Monitoring of Exhaled Carbon Monoxide to Promote Pre-operative Smoking Cessation: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Monitoring of Exhaled Carbon Monoxide to Promote Pre-operative Smoking Cessation (NCT01014455), a randomized na study. The primary outcome is Exhaled CO Level Measured Immediately Prior to Surgery.

2. Clinical Question

3. Trial Design

The registry describes a completed, randomized, parallel-group study with no masking. It lists the study condition as smoking and the intervention type as behavioral. The trial enrolled 169 participants. Its recorded phase is NA. The sponsor is Mayo Clinic.

Both groups received a brief intervention recommending preoperative abstinence from smoking. In the CO Reminder group, patients were also told that their smoking status would be checked before surgery using exhaled carbon monoxide monitoring. The no CO Reminder group received the brief advice without being told about the check. The comparison therefore concerns the addition of a monitoring reminder to advice, rather than advice versus no advice.

The primary assessment was exhaled CO measured immediately before surgery. The registry describes routine clinical measurement on the morning of surgery, alongside questions about typical cigarette consumption and whether the patient had smoked that day. The monitoring equipment was maintained by the Division of Respiratory Therapy, including regular calibration. The registry gives the median time from study assessment at POE to surgery as 1 day, with an interquartile range of 1 to 3 days.

The registry records a start of 2010-01, primary completion in 2011-03, and status as completed. It does not report the random-sequence generation procedure, allocation concealment, or the original design assumptions for sample size. It also does not report whether adherence to the intervention was assessed beyond the described group assignment. These omissions limit how fully the trial’s implementation and planning can be reconstructed from the registry record.

4. Participant Flow and Accounting

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

Before assignment: Inform consent was obtained after data collection(approved by the IRB) because informing patients of the study at time of enrollment would have required providing information making it impossible to evaluate the hypothesis. Study records from patients who declined informed consent (n=14) were destroyed and are not included in the analysis.

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 Exhaled CO Level Measured Immediately Prior to Surgery, measured over The median time from study assessment at POE to surgery was 1 day with an interquartile range of 1 to 3 days..

The primary endpoint was the mean exhaled carbon monoxide level measured immediately before surgery, recorded in parts per million. Carbon monoxide in exhaled breath can indicate recent exposure to cigarette smoke because it remains in the body for a few hours. In this trial, the measure was used to compare recent smoke exposure between patients who were told that monitoring would occur and those who received brief advice without that reminder.

The registry describes the measurement as part of routine clinical practice on the morning of surgery. Patients were asked about typical cigarette consumption and whether they had smoked that day, and their exhaled CO level was measured with a Micro Smokerlyzer. The record says the monitors were maintained by the Division of Respiratory Therapy, including regular calibration. These details support understanding what was measured and the setting in which it was collected.

The endpoint is a continuous measure, and the posted group summaries report means and standard deviations. A mean captures the group’s average level, while the standard deviation describes the spread of individual measurements around that average. The measure is a short-window indicator: it does not establish longer-term abstinence, quantify all smoking behavior, or measure postoperative complications. The registry’s primary outcome is therefore a behavioral and exposure-related measure rather than a clinical outcome such as a surgical complication.

7. Results Posted in the Registry

• no CO Reminder: mean 9.30 versus 9.70 on control; difference -0.40 (95% CI -2.65 to 1.85), standardized effect d = -0.05; z-test p = 0.728.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary endpoint is continuous, so the posted summaries describe each group using a mean and standard deviation. The mean gives the average measured exhaled CO level. The standard deviation describes how dispersed the individual values are around that average. A larger standard deviation means observations tend to vary more within the group; it is not the uncertainty of the group mean. The registry reports the mean and standard deviation for each randomized group.

What does the mean difference represent?

The recalculation compares the no CO Reminder group with the CO Reminder group, subtracting the latter mean from the former. The reported mean difference is -0.4 ppm. The sign depends on that stated direction of comparison: a negative estimate means the no-reminder group had the lower observed mean. Reversing the comparison would reverse the sign, without changing the underlying separation between the group averages.

A mean difference remains in the original measurement units, ppm. This makes it possible to interpret the contrast on the same scale as the endpoint. The recalculated standardized effect, d = -0.055, instead expresses the difference relative to the pooled standard deviation. As a general statistical concept, standardization can help compare effects measured on different scales, but it may be less directly interpretable than an original-unit difference. Neither statistic by itself determines whether the difference matters clinically.

Why use a t-test?

The registry reports a two-sided t-test. In general, a t-test assesses whether the observed difference in group means is large relative to the estimated variability and sampling uncertainty. A two-sided test considers departures in either direction rather than only a decrease or only an increase. The posted p-value is 0.672. The recalculation based on the group summaries gives an unadjusted p-value of 0.728.

A p-value describes how compatible the observed data are with a null model of no population mean difference, subject to the test’s assumptions. It is not the probability that the null model is true, and it does not quantify the probability that a finding will replicate. It also does not convey the range of effect sizes compatible with the data; that information is better addressed by the confidence interval.

What does the confidence interval add?

The recalculated 95% confidence interval for the difference is -2.651 to 1.851 ppm. In general, a confidence interval reflects both the estimated effect and its sampling uncertainty. Under repeated use of the same valid method, intervals constructed in this way would cover the true effect at the stated rate. For an individual interval, the practical reading is that the data are compatible with a range of values, subject to the statistical model and design.

Because this interval crosses zero, it does not clearly distinguish a negative mean difference from no difference or a positive mean difference. Its width also signals limited precision: plausible values extend across both sides of zero. The interval is not a guarantee that the true value lies within these particular limits, and it does not account for unrecorded sources of bias or errors in the study process.

Why can a recalculation differ from the registry test?

A recalculation from aggregate means, standard deviations, and group sizes is an unadjusted analysis. It cannot reproduce methods that use individual participant data, adjustments, or other details that are not present in the summary statistics. The registry’s own reported test is the primary record of the sponsor-reported analysis. Here, the registry reports p = 0.672, while the posted unadjusted recalculation gives p = 0.728. The registry record does not provide enough detail to explain the numerical difference. The discrepancy flag is false, and the contrast between these p-values does not change the broad interpretation of substantial uncertainty around a small estimated mean difference.

What assumptions matter in general?

As a general statistical matter, a t-test relies on an appropriate model for the data and the sampling process. The distribution of the outcome, the influence of extreme observations, independence of observations, and the way participants were assigned can all affect how well an analysis represents uncertainty. The registry does not report diagnostics for the CO distribution or details of any adjustments. It is therefore not possible to assess those features from the posted aggregate results alone.

Randomization supports a fair comparison when implemented appropriately, but it does not guarantee identical groups on every characteristic, especially in a study with limited information. The registry says the study was randomized but does not report its sequence-generation or concealment procedures. The analysis should consequently be described as a comparison of the recorded randomized groups, without assuming unreported design or analysis protections.

10. Confidence Intervals and Precision

The recalculated 95% confidence interval for the mean difference, defined as no CO Reminder minus CO Reminder, is -2.651 to 1.851 ppm. The point estimate is -0.4 ppm. The interval includes zero and extends to both negative and positive values. Thus, the posted summary statistics do not tightly identify the direction of the underlying mean difference.

In general, a confidence interval communicates the range of effect estimates that are reasonably compatible with the data under the analysis model. A wider interval represents less precision than a narrower interval. The interval’s width is important alongside the estimate near zero: the data do not establish exact equality, and they also do not identify a clearly separated effect in one direction. The interval should not be read as a probability statement about where the true value lies after observing this particular dataset.

The group standard deviations are 8.0 ppm for CO Reminder and 6.6 ppm for no CO Reminder, compared with a mean difference of -0.4 ppm. The observed variation within groups is substantial relative to the estimated separation between their averages. That context helps explain why the confidence interval permits a range of effects even though the observed means are close.

The precision statement is conditional on the analysis and data captured in the registry. It does not address possible measurement error, departures from statistical assumptions, missing-outcome mechanisms, or bias in implementation. The registry reports that 5 participants did not complete CO measures and were excluded from the primary-outcome analysis; it does not report their group distribution or the reasons for missing measurements. The posted interval therefore describes the analyzed outcome data, not all possible uncertainty about the study process.

11. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning figures are participants per arm, not total events. They are illustrative calculations based on the observed effect, not the trial’s original design assumptions, which are not reported in the registry record. The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, a normal approximation, and an effect to detect set at the stated fraction of the observed effect.

The 80% power column gives the planned number of participants per arm for a future study to have an 80% chance of detecting the chosen effect under those assumptions, if that effect is the true one. The 90% power column uses a higher target probability of detecting that effect, so it requires more participants per arm. These are design targets, not guarantees about what a future study will find.

The rows consider planning for 100%, 75%, and 50% of the observed effect. For example, the figures for 75% of the observed effect are 9,337 participants per arm for 80% power and 12,499 participants per arm for 90% power. For 50% of the observed effect, the corresponding figures are 21,007 and 28,122 participants per arm. For 100% of the observed effect, they are 5,252 and 7,031 participants per arm. Smaller effects require more participants because they are harder to distinguish from random variation under the same assumptions.

These calculations are not observed power. They are planning illustrations and should not be used to infer the probability that the completed trial detected an effect. They also inherit the uncertainty of the observed effect used as their basis: an imprecise estimate can produce unstable planning values, particularly when the estimated effect is small.

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, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = -0.400; pooled standard deviation σ = 7.316.
2. n = 2 × (1.96 + 0.842)² × 7.316² / -0.400² = 5,251.7, rounded up to 5,252 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
delta <- -0.399999999999999                      # observed difference in means
sigma <- 7.31625958540207                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^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 (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(1014455)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 5,252 (80% power), 7,031 (90% power)
 75% of observed effect: n per arm = 9,337 (80% power), 12,499 (90% power)
 50% of observed effect: n per arm = 21,007 (80% power), 28,122 (90% power)
power.t.test: 5252.6 
Simulated power at n = 5,252 per arm: 0.802 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-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. Analyses Reported by the Sponsor

The registry’s own reported analysis is a two-sided t-test with p = 0.672. The unadjusted recalculation from the posted group summaries gives p = 0.728 and a mean difference of -0.4 ppm, with a 95% confidence interval from -2.651 to 1.851 ppm. The sponsor-reported p-value and the recalculated p-value are not identical.

The record does not provide enough detail to determine why the p-values differ. A recalculation from aggregate summaries may not reproduce the original analysis if the original procedure used information or choices that are not captured in the summary statistics. The registry record does not report an adjusted estimate, the original analysis dataset, or a fuller explanation of the test. The discrepancy flag is false, so the posted analysis does not identify a discrepancy requiring a precedence statement.

Both p-values are consistent with considerable uncertainty about the mean difference, and neither should be read as a direct measure of effect size or clinical importance. The registry’s own analysis remains the sponsor-reported result; the recalculation is a simple unadjusted check using the posted summaries. The recalculated confidence interval is useful for describing precision, but it should not be presented as though it were a confidence interval reported by the sponsor. The registry does not give a sponsor-reported estimate or confidence interval for this outcome.

14. Multiplicity, Missing Data and Analysis Populations

The registry lists 2 outcome measures and 1 analysis. It does not report a multiplicity adjustment or a prespecified hierarchy for interpreting multiple outcomes. Multiplicity becomes relevant when several outcomes, time points, subgroups, or statistical tests are examined, because repeated testing can increase the chance of at least one apparently unusual result. Without a reported analysis plan, the extent to which this issue applies cannot be assessed. The record reports no secondary outcomes in the posted results.

The primary-outcome population description says that 5 participants did not complete CO measures and were excluded from the analysis of the primary outcome. The registry does not report which group those participants were assigned to, why measurements were unavailable, or whether any method was used to address missing outcomes. It also does not report a sensitivity analysis for missing data. The reported group summaries therefore describe the participants with observed CO measurements included in the analysis, rather than an explicitly documented analysis of every randomized participant.

The registry record does not report whether the analysis followed an intention-to-treat principle, whether participants were analyzed as randomized regardless of adherence, or whether any alternative analysis population was used. It does not report protocol deviations, exclusions beyond the stated missing CO measures, or the handling of other incomplete data. These omissions make it difficult to evaluate how missingness and analysis-population decisions could affect the estimate.

These are reporting limits, not evidence that a particular analysis was performed incorrectly. They matter because statistical interpretation depends not only on the test and estimate but also on which outcomes and participants enter the analysis, and on whether those choices were planned in advance. The posted results support description of the recorded endpoint analysis, while the broader analysis plan and missing-data assumptions are not reported in the registry record.

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

The posted results show that the observed mean exhaled CO levels were close: 9.7 ppm in the CO Reminder group and 9.3 ppm in the no CO Reminder group. The recalculated mean difference is -0.4 ppm for no CO Reminder minus CO Reminder. The uncertainty interval includes zero and allows values on either side, so these results do not establish a clear difference in average preoperative CO between the randomized groups.

This does not prove that the reminder had no effect. A result that does not clearly distinguish a difference from zero can arise when the underlying effect is small, when measurements vary, or when the estimate is imprecise. The confidence interval is more informative than a binary significant-or-not label because it communicates a range of compatible values. Even so, it describes statistical uncertainty under the analysis assumptions and does not resolve every possible source of bias or missingness.

The endpoint measures recent carbon monoxide exposure near surgery. It does not directly establish a person’s longer-term smoking pattern, whether they quit after surgery, or whether surgical complications changed. The registry does not report secondary outcome results, and it gives no result for postoperative complications. Claims about those outcomes are therefore not supported by the posted primary-endpoint data.

The comparison is specifically between brief advice plus a CO-monitoring reminder and brief advice without that reminder. It does not estimate the effect of brief advice compared with no advice. Nor does it establish that carbon monoxide monitoring itself, independent of informing patients about it, caused any observed difference. The most defensible reading is limited to the measured group means and the uncertainty around their contrast, within this trial’s recorded setting and analysis.

16. Why This Trial Matters Statistically

This trial raises a useful statistical question about separating an intervention’s behavioral message from its measurement procedure. Both groups received brief advice; the additional feature in one group was being told that CO would be checked. The primary endpoint was itself measured immediately before surgery. The design therefore focuses attention on whether awareness of monitoring is associated with a difference in a short-window exposure measure.

The result also illustrates why an estimate should be interpreted together with its uncertainty. The observed mean difference is small relative to the group standard deviations, while the confidence interval includes zero and permits effects in either direction. Reporting only a p-value would conceal that range. Reporting only the point estimate would give an incomplete impression of how much remains uncertain.

Aggregate results allow a transparent unadjusted comparison, but they do not reproduce every possible feature of an original analysis or reveal the distribution of individual measurements. The difference between the registry-reported and recalculated p-values is a reminder that numerical summaries may not fully document analysis choices. Clear reporting of the analysis population, missing measurements, prespecified outcomes, and methods would support a more complete assessment.

The planning illustrations further show the sensitivity of future sample-size requirements to the effect chosen for detection. Smaller target effects require more participants per arm under the stated assumptions. Because those illustrations use the observed effect, their large values should be understood as planning scenarios rather than as a judgment about the completed trial’s power or the value of the intervention. The trial’s statistical importance lies in the distinctions it makes possible: an observed estimate, its precision, the analysis used, and the limits of what the endpoint can answer.

17. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The primary result is based on the participants with completed CO measures; the registry says 5 participants were excluded from that analysis because they did not complete the measurements. The reasons for missing measures and their distribution across groups are not reported in the registry record. If missingness was related to group assignment or CO level, the analyzed comparison could differ from the comparison that would have been observed with complete data.

The registry describes a randomized, unmasked study, but it does not report sequence-generation or allocation-concealment procedures. Because patients knew whether they had been told about monitoring, the behavioral intervention was not masked. The record does not describe whether staff measuring or recording CO were aware of assignment, or whether that could have influenced measurement or data handling.

The endpoint captures recent exposure over a short period and was assessed on the morning of surgery. The registry reports a median interval of 1 day from study assessment at POE to surgery, with an interquartile range of 1 to 3 days. This outcome does not provide a longer-term measure of smoking behavior. The registry does not report follow-up results for quitting, complications, or other clinical outcomes.

The posted data are aggregate summaries. Individual measurements, distributional details, covariate balance, and diagnostic checks are not reported. This limits assessment of skewness, influential observations, or other features that could matter for a mean-based analysis. The record also does not report the original sample-size assumptions, a detailed statistical analysis plan, or a rationale for the analysis population. These limitations constrain interpretation but do not, on their own, establish that the observed comparison is biased.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Monitoring of Exhaled Carbon Monoxide to Promote Pre-operative Smoking Cessation — NCT01014455.
• Registry record last updated 2011-08-22; status last verified 2011-07.
• Results first posted 2011-08-22.
• 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

NCT01014455 is a completed randomized, parallel-group study of a CO-monitoring reminder for smokers scheduled for surgery. Both groups received brief advice recommending preoperative abstinence; one group was also told that CO monitoring would occur before surgery.

The primary endpoint was mean exhaled CO immediately before surgery, measured in ppm. The CO Reminder group had a mean of 9.7 ppm with a standard deviation of 8.0, and the no CO Reminder group had a mean of 9.3 ppm with a standard deviation of 6.6. The registry reports a two-sided t-test with p = 0.672. The unadjusted recalculation estimates a difference of -0.4 ppm, with a 95% confidence interval from -2.651 to 1.851 ppm.

The interval includes zero and indicates uncertainty about the direction and size of the mean difference. The result applies to this short-window CO endpoint; longer-term smoking behavior and surgical outcomes are not reported. The registry does not report the original sample-size assumptions, detailed missing-data methods, or a full statistical analysis plan.