Treatment of Hypochondriasis With CBT and/or SSRI: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Treatment of Hypochondriasis With CBT and/or SSRI (NCT00339079), a randomized phase 1, phase 2 study. The primary outcome is 25% Improvement on Both Whiteley Index and H-YBOCS-M.

2. Clinical Question

The clinical question is therefore about a specific joint response criterion, not a general judgment about all symptoms or longer-term outcomes. The Whiteley Index is described as a self-report measure of hypochondriasis, while H-YBOCS-M is described as an independent-evaluator structured assessment. Requiring improvement on both makes the endpoint depend on each measure meeting the stated criterion.

The registry describes the study as randomized, quadruple-masked, and parallel. It lists the trial as completed and identifies Brigham and Women's Hospital as sponsor. Its brief summary says the study compares CBT, antidepressant medication, and a combination of the two. The posted results also include the placebo comparison. They do not report additional details about how participants were recruited or the population's clinical profile, so conclusions should remain tied to the population and endpoint actually described.

3. Trial Design

The registry describes a randomized, quadruple-masked, parallel-group trial in Phase 1 and Phase 2. The record lists enrollment as 195. The study start is recorded as 2006-06, primary completion as 2011-05, and status as completed. These registry entries describe the design and timing; they do not establish how masking was implemented or whether it remained intact.

Four arms appear in the posted results: CBT, placebo, fluoxetine, and combined CBT and fluoxetine. The arm counts are 53 for CBT, 44 for placebo, 45 for fluoxetine, and 53 for the combined arm. The primary endpoint counts are 21, 13, 20, and 25, respectively. These are the participants recorded as meeting the endpoint, not counts of all participants who experienced any degree of improvement.

The CBT description specifies six weekly sessions lasting 60 minutes, followed by four bi-weekly sessions and three monthly booster sessions. The fluoxetine arm is described as using a fixed-flexible dosing regimen, with stated dose steps and a target dose of 80 mg/day thereafter. The combined arm is described as receiving both interventions in the same way as when each was administered alone. The placebo arm received placebo pills accompanied by medication-management supportive therapy, including non-specific encouragement, support, and explanation similar to that provided in a physician's office.

The primary outcome was measured at Week 24. The record lists one secondary outcome, Columbia Heightened Illness Concern - Obsessive-Compulsive Scale. It does not report the original sample-size justification, detailed randomization procedures, masking procedures, or the original statistical analysis plan. A planning illustration based on the observed effects is distinct from evidence about the trial's original design assumptions.

4. Participant Flow and Accounting

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

Overall Study

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

6. Primary Endpoint

The registry lists the primary outcome as 25% Improvement on Both Whiteley Index and H-YBOCS-M, measured over Measured at Week 24.

The primary endpoint is titled “25% Improvement on Both Whiteley Index and H-YBOCS-M.” It is a binary participant-level outcome: each participant either met the joint threshold or did not. The measure is reported as a count of participants, and the assessment timeframe is Week 24.

The two instruments capture hypochondriasis from different assessment perspectives. The registry describes the Whiteley Index as a self-report measure and H-YBOCS-M as an independent-evaluator structured assessment. A joint endpoint requires the stated improvement threshold on both instruments, rather than improvement on either one alone. This can make the endpoint more specific to participants whose changes are reflected across both measures, while also making it dependent on both assessments and their scoring.

Using a binary endpoint makes the arm comparison straightforward to express as the proportion meeting the criterion and as a risk difference between arms. It also discards gradations within the categories: participants who fall short of the threshold are grouped together, even if their measured changes differ. The posted primary results therefore address the frequency of meeting this defined response criterion, not the size of change on either scale, change in every symptom, or outcomes beyond the stated assessment time.

The registry description does not give further scoring details, explain how the threshold was selected, or specify how missing assessments were handled. Those details are not reported in the registry record.

7. Results Posted in the Registry

• Cognitive Behavioral Therapy (CBT): 39.6% (21/53) versus 29.5% (13/44) on control; risk difference 10.1 percentage points (95% CI -8.8 to 28.9), risk ratio 1.34 (0.76–2.36); unadjusted two-proportion z-test p = 0.300.
• Fluoxetine: 44.4% (20/45) versus 29.5% (13/44) on control; risk difference 14.9 percentage points (95% CI -4.9 to 34.7), risk ratio 1.50 (0.86–2.64); unadjusted two-proportion z-test p = 0.146.
• Combined CBT and Fluoxetine: 47.2% (25/53) versus 29.5% (13/44) on control; risk difference 17.6 percentage points (95% CI -1.4 to 36.7), risk ratio 1.60 (0.93–2.73); unadjusted two-proportion z-test p = 0.077.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is binary, so the analysis compares the number meeting the response criterion with the number assessed in each arm. A risk is the proportion of participants with the outcome in an arm. A risk difference subtracts the comparator-arm risk from the active-arm risk. It is an absolute contrast on the proportion scale: a positive estimate means the observed endpoint proportion was higher in the active arm, while a negative estimate would mean it was lower.

For CBT versus placebo, the posted risks are 0.396 and 0.295, giving a risk difference of 0.101. For fluoxetine versus placebo, the risks are 0.444 and 0.295, with a risk difference of 0.149. For combined CBT and fluoxetine versus placebo, the risks are 0.472 and 0.295, with a risk difference of 0.176. These are unadjusted estimates based on the posted arm-level summaries. They do not account for any other covariates.

What do the relative measures add?

A risk ratio compares the two arm risks by division. An odds ratio compares the odds of meeting the endpoint between arms; odds are formed from the number meeting the endpoint relative to the number not meeting it. These relative measures express contrasts differently from the risk difference. They can be useful alongside the absolute contrast, but they do not replace it. When outcomes are not rare, an odds ratio can be farther from the no-difference value than a risk ratio, so the measures should not be read as interchangeable.

The registry reports risk ratios and odds ratios with confidence intervals for each active arm versus placebo. For each relative measure, the value corresponding to no difference lies within its reported interval. That interval-based observation complements the risk-difference results. The record does not state that any of these contrasts was adjusted for baseline characteristics.

What does a confidence interval express?

A confidence interval describes uncertainty around an estimate under the method used to calculate it. It is not a range containing a known true value with certainty, and it does not assign a probability to each value inside the interval. In repeated use of a method under its assumptions, intervals constructed in this way have a stated long-run coverage property. In practical interpretation, interval width indicates how precisely the estimate is determined by the data and method.

For a risk difference, an interval spanning zero includes the no-difference value as well as values in both directions. This is the case for each posted primary contrast. The interval should be considered with the point estimate, the clinical meaning of the endpoint, and the design limitations; it is not a stand-alone verdict.

Why is the p-value not the effect size?

A p-value measures how unusual data at least as incompatible with a specified null model would be, if that model and the analysis assumptions held. It does not measure the size or importance of an effect, and it is not the probability that the null hypothesis is true. The three posted p-values are unadjusted. They should be read together with the effect estimates and intervals, especially because the registry record does not report a multiplicity adjustment.

Statistical significance is not a substitute for precision or clinical interpretation. A p-value alone cannot show which range of treatment differences is compatible with the data. Conversely, a result that does not meet a conventional threshold does not demonstrate equivalence. Equivalence or non-inferiority conclusions require specified margins and an analysis designed for that purpose; such details are not reported in the registry record.

What can the arm-level counts establish?

The posted counts describe participants recorded as meeting the endpoint in each arm. They do not explain what happened to participants without a posted endpoint assessment, whether every randomized participant was included in the comparison, or how missing outcome data were treated. The registry record does not report analysis-population definitions. These omissions matter because the interpretation of a binary outcome depends on how the assessed group relates to the randomized group and how unavailable measurements are handled.

The trial is described as randomized, which supports an intended comparison based on assignment to arms. However, the registry extract does not give detailed allocation procedures, baseline balance, adherence, or treatment exposure. The posted estimates can be interpreted as the unadjusted arm contrasts recorded, but details needed to evaluate implementation and analysis assumptions are not reported.

10. Confidence Intervals and Precision

Precision concerns how narrowly the data and analysis constrain an effect estimate. The risk-difference intervals are broad relative to their point estimates. For CBT versus placebo, the estimated difference is 0.101, with a 95% interval from -0.088 to 0.289. For fluoxetine versus placebo, it is 0.149, with an interval from -0.049 to 0.347. For combined CBT and fluoxetine versus placebo, it is 0.176, with an interval from -0.014 to 0.367.

Each interval includes zero and extends to positive differences. This means the posted data do not precisely pin down the direction and size of the difference. The intervals also include a range of possible positive effects, so the point estimates should not be treated as exact descriptions of the underlying contrast. The intervals are uncertainty summaries under their calculation method, not guarantees about the true effect.

Precision is affected by the amount of information in the comparison and by how common the outcome is. With a binary endpoint, the number of participants and the number meeting the criterion both contribute to the uncertainty. General statistical principle: more information often narrows an interval, while sparse outcome information can leave estimates imprecise. This general point does not establish why these particular intervals have their observed widths.

The registry record gives no separate precision target or clinically important difference for the endpoint. Without such a benchmark, the intervals can be described statistically but cannot be classified against a registry-defined threshold for importance. The relative-measure intervals are also wide and include their no-difference values. Together, the absolute and relative estimates suggest that the observed contrasts are uncertain rather than sharply estimated.

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 for the binary endpoint comparisons; they are not total event counts. Each comparison is planned separately against placebo, and the entries describe hypothetical future planning based on a stated fraction of the observed effect. The row labelled 100% uses the full observed effect as the planning target; the 75% and 50% rows use progressively smaller target effects.

The planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is the stated fraction of the observed effect. Under these assumptions, the 80% power column gives the estimated number of participants per arm for a design with an 80% chance of detecting the target effect if that effect and the assumptions hold. The 90% column describes the corresponding planning target with a higher chance of detection, and therefore generally requires more participants per arm.

Smaller effects are harder to distinguish from random variation, so planning for a smaller fraction of the observed effect requires more participants per arm. These figures are illustrations calculated from observed estimates, not the trial's original sample-size justification. The original design assumptions are not reported in the registry record. This is not observed power: it does not measure the probability that the completed trial detected an effect, and it should not be used to reinterpret the posted confidence intervals or p-values.

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 = 0.295; treatment proportion p₁ = 0.396; average p̄ = 0.346.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6727 = 1.3184.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6689 = 0.5629.
4. n = (1.3184 + 0.5629)² / (0.1008)² = 348.5, rounded up to 349 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 <- 0.295454545454545                          # control proportion
p1 <- 0.39622641509434                          # 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(339079)
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 = 349 (80% power), 467 (90% power)
 75% of observed effect: n per arm = 610 (80% power), 816 (90% power)
 50% of observed effect: n per arm = 1,346 (80% power), 1,802 (90% power)
power.prop.test: 348.5 
Simulated power at n = 349 per arm: 0.806 (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 posted results include three active-treatment comparisons against the placebo arm, and the registry record lists one secondary outcome. When several outcomes or comparisons are examined, the chance of at least one apparently unusual result can increase. Statistical approaches to multiplicity can control or otherwise address this issue, but the registry record does not report a multiplicity procedure, a hierarchy of hypotheses, or which outcome was designated for confirmatory inference beyond identifying the primary outcome.

The p-values are explicitly unadjusted. They should not be interpreted as multiplicity-corrected evidence across the comparisons. Nor does the record establish that every possible analysis was reported. The listed secondary outcome is Columbia Heightened Illness Concern - Obsessive-Compulsive Scale, but a secondary-outcome result is not included in the posted outcome details. The record lists zero posted analyses.

Important details about missing data are not reported in the registry record. It does not state whether participants with unavailable Week 24 assessments were excluded, imputed, or classified in another way. It also does not describe follow-up completeness by arm or sensitivity analyses for missing outcomes. As a result, the relationship between the arm counts in the primary result and all randomized participants cannot be fully evaluated from the record.

Analysis populations are also not reported. The record does not identify whether the primary comparison used an intention-to-treat population, a per-protocol population, or another definition. Details about interim analyses, prespecified stopping rules, and the statistical plan are not reported either. These are limits on what can be assessed from the registry record; they should not be filled in by assumption.

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 figures report zero affected participants in each arm: CBT, placebo, fluoxetine, and combined CBT and fluoxetine. The numbers at risk listed for these summaries are 53, 44, 45, and 53, respectively. No serious adverse event terms are listed. A zero in these posted counts means no serious adverse event was recorded in the corresponding summary; it does not establish that the underlying risk is zero or exclude events not captured by the record.

For other, non-serious adverse events, the record lists zero affected participants in CBT, one in placebo, zero in fluoxetine, and zero in the combined arm. These counts should not be treated as a complete description of tolerability: the record does not provide event details, severity gradations, exposure-adjusted rates, or a narrative explaining collection and coding. Death counts are not reported in the registry record.

Safety summaries are descriptive and should be kept distinct from the primary efficacy endpoint. The posted adverse-event counts do not establish comparative safety, particularly without additional event detail and information about ascertainment. The absence of a listed serious event is not a basis for a broader safety conclusion.

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

The posted results describe how many participants met a specific Week 24 response criterion in each randomized arm and estimate the unadjusted difference versus placebo. The point estimates are positive for CBT, fluoxetine, and the combination, while the confidence intervals for all three contrasts include zero and extend across a range of possible differences. This is evidence with uncertainty, not proof that the arms are equivalent and not proof that any active-treatment comparison differs from placebo.

The primary endpoint requires 25% improvement on both the Whiteley Index and H-YBOCS-M. It does not represent every possible definition of improvement, the average change on either scale, or an outcome beyond Week 24. The posted binary counts cannot show whether participants just below the threshold experienced no change or substantial change that did not meet the joint criterion.

The combined arm has the largest point estimate among the three active-versus-placebo comparisons, but no direct statistical comparison of the combined arm with CBT alone or fluoxetine alone is reported. The observed ordering therefore should not be read as a demonstrated advantage of the combination over either single intervention.

The record does not report adjustment for multiple comparisons, handling of missing outcomes, analysis-population definitions, or the original analysis plan. Those gaps limit assessment of how robustly the posted estimates represent the randomized groups and how the results were selected or interpreted. The findings should be understood as registry-posted arm counts and unadjusted contrasts for this endpoint, rather than as a comprehensive account of all outcomes or treatment effects.

No medical conclusion follows from these statistics alone. They describe the recorded outcome and its uncertainty; they do not establish what an individual should choose or predict an individual participant's outcome.

16. Why This Trial Matters Statistically

This trial illustrates the distinction between an observed difference and a precisely estimated difference. The three risk-difference estimates are positive, but each interval includes zero and spans values in both directions. Reading the point estimate without its interval would overstate how firmly the data determine the comparison. Reading only the p-value would also omit the scale and range of plausible contrasts.

The primary outcome is a joint binary criterion based on a self-report measure and an independent-evaluator assessment. This makes the endpoint definition central to interpretation: it asks whether participants met a threshold on both instruments, not whether scores changed on average. The design highlights a general statistical trade-off. A threshold can make a response easy to count, but it compresses graded measurements into a yes-or-no result and makes the conclusion dependent on the threshold and assessment time.

The trial also makes the comparator important to describe accurately. The placebo arm included medication-management supportive therapy. The contrast is therefore with that recorded condition, not with an absence of clinical contact as described in the registry.

Finally, the planning illustrations show how sample-size requirements depend on the target effect and planning assumptions. They are separate from the original trial design and from the results observed in the trial. Distinguishing those purposes prevents a planning calculation from being mistaken for evidence that the completed study was adequately powered or for a measure of observed 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 trial-specific limitations affect interpretation of the posted results. The primary endpoint is measured at Week 24 and is defined by a threshold on both the Whiteley Index and H-YBOCS-M. The registry record does not report further scoring or threshold-selection details, so the clinical interpretation of meeting the criterion cannot be evaluated beyond the stated definition.

The posted primary counts are modest in each arm, and the confidence intervals are broad. This leaves uncertainty about the size and direction of each active-versus-placebo contrast. The record does not report a direct contrast between the active arms, so the estimates cannot establish whether CBT, fluoxetine, and the combination differ from one another.

The comparator included supportive therapy with placebo pills. That context is relevant to what the comparisons mean. The registry does not provide detailed information about treatment adherence, treatment exposure, participant flow, baseline characteristics, or whether masking was maintained. The fluoxetine description is truncated after its medication-management wording, so further details should not be inferred.

Only one secondary outcome is named, and its result is not included in the posted outcome details. The registry record does not report handling of missing outcome data, analysis populations, adjustment for multiplicity, interim analyses, or the original analysis plan. These omissions prevent a full appraisal of the prespecification and robustness of the analyses. The record also does not give a clinically important difference against which to judge the interval estimates.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Treatment of Hypochondriasis With CBT and/or SSRI — NCT00339079.
• Registry record last updated 2017-04-25; status last verified 2017-03.
• Results first posted 2017-03-03.
• 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 registry describes a completed, randomized, quadruple-masked, parallel-group Phase 1 and Phase 2 study of treatment for hypochondriasis, sponsored by Brigham and Women's Hospital. It lists enrollment as 195 and compares CBT, fluoxetine, their combination, and placebo accompanied by medication-management supportive therapy.

The primary endpoint was the count of participants with 25% improvement on both the Whiteley Index and H-YBOCS-M at Week 24. Posted arm counts were 21 for CBT, 13 for placebo, 20 for fluoxetine, and 25 for combined CBT and fluoxetine. The unadjusted risk differences versus placebo were 0.101, 0.149, and 0.176, respectively; all reported risk-difference intervals include zero.

The registry lists Columbia Heightened Illness Concern - Obsessive-Compulsive Scale as a secondary outcome. Important details about missing data, analysis populations, multiplicity, and the original analysis plan are not reported in the registry record.