This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for Clinical Evaluation of Narrow Band Imaging Colonoscope (NCT00628147), a randomized phase 4 study. The primary outcome is Neoplasm Miss Rate.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT00628147 |
| Phase | Phase 4 |
| Condition | Colonic Neoplasms |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Single |
| Enrollment | 284 |
| Arms in the primary outcome | 2 |
| Lead sponsor | VA Palo Alto Health Care System |
| Results posted | Yes |
2. Clinical Question
Population
The registry identifies the condition as colonic neoplasms and describes a tandem colonoscopy study. Enrollment was 284. Eligibility criteria and further details about the participants are not reported in the registry record, so the population to which the results apply cannot be characterized more specifically from this record.
Intervention
The intervention was a narrow band imaging colonoscope. The study description frames the question as whether examination using narrow band imaging detects polyps more effectively than examination using standard full spectrum white light. Procedural details beyond the registry descriptions are not reported in the registry record.
Comparator
The comparator was conventional white light examination, also named White Light in the posted results. The registry describes the comparison as randomized and parallel, with single masking. It does not report details about who was masked or how the assigned examinations were delivered.
Primary question
The primary endpoint was neoplasm miss rate, expressed as a percentage and assessed on the same day. The central question is whether the posted miss-rate estimates differ between the narrow band imaging colonoscope and white light groups. The record gives a point estimate and confidence interval for each group, but does not post an analysis estimating the between-group difference.
The official study title describes a tandem colonoscopy study intended to compare neoplasia miss rates. In general, a miss-rate endpoint concerns abnormal findings not identified during an examination but identified through the study’s subsequent assessment. The exact operational definition, including how the denominator was formed and how findings were attributed, is not reported in the registry record. That missing detail matters because a miss rate may be defined at the level of findings, examinations, or participants, and those definitions do not necessarily answer identical questions.
The trial record identifies phase 4 and randomized allocation. Its brief summary states that the purpose was to determine whether narrow band imaging was more effective at detecting polyps than standard full spectrum white light. The statistical question is narrower than a general comparison of colonoscopy strategies: the reported primary outcome concerns the percentage of neoplasms missed, not all aspects of detection, examination completion, or procedure-related outcomes.
3. Trial Design
The record describes a phase 4, randomized, single-masked, parallel study. It lists 284 as the enrollment and names the VA Palo Alto Health Care System as sponsor. The study start is recorded as 2006-01, primary completion as 2007-02, and status as completed. These registry entries describe the trial’s administrative record; they do not establish how closely the conduct followed the planned design or whether all planned analyses were completed.
The two named groups were Narrow Band Imaging Colonoscope and White Light. The corresponding descriptions are narrow band imaging colonoscope and Conventional White Light Examination. The intervention type is listed as a device. The registry does not describe the sequence of examinations in sufficient detail to reconstruct the tandem procedure, the assignment process, or whether the same participant could contribute findings to both examination conditions. The parallel model identifies the design category but does not, by itself, answer those procedural questions.
Single masking is recorded, but the identity of the masked party or parties is not reported in the registry record. The record also does not explain how masking was maintained during the examinations, whether it could be broken, or whether outcome assessment was masked. These details are relevant because expectations and knowledge of the examination method could influence procedural conduct or assessment.
The posted primary outcome is Neoplasm Miss Rate, with a miss rate percentage, a 95% confidence interval, and a same-day timeframe. Three secondary outcomes are listed: Neoplasm Detection Rates, Completion of Examinations, and Procedure Complications. The record indicates four outcome measures, but the details for the additional measure are not specified in the listed outcome titles. It reports no analyses. Definitions, analysis populations, the handling of missing observations, and the planned statistical comparison are not reported in the registry record.
The registry’s summary states the broad purpose as a comparison of polyp detection using narrow band imaging and standard full spectrum white light. That statement should be distinguished from the outcome data: the posted primary endpoint is a neoplasm miss-rate percentage, while the secondary outcomes address other aspects of examination and complications. The record does not provide enough procedural detail to determine how findings were classified or how the endpoints relate to each other.
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
| Stage | Narrow Band Imaging Colonoscope | White Light |
|---|---|---|
| STARTED | 142 | 142 |
| COMPLETED | 135 | 141 |
| NOT COMPLETED | 7 | 1 |
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.
| Characteristic | Unit | Narrow Band Imaging Colonoscope | White Light | Total |
|---|---|---|---|---|
| Participants analysed | 142 | 142 | 284 | |
| Age, Continuous | years | 64 (SD 10) | 64 (SD 10) | 64 (SD 10) |
| Gender: Female | Participants | 5 | 4 | 9 |
| Gender: Male | Participants | 137 | 138 | 275 |
6. Primary Endpoint
The registry lists the primary outcome as Neoplasm Miss Rate, measured over same day.
| Element | Registry entry |
|---|---|
| Result type | Number |
| Measure of spread | 95% Confidence Interval |
| Unit | Miss rate percentage |
| Endpoint classification | not reduced to a single comparison |
The primary endpoint is Neoplasm Miss Rate, reported as a percentage for the same day. In general, a miss rate expresses the share of relevant findings that were not detected during an examination according to the study’s chosen reference or subsequent assessment. It is useful because an examination’s ability to identify neoplasia can be considered not only through findings detected, but also through findings that were initially overlooked.
The registry gives no written endpoint definition, denominator, or adjudication procedure. It therefore does not establish precisely what counted as a neoplasm, how an initially missed finding was identified, or how multiple findings from a participant were treated. Those choices affect interpretation: a finding-based rate and a participant-based rate describe different units, even when both are expressed as percentages. The record’s tandem-study title signals the context for assessing misses, but does not supply the operational details needed to reproduce the endpoint.
A confidence interval accompanies each reported percentage. In general, a confidence interval represents statistical uncertainty around an estimate under the method used to construct it. It is not a range of all possible outcomes, nor does it by itself test whether the groups differ. Because the registry does not report a direct comparison or its confidence interval, the group-specific intervals should be read as descriptions of each estimate’s precision, not as a substitute for an estimated treatment contrast.
The endpoint is specific to neoplasm misses; it is not a comprehensive measure of diagnostic performance, procedural completion, complications, or patient outcomes. The record lists some of these topics as secondary outcomes, but their results are not included in the posted outcome values.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The reported neoplasm miss-rate estimate for Narrow Band Imaging Colonoscope is 12.6%, with a 95% confidence interval from 7.5% to 19.4%. The estimate for White Light is 12.1%, with a 95% confidence interval from 7.2% to 18.6%. These are group-specific estimates. The posted record does not include a between-group estimate, a hypothesis test, or a p-value for the primary endpoint.
The point estimates are close in magnitude, and the confidence intervals overlap. This describes the reported estimates but does not establish equivalence, non-inferiority, or absence of a difference. Overlap between separate confidence intervals is not a formal test of the difference between groups. Conversely, the lack of a posted comparison does not establish that the groups are the same. A direct contrast with its uncertainty and the analysis used to obtain it would be needed to evaluate the comparative question statistically.
The intervals indicate that the reported point estimates are not perfectly precise. The uncertainty is compatible with a range of miss-rate values for each group under the interval method. The registry does not provide the method used to calculate the intervals, so assumptions behind their coverage cannot be checked from the record. The interval width also does not tell whether the trial was designed to detect a particular difference; the sample-size justification and planned effect size are not reported in the registry record.
The primary result should be interpreted alongside the endpoint’s incomplete definition. The record calls it a miss rate percentage and gives a same-day timeframe, but does not explain the denominator or the exact identification process. Without those details, readers cannot verify whether the outcome represents missed neoplasms per examination, per participant, or another unit. This uncertainty also limits comparisons with results from other studies that may define miss rates differently.
The posted arm sample sizes are 135.0 for Narrow Band Imaging Colonoscope and 141.0 for White Light. These are the Ns attached to the outcome estimates; the registry separately gives enrollment as 284. The record does not explain the relationship between enrollment and outcome-specific Ns, or whether exclusions, missing observations, or a particular analysis population account for the difference in the arm counts. No conclusion about attrition or completeness should be drawn without that information.
The record says that analyses are not posted. This is consequential: a pair of estimates with confidence intervals is not the same as a randomized comparison analysis. No adjustment, stratification, clustering approach, or handling of any within-participant correlation is described. If the study design produced multiple findings from a participant, the appropriate analysis could depend on how that dependence was handled; the registry record does not report whether it was relevant or addressed.
Accordingly, the posted results support a descriptive statement about the two arm-specific estimates and their reported uncertainty. They do not, by themselves, quantify the difference between examination methods or resolve the study’s comparative question. The result should not be described as evidence that one method is superior, equivalent, or ineffective. Such claims would go beyond the analyses reported in the registry record.
9. Statistical Methods Explained
Random allocation is intended, in general, to make comparison groups similar in expectation with respect to factors that could influence an outcome. It does not guarantee identical groups in a particular trial, and it does not remove the need to describe the actual allocation process and analysis. The record labels this study randomized, but the sequence-generation method, allocation concealment, and any restrictions on allocation are not reported in the registry record. The posted outcome Ns also do not explain which participants or findings were included in each estimate.
The parallel design label indicates that the record treats the groups as parallel arms. It does not explain the detailed structure of the tandem colonoscopy procedure or whether an individual could contribute outcome information related to both examination methods. That distinction matters statistically. If observations are correlated within a participant, an analysis treating every finding as independent could understate uncertainty. In general, methods for clustered or repeated observations account for such dependence, but the registry does not report whether the endpoint had this structure or how it was handled.
The primary outcome is a percentage, accompanied by a 95% confidence interval. A point estimate is a concise summary of the observed data under the study’s definition. The interval communicates uncertainty around that estimate under a statistical method and its assumptions. It is not a guarantee that the underlying value lies in the interval, and it does not supply a direct comparison of the two groups. Because the construction method is not given, the registry record does not allow readers to assess whether the interval method was appropriate for the endpoint’s denominator or data structure.
Comparative inference ordinarily focuses on a contrast, such as a difference in rates, and its uncertainty. The contrast can be evaluated against a prespecified question using an estimate, confidence interval, and possibly a test. The registry gives separate estimates for the two groups but no direct contrast, test statistic, or p-value. The absence of those items means that a comparative statistical conclusion cannot be recovered from the group-specific intervals alone. The record also does not identify whether a planned analysis would have adjusted for any baseline characteristics or design features.
In general, a confidence interval’s width depends on the amount and structure of information and on the method used. Over-dispersion can arise when observed variability exceeds what a simple model assumes; if ignored, uncertainty may be understated. This is a general consideration, not a statement that over-dispersion occurred in this trial. The registry record does not report the distribution of findings, the variance model, or diagnostics that would establish whether this issue arose.
The registry includes a sample-size planning assumption: the planning illustration uses the observed effect, a stated fraction of that observed effect, equal allocation, a two-sided significance level of 0.05, and a normal approximation. Such an illustration is a planning exercise, not evidence about the trial’s actual design or the probability that the trial detected an effect. The registry record does not report the trial’s own sample-size justification, its target effect, or any planning values from such an analysis. Planning calculations based on an observed effect are especially dependent on that estimate and should not be treated as observed power.
Missingness is another analytic concern for percentage outcomes. In general, an estimate can be affected if outcome information is unavailable for some participants or findings, especially if the chance of missing information is related to the outcome. The record does not report missing-data counts, reasons, assumptions, or methods. Nor does it define whether the posted Ns represent all randomized participants, participants with assessable outcomes, or another population. Without a participant flow or analysis-population description, the extent and potential effect of missing information cannot be assessed.
Multiplicity concerns arise when a study evaluates several outcomes or several analyses. The registry lists Neoplasm Detection Rates, Completion of Examinations, and Procedure Complications as secondary outcomes, in addition to the primary endpoint. In general, considering many outcomes can increase the chance of finding an apparently notable result by chance if inference is not planned or interpreted appropriately. The record does not report a multiplicity strategy, hierarchy among outcomes, or results for the listed secondary outcomes. It also does not report analyses, so there is no basis to determine which comparisons were conducted or how they were handled.
Single masking can reduce some forms of expectation-related influence, but the label alone does not reveal who was masked or whether outcome assessment was protected from knowledge of assignment. In general, masking is more informative when the masked roles and procedures are specified. Those details are not reported in the registry record. Likewise, the record does not describe protocol deviations, adherence, or whether the analysis followed randomized assignment. These omissions limit appraisal of how the design was implemented.
The statistical information posted is therefore limited to arm-specific point estimates, confidence intervals, and arm Ns for the primary outcome. No recalculated comparison is warranted from the available summary statistics, and the sponsor’s own analysis cannot be assessed because no analysis is posted. The most defensible statistical reading is descriptive: retain the stated estimates and uncertainty, avoid substituting visual interval comparison for a formal contrast, and distinguish planning assumptions from reported trial analysis.
10. Multiplicity, Missing Data and Analysis Populations
The record lists three secondary outcomes: Neoplasm Detection Rates, Completion of Examinations, and Procedure Complications. It also lists four outcome measures overall, while the primary endpoint is Neoplasm Miss Rate. The additional measure is not identified in the listed outcome titles. No secondary outcome results or analyses are posted. A multiplicity plan, outcome hierarchy, or explanation of which outcomes were intended to support confirmatory inference is not reported in the registry record.
In general, multiplicity becomes relevant when several outcomes, timepoints, subgroups, or comparisons are examined. Without a prespecified approach, repeated testing can make nominal evidence harder to interpret. This is a general statistical concern, not a claim that a particular multiplicity problem occurred in this trial. Because analyses are not posted, the record does not show how many comparisons were conducted or whether any adjustment was used.
The registry does not describe the analysis population for the primary endpoint, the handling of missing findings or participant data, or the reasons that the outcome Ns differ from the enrollment figure. It does not provide a participant flow, protocol-deviation summary, or missing-data strategy. These omissions prevent readers from determining whether the estimates reflect all randomized participants or a subset with assessable outcomes.
In general, complete-case analysis can be misleading when missingness is related to outcome or group assignment. Other methods also require assumptions, which should be made explicit. No missing-data assumptions or methods are reported in the registry record, so their appropriateness cannot be evaluated. The absence of these details does not prove that data were missing; it means the record does not describe their availability or handling.
The record reports no analyses. Consequently, it does not identify an intention-to-treat, per-protocol, or other analysis population, and none should be inferred from the randomized design label or the arm Ns alone.
11. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| Narrow Band Imaging Colonoscope | 0 | 142 | 0.0% |
| White Light | 0 | 142 | 0.0% |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 0% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| Narrow Band Imaging Colonoscope | 0 of 142 |
| White Light | 0 of 142 |
The adverse-event record lists 142 at risk in the Narrow Band Imaging Colonoscope group and 142 at risk in the White Light group. Serious adverse events affected 0 in each group. Other adverse events affected 0 in each group. Death counts are null, so death information is not reported in the registry record. No serious-event terms are listed.
These figures should be read as the registry’s recorded counts, not as proof that no adverse events could occur or that the methods have identical safety profiles. A zero count in the posted record means no affected participants were recorded for that category in the stated group; it does not establish the absence of risk. The record does not define the other-adverse-event category, give event-level descriptions, or explain the collection and reporting procedures.
The at-risk counts are separate safety figures and should not be assumed to match the outcome Ns or enrollment. The registry does not explain the relationship between the safety and efficacy populations, the observation period for adverse events, or whether all participants were followed for the same duration. Although the primary endpoint timeframe is same day, the adverse-event observation window is not reported in the registry record.
Because the posted safety data contain no event descriptions and no reported death counts, they offer limited information for comparing safety. The counts should be interpreted narrowly as registry entries, without converting them into percentages or drawing conclusions about relative risk.
12. What the Result Does, and Does Not, Mean
The result means that the registry reports a neoplasm miss-rate estimate of 12.6% for Narrow Band Imaging Colonoscope and 12.1% for White Light, each with a 95% confidence interval. Those values describe the posted group-specific estimates under the endpoint definition used by the study, though that definition is not fully reported. The intervals communicate uncertainty around each estimate according to an unreported calculation method.
The result does not provide a formal estimate of the difference between groups. It therefore does not show whether the observed separation in point estimates is statistically distinguishable from chance variation, nor whether a clinically meaningful difference was excluded. Separate intervals, even when overlapping, cannot establish equivalence or non-inferiority. Those conclusions require an appropriate design, a prespecified margin where relevant, and a direct analysis that is not reported in this record.
The results also do not establish how generalizable the estimates are. The eligibility criteria, participant characteristics, detailed examination setting, and exact endpoint procedures are not reported in the registry record. The record names a sponsor and condition, but that is not a substitute for a description of the population and procedures needed to judge applicability.
The listed secondary outcomes indicate that detection rates, examination completion, and procedure complications were of interest. Their results are not posted in the data described by the record. The primary miss-rate estimates should not be treated as a complete account of those other outcomes, or as evidence about long-term consequences. The study is described as same-day for the primary endpoint, and no longer-term outcome is reported.
In practical statistical terms, the safest conclusion is descriptive rather than comparative: the posted estimates are similar in magnitude, with overlapping reported intervals, but the record does not provide the direct analysis needed to answer the comparative question. This statement does not mean the methods are the same, and it does not mean one method is better. It identifies what the registry data can support and where interpretation stops.
13. Why This Trial Matters Statistically
This trial illustrates the difference between reporting an estimate for each arm and reporting a randomized comparison. A reader can see the neoplasm miss-rate point estimates and their confidence intervals, but the comparative question concerns the contrast between the arms. Without that contrast and its uncertainty, the main statistical question remains unresolved by the posted analysis.
It also highlights why endpoint definitions matter. A percentage is not self-explanatory: its meaning depends on what is counted in the numerator and denominator, how findings are identified, and whether the unit is a participant, examination, or lesion. The registry names the endpoint and timeframe but does not report those details. A clear operational definition is essential for reproducibility and for comparing estimates across studies.
The record provides a useful example of the limits of interpreting separate confidence intervals. Their overlap is descriptive, not a formal test of the difference. The interval widths express uncertainty around each group estimate, but do not directly quantify uncertainty in the contrast. Statistical interpretation should keep those roles distinct.
Finally, the registry entries demonstrate why complete reporting includes analysis populations, missing-data handling, multiplicity plans, and safety follow-up. Randomized allocation is an important design feature, but it cannot compensate for missing details about the endpoint or analysis. The posted estimates are informative as descriptive summaries, while the absent comparison and methodological details constrain stronger conclusions.
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 primary endpoint is not operationally defined in the registry record. The numerator, denominator, unit of analysis, reference procedure, and classification process are not described. The tandem-study title supplies context but does not resolve these details. As a result, readers cannot determine exactly what the reported percentage represents or reproduce its calculation from the registry information.
No direct statistical analysis is posted. The record does not provide a contrast, confidence interval for a contrast, test result, or p-value. It also does not report the statistical method used for the group-specific intervals. This prevents evaluation of the primary comparative question and the assumptions underlying the posted uncertainty estimates.
The registry gives enrollment as 284 and outcome Ns of 135.0 and 141.0, but does not explain how those Ns were determined. There is no reported participant flow, missing-data summary, analysis-population definition, or account of exclusions. The record does not state whether the outcome was assessed for every randomized participant or whether findings within a participant were treated as dependent.
Further limitations include the absence of eligibility criteria, participant characteristics, allocation procedures, masking details, examination-procedure details, and protocol-deviation information. The record lists secondary outcomes but does not post their results or describe a multiplicity plan. It does not report the trial’s sample-size justification or the intended effect size, and the planning assumptions are not a substitute for the protocol’s design rationale.
Safety reporting is also limited: serious and other adverse-event counts are zero in both listed groups, while death counts are not reported. Event definitions and follow-up duration are absent. These registry limitations constrain interpretation; they do not establish that the trial itself lacked procedures or analyses that were simply not recorded in this entry.
15. Related Tutorials
16. Related Statistical Calculators
17. Sources
- ClinicalTrials.gov: Clinical Evaluation of Narrow Band Imaging Colonoscope — NCT00628147.
- Registry record last updated 2017-03-08; status last verified 2017-01.
- Results first posted 2017-03-08.
- 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.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
18. Record Summary
NCT00628147 is a completed phase 4 randomized, single-masked, parallel study sponsored by VA Palo Alto Health Care System. It compares a narrow band imaging colonoscope with conventional white light examination in a tandem colonoscopy study. Enrollment is 284; the recorded start is 2006-01 and primary completion is 2007-02.
The primary outcome is same-day Neoplasm Miss Rate, expressed as a percentage with a 95% confidence interval. The posted estimate is 12.6% (95% CI 7.5% to 19.4%) for Narrow Band Imaging Colonoscope and 12.1% (95% CI 7.2% to 18.6%) for White Light. The registry does not post a direct comparison or analysis.
Neoplasm Detection Rates, Completion of Examinations, and Procedure Complications are listed as secondary outcomes, without results in the record described. Serious and other adverse-event counts are 0 in each group; death information is not reported. Key endpoint, population, missing-data, and analysis details are not reported in the registry record.