Chantix for Treating Cocaine Dependence: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Chantix for Treating Cocaine Dependence (NCT00567008), a randomized phase 2 study. The primary outcome is Evidence of Abstinence From Cocaine as Indicated by Qualitative Urinalysis for Benzoylecgonine..

2. Clinical Question

The comparison is therefore a question about a short-term, binary test result in a small randomized pilot, rather than a broad measure of every aspect of cocaine dependence or its course. The registry record does not report the primary outcome’s analysis population beyond the arm counts, nor does it describe additional eligibility or follow-up details that would clarify how broadly the result applies.

3. Trial Design

This was a Phase 2, randomized, quadruple-masked, parallel-group pilot trial comparing varenicline with placebo for cocaine dependence. The registry identifies the sponsor as the University of Pennsylvania. Its recorded status is completed; the recorded start is 2007-12 and primary completion is 2009-09. Total enrollment was 37.

In a parallel-group design, participants are assigned to one of the study groups and followed in that group. Randomization is intended to make treatment groups comparable on average, while masking is intended to limit the influence of knowledge about assignment on study conduct or assessment. The registry labels the masking as quadruple, but does not specify which people or roles were masked. The method used to generate or conceal the allocation sequence is not reported in the registry record.

The two intervention groups were varenicline and placebo. For the primary outcome, the posted group sizes were 18 and 19, respectively. The outcome was assessed at 8 weeks and counted participants with a negative qualitative urinalysis for benzoylecgonine. The trial title identifies the study as a pilot trial. The registry record does not report the original sample-size justification, recruitment targets beyond the enrollment figure, or the assumptions used when the trial was designed.

The record lists one outcome measure and no posted secondary outcomes. It does not report details of treatment administration, adherence, the exact visit schedule, or follow-up after the primary assessment. These omissions limit the amount of design context that can be brought to interpretation of the posted result; they do not alter the recorded group counts or the stated endpoint.

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 Evidence of Abstinence From Cocaine as Indicated by Qualitative Urinalysis for Benzoylecgonine., measured over 8 weeks.

The primary endpoint counted participants whose qualitative urinalysis assessment was negative for benzoylecgonine. The registry describes this as evidence of abstinence from cocaine, with an assessment timeframe of 8 weeks. It is a binary outcome at the participant level: each assessed participant contributes either a negative or a non-negative test result. The posted counts of negative results were 2 in the varenicline group and 1 in the placebo group.

A qualitative test reports a category such as negative or positive rather than a continuous concentration. Binary outcomes can be useful when the research question concerns whether a defined event or status was observed, and when the test and threshold are specified. Here, a negative result is the recorded endpoint event; it should not be read as a complete account of a participant’s experience over the entire study period. The registry record does not provide details about the test’s operating characteristics, the timing of specimen collection within the 8-week timeframe, or how missing assessments were handled.

The endpoint is not a measure of the amount of cocaine use, duration of abstinence, or change in dependence symptoms. Nor does a single assessment by itself describe the pattern of results across repeated measurements. Those distinctions matter when interpreting a short-term test result: the outcome captures the specific status measured at the stated assessment, and conclusions should remain within that scope.

7. Results Posted in the Registry

• Varenicline: 11.1% (2/18) versus 5.3% (1/19) on control; risk difference 5.8 percentage points (95% CI -11.8 to 23.5), risk ratio 2.11 (0.21–21.32); unadjusted two-proportion z-test p = 0.515.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is binary: a participant either had a negative qualitative urinalysis for benzoylecgonine or did not. For a binary outcome, the basic group summary is the number of participants with the event relative to the number assessed. The posted counts are 2 of 18 for varenicline and 1 of 19 for placebo. A comparison can be expressed in absolute terms as a risk difference, or in relative terms as a risk ratio or odds ratio. These measures answer related but distinct questions, and their interpretation depends on the outcome and the populations being compared.

What does a risk difference describe?

A risk difference subtracts one group’s event proportion from the other group’s event proportion. It is an absolute comparison, expressed on the probability scale. A value above zero indicates a higher observed event proportion in the group placed first in the comparison; a value below zero indicates a lower observed proportion. For this trial, the posted estimate is above zero, while its confidence interval extends across zero. The interval therefore includes both a possible positive difference and a possible negative difference.

Absolute measures are often easier to relate to the observed event frequencies than relative measures, but they can be unstable when the outcome is rare and the groups are small. Here, only 2 and 1 participants, respectively, had the event. A change in the outcome status of one participant would have a noticeable influence on these group summaries. That is a general feature of sparse binary data, not evidence that any particular participant’s result was incorrect.

Why are there several effect measures?

A risk ratio compares event proportions by division, while an odds ratio compares the odds of the event. Ratios are expressed relative to the comparison group, whereas the risk difference is expressed as an absolute contrast. The registry-derived comparison includes ratio estimates as well as the risk difference. Their confidence intervals are broad, reflecting the limited information in the small number of observed events. No single measure removes the underlying uncertainty, and a ratio should not be interpreted as an absolute increase or decrease in the outcome.

When events are uncommon, the odds and risk can be numerically similar, but they are not the same quantity. More generally, a ratio can look large even when the absolute event proportions remain small. For that reason, interpretation should consider the event counts and absolute difference alongside any relative measure. In this trial, the counts are particularly important context for the estimated comparison.

What does the p-value say?

The unadjusted p-value is 0.515. In general, a p-value describes 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 state the probability that the null model is true, nor does it measure the chance that the result will reproduce. It also does not quantify clinical importance. A p-value should be read together with the estimated effect and its confidence interval, especially when the data are sparse.

The registry record does not report a statistical analysis plan or a prespecified test for this endpoint. The posted comparison is an unadjusted recalculation from the group-level summary data. It does not incorporate adjustment for additional variables. The record does not report whether other planned analyses were considered or how any such analyses would have been handled.

Why does sparse data matter?

General statistical methods for binary outcomes rely on assumptions about how the observed events arise and how well the chosen approximation represents the data. With very few events, standard approximations may be less stable, and estimates can have wide uncertainty. Over-dispersion is a general issue in count models when variability exceeds the model’s assumptions; it is not a reported finding or a stated analysis feature of this trial. The outcome is a binary participant count, and the registry does not report a model that addresses over-dispersion.

Randomization supports a comparison between assigned groups, but it does not guarantee that small groups will be balanced on every characteristic. No baseline characteristics or adjusted estimates are included in the registry result. The posted counts consequently support a simple unadjusted comparison, while leaving questions about covariate balance, missing observations, and analysis populations unresolved.

10. Confidence Intervals and Precision

The confidence interval for the risk difference runs from -0.118 to 0.235. It includes zero and values below and above zero. In general, a confidence interval is a range of values compatible with the observed data and the method’s assumptions. It conveys uncertainty around an estimate; it is not a guarantee that the true effect lies inside the particular interval, nor does it assign a probability to each value within it.

The width and span of the interval are important in this trial. The interval allows for a negative difference as well as a positive one, so the observed direction is not estimated precisely enough to rule out a difference in the opposite direction. The interval also includes differences larger than the point estimate. This broad range is consistent with the limited information in the endpoint counts, which were 2 and 1 participants in groups of 18 and 19.

The reported risk ratio and odds ratio also have wide confidence intervals. In general, ratio intervals are interpreted relative to 1, the value corresponding to no relative difference. The posted intervals include values below and above that reference, so they likewise do not pinpoint the direction or size of a relative effect. The risk-difference interval is useful for seeing uncertainty in absolute terms, while relative measures present the contrast on a different scale.

Precision is distinct from whether a result meets a significance threshold. A p-value alone does not show the range of effect sizes consistent with the data, while the interval makes that range visible. For this small pilot result, the interval is more informative when read alongside the sparse event counts than when treated as a standalone verdict. Greater precision would require more information; the registry record does not establish what effect or level of precision a future study should target.

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 display gives planning illustrations in participants per arm, not total events. Its calculations use the observed effect as a starting point and show scenarios for detecting 100%, 75%, and 50% of that effect. The listed planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The original design assumptions for this trial are not reported in the registry record.

The 80% power column describes the number of participants per arm in a hypothetical design that would have an 80% chance of detecting the specified effect if the planning assumptions and effect were correct. The 90% power column gives the corresponding per-arm planning number for a higher chance of detecting that effect. These are design illustrations, not observed outcomes or evidence about the probability that the trial’s result is true.

For planning at 100% of the observed effect, the display gives 344 participants per arm for 80% power and 460 participants per arm for 90% power. Planning for 75% of the observed effect gives 562 participants per arm for 80% power and 752 participants per arm for 90% power. Planning for 50% of the observed effect gives 1,151 participants per arm for 80% power and 1,540 participants per arm for 90% power. The progressively larger planning numbers reflect the general principle that smaller effects are harder to distinguish from random variation and require more participants under the same assumptions.

These are not observed power calculations. They are forward-looking illustrations based on the observed effect, which is itself uncertain because the trial recorded few endpoint events. They should not be read as a recommendation or as the original trial’s sample-size target. The associated curve presents the same planning relationship visually; it does not add evidence about the treatment comparison.

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 = 0.053; treatment proportion p₁ = 0.111; average p̄ = 0.082.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.3877 = 0.7599.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.3855 = 0.3245.
4. n = (0.7599 + 0.3245)² / (0.0585)² = 343.9, rounded up to 344 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 <- 0.0526315789473684                          # control proportion
p1 <- 0.111111111111111                          # 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(567008)
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 = 344 (80% power), 460 (90% power)
 75% of observed effect: n per arm = 562 (80% power), 752 (90% power)
 50% of observed effect: n per arm = 1,151 (80% power), 1,540 (90% power)
power.prop.test: 343.9 
Simulated power at n = 344 per arm: 0.803 (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 registry lists one outcome measure and no posted secondary outcomes. This means the record does not present a set of additional outcome results for comparison alongside the primary endpoint. It does not establish whether other outcomes were considered during the study or whether any were collected but not posted. A statistical plan describing intended analyses or multiplicity procedures is not reported in the registry record.

Multiplicity is a general concern when several hypotheses are tested: the chance of at least one apparently notable result can increase as more comparisons are made. With no secondary outcomes posted and no analysis plan described, it is not possible to assess whether multiple outcomes or analyses were planned, or how such testing would have been addressed. The unadjusted p-value for the primary comparison should not be mistaken for a documented multiplicity-adjusted result.

The registry record does not report how missing primary outcome assessments were handled, whether all randomized participants were included in the analysis, or the definition of an analysis population. The outcome arm counts total the stated enrollment, but those counts alone do not establish the handling of missing data, participant exclusions, or deviations from assigned treatment. No imputation method, sensitivity analysis, or per-protocol analysis is reported.

These omissions matter because different approaches to missing observations and analysis populations can affect a binary outcome comparison. The result available in the registry is therefore interpretable as a comparison of the posted group counts, but it cannot answer whether the analysis followed an intention-to-treat principle or another prespecified approach. Details of interim analyses and any stopping rules are not reported in the registry record either.

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 5% reporting threshold.

Most frequent non-serious adverse event terms

The serious adverse event figures show 0 affected participants among 17 at risk in the varenicline group and 0 among 18 at risk in the placebo group. This describes the counts posted for the listed risk sets; it does not establish that no serious events occurred outside the recorded reporting scope or that the treatments have equivalent safety profiles. The registry lists no serious adverse event terms.

For other adverse events, the record lists 15 affected participants among 17 at risk for varenicline and 17 among 18 at risk for placebo. These are counts of participants with other adverse events, not counts of separate event episodes. The registry record does not provide the specific non-serious event terms in the information, so the figures cannot show which events occurred, their severity, timing, or relationship to treatment.

Deaths are not reported in the registry record. The adverse-event risk sets differ from the primary outcome arm counts, and the record does not explain the relationship between those denominators. Safety counts are descriptive and should be interpreted with the small group sizes and limited event detail in mind.

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

The posted result shows that 2 participants in the varenicline group and 1 in the placebo group had a negative qualitative urinalysis for benzoylecgonine at the 8-week assessment. The risk-difference estimate is positive, but its confidence interval extends across zero, from -0.118 to 0.235, and the unadjusted p-value is 0.515. Together, these statistics describe a numerical difference with substantial uncertainty, rather than a precise estimate of a between-group effect.

The result does not establish that varenicline caused a higher chance of a negative test, and it does not establish that the groups were equivalent. The confidence interval includes possible differences in either direction. The small number of negative tests means that the estimate is sensitive to individual outcomes and does not tightly constrain the range of plausible effects.

The endpoint also has a limited scope. It records a qualitative test result at the stated 8-week timeframe. It does not, by itself, measure the duration or pattern of abstinence, the amount of cocaine use, symptoms of dependence, or outcomes beyond that assessment. The registry record does not report enough detail about assessment timing, missing data, or analysis populations to extend interpretation beyond the posted counts.

As a pilot trial, this study contributes a small randomized comparison and an indication of the uncertainty around its measured endpoint. The result is not a basis for a treatment recommendation, and it should not be generalized automatically to other populations or outcome definitions. The sample-size illustrations are planning calculations based on the observed effect, not evidence that a larger study would necessarily reproduce the same estimate.

16. Why This Trial Matters Statistically

Statistically, this trial illustrates why an observed difference and the precision of that difference need to be considered together. The varenicline group had more negative tests in absolute count, but only three participants across both groups had the endpoint. A positive point estimate can attract attention, yet the confidence interval extending across zero shows that the direction and size of the underlying difference remain uncertain.

The study also demonstrates how sparse binary outcomes can limit what a randomized trial can resolve. Random allocation supports a fair comparison in principle, but a small pilot cannot ensure close balance in every relevant characteristic, and few events yield imprecise estimates. The reported p-value does not substitute for the interval or the event counts; each provides a different part of the statistical picture.

The sample-size illustrations make a related planning point: under the stated assumptions, aiming to detect a smaller fraction of the observed effect requires more participants per arm. That relationship is general, but the illustrations depend on the observed effect, which is uncertain in a small study. They should therefore inform an understanding of planning sensitivity rather than be treated as definitive design requirements.

Finally, the record’s limited reporting of analysis plans, missing-data handling, analysis populations, and secondary outcomes constrains evaluation of how the comparison was conducted. Recognizing these gaps is part of statistical interpretation. The result is useful as a bounded report of an eight-week endpoint in a pilot trial, while remaining too imprecise to settle the broader treatment question.

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.

This trial’s main limitation for interpreting the primary comparison is the small number of participants and observed endpoint events. Enrollment was 37, and the posted primary outcome counts were 2 and 1. The resulting estimate is imprecise, as reflected by a confidence interval that includes differences in both directions. Any interpretation centered only on the positive point estimate would omit this uncertainty.

The primary endpoint captures a qualitative urinalysis result at 8 weeks, not a continuous measure or a longer-term pattern. The registry record does not report the timing of specimen collection within that timeframe, test performance characteristics, or repeated assessment results. It also does not explain how missing assessments or other departures from follow-up were handled.

Several design and analysis details are not reported in the registry record: the original sample-size justification, the statistical analysis plan, the method of allocation concealment, the analysis population, and any adjustment for covariates. The registry does not report interim analyses or stopping rules. The absence of these details limits assessment of prespecification and analytic choices; it should not be taken as evidence that a particular method was or was not used.

The safety information is also limited. Serious adverse event counts are posted, but serious event terms are not listed, and the record does not describe the specific other adverse events. Deaths are not reported in the registry record. These constraints do not change the stated counts, but they limit the detail available for contextual interpretation.

As a pilot study, the result is also limited in its ability to characterize a precise treatment-group difference. The registry record does not supply additional population details that would support broad generalization. Any inference should remain tied to the recorded groups, endpoint, and timeframe.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Chantix for Treating Cocaine Dependence — NCT00567008.
• Registry record last updated 2013-11-07; status last verified 2013-10.
• Results first posted 2013-11-07.
• 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 identifies NCT00567008 as a completed Phase 2 randomized, quadruple-masked, parallel-group pilot trial of varenicline versus placebo for cocaine dependence. The sponsor is the University of Pennsylvania. Enrollment was 37; the recorded start was 2007-12 and primary completion was 2009-09.

The primary outcome was the number of participants with a negative qualitative urinalysis for benzoylecgonine at 8 weeks. The posted counts were 2 of 18 for varenicline and 1 of 19 for placebo. The unadjusted risk-difference confidence interval extends from -0.118 to 0.235, and the unadjusted p-value is 0.515. The estimate is imprecise, with few endpoint events.

The registry lists no secondary outcomes or analysis entries. It does not report the original sample-size justification, analysis plan, missing-data handling, or analysis population. Serious adverse event counts were 0 among 17 at risk for varenicline and 0 among 18 at risk for placebo; deaths are not reported in the registry record.