Comparison Of Chest Ultrasound Techniques To Identify Clinically Significant Pneumothorax: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Comparison Of Chest Ultrasound Techniques To Identify Clinically Significant Pneumothorax (NCT02619591), a randomized na study. The primary outcome is Pneumothorax (Positive or Negative).

2. Clinical Question

3. Trial Design

The registry describes a randomized, single-masked, parallel-group trial comparing two chest ultrasound approaches. One arm used a single view of each hemithorax, and the other used multiple views of each hemithorax. The brief summary characterizes the study as prospective. The registry lists ultrasound imaging as the intervention and pneumothorax as the condition.

Enrollment was 260. The posted outcome counts are 17 participants with a positive outcome among 139 in the single-view arm and 21 among 121 in the multiple-view arm. The primary outcome was recorded as positive or negative, with the presence of pneumothorax assessed at the time of ultrasound and a timeframe of up to 20 minutes. These are the registry’s outcome counts; they do not by themselves describe how the outcome was verified.

The trial is marked completed. The registry lists the start as 2011-09 and primary completion as 2014-03. The phase is listed as NA, which is consistent with the registry’s classification of an imaging-technique comparison rather than a drug-phase study. The sponsor is University of Massachusetts, Worcester.

The registry identifies the study as randomized but does not report the method used to generate or conceal allocation. It identifies single masking, but who was masked is not reported in the registry record. Details of ultrasound operators, training, recruitment setting, participant eligibility, and handling of protocol deviations are also not reported. The registry lists one primary outcome measure and no posted analyses. Those omissions limit assessment of how the recorded counts were obtained and which analysis population was used.

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 were patients who had recieved a chest tube already, or those with known pneumothoraces.

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 Pneumothorax (Positive or Negative), measured over up to 20 minutes.

The primary endpoint records whether a participant had a pneumothorax at the time of ultrasound. It is a binary outcome: positive or negative. The registry reports the number of participants with an event, rather than a continuous measurement, and gives a timeframe of up to 20 minutes. This format permits comparison of the observed event counts between the single-view and multiple-view techniques.

A binary endpoint is useful when the clinical question concerns whether a condition was detected or present, rather than its measured size or severity. For this comparison, the count of positive outcomes describes how often pneumothorax was recorded in each arm. It does not alone measure diagnostic accuracy. To assess accuracy, readers would need information about how the true presence or absence of pneumothorax was established, including the reference standard and the timing of its assessment. Those details are not reported in the registry record.

The title refers to clinically significant pneumothorax, but the primary outcome description says presence of pneumothorax at the time of ultrasound. The registry does not separately define clinical significance in the endpoint description. The results should therefore be interpreted as counts for the endpoint as registered, not as evidence about a distinct severity classification or about patient outcomes beyond that endpoint.

7. Results Posted in the Registry

• "Device" - Ultrasound Imaging - Multiple Views: 17.4% (21/121) versus 12.2% (17/139) on control; risk difference 5.1 percentage points (95% CI -3.5 to 13.8), risk ratio 1.42 (0.79–2.56); unadjusted two-proportion z-test p = 0.243.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is binary: each participant is recorded as positive or negative for pneumothorax at the time of ultrasound. For a binary outcome, an event proportion describes the share of participants in a group who experienced or had the recorded event. The registry’s arm counts are 17 of 139 for single-view ultrasound and 21 of 121 for multiple-view ultrasound. The corresponding posted values are 0.122 and 0.174. The comparison concerns these observed proportions.

What does the risk difference express?

The risk difference subtracts one group’s event proportion from the other’s. The posted estimate is 0.051, with multiple views as the comparison arm and single view as the control. A positive estimate means the observed event proportion was higher in the multiple-view group. A risk difference expresses an absolute contrast on the proportion scale. It does not establish why the groups differed, and it does not describe diagnostic accuracy unless the outcome classification has been validated against a reference standard.

How do the relative measures differ?

The risk ratio compares the event proportions as a ratio; the posted estimate is 1.419. The odds ratio compares event odds rather than event proportions; the posted estimate is 1.507. Ratios use different scales from the risk difference, so their numerical values should not be compared directly with it. All three measures describe the same arm comparison from different perspectives. The registry does not report adjusted estimates, so the posted comparison should not be described as controlling for other participant or study characteristics.

What does the p-value tell a reader?

The posted unadjusted p-value is 0.243. A p-value describes how unusual the observed data, or data more extreme, would be under a specified null model and test procedure. It is not the probability that the null hypothesis is true, and it does not quantify the size or clinical importance of a difference. The statistical test used to obtain the p-value is not reported in the registry record. The unadjusted result should therefore be read as the registry’s posted comparison, not as a fully documented analysis plan.

Why consider an interval as well as a point estimate?

A point estimate gives one value for the comparison, while an interval describes uncertainty around that estimate under a statistical method. The posted risk-difference interval runs from -0.035 to 0.138. Because it spans zero, the data are compatible with no difference under the interval method as well as with effects on either side of zero. The reported intervals for the risk ratio and odds ratio likewise include their no-association values. The confidence level and method used to construct these intervals are not reported in the registry record, so those details cannot be evaluated from the record.

Why might a binary-outcome analysis need more detail?

General statistical context: comparisons of event proportions may be analyzed using several methods, and their assumptions can differ. Some approaches rely on approximate sampling distributions; others use exact or model-based procedures. With binary data, the number of events and the group sizes affect the precision of estimates. The registry does not state which test or interval method was used for the comparison, whether any adjustment was planned, or whether the analysis accounted for features such as clustering or stratification. Those features should not be assumed to have been present.

Random assignment is intended to make treatment groups comparable on average, but it does not eliminate the need to document allocation procedures, adherence, missing outcome data, and analysis populations. The registry says the trial was randomized and parallel-group; it does not report the allocation sequence method or concealment procedures. It reports single masking but does not identify who was masked. It also does not report whether all randomized participants were included in the outcome analysis or how missing observations were handled.

The posted summary is therefore useful for describing the observed event counts and the unadjusted comparison, but it does not permit a complete audit of the statistical analysis. In particular, the test specification, interval construction, confidence level, covariate adjustments, and analysis population are not reported in the registry record. No analysis should be attributed to the investigators beyond what the registry explicitly records.

10. Confidence Intervals and Precision

Precision concerns how tightly an estimate is determined by the available data and the statistical method. The risk-difference estimate is 0.051, and its posted interval extends from -0.035 to 0.138. The interval includes zero, the value corresponding to no difference on the risk-difference scale. It also extends to positive values, so the data are compatible with a higher event proportion in the multiple-view group. The interval communicates uncertainty that a point estimate alone cannot convey.

The relative measures have the same broad interpretation with their own reference value: the risk ratio is 1.419, with an interval from 0.786 to 2.563, and the odds ratio is 1.507, with an interval from 0.754 to 3.011. Their intervals include 1, the no-association value for a ratio. This is consistent with the uncertainty indicated by the risk-difference interval. These intervals are not proof that all values within them are equally plausible, and they do not identify the true effect with certainty.

The confidence level and construction method for these intervals are not reported in the registry record. Their precision also depends on how the endpoint was measured and whether observations were missing or excluded; those details are not reported either. The intervals therefore describe the uncertainty conveyed by the posted comparison, while the unavailable methods and outcome-verification details limit a fuller assessment of that uncertainty.

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. Each row considers a different fraction of the observed effect: 100%, 75%, or 50%. For each assumed effect, the two power columns give the participants per arm for 80% power and for 90% power. Power is the chance, under the specified planning model and assumed effect, of detecting an effect at the stated significance threshold if that effect is present.

The planning illustration uses a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is the stated fraction of the observed effect. For 100% of the observed effect, the planning figures are 753 participants per arm for 80% power and 1,007 participants per arm for 90% power. For 75% of the observed effect, they are 1,290 participants per arm for 80% power and 1,726 participants per arm for 90% power. For 50% of the observed effect, they are 2,793 participants per arm for 80% power and 3,738 participants per arm for 90% power.

The curve should be read as expressing the same planning relationship across assumed effect sizes: smaller effects require more participants to detect under the stated assumptions. These are planning illustrations based on the observed effect, not observed power and not the trial’s original sample-size calculation. The trial’s own design assumptions and sample-size justification are not reported in the registry record. The illustration does not show that the completed trial had any particular probability of detecting its result.

12. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.122; treatment proportion p₁ = 0.174; average p̄ = 0.148.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.5021 = 0.9841.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5008 = 0.4215.
4. n = (0.9841 + 0.4215)² / (0.0513)² = 752.1, rounded up to 753 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 <- 0.122302158273381                          # control proportion
p1 <- 0.173553719008264                          # 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(2619591)
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 = 753 (80% power), 1,007 (90% power)
 75% of observed effect: n per arm = 1,290 (80% power), 1,726 (90% power)
 50% of observed effect: n per arm = 2,793 (80% power), 3,738 (90% power)
power.prop.test: 752.1 
Simulated power at n = 753 per arm: 0.800 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

13. Multiplicity, Missing Data and Analysis Populations

The registry lists one primary outcome measure and no posted secondary outcome measures. It also lists no posted analyses. With only one registered primary endpoint in the record, the available information does not show a family of reported endpoint tests, but it is not enough to establish the investigators’ complete analysis plan or whether additional analyses were conducted.

Multiplicity arises when several hypotheses are tested and the chance of a false-positive result across those tests may increase. Statistical approaches can address that issue, but the registry does not report whether multiplicity was considered, whether any adjustment was planned, or which comparisons were designated as primary in an analysis plan. No adjustment should be inferred from the posted unadjusted p-value.

Missing outcome data and analysis populations are also not described. The record does not report whether any participants lacked an evaluable ultrasound result, whether all randomized participants were included, or how exclusions and protocol deviations were handled. The arm counts and positive-event counts are available, but their relationship to a prespecified analysis population is not documented. It is therefore not possible to determine from the registry record whether the comparison followed an intention-to-treat approach, a per-protocol approach, or another population definition.

The registry does not report imputation procedures, sensitivity analyses for missing outcomes, or an approach to missingness. These omissions matter because the interpretation of an event-proportion comparison can depend on who was included and how unavailable outcomes were handled. They do not show that data were missing or that any particular analysis was performed; they mean those details are not reported in the registry record.

14. Safety Results

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

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

The registry reports zero serious adverse events in the single-view arm, with 139 at risk, and zero in the multiple-view arm, with 121 at risk. It also reports zero other adverse events in each arm, using the same at-risk counts. No serious adverse-event terms are listed. Death counts are not reported in the registry record.

These entries describe the adverse-event information posted for this trial. They should be read as registry counts, not as proof that no adverse events of any kind occurred or that the techniques have established safety. The record does not describe how adverse events were collected, what observation period was used for safety reporting, or how events were classified. It also does not report an analysis comparing adverse-event risks. The zero counts therefore provide limited descriptive information, and the missing collection and follow-up details constrain their interpretation.

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

The registry records a randomized comparison of two approaches to chest ultrasound: a single view and multiple views of each hemithorax. Its primary endpoint is a positive or negative pneumothorax finding at the time of ultrasound, with a timeframe of up to 20 minutes. The posted counts and unadjusted estimates describe how often that endpoint was recorded in each arm.

The point estimate for the risk difference is positive, while the interval includes zero and the unadjusted p-value is 0.243. The relative-effect intervals also include their no-association value. Taken together, these figures do not provide a statistically clear separation between the groups in the posted comparison. They do not prove equivalence or show that the techniques produce identical results; the estimates remain uncertain.

The endpoint does not by itself establish whether either approach correctly identified pneumothorax. The registry does not report a reference standard, sensitivity, specificity, or a separate result for clinically significant pneumothorax. Although clinical significance appears in the title, the outcome description concerns the presence of pneumothorax at the time of ultrasound. The posted outcome should not be treated as a measure of severity or as a patient outcome beyond detection status.

The record also does not describe the ultrasound protocol in enough detail to assess reproducibility, identify who was masked, or explain the analysis population and handling of missing data. Those uncertainties limit conclusions about how broadly the results apply. The trial’s completion status and its posted counts establish that results are recorded, but they do not resolve these measurement and analysis questions. The results are best understood as a comparison of the registered positive-or-negative endpoint in the two trial arms, with uncertainty reflected by the reported intervals.

16. Why This Trial Matters Statistically

This trial is statistically informative because it compares two ways of collecting an imaging assessment while holding the general intervention category constant. The distinction is the number of views used for each hemithorax. A randomized, parallel-group design provides a framework for comparing outcome counts between assigned techniques, though the registry does not report enough allocation detail to assess how randomization was implemented.

The endpoint is binary, so the event counts can be expressed as group proportions and compared using absolute and relative measures. The registry posts a risk difference, risk ratio, odds ratio, and an unadjusted p-value. Reading these measures together helps distinguish the estimated direction from the uncertainty in that estimate. Here, the point estimates favor a higher recorded event proportion in the multiple-view arm, but the intervals include the no-difference values. That combination illustrates why a single point estimate or p-value is not a complete account of evidence.

The case also highlights a measurement issue central to diagnostic research: a recorded positive finding is not automatically a verified true positive. Without a reported reference standard, the event counts do not establish diagnostic accuracy. The distinction between the title’s reference to clinically significant pneumothorax and the endpoint’s description of pneumothorax presence further underlines the importance of precise outcome definitions.

Finally, the registry’s omissions are relevant to interpretation, not just reporting completeness. Details about masking, missing outcomes, analysis populations, and interval methods affect how confidently readers can evaluate the comparison. The posted record supports a careful description of the observed endpoint and its uncertainty, while leaving important aspects of measurement and analysis unresolved.

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 registry’s endpoint description does not specify a reference standard for determining whether pneumothorax was truly present. It also does not report how a positive or negative ultrasound result was verified. As a result, the posted counts cannot by themselves quantify diagnostic accuracy, such as the ability to identify true cases or rule out absent cases.

The title refers to clinically significant pneumothorax, but the primary outcome description records pneumothorax presence at the time of ultrasound. A separate definition or result for clinical significance is not reported in the registry record. This limits conclusions about whether the registered outcome corresponds to the clinically significant cases named in the title.

Information about the clinical setting, eligibility criteria, ultrasound operators, training, and detailed scanning protocols is not reported. The record says single masking, but does not identify who was masked. The method of randomization and allocation concealment is also not reported. These gaps make it difficult to judge implementation, potential sources of bias, and how well the findings might apply to other settings.

The registry does not report the statistical test, interval construction method, confidence level, analysis population, handling of missing outcomes, or a prespecified analysis plan. It lists no posted analyses. These limitations prevent a full audit of the comparison and leave uncertainty about whether all eligible randomized participants contributed to the endpoint summary. The posted estimates remain useful as descriptive registry results, but these missing details narrow what can be concluded from them.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Comparison Of Chest Ultrasound Techniques To Identify Clinically Significant Pneumothorax — NCT02619591.
• Registry record last updated 2016-11-07; status last verified 2015-12.
• Results first posted 2016-11-07.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

21. Record Summary

NCT02619591 is a completed randomized, single-masked, parallel-group trial sponsored by University of Massachusetts, Worcester. It compares ultrasound using a single view of each hemithorax with ultrasound using multiple views of each hemithorax. The registry lists enrollment of 260 and a primary outcome of pneumothorax, recorded as positive or negative at the time of ultrasound, with a timeframe of up to 20 minutes.

The posted event counts are 17 among 139 in the single-view arm and 21 among 121 in the multiple-view arm. The risk-difference estimate is 0.051, with an interval from -0.035 to 0.138; the unadjusted p-value is 0.243. The record lists zero serious and zero other adverse events in both arms; death counts are not reported in the registry record.

The registry does not report a reference standard, detailed analysis plan, analysis population, missing-data handling, or who was masked. It lists one primary outcome measure and no posted analyses.