Does Capnography Prevent Hypoxemia During Endoscopic Retrograde Cholangiopancreatography and Endoscopic Ultrasound?: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Does Capnography Prevent Hypoxemia During Endoscopic Retrograde Cholangiopancreatography and Endoscopic Ultrasound? (NCT00675415), a randomized na study. The primary outcome is Number of Participants Experiencing Hypoxemia During Endoscopy.

2. Clinical Question

3. Trial Design

The registry record describes a randomized, controlled, parallel-group trial with triple masking. Participants were assigned to capnography-titration or standard monitoring. In the titration arm, the endoscopy team had access to capnographic findings; in the blinded arm, the team did not. The record labels masking as triple, but does not identify which groups of people were masked. The trial phase is listed as NA.

The registry lists an enrollment of 263 and a completed status. The study start is recorded as 2006-09, with primary completion in 2008-09. These registry entries describe the trial record; they do not, by themselves, establish which participants were included in each posted outcome analysis. The primary outcome counts are reported for 124 participants in the capnography arm and 123 in standard monitoring.

The intervention was capnography, categorized in the registry as a diagnostic test. The clinical contrast was whether the endoscopy team could respond to capnographic abnormalities as they arose, against a comparator in which that information was unavailable to the team. Both arms were described as receiving standard monitoring and sedation. The study summary states that the investigators hypothesized that using capnography could prevent hypoxemia.

The registry record does not report the original sample-size justification, the detailed randomization procedure, the specific parties included in triple masking, or the planned statistical analysis. It lists The Cleveland Clinic as sponsor. The description supports a comparison of the two monitoring strategies in elective ERCP and EUS, but does not give enough detail to characterize additional aspects of trial conduct.

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 Number of Participants Experiencing Hypoxemia During Endoscopy, measured over Continuously Assessed during a single endoscopic procedure, typically about 2 hours.

The primary endpoint is the count of participants who experienced hypoxemia during endoscopy. The registry definition requires oxygen saturation less than 90 percent for at least 15 seconds. Continuous assessment during the procedure means the outcome concerns whether a participant crossed this threshold for the stated duration at any point during the procedure, rather than a single measurement at a fixed time.

A threshold-based binary endpoint turns a sequence of oxygen-saturation observations into an event status for each participant: the registry reports whether the defined hypoxemia event occurred. This makes the outcome directly interpretable as a comparison of the number of participants meeting the definition in each arm. It also means the result does not describe the duration, timing, or full pattern of oxygen-saturation changes for each participant.

The endpoint is relevant to the trial’s stated question because capnographic abnormalities could alert the endoscopy team to respiratory changes while a procedure is underway. The primary outcome captures low oxygen levels under a prespecified threshold definition, not all respiratory events or every possible consequence of monitoring. The registry record gives the assessment timeframe as continuous during a single endoscopic procedure, typically about 2 hours. It does not report further details about how missing or unreadable measurements were handled.

7. Results Posted in the Registry

• Capnography: 46.0% (57/124) versus 69.1% (85/123) on control; risk difference -23.1 percentage points (95% CI -35.1 to -11.2), risk ratio 0.67 (0.53–0.83); unadjusted two-proportion z-test p = <0.001.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Proportion of Participants Requiring Supplemental Oxygen

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Capnography: 52.4% (65/124) versus 66.7% (82/123) on control; risk difference -14.2 percentage points (95% CI -26.4 to -2.1), risk ratio 0.79 (0.64–0.97); unadjusted two-proportion z-test p = 0.023.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically lower for Capnography than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Proportion of Participants Experiencing Severe Hypoxemia

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Capnography: 15.3% (19/124) versus 30.9% (38/123) on control; risk difference -15.6 percentage points (95% CI -25.9 to -5.2), risk ratio 0.50 (0.30–0.81); unadjusted two-proportion z-test p = 0.004.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically lower for Capnography than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Proportion of Participants Experiencing Apnea

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Capnography: 41.1% (51/124) versus 62.6% (77/123) on control; risk difference -21.5 percentage points (95% CI -33.6 to -9.3), risk ratio 0.66 (0.51–0.84); unadjusted two-proportion z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically lower for Capnography than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The registry lists three secondary outcomes: requiring supplemental oxygen, severe hypoxemia, and apnea. Across these outcomes, the posted comparisons follow the same general direction as the primary outcome: the estimated event proportions are lower in the capnography arm than in the standard-monitoring arm. This coherent direction offers context for the primary finding because the secondary measures relate to oxygen support and respiratory events during endoscopy.

These outcomes are related but not interchangeable with the primary endpoint. Supplemental oxygen describes a clinical response, severe hypoxemia is a more restrictive outcome, and apnea concerns a different respiratory event. Agreement in direction does not establish that the measures capture the same underlying process, nor does it independently establish a causal mechanism. They add descriptive context rather than replacing the primary comparison.

The secondary p-values are unadjusted. When several outcomes are examined, the chance of obtaining at least one apparently strong result by chance can increase. Without a reported multiplicity strategy, the secondary comparisons should therefore carry less inferential weight than a prespecified primary outcome. The registry record does not report a testing hierarchy, multiplicity adjustment, or analyses explaining how the secondary results were incorporated into the study’s conclusions. The secondary pattern is informative, but it should be interpreted as supportive descriptive evidence rather than as a set of independent confirmations.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Proportion of Participants Requiring Supplemental Oxygen
• Proportion of Participants Experiencing Severe Hypoxemia
• Proportion of Participants Experiencing Apnea

9. Statistical Interpretation

10. Statistical Methods Explained

The outcome is binary at the participant level: each participant either met the hypoxemia definition or did not, and the registry reports the number meeting it. The comparison is between randomized arms in a parallel-group trial. This structure permits a direct comparison of event proportions, provided the analysis uses appropriate participant-level outcome data and handles missing observations transparently. The registry record does not report the detailed analysis population or missing-data approach.

What does the risk difference express?

The risk difference compares the event proportion in the capnography arm with the event proportion in the standard-monitoring arm. A value below zero means the observed proportion is lower in the capnography arm under this direction of comparison; a value above zero would mean it is higher. Unlike a relative measure, the risk difference expresses the contrast on the absolute probability scale. Here, the posted risk difference is -0.231. It should be interpreted with its confidence interval, not as an isolated point estimate.

What does the confidence interval add?

A confidence interval conveys uncertainty around an estimated effect under the statistical procedure used to form it. Narrower intervals indicate greater precision, while wider intervals indicate more uncertainty about the location of the effect. The interval for the primary risk difference extends from -0.351 to -0.112. Its endpoints remain on the negative side of zero, but the range still reflects uncertainty about the magnitude of the contrast. The registry record does not specify the confidence level or the method used to calculate the interval, so those details cannot be assessed from the record.

What does the p-value say?

The reported p-value is an unadjusted test result. In general, a p-value evaluates how incompatible the observed data are with a stated null model, assuming that model and the test’s conditions apply. A small p-value does not state the probability that the null is true, measure the importance of the effect, or account for all sources of uncertainty. Because the posted p-value is unadjusted, it does not account for the fact that the record also lists secondary outcomes. The registry does not report a multiplicity procedure or the specific test used.

Why is randomization relevant?

Random assignment is intended to make the trial arms comparable, on average, with respect to factors that could affect outcomes. It strengthens the basis for comparing assigned strategies, but it does not guarantee that every relevant characteristic was balanced in this particular trial. The registry identifies the allocation as randomized but does not describe sequence generation, concealment, stratification, or any adjustment for baseline characteristics. Those features therefore cannot be evaluated from the registry record.

Masking can reduce the influence of knowledge about assignment on care or outcome assessment. The record labels the trial triple-masked and describes the capnography-blinded arm as one in which the endoscopy team could not access capnographic information. It does not identify the three masked parties, or explain how masking was maintained for outcome assessment. The intervention itself necessarily differed in whether the team could see the tracing, so the record’s label alone is not enough to characterize all potential sources of bias.

Finally, the endpoint definition determines what the analysis can address. A participant is counted as having hypoxemia if oxygen saturation is below the stated threshold for the stated duration during the procedure. The count does not preserve details about event timing, recurrence, or duration. The registry record lists continuously assessed outcomes but does not specify the full measurement protocol, data-quality rules, or statistical model. Accordingly, the posted comparison is best read as an unadjusted contrast of the recorded participant counts, not as a fully documented account of the trial’s prespecified analysis.

11. Confidence Intervals and Precision

The primary risk-difference estimate is -0.231, and its confidence interval ranges from -0.351 to -0.112. The interval describes the uncertainty around the estimated contrast under the method used to calculate it. It is not a range of individual participant outcomes, and it does not mean that every value in the range is equally likely. The registry record does not report the confidence level or interval-calculation method.

The entire interval lies below zero. This indicates that the posted uncertainty interval does not include no difference for this unadjusted estimate. The estimate itself is more informative when considered together with the interval: the point estimate gives the central estimated contrast, while the limits indicate a span of values consistent with the data and calculation method. The interval also makes clear that the exact magnitude is not known with certainty.

Precision is distinct from statistical significance. A p-value can summarize evidence against a null model, while a confidence interval makes the range of estimated effects easier to inspect. Neither resolves potential limitations in measurement, missing-data handling, analysis selection, or generalizability. The registry record does not report those analysis details, and the confidence interval should therefore be understood as precision for the posted unadjusted comparison rather than a complete expression of every uncertainty about the trial.

12. 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 a total across arms. They are illustrations based on the observed effect, not the trial’s original design assumptions. The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set as the stated fraction of the observed effect. The original trial’s sample-size justification and assumptions are not reported in the registry record.

The 80% power column describes the number of participants per arm estimated to give an 80% chance of detecting the selected effect under the planning assumptions, if that effect and those assumptions hold. The 90% power column describes the corresponding planning target for a 90% chance. These are design calculations, not guarantees that a future study will detect an effect. They also do not show the power achieved by this completed trial; this is not observed power.

The rows consider planning for 100%, 75%, and 50% of the observed effect. A smaller effect is harder to distinguish from random variation, so the planning calculations call for more participants per arm as the target effect becomes smaller. For example, the primary-outcome illustration for 75% of the observed effect lists 124 participants per arm for 80% power and 165 for 90% power. Such figures help show how design requirements depend on the effect targeted, but they should not be mistaken for the trial’s enrolled or analyzed group sizes.

The planning calculations use an observed effect as their starting point, which can make them optimistic if that estimate is larger than the effect a future study should reasonably expect. They do not replace a prospective sample-size justification based on a clinically relevant target and a documented analysis plan.

13. 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 = 0.691; treatment proportion p₁ = 0.460; average p̄ = 0.575.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6990 = 1.3701.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6796 = 0.5720.
4. n = (1.3701 + 0.5720)² / (-0.2314)² = 70.4, rounded up to 71 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 <- 0.691056910569106                          # control proportion
p1 <- 0.459677419354839                          # 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(675415)
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 = 71 (80% power), 94 (90% power)
 75% of observed effect: n per arm = 124 (80% power), 165 (90% power)
 50% of observed effect: n per arm = 272 (80% power), 363 (90% power)
power.prop.test: 70.4 
Simulated power at n = 71 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.

14. Multiplicity, Missing Data and Analysis Populations

The registry record identifies a primary endpoint and lists three secondary outcomes. It does not report whether the secondary outcomes were prespecified in a hierarchy, whether a family-wise error or other multiplicity procedure was planned, or whether any adjustment was applied. The posted secondary p-values are unadjusted. When multiple outcomes are tested, unadjusted results do not account for the increased opportunity to observe a small p-value across the set. This makes the secondary findings less conclusive than a single prespecified primary comparison.

No reported analyses are listed in the registry record. The record does not provide a full statistical analysis plan, the test used for the comparisons, or details of any covariate adjustment. The posted arm-level comparisons are unadjusted calculations from the outcome summaries. They can be read as direct contrasts of the recorded counts, but cannot reveal whether the trial team conducted additional analyses or how those analyses were selected.

The record does not report how missing outcome measurements were handled, whether all randomized participants were included in each analysis, or what analysis populations were intended. It also does not describe exclusions after randomization, protocol deviations, or sensitivity analyses. These details matter because the interpretation of a randomized comparison depends partly on which participants contribute outcome data and how incomplete observations are treated. No particular missing-data method or analysis population should be assumed from the posted counts alone.

Interim analyses and any associated stopping rules are also not reported in the registry record. This absence does not establish that such procedures did or did not occur; it means the record does not allow them to be assessed. Overall, the outcome summaries support a straightforward unadjusted comparison, while the lack of reported multiplicity, missing-data, and analysis-population details limits assessment of the broader analysis process.

15. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

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 adverse-event summary lists zero serious adverse events in the capnography arm and zero in standard monitoring. It also lists zero deaths in each arm. These figures describe the counts recorded in the registry; they do not establish that serious events or deaths are impossible, or that the groups have identical safety profiles. Rare outcomes can be difficult to characterize from a trial of this size, and a zero count is not proof of zero risk.

The registry also lists other adverse events for 101 participants in the capnography arm and 95 in standard monitoring. The record does not give the terms or definitions for these other events in the posted adverse-event summary, so their clinical nature and severity cannot be inferred from the counts. No serious adverse-event terms are listed. The figures should be read as registry reporting, not as a complete description of every event or as a causal comparison of safety.

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

The posted primary result shows a lower observed proportion of participants meeting the registry’s hypoxemia definition in the capnography-titration arm than in the standard-monitoring arm. The unadjusted risk difference is -0.231, with a confidence interval from -0.351 to -0.112, and the unadjusted p-value is <0.001. The interval does not include zero, indicating a clear difference in the direction of the posted estimate under the reported calculation.

This result concerns a specific monitoring strategy, population, and procedural context: capnographic information was made available to the endoscopy team during elective ERCP and EUS. It does not isolate the effect of measuring carbon dioxide from the effect of the team being alerted to abnormalities and potentially responding to them. The comparison is between the assigned strategies as described in the registry, not a test of every possible way of using capnography.

The result does not establish effects on outcomes that were not measured or reported, nor does it show how long hypoxemia lasted, when it occurred, or what response followed each event. A lower count meeting the threshold is not equivalent to proof that all respiratory problems were prevented. It also does not establish that the findings apply to other procedures, other patient groups, or different monitoring practices. Participant characteristics and detailed setting information are not reported in the registry record.

Finally, statistical evidence is not the same as a complete assessment of clinical importance. The posted estimate and its interval describe the contrast and its uncertainty for the recorded endpoint; they do not determine how the result should affect clinical decisions. The registry does not report the original analysis plan, missing-data handling, or multiplicity approach. Those omissions leave uncertainty about aspects of the analysis beyond the uncertainty represented by the confidence interval.

17. Why This Trial Matters Statistically

This trial illustrates why a monitoring technology should be evaluated by the comparison it actually implements. Capnography was not simply measured in both arms and compared afterward: the endoscopy team could see and respond to abnormalities in the titration arm, while the information was unavailable in the blinded arm. The trial therefore addresses the strategy of making capnographic information available during the procedure, rather than measurement alone.

The primary outcome also demonstrates the importance of an operational endpoint definition. Hypoxemia was defined using an oxygen-saturation threshold sustained for a stated duration, and was assessed throughout the procedure. This makes the event count interpretable, but it compresses a continuous physiologic course into a participant-level outcome. The count does not convey the timing or duration of events, and those details could be relevant to understanding what the monitoring strategy changed.

Statistically, the risk difference and confidence interval place the primary comparison on an absolute scale and show uncertainty around its magnitude. The unadjusted p-value adds evidence against a no-difference model, but does not replace effect estimation or address the secondary-outcome multiplicity. The pattern across related secondary outcomes can inform interpretation, while their unadjusted status calls for restraint.

The registry record also illustrates the limits of interpreting outcome summaries without a full analysis plan. Randomization and the posted results are informative, but the record does not describe missing-data handling, analysis populations, or multiplicity procedures. These are not minor reporting details: they shape how completely readers can judge the robustness and scope of an otherwise clear numerical comparison.

18. 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 registry record gives limited detail about the enrolled population beyond elective ERCP and EUS. Participant characteristics, eligibility criteria, clinical setting, and reasons for exclusion are not reported in the registry record. This makes it difficult to assess how closely the studied participants resemble other populations undergoing these procedures or to identify groups for whom the comparison may differ.

The intervention combines access to capnographic information with the endoscopy team’s response to abnormalities. The record does not specify a detailed response protocol, so the contribution of the tracing itself cannot be separated from the actions it may prompt. The description of standard monitoring is also incomplete in the posted record. In addition, although the trial is labeled triple-masked, the identities of the masked parties and the procedures used to maintain masking are not reported.

The primary endpoint uses a clear threshold definition, but the registry does not report details of data capture, event adjudication, or handling of incomplete readings. Nor does it provide the analysis population, missing-data methods, full statistical plan, or any interim-analysis information. These gaps limit assessment of potential measurement and analysis biases.

The trial concerns outcomes during a single endoscopic procedure, typically about 2 hours. The posted endpoint does not characterize longer-term outcomes or the full respiratory course. The registry reports no serious adverse-event terms, and the other adverse-event counts are not accompanied by event descriptions in the available summary. These limitations do not negate the posted comparison, but they constrain what can be concluded from it.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Does Capnography Prevent Hypoxemia During Endoscopic Retrograde Cholangiopancreatography and Endoscopic Ultrasound? — NCT00675415.
• Registry record last updated 2017-11-28; status last verified 2017-10.
• Results first posted 2017-05-30.
• 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.

22. Record Summary

NCT00675415 is a completed randomized, controlled, parallel-group trial sponsored by The Cleveland Clinic. The registry lists enrollment as 263, a start date of 2006-09, and primary completion in 2008-09. The trial is labeled triple-masked, with phase listed as NA.

The comparison was capnography-titration versus standard monitoring during elective ERCP and EUS. The primary outcome was participants experiencing hypoxemia, defined as oxygen saturation less than 90 percent for at least 15 seconds during a single procedure, typically about 2 hours. The posted primary comparison favors a lower observed hypoxemia proportion in the capnography arm, with an unadjusted risk difference of -0.231 and a confidence interval from -0.351 to -0.112.

The record lists three secondary outcomes and adverse-event counts, but does not report a detailed analysis plan, missing-data approach, or multiplicity procedure. The primary result is therefore best read as an unadjusted comparison of the posted arm summaries.