A Randomized Controlled Trial of Lung Ultrasound Compared to Chest X-ray for Diagnosing Pneumonia in the Emergency Department: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Randomized Controlled Trial of Lung Ultrasound Compared to Chest X-ray for Diagnosing Pneumonia in the Emergency Department (NCT01654887), a randomized na study. The primary outcome is Percentage of Participants For Whom CXR Was Not Needed to Diagnose Pneumonia.

2. Clinical Question

3. Trial Design

The registry describes a randomized, parallel-group trial comparing two diagnostic sequences in the emergency department. In the lung-ultrasound arm, lung ultrasound was performed first, with the option of obtaining chest X-ray afterward. In the comparator arm, chest X-ray was performed first, followed by lung ultrasound. The study was unmasked. Whether allocation was concealed before assignment is not reported in the registry record.

The registry lists an enrollment of 191 and shows 103 participants in the lung-ultrasound arm and 88 in the chest-X-ray arm for the posted outcomes. These outcome-specific group counts are the denominators associated with the reported measures; the record does not explain whether exclusions, missing observations, or differences between enrollment and analyzed groups account for their relationship to the enrollment figure.

The study is marked completed. Its recorded start was 2012-08, and its primary completion was 2013-07. The phase is listed as NA. The sponsor is Icahn School of Medicine at Mount Sinai. These registry entries describe the study structure and status, but do not supply a protocol rationale or a full account of trial conduct.

The primary outcome was the percentage of participants for whom chest X-ray was not needed to diagnose pneumonia, measured up to 5 hours. The registry also identifies follow-up outcomes concerning unscheduled healthcare visits, antibiotic use, and hospital admission during weeks 1-2, as well as emergency-department length of stay. Results for the length-of-stay outcome are not reported in the registry record. Details of the diagnostic criteria, outcome adjudication, and timing of the follow-up ascertainment beyond the listed timeframes are not reported in 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.

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 Percentage of Participants For Whom CXR Was Not Needed to Diagnose Pneumonia, measured over up to 5 hours.

The primary endpoint measures the percentage of participants for whom chest X-ray was not needed, or who received only lung ultrasound, to diagnose pneumonia. It is reported for the lung-ultrasound-first and chest-X-ray-first groups, with a 95% confidence interval for each percentage. The timeframe is up to 5 hours.

This endpoint is aligned with the study’s stated objective: evaluating whether beginning with lung ultrasound could reduce the need for chest X-ray during assessment for possible pneumonia. It captures the frequency of the specified diagnostic pathway outcome, rather than directly reporting a measure of diagnostic accuracy. The registry record does not provide sensitivity, specificity, or a reference-standard comparison for pneumonia diagnosis, so the endpoint should not be read as an estimate of those properties.

The reported percentage was 38.8% in the lung-ultrasound group and 0.0% in the chest-X-ray group. The corresponding 95% confidence intervals were 30.0% to 48.9% and 0.0% to 3.6%. The comparison is reported as a risk difference with an unadjusted p-value of <0.001. These figures describe the endpoint as defined in the registry; they do not by themselves establish that the diagnostic strategies identify pneumonia equally well or that one strategy improves clinical outcomes.

7. Results Posted in the Registry

• Chest X-Ray: 0.0% (0/88) versus 38.8% (40/103) on control; risk difference -38.8 percentage points (95% CI -48.2 to -29.4); 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.

Comparison of Unscheduled Healthcare Visits

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Chest X-Ray: 11.4% (10/88) versus 8.7% (9/103) on control; risk difference 2.6 percentage points (95% CI -6.0 to 11.2), risk ratio 1.30 (0.55–3.06); unadjusted two-proportion z-test p = 0.546.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Chest X-Ray and the control. The interval gives the range of effects that remain compatible with the data.

Percentage of Participants With Antibiotic Use

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Chest X-Ray: 27.3% (24/88) versus 37.9% (39/103) on control; risk difference -10.6 percentage points (95% CI -23.8 to 2.6), risk ratio 0.72 (0.47–1.10); unadjusted two-proportion z-test p = 0.121.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Chest X-Ray and the control. The interval gives the range of effects that remain compatible with the data.

Percentage of Participants Who Had Hospital Admission.

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Chest X-Ray: 17.0% (15/88) versus 19.4% (20/103) on control; risk difference -2.4 percentage points (95% CI -13.3 to 8.6), risk ratio 0.88 (0.48–1.61); unadjusted two-proportion z-test p = 0.673.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Chest X-Ray and the control. The interval gives the range of effects that remain compatible with the data.

Across the posted secondary outcomes, the estimates do not reproduce the clear separation seen in the primary process measure. The intervals for the reported secondary risk differences include zero, and their unadjusted p-values do not provide clear evidence of group differences. This overall pattern is compatible with uncertainty about whether the strategies differed on those outcomes; it does not establish that the groups were equivalent.

These outcomes address different questions from the primary endpoint. Unscheduled healthcare visits, antibiotic use, and hospital admission concern events after the diagnostic encounter, whereas the primary measure concerns whether chest X-ray was needed within the stated diagnostic timeframe. A difference in the primary pathway measure need not be accompanied by a detectable difference in subsequent outcomes, particularly when those outcomes are less frequent or estimated imprecisely. The registry does not report results for the emergency-department length-of-stay outcome, so that part of the secondary objective cannot be assessed from the posted results.

The registry presents the secondary comparisons as unadjusted. In general, testing multiple outcomes increases the chance that at least one comparison will appear unusual by chance if each is interpreted as confirmatory. Without a reported multiplicity strategy or outcome hierarchy, secondary p-values should carry less inferential weight than a result tied to a clearly stated primary endpoint. No such strategy is reported in the registry record. The secondary findings are therefore useful for describing the available follow-up data, but they should not be treated as definitive evidence for or against effects on those outcomes.

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

• Comparison of Unscheduled Healthcare Visits
• Percentage of Participants With Antibiotic Use
• Percentage of Participants Who Had Hospital Admission.
• Comparison of the Length of Stay in the ED

9. Statistical Interpretation

10. Statistical Methods Explained

The posted comparisons concern binary outcomes: each participant either met the outcome definition or did not. For the primary measure, the event was being a participant for whom chest X-ray was not needed to diagnose pneumonia. The registry reports the percentage in each group and a risk difference comparing chest-X-ray-first with lung-ultrasound-first. The corresponding reported counts were 40 of 103 and 0 of 88, respectively.

What does a risk difference express?

In general, a risk difference subtracts the event proportion in one group from the event proportion in the other. Its sign depends on which group is subtracted from which, so the comparison order matters. In the posted primary comparison, the estimate is -0.388 because the chest-X-ray-first group is compared with the lung-ultrasound-first group for an outcome that occurred in 0.0% and 38.8%, respectively. The negative sign therefore indicates a lower observed event proportion in the chest-X-ray-first group for this endpoint. It should not be interpreted as a difference in diagnostic accuracy, which is not reported.

What does the p-value say?

In general, a p-value assesses how incompatible the observed data are with a specified null model, assuming that model and the analysis assumptions hold. It is not the probability that the null hypothesis is true, nor does it measure the size or clinical importance of an effect. The primary comparison has an unadjusted p-value of <0.001. The registry does not report a multiplicity adjustment for this comparison or a detailed statistical analysis plan, so the p-value should be read as the posted unadjusted result rather than as evidence that all sources of uncertainty have been addressed.

Why do confidence intervals matter?

In general, a confidence interval gives a range of estimates compatible with the data under the method used. It helps readers assess precision as well as direction. For the primary risk difference, the 95% confidence interval runs from -0.482 to -0.294. Because the interval is entirely below zero, it aligns with the reported difference in the endpoint between groups. The interval does not address limitations in the outcome definition or show whether the strategies have equal diagnostic accuracy.

What assumptions underlie a binary comparison?

In general, a simple comparison of binary proportions treats each participant’s outcome as a binary observation and relies on an appropriate model or approximation for the uncertainty. The registry reports the estimates and confidence intervals, but does not report the specific method used to calculate the primary interval or p-value. It also does not describe adjustment for baseline characteristics, clustering, or other covariates. Those methods should not be inferred from the summary statistics alone.

How should the planning illustrations be read?

In general, a sample-size planning illustration asks how many participants per arm might be needed under stated assumptions to detect a specified effect with a chosen power and significance level. It is not the number actually analyzed, and it is not observed power. The registry labels its illustrations as based on the observed effect rather than the trial’s own design. The stated assumptions are a two-sided significance level of 0.05, equal allocation between arms, a normal approximation, and an effect equal to the stated fraction of the observed effect. Such illustrations depend strongly on the assumed effect size. They do not establish that the completed trial was designed with those assumptions or that its actual design had a particular power. The trial’s own sample-size justification is not reported in the registry record.

The analyses listed for the secondary outcomes are unadjusted. For those outcomes, the registry gives risk differences, confidence intervals, and unadjusted p-values. The record does not report adjusted analyses or a prespecified procedure for controlling the chance of false-positive findings across multiple outcomes. Thus, secondary comparisons are most appropriately treated as exploratory descriptions of the posted data rather than as independent confirmatory claims.

11. Confidence Intervals and Precision

The primary group estimates have 95% confidence intervals that convey the uncertainty around each reported percentage. For lung ultrasound, the estimate is 38.8%, with an interval from 30.0% to 48.9%. For chest X-ray, it is 0.0%, with an interval from 0.0% to 3.6%. These intervals show that the observed percentage is estimated with uncertainty, even when the reported point estimate is at a boundary such as 0.0%.

The comparison-level interval is especially useful for assessing the difference between strategies. The primary risk difference is -0.388, and its 95% confidence interval extends from -0.482 to -0.294. The interval does not include zero, consistent with the reported unadjusted p-value of <0.001. This supports a difference in the specified endpoint in these data. The width of the interval also shows that the exact magnitude is not known without uncertainty.

In general, a 95% confidence interval describes the performance of a procedure over repeated samples under its assumptions; it is not a statement that there is a 95% probability that this particular interval contains the true value. Its interpretation also depends on the validity of the statistical method and the quality of the underlying data. The registry does not report the interval-calculation method, missing-data approach, or analysis population in enough detail to independently assess those assumptions.

Confidence intervals do not resolve limitations in what the endpoint measures. The primary outcome captures whether chest X-ray was not needed under the recorded diagnostic pathway, not the accuracy of either imaging method against a reference standard. Precision in estimating that endpoint is distinct from certainty about the broader clinical meaning. The registry also does not provide a confidence interval for emergency-department length of stay because its result is not reported in the registry record.

12. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when a study evaluates several outcomes or makes several statistical comparisons. In general, the more comparisons treated as confirmatory, the greater the chance of at least one apparently significant result arising by chance. A prespecified hierarchy or adjustment method can clarify how to interpret a family of tests. The registry lists a primary outcome and multiple secondary outcomes, but a multiplicity procedure, testing hierarchy, or prespecified distinction between confirmatory and exploratory analyses is not reported in the registry record. The secondary comparisons are reported as unadjusted.

The record also does not describe missing-data handling. It does not say whether participants with unavailable outcome information were excluded, imputed, or handled through another approach. Although outcome-specific arm denominators are posted, those figures alone do not establish how missing data were treated or whether all randomized participants contributed to each analysis. The analysis population—such as an intention-to-treat or per-protocol population—is not reported in the registry record.

Other aspects of the statistical plan are also not reported in the registry record, including the planned model for the primary comparison, covariate adjustment, the sample-size justification, and any interim analyses. The absence of these details does not show that the procedures were or were not used; it limits what can be verified from the registry entry. The record lists no reported analyses beyond the posted outcome summaries.

These omissions matter because the statistical interpretation depends not only on the estimates, but also on how outcomes and analyses were selected and how incomplete observations were treated. The primary endpoint remains interpretable as a posted comparison, but a complete audit of the analysis plan and the potential effects of missingness or multiple testing is not possible from the registry record alone.

13. 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 records zero serious adverse events affecting participants in the lung-ultrasound group and zero in the chest-X-ray group. It also records zero other adverse events affecting participants in each group. The risk groups listed for these entries are 103 and 88, respectively. No serious adverse event terms are listed.

Deaths are recorded as null, which is not the same as a reported count of zero. Death information is therefore not reported in the registry record. The absence of listed serious adverse event terms should also be read as a feature of the posted safety record, not as proof that no adverse events occurred. The registry entry does not provide further detail on event collection, definitions, severity grading, follow-up completeness, or how safety events were assessed. These summary entries cannot establish comparative safety or rule out uncommon events.

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

The primary result shows a difference between the randomized groups on the registry’s stated endpoint: the percentage of participants for whom chest X-ray was not needed to diagnose pneumonia. The reported percentage was higher in the lung-ultrasound-first group than in the chest-X-ray-first group, and the primary comparison’s confidence interval does not include zero. This describes the outcome as recorded under the study’s diagnostic sequences.

That finding does not establish that lung ultrasound and chest X-ray are equally accurate for diagnosing pneumonia. The registry does not report sensitivity, specificity, or results against a diagnostic reference standard. Nor does the primary endpoint establish that using one sequence rather than the other improves health outcomes, reduces complications, or changes later care. Those are different questions from whether chest X-ray was not needed during the initial diagnostic pathway.

The secondary outcome summaries do not clearly establish group differences, and the emergency-department length-of-stay result is not reported in the registry record. The lack of a clear secondary difference should not be treated as proof of equivalence. The available confidence intervals retain uncertainty, and the record does not describe an analysis plan for confirming equivalence or non-inferiority.

More broadly, the results apply to the endpoint definitions, participants, and care setting represented in this trial. The registry does not report enough eligibility, implementation, or diagnostic-process detail to determine how closely the findings would match other settings. The most defensible interpretation is therefore focused: the posted data indicate a difference in the primary process measure, while the diagnostic accuracy and broader clinical consequences remain unanswered by the information reported.

15. Why This Trial Matters Statistically

This trial is statistically informative because its primary endpoint is closely connected to the diagnostic sequence being compared. Rather than treating imaging use as an incidental measure, it asks whether the chest X-ray step was needed under a lung-ultrasound-first approach. The posted group percentages and risk difference make the contrast in this process outcome explicit.

The result also illustrates why an endpoint must be interpreted according to its definition. A pronounced difference in whether chest X-ray was needed is not automatically evidence about diagnostic accuracy, patient outcomes, or the safety of a strategy. Those conclusions require outcomes and methods suited to those questions. The registry’s stated objective and primary measure do not substitute for accuracy metrics or a reference-standard analysis that is not reported.

The secondary findings provide a useful counterpoint: evidence of a difference in a primary process measure does not guarantee clear differences in follow-up outcomes. Those outcomes have their own uncertainty, and their unadjusted status makes them less suitable as stand-alone confirmatory claims. The unavailable length-of-stay result further limits the assessment of the full secondary objective.

Finally, the record highlights the value of transparent analysis details. Without a reported analysis population, missing-data approach, multiplicity strategy, and sample-size justification, readers cannot fully assess how the posted summaries relate to the planned analysis. The result can still be described precisely, but its inferential scope should remain tied to what the registry actually reports.

16. 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.
• Participant counts were reconstructed from posted percentages and arm sizes, so small rounding differences are possible.

The registry gives limited detail about who was eligible and how participants were recruited. Apart from the emergency-department context and possible pneumonia, further participant characteristics and eligibility criteria are not reported in the registry record. This makes it difficult to judge how closely the study population matches other clinical settings.

The trial was unmasked, and the registry does not provide details of allocation concealment or of how the diagnostic pathway was implemented. Because the primary outcome concerns whether chest X-ray was needed, the process used to decide whether to obtain it is important. Criteria for that decision, diagnostic adjudication procedures, and use of a reference standard are not reported in the registry record. The posted endpoint therefore cannot be treated as a direct assessment of diagnostic accuracy.

Outcome completeness and analysis populations are also unclear. The record does not describe missing observations or explain how each outcome’s analyzed participants were selected. It does not report results for emergency-department length of stay, despite listing that outcome among the secondary outcomes. The reported safety entries likewise lack details about collection methods and follow-up, and death information is not reported in the registry record.

Finally, the statistical analysis plan, multiplicity handling, and trial-specific sample-size justification are not reported in the registry record. These limitations do not invalidate the posted summaries, but they constrain independent evaluation of their robustness and their broader applicability.

17. Related Tutorials

18. Related Statistical Calculators

19. Sources

• ClinicalTrials.gov: A Randomized Controlled Trial of Lung Ultrasound Compared to Chest X-ray for Diagnosing Pneumonia in the Emergency Department — NCT01654887.
• Registry record last updated 2017-03-10; status last verified 2017-01.
• Results first posted 2017-03-10.
• 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.

20. Record Summary

NCT01654887 is a completed randomized, unmasked, parallel-group trial of lung ultrasound compared with chest X-ray for evaluating possible pneumonia in the emergency department. The registered enrollment is 191. The primary endpoint was the percentage of participants for whom chest X-ray was not needed to diagnose pneumonia, assessed up to 5 hours.

The registry reports 38.8% for lung ultrasound first and 0.0% for chest X-ray first, with a risk difference of -0.388 and an unadjusted p-value of <0.001. Secondary results are reported for unscheduled healthcare visits, antibiotic use, and hospital admission; the emergency-department length-of-stay result is not reported in the registry record. Important analysis details, including missing-data handling, the analysis population, and multiplicity procedures, are also not reported in the registry record.