Increasing Contingency Management Success in Smoking Cessation: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Increasing Contingency Management Success in Smoking Cessation (NCT00273793), a randomized na study. The primary outcome is Breath Carbon Monoxide Levels Indicating Smoking Abstinence During the Study, i.e., the Number of Breath Samples With Carbon Monoxide (CO) Levels Less Than 3 Parts Per Million (Ppm).

2. Clinical Question

3. Trial Design

This was a randomized, parallel-group behavioral study of contingency management for smoking cessation. The registry lists the allocation as randomized and masking as none. It reports enrollment of 328 participants and marks the study completed. The study began in 2005-06, with primary completion in 2010-06. The registry lists the phase as NA.

The six intervention arms fall into two labeled groups. Among hard-to-treat smokers, the approaches were shaping, a fixed criterion, and non-contingent incentives. Among smokers with early success, the approaches were ascending incentive values, fixed values, and non-contingent incentives. The trial description says it investigates two features of incentives: whether incentives follow cessation or smoking reductions, and whether incentive amounts remain fixed or increase with continued success. The registry does not give further operational detail about these approaches, such as the specific incentive schedule or eligibility rules for the labeled groups.

Random allocation supports a planned comparison of assigned approaches, but the registry record does not report the allocation ratio, sequence-generation method, concealment procedures, or whether any stratification was used. It also does not report a protocol-based sample-size justification. The primary endpoint is reported for participants randomized to condition, with the number of participants varying across arm summaries. The record does not explain the reasons for those differences.

Results include the primary breath carbon monoxide endpoint and a secondary measure of cigarettes reported per day at follow-up six months after study entry. The registry lists no reported analyses. It does not describe a prespecified primary contrast, a statistical model, an interim-analysis plan, or a procedure for handling missing outcomes. Those omissions limit how completely the reported results can be linked to the trial’s intended analysis.

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: Five weekdaily visits that took about 5 minutes occurred before randomization. 328 individuals consented and of these 239 completed baseline and were randomized. Which set of three study arms individuals were randomly assigned to depended on whether they successfully delivered a C) \< 3 ppm during this baseline for which they earned an incentive.

Overall Study

Analysis population for the primary outcome, as defined in the registry: all subjects randomized to condition

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.

Only the first 6 of 7 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Breath Carbon Monoxide Levels Indicating Smoking Abstinence During the Study, i.e., the Number of Breath Samples With Carbon Monoxide (CO) Levels Less Than 3 Parts Per Million (Ppm), measured over daily for breath CO.

The primary endpoint is the number of breath samples with carbon monoxide (CO) levels below 3 ppm during the study. A breath CO measurement is an objective indicator used to assess recent smoking status. Counting samples below a specified threshold turns repeated measurements into a count of occasions meeting the endpoint definition; a larger count means more samples met that definition, but does not by itself show the timing or continuity of abstinence.

The registry describes the endpoint as measured daily for breath CO and identifies the analysis population as all subjects randomized to condition. It summarizes the count using a median and interquartile range. The median describes the middle of the observed distribution, while the interquartile range conveys the spread of the central half of the observations. These summaries are useful when a count outcome may be unevenly distributed or contain many low values, but they do not estimate a treatment contrast on their own.

The posted arm medians and interquartile limits are 2.0 (0.0 to 12.5) for shaping, 3.5 (0.0 to 39.0) for the fixed criterion, and 0.0 (0.0 to 12.0) for non-contingent incentives among hard-to-treat smokers. Among smokers with early success, the corresponding summaries are 58.5 (15.0 to 64.0), 49.5 (25.0 to 61.0), and 28.0 (6.0 to 59.0). These are descriptive arm summaries; no primary inferential comparison is reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What do the primary summaries represent?

The primary outcome is a count: the number of breath samples with CO below 3 ppm. The registry summarizes that count with a median and interquartile range. The median is the midpoint of an ordered set of observations, not an arithmetic average. The interquartile limits describe the central half of the observed distribution, with one limit below and one above the median. This pair of summaries gives information about both a typical value and the spread of observations without assuming that the outcome is normally distributed.

For example, the hard-to-treat shaping arm has a median of 2.0 and interquartile limits from 0.0 to 12.5. The lower limit is not a minimum, and the upper limit is not a maximum: they mark the reported interquartile range. The summaries therefore do not give the full distribution, the frequency of each count, or the number of participants who met a separate abstinence definition. Those details are not reported in the registry record.

Why are arm summaries not the same as a treatment comparison?

An arm-specific median describes one group. A treatment comparison estimates how an outcome differs between groups and quantifies uncertainty around that difference. The primary endpoint is posted as six arm summaries, but no primary contrast, confidence interval, or test is reported. It is not appropriate to infer statistical significance by comparing medians visually or by treating the difference between them as an estimated treatment effect. The primary endpoint is also a count, while the secondary recalculations concern a continuous mean outcome; the methods and interpretation are not interchangeable.

How should the secondary recalculations be read?

The registry posts means and standard deviations for cigarettes reported per day at six-month follow-up. The recalculated contrasts are simple, unadjusted comparisons of the shaping arm for hard-to-treat smokers with the fixed-criterion arm and the non-contingent arm in that labeled group. For the fixed-criterion contrast, the recalculated mean difference is -0.2, with a 95% confidence interval from -5.14 to 4.74 and p = 0.937. For the non-contingent contrast, the recalculated mean difference is 2.3, with a 95% confidence interval from -3.592 to 8.192 and p = 0.444.

A mean difference expresses the difference between the arm means using the stated comparison. A confidence interval gives a range of values compatible with the estimate under the calculation’s assumptions. Both intervals include 0, a value corresponding to no mean difference. The p-values are unadjusted, and these calculations are not identified as the trial’s prespecified analyses. The sponsor’s own analysis is not posted, so the recalculations should be treated as descriptive supplementary analyses rather than as a substitute for a full trial analysis.

Standard deviation describes how spread out individual observations are around a group mean. It is not the same as a standard error, which describes uncertainty in an estimate. The posted secondary arm summaries include standard deviations, and the recalculation uses those summary statistics to estimate mean differences and uncertainty. These methods do not account for any covariates, repeated observations, or other features beyond the posted summary data. Whether the original analysis used additional adjustment is not reported in the registry record.

What do the planning illustrations mean?

The planning illustrations estimate participants per arm for 80% or 90% power using the observed secondary effect as a reference. The stated assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The rows illustrate planning for 100%, 75%, or 50% of the observed effect. For the fixed-criterion contrast, the illustration gives 41,553 participants per arm for 80% power when planning for 100% of the observed effect, 73,871 for 75%, and 166,209 for 50%. For the non-contingent contrast, the corresponding figures are 363, 646, and 1,452 participants per arm. The 90% power figures are also reported in the planning results.

These are planning illustrations based on the observed effects, not the sample-size justification for the completed trial. They are not observed power and do not indicate the probability that the trial’s findings are true. Because the observed effect is uncertain, using it as a planning anchor can produce unstable estimates, especially when the observed contrast is small. Planning for a fraction of that effect illustrates how the required sample size changes under the stated assumptions, but does not validate those assumptions or recover the trial’s original design rationale.

Why can over-dispersion matter for a count?

In general, over-dispersion means that a count outcome varies more than a simple model would expect. If that feature is present and ignored, standard errors and tests from an unsuitable model can be unreliable. A count endpoint with many low values and a broad spread may call for careful assessment of its distribution and analysis model. The registry record does not report a model for the primary count outcome, so it is not possible to determine whether over-dispersion was assessed or how it was handled.

The registry also does not report a detailed statistical analysis plan, model diagnostics, an interim-analysis plan, or a multiplicity procedure. Those omissions matter because a result’s interpretation depends not only on its summary statistic, but also on the planned contrast, model, assumptions, and handling of multiple comparisons and incomplete observations.

10. Multiplicity, Missing Data and Analysis Populations

The registry record does not report a multiplicity procedure. It lists a primary endpoint, a secondary endpoint, six intervention arms, and two recalculated secondary comparisons, but does not state which comparisons were prespecified as primary contrasts or how the family of possible tests was defined. When multiple outcomes or comparisons are evaluated, the chance of at least one apparently unusual result can increase. Procedures such as adjustment, hierarchical testing, or designation of a limited set of primary contrasts can address that issue, but the record does not say whether any such approach was used.

The registry lists no reported analyses. In particular, it does not provide a primary endpoint test, an inferential model for the count outcome, or a complete comparison of the arms. The available secondary p-values are explicitly unadjusted recalculations, not evidence of the trial’s full statistical analysis. Whether the investigators conducted additional analyses or adjusted for multiplicity is not reported in the registry record.

Missing-data handling is also not reported. The registry gives different participant counts across the posted outcomes and arms, but does not explain the reasons for those counts or describe imputation, exclusion rules, or sensitivity analyses. A difference in the number contributing to a result does not by itself identify why data are absent. Without further information, it is not possible to assess whether missing observations could have affected the estimates.

The primary endpoint description identifies all subjects randomized to condition as the population of interest. However, the record does not explain how that population was operationalized when outcome data were unavailable, or whether analyses followed an intention-to-treat principle. The secondary results likewise do not identify a specific analysis population beyond the posted arm counts. Interim analyses, stopping rules, and a prespecified analysis plan are not reported in the registry record. These omissions limit assessment of whether the posted results reflect the planned analysis and how robust they are to alternative assumptions.

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 1% reporting threshold.

The registry reports 0 serious adverse events affected and 0 other adverse events affected in each of the six arms. The at-risk counts match the posted primary outcome arm counts: 36, 38, 18, 60, 58, and 29 across the listed groups. These are the registry’s reported event figures; they should be read as safety reporting for the recorded categories, not as a complete description of every possible harm or symptom.

The record lists no serious adverse event terms. Death counts are null rather than reported numerical values, so deaths are not reported in the registry record. A zero affected count does not establish that no event occurred outside the categories or reporting procedures captured in the registry. The record does not describe the adverse-event collection period, definitions, severity grading, or attribution rules. These figures therefore provide limited information about recorded adverse events and should not be interpreted as a comprehensive safety assessment.

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

The posted results show the distributions of the primary breath CO count in six intervention arms and arm-level summaries of cigarette consumption at follow-up six months after study entry. They describe what was recorded for these groups. The primary medians differ across arms, and the interquartile limits show variation within arms. The secondary means also vary across groups. These summaries can guide questions about patterns and uncertainty, but they do not by themselves establish that one incentive approach caused a better outcome than another.

The primary outcome has no posted test or confidence interval, and the registry does not report the investigators’ analysis of it. The secondary calculations are limited to two contrasts and are unadjusted. Their confidence intervals include 0, and they do not provide a complete comparison across the six-arm design. Because these calculations are simple re-estimations from posted summary statistics, they should not be treated as the trial’s definitive analysis.

The arm labels refer to hard-to-treat smokers and smokers with early success, but the record does not provide enough detail to define those categories or to compare them as if they were randomized groups. The difference between their primary outcome summaries may reflect differences between the labeled populations as well as other factors; the registry does not provide an analysis that separates those possibilities.

These findings do not establish that an intervention works or does not work, and they do not show what an individual participant would experience. The primary count measures samples below a CO threshold, not a complete pattern of long-term abstinence. The secondary measure is self-reported consumption during a past week at follow-up, not a direct substitute for the primary endpoint. The record also does not report missing-data methods, multiplicity handling, or enough of the original statistical plan to evaluate the full inferential process. The soundest reading is therefore descriptive and cautious: use the posted summaries to understand observed patterns, while recognizing that comparative conclusions remain uncertain.

13. Why This Trial Matters Statistically

This trial is statistically interesting because its intervention question is organized around two different incentive features: the criterion for earning incentives and whether incentive amounts are fixed or increase with continued success. It also separates arms under two labels, hard-to-treat smokers and smokers with early success. That design raises an important interpretive issue: arm summaries across those labels are not automatically comparisons of incentive strategies, since the registry does not explain how the categories were defined or provide a direct cross-category analysis.

The primary outcome is a repeated-measure count summarized with medians and interquartile limits. Such summaries preserve information about the center and spread of a potentially uneven count distribution, but they do not answer the comparative question without a formal analysis. The absence of a reported primary contrast illustrates why descriptive results and inferential evidence must be distinguished. A visible difference in arm medians is not equivalent to an estimated treatment effect with quantified uncertainty.

The secondary results offer a separate example of the limits of post hoc interpretation. Recalculations from summary means and standard deviations can make a contrast and its uncertainty more interpretable, but only for the comparisons that can be reconstructed. They cannot recover an unreported model, multiplicity adjustment, missing-data strategy, or prespecified analysis hierarchy. Planning illustrations based on observed effects can be informative as hypothetical design exercises, but they are not evidence of the trial’s original sample-size rationale or of observed power.

Overall, the record highlights the value of reporting a complete statistical plan alongside results: the planned estimand, comparison set, analysis population, model, uncertainty measures, and handling of missingness and multiple testing. Without those details, even a randomized design and arm-level outcomes leave important questions unresolved about the strength and scope of comparative conclusions.

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.

A central limitation is that the registry does not report inferential results for the primary endpoint. It provides median and interquartile summaries by arm but no formal primary comparison, confidence interval, or test. This makes it difficult to distinguish observed differences from sampling uncertainty or to determine which contrast was intended to answer the study’s principal question.

The record also does not report the definitions of hard-to-treat smokers or smokers with early success. Since the primary summaries differ between these labeled groups, the lack of criteria restricts interpretation of those patterns and limits the ability to determine how the groups relate to the randomized comparisons. The registry does not provide an allocation ratio, randomization implementation details, or the original sample-size justification.

Outcome counts vary across arms and between outcomes. The reasons are not reported, and missing-data handling is not described. The registry does not provide a statistical analysis plan, multiplicity procedure, primary model, or interim-analysis information. The available secondary recalculations are unadjusted and cover only two contrasts, leaving other comparisons unreported. The primary count outcome also lacks a reported distributional model, so the approach to any skewness or over-dispersion cannot be assessed.

Finally, the primary endpoint counts breath samples under a CO threshold, while the secondary endpoint is self-reported cigarette consumption at a six-month follow-up. They capture different aspects and time frames of smoking behavior. Adverse-event reporting gives zero affected counts in the listed categories but does not describe collection methods or reporting periods, and death counts are not reported in the registry record. These features constrain both efficacy-related interpretation and safety assessment.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Increasing Contingency Management Success in Smoking Cessation — NCT00273793.
• Registry record last updated 2012-06-21; status last verified 2012-06.
• Results first posted 2012-06-07.
• 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.

18. Record Summary

NCT00273793, titled “Increasing Contingency Management Success in Smoking Cessation,” is a completed randomized, parallel-group behavioral study sponsored by The University of Texas Health Science Center at San Antonio. The registry reports enrollment of 328, study start in 2005-06, and primary completion in 2010-06. Masking is listed as none, and phase is listed as NA.

The six arms compare shaping, fixed criteria or values, and non-contingent incentives among groups labeled hard-to-treat smokers or smokers with early success. The primary outcome is the number of breath samples with CO below 3 ppm during the study, summarized by median and interquartile range. The secondary outcome is average cigarettes reported per day in the past week at follow-up six months after study entry.

The registry provides arm-level outcomes but no reported analyses. It does not report a primary inferential comparison, detailed analysis plan, missing-data method, or multiplicity procedure. Two unadjusted secondary contrasts are recalculated from the posted summary statistics. These results support descriptive interpretation, while important questions about the planned and complete comparative analysis remain unresolved.