Evaluation of Atomoxetine for Cocaine Dependence: A Pilot Trial: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Evaluation of Atomoxetine for Cocaine Dependence: A Pilot Trial (NCT00617201), a randomized phase 2 study. The primary outcome is % Urine Samples Negative for Cocaine.

2. Clinical Question

This is a pilot study, so the posted record is most useful for describing the trial’s design and the results it reports, rather than for making a definitive comparative claim. The primary endpoint is a repeated-measure summary: urine samples were collected 3 times per week, and the endpoint combines the negative-sample information over the trial period. The registry states that the intent-to-treat analysis included all subjects and counted a missing urine sample as positive.

The endpoint title and description refer to negative samples, but the unit is worded as “percentage of positive urine samples.” That inconsistency in labeling makes it important to retain the record’s wording when interpreting the posted values. The record reports 0.67 for placebo and 0.74 for atomoxetine. It does not report an estimable two-arm comparison for this count-type outcome, so the numerical group values alone do not establish the precision or statistical uncertainty of the between-arm contrast.

3. Trial Design

The trial is registered as a randomized, parallel-group Phase 2 study of atomoxetine for cocaine dependence. It is listed as completed, with an enrollment of 50 participants. The record gives 25 participants in the placebo arm and 25 in the atomoxetine arm. The comparator was matched placebo, and the registry classifies masking as quadruple. The brief summary separately describes the study as double-blind. These descriptions indicate masking, but the record does not explain the roles covered by the quadruple classification.

The brief summary describes a 12-week trial. It says participants were assigned to atomoxetine or matched placebo, with an ascending-dose lead-in for the active treatment before maintenance on the assigned dose. It also describes contingency management procedures to reinforce attendance and compliance with study procedures. Details of how those procedures were implemented are not reported in the registry record.

The primary outcome was the total percentage of urine samples negative for benzoylecgonine during the trial. Urine was collected 3 times per week, with Monday, Wednesday and Friday given as examples. The stated analysis population was intent-to-treat, and missing urine samples were counted as positive. Retention was the posted secondary outcome and was measured in days over 12-weeks.

The record lists the start as 2007-07 and primary completion as 2011-08. It does not report the sample-size justification, the planned statistical analysis in sufficient detail to reconstruct it, or whether interim analyses were planned. The registry record also does not describe other design features beyond those stated above, so the study’s operational details should not be inferred from its general design labels.

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: Exclusions included dependence on any drug requiring detoxification, other Axis I disorders, medical problems, pregnant or lactating females, Cytochrome P450 2D6 inhibitors/inducers, monoamine oxidase inhibitors or selective serotonin reuptake inhibitors. Enrollment in other drug treatment programs urine samples required for parole/probation.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Intent-to-treat analysis included all subjects with missing urine sample counted as positive.

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 % Urine Samples Negative for Cocaine, measured over Urines were collected 3 times per week (e.g., Monday, Wednesday and Friday) for 12 weeks.

The primary endpoint is titled “% Urine Samples Negative for Cocaine.” Its description defines the measure as the total percentage of urine samples negative for benzoylecgonine over the 12-week trial. Urine was collected 3 times per week, with Monday, Wednesday and Friday given as examples. The endpoint therefore represents a summary across repeated urine assessments rather than a single measurement at the end of treatment.

In general, a urine-based measure can provide an objective indicator relevant to substance use, complementing information based on participant report. Repeated collection can describe the pattern across a treatment period rather than relying on one sample. The registry’s stated intent-to-treat approach included all subjects and counted missing urine samples as positive, which defines how missing observations contributed to the endpoint.

The record’s wording is not fully consistent: the title and description refer to negative samples, while the unit is “percentage of positive urine samples.” The posted arm values are 0.67 for placebo and 0.74 for atomoxetine. They should be interpreted with that labeling ambiguity in mind. No estimable two-arm comparison is posted for the primary outcome, so the record does not provide a confidence interval or p-value for the difference between arms. As a result, the endpoint values describe the posted groups but do not, by themselves, show how precisely the groups differ.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Interpreting these results requires distinguishing a group summary from a comparison, and a registry-posted analysis from a simple recalculation. The primary outcome is a count-derived percentage summary across urine samples. The secondary outcome is a continuous measure reported as a mean with a standard error. Those outcome types need not have the same analysis or the same available summary statistics.

What does a confidence interval describe?

In general, a confidence interval describes the uncertainty around an estimated quantity under the statistical method used to construct it. A narrower interval indicates greater precision, while a wider interval indicates less precision. For a mean difference, an interval that includes zero is compatible with no difference under the model used, but it may also be compatible with differences in either direction. It does not prove equality or equivalence.

For retention, the registry provides arm means and standard errors, and a simple unadjusted recalculation gives a mean difference of -5.8 days with a 95% confidence interval from -24.79 to 13.19. The interval is wide enough to include zero as well as values on both sides of zero. The recalculated p-value is 0.549. These statistics describe the recalculated, unadjusted comparison; the registry lists no reported analyses. The confidence interval and p-value should not be mistaken for a sponsor-reported model result.

How should a standard error be read?

In general, a standard error describes the estimated sampling variability of a statistic, such as a mean. It is not the spread of individual participants’ retention times. The retention means are reported with standard errors of 7.0 days for placebo and 6.7 days for atomoxetine. The record does not provide individual-level data or a standard deviation for retention, so the range of participant-level retention values cannot be recovered from these summaries alone.

For general context, standard errors depend on both variability in the observations and the amount of information available. They express uncertainty about an estimated mean, rather than the degree to which individual observations differ from one another. Keeping those meanings distinct prevents interpreting the reported standard errors as ranges containing most participants.

Why is the primary comparison not estimated?

The primary outcome is classified as a count-type endpoint, and the registry record indicates that values are missing or not estimable for a two-arm comparison. It reports the two arm values, but no primary comparison estimate, confidence interval or p-value. A group-level summary does not automatically provide the information needed to estimate uncertainty around a contrast. For example, a comparison may require a specified model and information about the distribution and dependence of repeated observations. The record does not report those details for the primary analysis.

In general, repeated measurements from one participant are related rather than independent. An analysis of repeated urine results may need to account for that within-participant structure, depending on how the endpoint is defined and modeled. The registry does not report the primary analysis model, so the method used to address repeated observations cannot be determined from the record.

What do the sample-size illustrations mean?

The retention comparison includes planning illustrations based on the observed effect, using a two-sided significance level of 0.05, equal allocation, and a normal approximation. These are planning calculations, not observed power and not evidence about the probability that the trial detected an effect. They are not the trial’s documented design justification; that justification is not reported in the registry record. The illustrations are associated with the retention comparison, rather than an estimable primary-outcome contrast.

• Planning for 100% of the observed effect gives 548 participants per arm for 80% power and 734 participants per arm for 90% power.

• Planning for 75% of the observed effect gives 974 participants per arm for 80% power and 1,304 participants per arm for 90% power.

• Planning for 50% of the observed effect gives 2,191 participants per arm for 80% power and 2,933 participants per arm for 90% power.

In general, planning estimates that use an observed effect can be sensitive to the estimate’s uncertainty. A small pilot study can produce an unstable observed effect, so these illustrations should not be treated as fixed requirements for a future study. They show the consequences of the stated assumptions and effect fractions, not the design rationale for this completed trial.

Why can outcome distributions matter?

In general, count outcomes may have variability that exceeds what a simple model assumes; this is called over-dispersion. When present, over-dispersion can make uncertainty larger than a model that assumes a tighter distribution would suggest. The posted primary outcome is a percentage summary based on repeated urine samples, but the record does not report a model, a dispersion estimate, or a comparison analysis. It is therefore not possible to determine whether over-dispersion was assessed or how it would affect a formal primary comparison.

For continuous outcomes, a mean difference is naturally expressed in the units of the measurement, which for retention are days. A standardized effect can also describe a difference relative to variability, but it does not replace the original-unit estimate. The registry’s planning illustrations are calculated from the observed retention effect; their interpretation depends on the stated assumptions and should not be carried over to the primary endpoint.

Finally, the distinction between planned and reported analysis matters. The record contains outcome summaries and one secondary comparison that can be recalculated from summary statistics, but it lists no reported analyses. The statistical analysis plan, handling of multiplicity, and any interim-analysis approach are not reported in the registry record. These omissions limit how fully the inferential choices can be evaluated.

10. Multiplicity, Missing Data and Analysis Populations

The registry record does not report whether the statistical plan included a multiplicity adjustment, an outcome hierarchy, or a strategy for interpreting secondary outcomes. This matters because testing or evaluating multiple outcomes can increase the chance of a result appearing notable by chance. In general, a prespecified primary endpoint is often given the main confirmatory role, with secondary outcomes interpreted more cautiously, but the registry does not state the planned inferential framework for this trial.

The record identifies one secondary outcome, retention, and lists no reported analyses. A simple unadjusted comparison can be recalculated from the posted arm means and standard errors, but this is not evidence that the same comparison was prespecified or used by the investigators. The record does not report the analysis population or model details for that secondary comparison beyond the arm summaries. As a result, the recalculation should remain clearly labeled as unadjusted and exploratory.

For the primary outcome, the record explicitly states that the intent-to-treat analysis included all subjects and counted missing urine samples as positive. This is an important rule because missing outcomes are not simply omitted: the record describes how missing samples contribute to the endpoint. The record does not provide the number of missing samples, reasons for missingness, or sensitivity analyses that might show how other assumptions about missing observations affect the results.

The primary endpoint’s comparative values are listed as not estimable for a two-arm comparison. The registry therefore does not provide a primary-outcome confidence interval or p-value. Nor does it report enough detail to determine the model, the treatment of repeated samples beyond the stated missing-data rule, or any interim analyses. These features should be regarded as unreported, rather than assumed absent. Without them, readers cannot fully assess whether the posted summaries reflect all planned analyses or how uncertainty and multiplicity were handled.

11. 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.

The registry lists 25 participants at risk in each arm for serious and other adverse-event reporting. It records 0 participants affected by serious adverse events in the atomoxetine group and 0 in the matched-placebo group. It also records 0 participants affected by other adverse events in each group. These are the posted participant counts; they do not describe event types or provide detailed event-level information.

The death entries are not populated. Death information is therefore not reported in the registry record, rather than being established as zero by the posted figures. The record lists no serious adverse-event terms, but that does not supply details about how events were collected, classified, or monitored.

In general, a lack of recorded affected participants in a small trial cannot establish that an intervention is free of risk. The safety entries should be read as a limited report of what is recorded for this trial, not as a comprehensive assessment of uncommon events or safety in other populations. No medical conclusion follows from these figures alone.

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

The posted results show that the registry contains primary outcome values for both randomized groups and a secondary retention summary. The primary values are 0.67 for placebo and 0.74 for atomoxetine, but the registry does not provide an estimable two-arm comparison for this endpoint. The title and description define the endpoint in terms of negative urine samples, while the unit wording refers to positive urine samples. That inconsistency adds a reason to avoid interpreting the values beyond the definition stated in the record.

The record does not establish the precision of the primary comparison because it gives no comparative confidence interval or p-value. A numerical difference between group summaries is not, by itself, proof of a treatment effect. The retention comparison is also uncertain: its unadjusted recalculation has a confidence interval that includes zero and a p-value of 0.549. Neither result establishes equivalence, and neither can substitute for a reported primary analysis.

In general, a randomized comparison is designed to help distinguish treatment differences from differences that could arise by chance, but the strength of the resulting inference depends on the analysis and its precision. The registry does not report the primary analysis method, a sample-size justification, or a detailed statistical plan. Those omissions limit the conclusions that can be drawn from the posted data.

The trial is described as a pilot, with 50 participants and a 12-week treatment period. Its results can inform understanding of what was measured and how the outcomes were reported, but they should not be treated as a definitive estimate of a treatment effect. The record does not establish whether findings apply beyond the enrolled population or trial setting. It also does not support medical advice. The appropriate interpretation is limited to the statistics and design details recorded for this study, with the uncertainty and missing analysis information kept in view.

13. Why This Trial Matters Statistically

This trial is statistically informative because it illustrates the difference between reporting group values and reporting an estimable comparison. For the primary urine-sample outcome, arm values are available, but the registry states that values are missing or not estimable for a two-arm comparison. Without a comparative estimate and uncertainty interval, readers cannot judge how precisely the groups differ on the primary endpoint.

It also shows why endpoint definitions and labels matter. The title and description refer to negative urine samples, while the unit wording refers to positive samples. Clear agreement between an endpoint’s title, unit, and analysis description is essential for a reader to know exactly what a value represents. Here, the stated intent-to-treat rule for missing urine samples is useful detail, but the absence of a reported comparison and a full analysis plan still limits interpretation.

The retention result offers a second statistical lesson: a recalculated estimate and its uncertainty should be distinguished from a reported analysis. The unadjusted retention comparison is imprecise, and the record lists no reported analyses. Planning illustrations based on the observed retention effect can also be mistaken for post hoc power if their purpose is not made explicit. They are planning scenarios under stated assumptions, not observed power and not the trial’s recorded sample-size justification.

Finally, a pilot trial can produce useful descriptive information without resolving a comparative question. A modest enrollment, repeated outcome collection, missing-data rules, and incomplete reporting of inferential methods all affect how much weight the posted results can carry. The record is most useful when its descriptive findings, recalculated statistics, and unreported analysis details are kept separate.

14. 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 trial constrain interpretation. The study is described as a pilot and enrolled 50 participants, split evenly between atomoxetine and matched placebo. That scale provides limited information for precise estimates, particularly when the primary outcome has no estimable comparative analysis in the registry. The record does not report a sample-size justification, so it is not possible to assess the intended precision or power from a stated design calculation.

The primary endpoint was based on repeated urine collection, with samples collected 3 times per week over 12 weeks. The registry describes counting missing samples as positive in the intent-to-treat analysis, but it does not report how many samples were missing, why they were missing, or whether alternative missing-data assumptions were explored. Such details would help assess how sensitive the endpoint summary is to the missing-data rule.

The outcome labeling is inconsistent: the endpoint title and description concern samples negative for cocaine or benzoylecgonine, while the unit says percentage of positive urine samples. The reported primary group values are therefore best interpreted with caution and without converting them into a more specific percentage statement than the record supports. The record also does not provide a primary comparative estimate, confidence interval, or p-value.

For retention, the comparison is an unadjusted recalculation from summary statistics, rather than a listed reported analysis. The registry does not report a multiplicity plan, detailed analysis methods, or interim-analysis information. Details of the contingency management procedures and the specific roles included in the quadruple-masking classification are also not reported in the registry record. These gaps do not show that such methods were absent; they mean the record does not allow them to be evaluated.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Evaluation of Atomoxetine for Cocaine Dependence: A Pilot Trial — NCT00617201.
• Registry record last updated 2017-08-31; status last verified 2017-08.
• Results first posted 2012-12-11.
• 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.

18. Record Summary

NCT00617201 is a completed, randomized, parallel-group Phase 2 pilot trial titled “Evaluation of Atomoxetine for Cocaine Dependence: A Pilot Trial.” The registry lists Sharon Walsh as sponsor, an enrollment of 50, a start of 2007-07, and primary completion in 2011-08. The trial compared atomoxetine with matched placebo, with 25 participants in each arm. The masking classification is quadruple, while the brief summary describes a double-blind trial.

The primary endpoint was the total percentage of urine samples negative for benzoylecgonine over 12 weeks, with missing samples counted as positive in the stated intent-to-treat analysis. Primary arm values are posted, but the comparison is not estimable in the registry record. Retention is the one secondary outcome. The record lists no reported analyses and does not report the sample-size justification, detailed statistical plan, multiplicity approach, or interim-analysis information.