Laparoscopic Narrow Band Imaging for Detection of Occult Cancer Metastases: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Laparoscopic Narrow Band Imaging for Detection of Occult Cancer Metastases (NCT01944930), a randomized phase 1 study. The primary outcome is Rate of Detectable Peritoneal Metastases.

2. Clinical Question

3. Trial Design

This was a randomized, unmasked, Phase 1 feasibility trial with a crossover model. The registry lists 23 enrolled participants and a completed study. It identifies Lahey Clinic as the sponsor. The study began in 2013-01, with primary completion in 2014-09.

The crossover structure is central to the design: participants were to undergo both laparoscopic methods, while the order of NBI and standard white-light laparoscopy was randomized for each patient. The registry describes two sequence groups, NBI first and standard white-light first. These are sequences within a single cohort, not descriptions of separate treatment populations receiving only one technique. The registry does not report how the sequence assignment was generated or concealed.

Suspicious-appearing abnormalities were to be biopsied for frozen-section histopathology using best clinical practices. The study aim was to compare the number of detected peritoneal metastases between the techniques. Recurrence surveillance through a one-year follow-up was described as a way to gauge the rate of potentially missed metastases.

The primary outcome is reported with an NBI arm value of 0 and a white-light arm value of 2, with 20 participants recorded for each. The registry does not explain how these arm-level entries relate to the 23 enrolled participants or how the paired crossover observations were handled. It also does not post participant-level data, an analysis population definition, or an inferential analysis. No masking was used, and the registry record does not report details of any additional blinding procedures.

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 Rate of Detectable Peritoneal Metastases, measured over After each Laparoscopy, at time of procedure.

The primary outcome is titled “Rate of Detectable Peritoneal Metastases.” Its stated timeframe is after each laparoscopy, at the time of the procedure, and its description is the number of additionally visualized peritoneal lesions. The registry lists the unit as a percentage of lesions, while identifying the outcome parameter as a number. It does not explain this difference in wording or provide a denominator for a percentage.

The posted values are 0 for NBI laparoscopy and 2 for standard white-light laparoscopy, with 20 recorded in each arm. These are the values as registered; the record does not provide an estimate of the paired contrast between techniques, uncertainty around an estimate, or a statistical test. The endpoint therefore conveys the recorded lesion counts but not the precision of a comparative effect.

In general, recording lesions visualized during laparoscopy can address whether one inspection method identifies abnormalities that are not seen with another. Histopathology of suspicious findings can help characterize those abnormalities. The registry describes biopsies but does not post pathology results or define how a lesion was counted as a metastasis for this outcome. It also describes one-year recurrence surveillance as a way to assess potentially missed disease, but does not post a recurrence outcome.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry identifies a randomized crossover design, but it does not report a statistical analysis plan or an inferential analysis. The methods below explain general statistical considerations relevant to such a design; they should not be read as methods confirmed to have been used in this trial.

What does randomization of order address?

In general, randomizing procedure order helps prevent a consistent assignment of one technique to the first inspection and the other to the later inspection. Order can matter because the surgeon may have already inspected the cavity, or because a finding seen during the first procedure could influence the second. Randomization can distribute such order effects across sequences in expectation, but it does not eliminate them automatically. The registry says each patient’s order was randomized and identifies NBI-first and white-light-first sequences. It does not describe the randomization method, allocation concealment, or an analysis of order effects.

Why is the crossover structure important?

In general, crossover observations are paired because the same participant undergoes both methods. A paired comparison accounts for the fact that observations from one person may be more alike than observations from different people. For a count outcome, an analysis also needs to match the distributional assumptions to how the counts arise. A method that ignores the pairing may misstate uncertainty. The registry does not post paired lesion counts or state whether the comparison accounted for participant-level pairing, sequence, or period. The arm-level entries alone cannot establish which approach was used.

How should a count outcome be analyzed?

In general, count data are non-negative and may be concentrated at low values or include many zeros. A suitable model depends on the outcome’s distribution and the study design. Over-dispersion matters when the observed variability in counts exceeds what a simple count model assumes; if ignored, uncertainty estimates may be unreliable. This is a general consideration, not a finding about this trial. The registry does not report a distributional assessment, a count model, or any other analysis method. The outcome is also described using both count and percentage language, and the record does not specify a denominator or clarify the intended scale.

What do uncertainty measures add?

In general, a point estimate summarizes the comparison, while a confidence interval describes its statistical uncertainty under the chosen model. A p-value, when reported, addresses compatibility with a specified null hypothesis under stated assumptions; it is not a measure of the size or clinical importance of an effect. The registry has no posted analysis results, confidence interval, or p-value. Thus, the outcome values do not show how stable a comparative estimate would be across repeated samples or what range of effects is compatible with the observed data.

What can a planning illustration tell readers?

The registry describes a planning illustration based on the observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. It states that the effect to detect is a stated fraction of the observed effect. These assumptions describe a planning exercise, not a trial-specific sample-size justification. The record does not report the trial’s own design justification. In general, a sample-size calculation is conditional on its assumed effect, variability, allocation, and analysis method; it does not establish the probability that a completed trial found a real effect. A planning illustration must not be interpreted as observed power or as evidence that the study was adequately powered.

For crossover trials, planning and analysis also depend on within-participant variability and the correlation between measurements. The registry does not report assumptions about those features. Without them, the planning description cannot establish whether the trial’s enrollment was selected to detect a particular paired difference. No additional numerical sample-size result or analysis detail is reported in the registry record.

What does the registry record establish statistically?

It establishes the design description, enrollment, endpoint wording, and the posted arm-level values. It does not report an inferential comparison, an analysis population, treatment of missing observations, or an adjustment for sequence or period. General statistical principles can clarify why these details matter, but they cannot fill in methods that the registry does not report.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists one primary outcome and no posted secondary outcomes. It reports zero posted analyses. The record does not describe a multiplicity strategy, such as how inferential error would be handled if several outcomes or comparisons were tested. With no posted analysis results, there is no reported set of tests from which to assess multiplicity in practice.

The registry does not report a missing-data plan, the number of participants with complete outcome observations, or how unavailable observations were handled. Total enrollment is 23, while the primary outcome lists 20 in each arm. The record does not explain the relationship between those figures, so they do not identify how many participants contributed a complete crossover pair. It would be speculative to attribute the difference to missing data or any particular exclusion.

For a crossover study, the analysis population could be defined using participants with observations from both procedures, but the registry does not state such a definition. It also does not report whether analyses followed randomized sequence assignment, included all participants who began a procedure, or used another population. These are important distinctions because incomplete paired observations can change which participants inform a within-person comparison.

The planned one-year recurrence survey is described in the study summary, but no recurrence results or missing follow-up information are posted. The registry therefore does not establish how many participants were followed or how incomplete follow-up was addressed. The endpoint and population details needed to assess selection bias, missingness, or the handling of multiple analyses are not reported in the registry record.

11. 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 0% reporting threshold.

The registry reports safety figures for the two procedure-order groups. In the NBI-first group, 10 were listed as at risk for serious adverse events and 0 as affected; 10 were listed as at risk for other adverse events and 0 as affected. In the standard white-light-first group, the corresponding figures are 10 at risk and 0 affected for serious adverse events, and 10 at risk and 0 affected for other adverse events. Deaths are listed as 0 in each group. No serious adverse event terms are listed.

These entries describe the posted safety counts for the listed groups. They do not establish that no adverse events occurred outside the registry’s reported categories or observation period, nor do they provide detail about event definitions, severity, relatedness, or follow-up completeness. The registry does not report a safety comparison or an estimate with uncertainty. Because the study is a crossover trial, the group labels identify procedure order, not separate groups exposed to only one technique. The figures should therefore be read as registry-reported counts rather than as evidence of a comparative safety effect.

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

The posted primary outcome entries are 0 for NBI laparoscopy and 2 for standard white-light laparoscopy, with 20 recorded in each arm. They describe the values entered for an outcome whose description is the number of additionally visualized peritoneal lesions. The record does not provide an inferential comparison, an uncertainty interval, or participant-level paired counts. It therefore does not establish the magnitude or precision of a difference between methods.

A value of 0 for this endpoint should not be read as proof that no metastases were present among participants. The endpoint concerns lesions visualized in the procedure, not the prevalence of disease in the enrolled population. Similarly, the value of 2 does not by itself identify the number of participants with metastases, the number of biopsies confirming metastases, or the number of lesions missed by the other method. Those interpretations require definitions and data that are not posted.

The study summary describes biopsy of suspicious findings and one-year recurrence surveillance to gauge potentially missed metastases. The registry does not post pathology results or recurrence outcomes. It also does not clarify how the listed values were derived from the crossover observations, or how the 23 enrolled participants relate to the 20 recorded in each arm.

In general, feasibility results can describe whether a study procedure was undertaken and can inform the design of later research. They are not automatically precise estimates of comparative performance. The posted record supports a description of the trial design and outcome entries, but not a conclusion about whether either technique is superior, equivalent, or more accurate. It also does not support a clinical recommendation. Such conclusions would require a defined analysis, adequate outcome and follow-up reporting, and an estimate of comparative uncertainty.

13. Why This Trial Matters Statistically

This trial raises a useful statistical issue: the measurement methods were compared within a randomized crossover design, but the registry presents arm-level outcome entries without a paired estimate. In general, within-person comparisons can account for stable differences between participants, but the analysis needs to preserve the pairing. Reporting only group-level values makes it difficult to assess how individual observations contributed to the comparison.

The outcome is also described in terms that do not fully align: the title refers to a rate, the unit refers to a percentage of lesions, and the outcome parameter is recorded as a number. The description specifies additionally visualized lesions. Clear definitions of the outcome, denominator, and lesion confirmation process are important for interpreting what the counts represent and for comparing methods consistently.

The registry’s enrollment and outcome entries also illustrate why analysis populations matter. The record lists 23 enrolled participants and 20 in each primary-outcome arm, but does not explain their relationship. In a crossover study, readers need to know how many participants contributed data from both procedures and how incomplete pairs were handled.

Finally, no inferential analyses are posted. Descriptive entries without an estimate of uncertainty cannot show the precision of a comparison. For a feasibility trial, transparent reporting of paired data, outcome definitions, missingness, and analysis methods is particularly useful for planning future studies. These issues are relevant to the interpretation of this record without implying that the techniques differ in effectiveness.

14. Limitations

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

The registry record leaves several trial-specific details unreported. It does not give participant characteristics beyond the eligibility description, explain how the 23 enrolled participants relate to the 20 recorded for each primary-outcome arm, or define the population included in the outcome summaries. The record does not identify how many participants contributed observations from both procedures.

The primary outcome wording is not fully aligned: it is titled a rate, has a unit of percentage of lesions, and is recorded as a number described as additionally visualized lesions. No denominator, operational definition of an additional lesion, or posted pathology result is available. This limits interpretation of the values and makes it unclear what exact comparison they represent.

Although the design is crossover, the registry does not report paired observations, a paired analysis, or an adjustment for procedure order or period. It does not report an inferential analysis, confidence interval, p-value, or analysis plan. The registry also does not report a trial-specific sample-size justification, missing-data methods, or details of follow-up completeness. The planned one-year recurrence surveillance has no posted outcome.

The study is described as a feasibility trial with 23 enrolled participants. The available arm-level counts do not provide enough information to assess the precision of a comparative effect. No masking was used. The registry does not report whether outcome assessment or pathology interpretation was blinded. These limitations concern what can be established from the record; they do not indicate why any particular design or reporting choice was made.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Laparoscopic Narrow Band Imaging for Detection of Occult Cancer Metastases — NCT01944930.
• Registry record last updated 2021-01-08; status last verified 2020-12.
• Results first posted 2021-01-08.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

18. Record Summary

NCT01944930, “Laparoscopic Narrow Band Imaging for Detection of Occult Cancer Metastases,” is a completed randomized, unmasked, Phase 1 crossover feasibility trial sponsored by Lahey Clinic. It enrolled 23 participants with known or suspected gastrointestinal or gynecologic malignancies and an indication for diagnostic laparoscopy. Participants were to undergo both NBI and standard white-light laparoscopy, with procedure order randomized.

The primary outcome concerns additionally visualized peritoneal lesions at the time of each procedure. The registry records a value of 0 for NBI laparoscopy and 2 for standard white-light laparoscopy, with 20 listed in each arm. It posts no inferential analyses and no secondary outcomes. It does not report a paired comparison, uncertainty estimate, recurrence result, or explanation of how the outcome entries relate to total enrollment. Safety entries list zero serious adverse events, zero other adverse events, and zero deaths in both procedure-order groups.