Evaluating the Effectiveness of Escitalopram to Prevent Depression in Head and Neck Cancer Patients Receiving Treatment: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Evaluating the Effectiveness of Escitalopram to Prevent Depression in Head and Neck Cancer Patients Receiving Treatment (NCT00536172), a randomized phase 4 study. The primary outcome is Depression as Assessed by the Quick Inventory of Depressive Symptomatology-Self Rated 16 (QIDS-SR-16).

2. Clinical Question

3. Trial Design

This was a randomized, parallel-group Phase 4 trial comparing escitalopram with placebo. The registry record lists quadruple masking, meaning that the design is recorded as masked at four levels; it does not identify those levels in the information summarized for this trial. The sponsor was the University of Nebraska. The record lists the study as completed, with a start date of 2007-12-01 and a primary completion date of 2012-06-01.

Overall enrollment was 160. For the primary outcome, the reported arm denominators were 60 for escitalopram and 65 for placebo, in an evaluable-subject population. Those outcome denominators do not add up to the overall enrollment, and the registry record does not report why they differ or how participants outside the outcome denominators were handled. The adverse-event summaries use 74 participants at risk in each arm; the record does not explain how these safety denominators relate to the enrollment or primary-outcome denominators.

The primary outcome was a count of participants reaching the defined QIDS-SR-16 threshold, assessed at several scheduled times from pre-treatment through Week 28. Although the record lists one outcome measure, it does not make clear which of those assessment times is represented by the posted group counts and comparison. No secondary outcome measures or analyses are posted in the registry record. The trial's original sample-size justification, detailed statistical analysis plan, and any interim-analysis plan are not reported in the registry record.

4. Participant Flow and Accounting

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

Before assignment: 12 of the 160 subjects who agreed to participate were not eligible: 9 QIDS or QIDS-C \>/=11; 1 not able/willing to return for follow-up; 1 received antidepressants in the past week; 1 not diagnosed with epidermoid cancer and meets MINI criteria for depression. The remaining 148 subjects were randomized.

Overall Study

Analysis population for the primary outcome, as defined in the registry: Evaluable subjects

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 Depression as Assessed by the Quick Inventory of Depressive Symptomatology-Self Rated 16 (QIDS-SR-16), measured over Measured pre-treatment and at Weeks 2, 4, 6, 8, 10, 12, 16, 20, 24, and 28.

The primary endpoint counts participants reaching a pre-defined threshold on the Quick Inventory of Depressive Symptomatology-Self Rated 16 (QIDS-SR-16). The specified threshold is a score >=11. The outcome description says that scores from 0 to 10 correspond with no to mild depression, while scores >=11 correspond with moderate to severe depression. The total QIDS-SR-16 score ranges from 0-27.

In general, converting a symptom score into a threshold-based outcome gives a binary result: each participant either reaches the stated threshold or does not. That can make the endpoint a direct way to describe how many participants crossed the defined symptom boundary, while not preserving all the detail in the underlying score. It does not show the size of score changes among participants who remain on the same side of the threshold.

The registry record lists assessments before treatment and at Weeks 2, 4, 6, 8, 10, 12, 16, 20, 24, and 28. However, it does not identify which assessment occasion generated the posted counts, or explain whether the counts represent a particular time point or another summary across assessments. That limits interpretation of the timing represented by the primary result.

7. Results Posted in the Registry

• Escitalopram: 10.0% (6/60) versus 24.6% (16/65) on control; risk difference -14.6 percentage points (95% CI -27.5 to -1.7), risk ratio 0.41 (0.17–0.97); unadjusted two-proportion z-test p = 0.032.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is binary: participants are counted according to whether they reach the stated QIDS-SR-16 threshold. For a binary outcome, each arm's count is interpreted in relation to the number of participants in that arm with outcome data. The registry lists 6 participants reaching the threshold among 60 in the escitalopram arm and 16 among 65 in the placebo arm. These counts and denominators are the basis for the posted group comparison. The record describes the population as evaluable subjects.

What does the risk difference express?

The risk difference is an absolute contrast between the proportions with the outcome in the two groups. The reported estimate is -0.146 for escitalopram compared with placebo. Its negative sign indicates that the observed proportion was lower in the escitalopram arm. An absolute contrast is expressed on the probability scale, so it is distinct from a ratio measure. No additional percentage or converted estimate is needed to interpret the reported value.

How do the risk ratio and odds ratio differ?

The risk ratio compares the outcome proportions in the two groups as a ratio. The posted estimate is 0.406. The odds ratio compares the odds of reaching the threshold rather than comparing the proportions directly; its posted estimate is 0.340. Both estimates point in the same direction as the risk difference, but their scales differ. In general, odds ratios and risk ratios are not interchangeable, and an odds ratio should not be described as a percentage-point difference.

What does the p-value add?

The unadjusted p-value is 0.032. In general, a p-value evaluates how surprising the observed data would be under a specified null hypothesis and statistical model. It is not a probability that the treatment effect is absent, and it does not measure the size of the effect. The word unadjusted indicates that the posted p-value is not presented as adjusted for other factors. The registry record does not report further details of the model or analysis procedure behind it.

Why use an interval estimate?

The reported interval for the risk difference extends from -0.276 to -0.017. In general, an interval estimate communicates the uncertainty associated with an estimate under the method used to construct it. The record does not state the interval's confidence level in the information summarized for this trial, so no confidence level should be inferred from the registry summary. The interval provides a range of values associated with the estimate, rather than a guarantee that the true effect lies within it.

The risk-ratio interval is 0.170 to 0.970, and the odds-ratio interval is 0.123 to 0.939. These intervals reflect uncertainty for their respective measures and should be read on their own scales. In general, intervals that cover a wider range indicate less precision than narrower intervals, but comparisons of width across different measures need care because the measures use different scales.

Why can over-dispersion matter?

Over-dispersion is a general concern when observed variation exceeds what a statistical model assumes. For binary outcomes, the assumed variability depends on the model and design. If that variability is understated, standard errors and interval estimates may be too small, which can make results appear more precise than warranted. The registry record does not report an over-dispersion assessment or a model that addresses it. This is a general statistical consideration, not a claim that over-dispersion occurred in this trial.

The posted comparison is an unadjusted group contrast, and the record does not report covariate-adjusted estimates. Randomization is recorded, but the registry summary does not provide further details about allocation procedures, analysis populations beyond evaluable subjects, or handling of missing outcomes. It also lists assessments at multiple occasions without specifying which one underlies the group counts. These details matter when judging the assumptions and scope of any statistical result.

10. Confidence Intervals and Precision

The estimated risk difference is -0.146, with a reported interval from -0.276 to -0.017. This interval gives a range of values around the estimate and indicates that the result is not known with exact precision. The registry record does not report the confidence level associated with this interval in the information summarized for the trial; therefore, its confidence level should not be assumed.

The interval is entirely on the negative side of zero. In general, an interval for a difference that excludes zero is consistent with a nonzero contrast under the interval method used. That observation does not describe the clinical importance of the contrast, establish the cause of the observed difference, or resolve the reporting limitations. The posted unadjusted p-value is 0.032, which provides a separate summary of compatibility with a null hypothesis under the analysis assumptions.

The risk ratio estimate of 0.406 has an interval from 0.170 to 0.970. The odds ratio estimate of 0.340 has an interval from 0.123 to 0.939. Each interval is on a different scale, so the values should not be compared directly with the risk-difference interval. Taken together, the estimates describe a contrast in the same direction, but uncertainty remains about its magnitude.

Precision also depends on what outcome and population the estimates represent. The result uses evaluable-subject denominators of 60 and 65, while overall enrollment was 160. The registry record does not explain the difference, identify the assessment occasion represented, or report how missing data were handled. These gaps limit the precision with which readers can interpret the result's scope, even though numerical intervals are 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 planning figures are participants per arm, not total participants, because the outcome is not a time-to-event outcome. The illustration assumes a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. It uses the observed effect as a planning reference and evaluates scenarios targeting 100%, 75%, or 50% of that observed effect. These are planning assumptions for an illustration; the trial's original design assumptions and sample-size justification are not reported in the registry record.

For planning around 100% of the observed effect, the figures are 104 participants per arm for 80% power and 139 participants per arm for 90% power. For 75% of the observed effect, they are 201 participants per arm for 80% power and 269 participants per arm for 90% power. For 50% of the observed effect, they are 486 participants per arm for 80% power and 650 participants per arm for 90% power.

In general, power is the probability that a planned study will detect an effect of the specified size under the assumed model and conditions. The 80% and 90% columns represent different target probabilities of detecting the chosen effect, not results measured in the completed trial. Smaller effects require more participants to detect under otherwise similar planning assumptions. These planning calculations are not observed power.

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

Worked calculation for this trial

1. Control proportion p₀ = 0.246; treatment proportion p₁ = 0.100; average p̄ = 0.173.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.5350 = 1.0486.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5249 = 0.4418.
4. n = (1.0486 + 0.4418)² / (-0.1462)² = 104.0, rounded up to 104 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 <- 0.246153846153846                          # control proportion
p1 <- 0.1                          # 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(536172)
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 = 104 (80% power), 139 (90% power)
 75% of observed effect: n per arm = 201 (80% power), 269 (90% power)
 50% of observed effect: n per arm = 486 (80% power), 650 (90% power)
power.prop.test: 104 
Simulated power at n = 104 per arm: 0.808 (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.

13. Multiplicity, Missing Data and Analysis Populations

The primary outcome was scheduled for assessment before treatment and at Weeks 2, 4, 6, 8, 10, 12, 16, 20, 24, and 28. Multiple assessments create a general multiplicity question when several time points or comparisons are examined: the chance of at least one apparently unusual result can change as the number of tests increases. The registry record does not identify which assessment occasion corresponds to the posted primary comparison, whether the occasions were analyzed separately, or whether any multiplicity adjustment was planned or used.

One outcome measure is listed, and no secondary outcome measures or analyses are posted. The registry record reports zero analyses. This does not establish that no additional analyses were conducted; it means that analysis results are not reported in the registry record summarized for this trial. The analysis plan, including any planned hierarchy of outcomes or multiplicity procedure, is not reported in the registry record.

The primary outcome population is described as evaluable subjects, with 60 participants in the escitalopram arm and 65 in the placebo arm. Overall enrollment was 160. The record does not report how evaluability was defined, why the outcome denominators differ from enrollment, or how missing measurements were handled. It also does not report whether participants were analyzed according to their randomized assignment or under another analysis-population definition. Those details should not be inferred from the word randomized alone.

In general, missing outcome data can affect an estimate if the availability of measurements is related to participant characteristics or outcomes. The direction and size of any such effect cannot be assessed from the posted information without details about missingness and the analysis approach. The record does not report imputation methods, sensitivity analyses, or a missing-data assumption.

14. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

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: dehydration (5.4%), nausea and vomiting (2.7%), hyponatremia (2.7%), atrial fibrillation (2.7%), chest pain (2.7%).

The serious adverse-event summary lists 74 participants at risk in each arm. In the escitalopram group, 16 were recorded as affected by serious adverse events; in the placebo group, 13 were recorded as affected. The other adverse-event count was 47 for escitalopram and 49 for placebo. Death counts were 16 and 13, respectively. These are separate reported fields; the registry record does not explain whether the categories overlap or provide further detail about their relationship.

The most frequently listed serious-event terms were Dehydration at 5.4%, followed by Nausea and vomiting, Hyponatremia, Atrial Fibrillation, Chest Pain, and Pain, each at 2.7%. The registry summary does not specify the denominator or event-counting convention for these percentages in the information, nor does it provide details about event timing, severity, attribution, or participant-level overlap. The safety denominators of 74 in each group differ from the primary-outcome denominators, and the record does not explain why. These figures should therefore be read as registry-reported safety summaries, not as a complete account of comparative risk.

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

The posted primary outcome result describes fewer participants reaching the QIDS-SR-16 threshold in the escitalopram arm than in the placebo arm among evaluable subjects: 6 of 60 compared with 16 of 65. The reported risk difference is -0.146, with an interval from -0.276 to -0.017, and the unadjusted p-value is 0.032. The risk ratio and odds ratio also indicate a lower observed outcome frequency in the escitalopram arm on their respective scales.

These findings describe the posted comparison; they do not by themselves establish a complete account of the trial's results. The registry record does not identify which scheduled assessment occasion produced the arm counts, despite listing pre-treatment and multiple follow-up assessments through Week 28. The outcome population is evaluable subjects, and the record does not explain why the denominators of 60 and 65 differ from the overall enrollment of 160 or how missing outcome data were addressed.

In general, statistical estimates summarize observed data under an analysis method and its assumptions. They do not prove that an observed difference was caused by the intervention, and a p-value does not quantify the probability that a treatment effect is real. The posted comparison also does not communicate the full distribution of QIDS-SR-16 scores because the primary endpoint counts threshold crossings rather than reporting score changes for each participant.

The record does not post secondary outcome results or analyses, details of covariate adjustment, or the original sample-size justification. The safety summary contains serious adverse-event counts and event terms but lacks further information about event attribution and overlap. Consequently, the registry results support a description of the recorded endpoint contrast and its reported uncertainty, while leaving important questions about timing, missing data, analysis details, and safety context unanswered. They should not be extended beyond the population and outcome information actually recorded.

16. Why This Trial Matters Statistically

This trial illustrates why a clinical result needs to be read together with its endpoint definition, denominators, and analysis details. The primary measure is not a continuous symptom-score comparison; it counts participants crossing a specified QIDS-SR-16 threshold. That choice emphasizes the frequency of threshold attainment, while not showing the degree of symptom change among participants who do not cross the threshold.

The registry lists several assessment occasions, from pre-treatment through Week 28, but does not connect the posted outcome comparison to a specific occasion. When a result's timing is unclear, its interpretation is necessarily limited: readers cannot tell from the registry summary whether it represents an early assessment, a later assessment, or another summary across time. This is a reporting issue distinct from the numerical uncertainty captured by the interval.

The record also makes denominator interpretation important. Enrollment is 160, while the primary outcome groups include 60 and 65 evaluable subjects, and the safety summaries list 74 at risk in each group. Without an explanation of these population differences or the handling of missing measurements, it is difficult to assess how closely each summary corresponds to the enrolled randomized population.

Finally, statistical significance and effect size answer different questions. The unadjusted p-value is 0.032, while the risk-difference estimate and its interval describe the direction and range of the reported contrast. Neither alone communicates all aspects of evidence, and the registry does not report the analysis plan or multiplicity handling for the repeated assessments. The trial is therefore a useful example of how a posted estimate can be numerically interpretable while key design and reporting details remain unavailable.

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.

The primary outcome description lists assessments before treatment and at Weeks 2, 4, 6, 8, 10, 12, 16, 20, 24, and 28, but the registry record does not state which assessment produced the posted counts. That makes the timing of the comparison uncertain. Although the outcome is described as a count of participants reaching a threshold, the record does not explain whether the reported result refers to one time point or a summary across assessments.

The primary outcome uses an evaluable-subject population with 60 participants in the escitalopram arm and 65 in the placebo arm, while overall enrollment was 160. The record does not explain the difference or report the number and pattern of missing measurements. It also does not define evaluability, provide a participant flow summary, or describe sensitivity analyses for missing data. The safety summaries use 74 participants at risk in each group, without an explanation of how those denominators relate to enrollment and outcome analysis.

The registry record does not report the original sample-size justification, detailed analysis plan, interim analyses, multiplicity procedure, or covariate-adjusted results. No secondary outcome measures or analyses are posted. These omissions limit assessment of how the reported comparison relates to planned analyses, other outcomes, or repeated assessments. The record also does not provide additional eligibility details, dose, or treatment duration in the information summarized for this trial. These limitations concern what is reported; they do not establish that a particular design or analysis procedure was or was not used.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Evaluating the Effectiveness of Escitalopram to Prevent Depression in Head and Neck Cancer Patients Receiving Treatment — NCT00536172.
• Registry record last updated 2023-09-21; status last verified 2023-09.
• Results first posted 2014-02-14.
• 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

NCT00536172 is a completed Phase 4 randomized, quadruple-masked, parallel-group trial sponsored by the University of Nebraska. It compared escitalopram with placebo for prevention of depression in people receiving treatment for head and neck cancer. The record lists enrollment of 160, a start date of 2007-12-01, and a primary completion date of 2012-06-01.

The primary outcome counts evaluable subjects reaching a QIDS-SR-16 score >=11. The posted counts are 6 of 60 for escitalopram and 16 of 65 for placebo. The reported risk difference is -0.146, with an interval from -0.276 to -0.017, and the unadjusted p-value is 0.032. The record lists one outcome measure, no secondary outcome measures, and zero analyses. It does not identify the assessment occasion represented by the outcome counts or explain the outcome and safety denominators.