whether the endpoint summaries include all randomized participants or another analysis population.

The registry entry provides no baseline score summaries, final score summaries, or individual-level outcome distributions. As a result, the posted mean changes cannot be independently examined for starting-score differences, unusual observations, or variation in response. The outcome is self-reported, and the record does not report an interpretation threshold for clinically important change. It also does not report adherence or other information that would clarify exposure to assigned study drug.

The statistical analysis is incompletely documented. The exact inferential method, any covariate adjustment, missing-data approach, and multiplicity plan are not reported in the registry record. The comparison has a confidence interval and unadjusted p-value, but without the analysis specification its assumptions cannot be fully assessed. No analyses are posted for the listed secondary outcomes.

The safety summary is limited to counts of affected participants. Detailed event descriptions, timing, severity, relatedness, and follow-up are not reported in the registry record; deaths are also not reported. The primary endpoint covers approximately 12 weeks, and longer-term outcome information is not reported. These limitations do not erase the posted comparison, but they restrict what can be concluded about outcome completeness, participant-level variation, broader safety characterization, and durability.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Efficacy and Safety of Temazepam in HIV Seropositive Patients With Insomnia — NCT02153788.
• Registry record last updated 2015-03-10; status last verified 2015-01.
• Results first posted 2015-03-10.
• 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.

21. Record Summary

The record describes a completed phase 4 randomized, triple-masked, parallel, placebo-controlled study sponsored by Duke University. It lists enrollment of 44, a recorded start of 2009-09, and primary completion of 2010-09. The intervention arm received temazepam 15 mg for 12 weeks after an initial period, and the comparator arm received placebo.

The primary endpoint was mean change in Insomnia Severity Index score from randomization to the final study visit, approximately 12 weeks. Posted mean changes were -4.01 for temazepam and 0.72 for placebo. The reported mean difference was -4.73, with a 95% confidence interval from -8.284 to -1.176 and an unadjusted p-value of 0.009. Six secondary outcomes are listed, but their analysis results are not reported in the registry record.

The safety summary records no serious adverse events in either arm and lists other adverse events affecting 4 participants in the temazepam arm and 2 in the placebo arm. Deaths are not reported in the registry record. Analysis populations, missing-data handling, the inferential method, and multiplicity procedures are not reported, limiting the detail available for evaluating the posted estimates.