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 Continuous Quantified EEG in NeuroIntensive Care (NCT02901262), a registered phase not stated study. The primary outcome is Number of Correct qEEG Evaluations.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT02901262 |
| Phase | Phase not stated |
| Condition | Coma, Traumatic Brain Injury, Subarachnoid Hemorrhage |
| Allocation | Not Stated |
| Design model | Not Stated |
| Masking | Not Stated |
| Enrollment | 30 |
| Arms in the primary outcome | n/a |
| Lead sponsor | Azienda Ospedaliera San Gerardo di Monza |
| Results posted | Yes |
2. Clinical Question
Population
Neuro-intensive care staff interpreting continuous EEG and quantitative EEG tracings. The registry describes patients with continuous EEG lasting >24 hours in the two study units. The conditions listed include coma, traumatic brain injury, subarachnoid hemorrhage, and ischemic stroke.
Intervention
An in-house training period for neuro-intensivists in interpreting EEG-derived information. The record describes assessment before and after training, covering symmetry, sedation level, seizure activity, and artefact.
Comparator
The primary outcome compares intensivists’ evaluations with those of a neurophysiologist. The record also describes interpretation before and after training, but does not state a randomized allocation or identify a separate control arm. These are not described as like-for-like treatment groups.
Primary question
After in-house training, what percentage of quantitative EEG evaluations by neuro-intensivists is correct when assessed against neurophysiologist evaluations? The primary outcome timeframe is 2 months. The registry record does not report the resulting percentage or the number of evaluations used to calculate it.
This is a study of staff interpretation rather than a trial comparing clinical treatments. Its focus is whether trained staff can interpret quantitative EEG tracings across specified domains. The registry describes patients as the source of tracings, while the primary outcome population is neuro-intensivists after training. Those populations serve different roles: patients contribute the recordings, and clinicians perform the evaluations being assessed.
The question is relevant to interpretation of EEG information in neuro-intensive care, but the record does not establish that any evaluated reading changed patient management or outcomes. It reports an intended comparison of staff assessments with neurophysiologist assessments, not a patient-level clinical benefit question. The numerical result needed to judge the observed accuracy is not reported in the registry record.
3. Trial Design
The registry identifies this study as completed, with enrollment of 30. The record gives a start date of 2016-01 and a primary completion date of 2017-01. The sponsor is Azienda Ospedaliera San Gerardo di Monza. The title concerns continuous quantified EEG in neuro-intensive care, and the official title describes the utility of continuous electroencephalogram and derived variables in that setting.
The stated purpose is to evaluate neuro-intensive care unit staff interpretation of continuous EEG and quantitative EEG tracings before and after a training period. The domains named are symmetry, sedation level, seizure activity, and artefact. The primary outcome is the percentage of correct evaluations by intensivists compared with a neurophysiologist, with the primary population specified as neuro-intensivists after in-house training and a timeframe of 2 months.
The arm description refers to patients with continuous EEG lasting >24 hours in the two study units. The registry does not describe this as a randomized treatment comparison. Allocation is not stated, and the record cautions that the arms are not a like-for-like control comparison. Accordingly, a contrast between groups, if considered, should not be read as a randomized estimate of a treatment effect. The record does not report the operational details of how clinicians, tracings, or evaluation occasions were selected or matched.
A secondary outcome is the kappa coefficient of agreement between intensivists and neurophysiologists after training. The registry lists 2 outcome measures but 0 analyses. It does not report numerical outcome estimates, uncertainty intervals, or analysis results. The design description therefore clarifies the evaluation goal, but does not provide the results needed to quantify accuracy or agreement.
Several design details are not reported in the registry record, including a sample-size justification, the number of clinicians, the number of tracings evaluated, the distribution of observations among clinicians or patients, and whether the same staff contributed evaluations before and after training. The record also does not specify masking, the training content, or the statistical analysis plan. These omissions limit assessment of how the evaluation was conducted and how broadly its results might apply.
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: After the evaluation of qEEG (quantitative electroencephalography) from 30 patients, other intensivists and neurophysiologists were engaged to use the customized web-based online application that was downloaded on computers and cell phones to evaluate qEEG.
Overall Study
| Stage | Study Cohort |
|---|---|
| STARTED | 30 |
| COMPLETED | 30 |
| NOT COMPLETED | 0 |
Analysis population for the primary outcome, as defined in the registry: Neuro-intensivists, after the in-house training
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 | qEEG Patients |
|---|---|---|
| Participants analysed | 30 | |
| Age, Categorical: <=18 years | Participants | 0 |
| Age, Categorical: Between 18 and 65 years | Participants | 30 |
| Age, Categorical: >=65 years | Participants | 0 |
| Sex: Female, Male: Female | Participants | 18 |
| Sex: Female, Male: Male | Participants | 12 |
| Race and Ethnicity Not Collected | Participants | — |
| Region of Enrollment: Italy | participants | 30 |
6. Primary Endpoint
The registry lists the primary outcome as Number of Correct qEEG Evaluations, measured over 2 months.
| Element | Registry entry |
|---|---|
| Result type | Number |
| Measure of spread | not applicable |
| Unit | percentage of correct evaluation |
| Endpoint classification | not reduced to a single comparison |
The primary endpoint is the percentage of correct quantitative EEG evaluations by neuro-intensivists after the in-house training, compared with evaluations by a neurophysiologist. The description says results are expressed as a percentage of correct evaluations for each domain. The domains introduced in the study summary are symmetry, sedation level, seizure activity, and artefact.
A percentage expresses how often evaluations meet a stated reference standard relative to the evaluations assessed. In general, this can make performance easier to interpret than a raw count when the number of assessed readings is known. The registry does not state the denominator, the number of recordings or assessments, or the rule used to define a correct evaluation. The reported percentage and its precision are also not reported in the registry record.
Comparing an intensivist’s assessment with a neurophysiologist’s assessment is intended to describe how closely the interpretations align with that reference. The endpoint is therefore about classification or interpretation, not patient recovery, safety, or treatment effectiveness. A percentage of correct evaluations can summarize performance, but it does not by itself show whether errors have different consequences across domains, whether accuracy varies between clinicians, or whether observed performance would generalize to other settings.
Using domain-specific percentages can preserve differences among the kinds of features being interpreted rather than collapsing all evaluations into a single summary. However, the registry does not report domain-specific estimates, uncertainty, or the procedures used to combine observations. Without those details and the actual results, the endpoint’s intended meaning is clear, but the magnitude and precision of performance cannot be assessed.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The registry record describes an evaluation of interpretation after training, with neurophysiologist assessments serving as the comparison for the primary endpoint. It also describes staff assessment before and after training. These statements establish the question being investigated, but they do not provide a numerical estimate of the percentage of correct evaluations. The primary result is not reported in the registry record, so no conclusion about the observed level of accuracy can be drawn from the available outcome information.
The record lists agreement between intensivists and neurophysiologists after training as a secondary outcome, expressed using a kappa coefficient. No kappa estimate is reported. Agreement and correctness are related but distinct statistical ideas. Correctness evaluates an assessment against a designated reference, while agreement describes how often assessors give the same classification after accounting for agreement that might occur by chance under the chosen calculation. Neither measure alone describes the clinical consequences of a mistaken interpretation.
The registry marks the study completed and records enrollment of 30, but does not report how many staff, tracings, or evaluations contributed to the primary endpoint. Enrollment should not be treated as the denominator for an accuracy percentage: patients may provide recordings, while staff make the interpretations. The relationship between those observations and the enrolled participants is not reported. Nor does the record state whether repeated evaluations from a clinician or repeated tracings from a patient were handled as dependent observations.
That dependence matters in general. If the same clinician evaluates multiple tracings, those assessments may be more alike than assessments from different clinicians. Similarly, multiple readings from a patient may share features. Analyses that treat every reading as independent can understate uncertainty when observations are clustered. The registry does not report whether clustering was assessed or addressed, so the precision of any unreported estimate cannot be evaluated.
The study summary’s before-and-after framing does not establish that training caused a change in interpretation. The registry does not report random allocation, and it does not identify a like-for-like control comparison. A difference over time could reflect training or other influences, but the record lacks the design and results needed to distinguish among explanations. Since numerical before-and-after results are not reported, even the size and direction of any change are unknown.
The record lists 0 analyses and no numerical primary or secondary outcome findings. It also does not report confidence intervals, standard errors, or hypothesis-test results. This absence is not evidence of no effect or no agreement; it means the posted registry information does not quantify them. The study’s completed status and stated objective should therefore be kept separate from evidence about what the evaluations found.
Planning illustrations, when interpreted, are not observed power or trial results. The stated planning assumptions concern an illustrative calculation based on an observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, a normal approximation, and effects expressed as a stated fraction of the observed effect. Those assumptions do not establish that the study was designed using them, and the record says the illustration is not the trial’s own design. They cannot substitute for the missing outcome estimates or an account of the actual analysis.
9. Statistical Methods Explained
What does a percentage of correct evaluations represent?
In general, a percentage of correct evaluations is a proportion: the number of assessments meeting a specified correctness rule relative to the assessments eligible for evaluation. It summarizes performance in an intuitive scale, but interpretation depends on both the numerator and denominator. The registry states that results will be expressed as percentages for each domain, but does not report the counts, the denominator, or the precise correctness rule. Consequently, the reported outcome cannot be reconstructed from the registry record.
A percentage is an estimate from the observations assessed, not necessarily a fixed description of all future interpretations. In general, uncertainty intervals can show how compatible the data are with a range of underlying accuracy values. Their width depends on the amount of information and on the analysis assumptions. No interval estimate is reported. The count of enrolled patients does not reveal the amount of information for this endpoint because the primary population is neuro-intensivists and the assessment unit is an evaluation. The number of evaluations and how they relate to patients and clinicians are not reported.
Why distinguish accuracy from agreement?
Accuracy compares a judgment with a reference classification. In this study description, the primary comparison is between intensivists’ quantitative EEG evaluations and neurophysiologist evaluations. Agreement instead asks how often assessors assign the same classification. Agreement can be high when a category is common even if the assessment process has limitations, and disagreement can occur for several reasons. In general, a reference assessment is useful for defining correctness, but it is not automatically a perfect measure of an underlying truth.
The listed secondary outcome uses a kappa coefficient of agreement after training. In general, kappa compares observed agreement with the agreement expected under a particular chance-agreement model. Its interpretation depends on category frequencies, the assessment design, and the chosen form of the coefficient. A kappa value should be interpreted alongside the underlying agreement information and the way categories were defined. The registry does not report a kappa result, the underlying cross-classification of ratings, or details of the coefficient calculation.
How should repeated observations be considered?
General statistical reasoning calls for attention to clustering when assessments are not independent. Evaluations may be grouped within clinicians, patients, or recordings. When observations within a cluster resemble each other, counting each evaluation as wholly independent can make uncertainty appear smaller than it is. Methods that account for clustering can model or otherwise adjust for those dependencies. The registry does not state the number of clinicians or readings, whether observations were repeated, or whether a clustered analysis was used. It would therefore be inappropriate to assume a particular model or describe an adjustment as part of the study’s analysis.
Domain-specific outcomes also raise a general question about how results are summarized. Separate percentages can make performance in symmetry, sedation level, seizure activity, and artefact easier to distinguish. A combined percentage might conceal differences between domains, while examining multiple domain results may create multiplicity concerns. The registry does not say whether a combined measure was planned, whether all domains were treated as primary, or how any multiple testing was handled.
What can confidence intervals and tests add?
In general, a confidence interval communicates uncertainty around an estimate under the assumptions of the method used. A narrow interval indicates greater precision than a wide interval, but does not by itself establish practical importance or eliminate bias. A hypothesis test addresses compatibility with a specified statistical null model, not whether an outcome is clinically meaningful. Neither an interval nor a p-value can remedy limitations in sampling, measurement, or study design. No confidence intervals, p-values, or standard errors for the primary or secondary outcomes are reported in the registry record.
The stated planning assumptions describe an illustrative sample-size calculation based on an observed effect. They specify a two-sided significance level of 0.05, equal allocation between the two arms, a normal approximation, and detection of a stated fraction of the observed effect. This is a planning illustration, not the study’s own design justification and not observed power. In general, planning calculations depend on assumptions about the target effect, variability, allocation, and analysis. The registry does not report a sample-size justification for this study, and the illustrative assumptions should not be mistaken for proof that the design achieved a particular probability of detecting an effect.
When an illustration is based on the observed effect, its output is conditional on that estimate and should be read as a planning aid rather than a retrospective measure of evidential strength. Sample-size illustrations describe hypothetical future or alternative planning choices under stated assumptions. They do not turn an unreported endpoint into a reported result. The registry record gives no numerical analysis result to assess alongside such planning information.
What analysis details are missing?
The registry does not report a statistical analysis plan, the methods for handling missing evaluations, an analysis population, or whether comparisons were adjusted for any variables. It also does not identify how observations were assigned to staff, whether assessors were masked to other ratings, or how the neurophysiologist reference was established. These details affect interpretation of both the estimated percentage and its uncertainty. Their absence does not show that no such procedures were used; it means they are not reported in the registry record.
10. Multiplicity, Missing Data and Analysis Populations
The registry describes several interpretation domains: symmetry, sedation level, seizure activity, and artefact. It says that the percentage of correct evaluations will be expressed for each domain. The record does not explain whether those domain-specific measures were all treated as primary, whether one was designated as the main inferential target, or whether a combined result was also planned. A secondary outcome is the kappa coefficient of agreement between intensivists and neurophysiologists after training. The approach to multiplicity is not reported in the registry record.
In general, evaluating multiple outcomes or domains can increase the chance that at least one result appears noteworthy by chance when separate tests are used. Statistical approaches to multiplicity can control or describe that issue, but the appropriate approach depends on the intended questions and analysis plan. Since no analysis plan or numerical results are reported, it is not possible to determine whether multiplicity adjustments were planned or used. The registry lists 0 analyses.
The registry does not report how missing or unusable tracings, incomplete assessments, or missing domain ratings were handled. It does not state whether all eligible evaluations were included, whether exclusions occurred, or whether any method was used to address missing data. In general, missingness can affect an accuracy estimate if the evaluations that remain differ systematically from those not assessed. The record does not give enough information to assess that possibility.
The analysis population is also not reported in detail. The primary population is described as neuro-intensivists after in-house training, while the arm description refers to patients with continuous EEG lasting >24 hours. The number of contributing clinicians and evaluations, and the connection between patient recordings and staff assessments, are not reported. Thus, the registry does not identify a denominator or provide a clear accounting of which observations entered each outcome calculation.
11. What the Result Does, and Does Not, Mean
The study addresses a practical measurement question: how neuro-intensivists interpret quantitative EEG tracings after training, compared with neurophysiologist assessments. The focus on specified interpretation domains makes the question more concrete than a broad claim about the usefulness of EEG. The record also lists agreement after training as a secondary outcome, which is conceptually distinct from the primary percentage of correct evaluations.
However, the registry record does not report the primary percentage, its domain-specific values, the kappa coefficient, or uncertainty estimates. It therefore does not establish what accuracy or agreement was observed, how precise those estimates were, or whether performance differed across domains. A completed status indicates the study is marked completed in the registry, but completion alone is not an outcome result.
Even with results, an accuracy estimate would describe the evaluations assessed under the study’s procedures. It would not automatically establish how staff would perform in other units, with different recordings, or under different training conditions. Generalizability depends on who was assessed, which tracings were included, how ratings were obtained, and whether observations reflect the broader clinical setting. Those details are not fully reported in the registry record.
The study is not described as a randomized comparison of treatment groups, and the record explicitly notes that its arms are not a like-for-like control comparison. The stated before-and-after assessment therefore should not be interpreted as definitive evidence that training caused any change. Nor does an evaluation of interpretation by itself show an effect on clinical decisions or patient outcomes. Those are different questions and are not the stated primary endpoint.
The registry lists zero adverse-event counts, but also reports zero participants at risk in the adverse-event group. Those entries do not provide a basis for concluding that adverse events were absent among an adequately characterized population. Safety interpretation requires clear reporting of the population assessed and the observation process, which are not established by these entries.
In short, the record identifies the intended evaluation and its outcomes, but the numerical findings and key analytic details are not reported. The most defensible reading is that the trial was registered as completed and was designed to assess interpretation performance and agreement; the posted registry information does not quantify either result or demonstrate a clinical effect.
12. Why This Trial Matters Statistically
This trial is statistically interesting because its outcome is based on repeated interpretation tasks rather than a simple comparison of patient outcomes between treatment groups. An accuracy percentage can appear straightforward, yet its meaning depends on how the reference assessment is defined, how many evaluations are included, and whether assessments are clustered within clinicians, patients, or recordings. Those features determine both the estimate and the uncertainty around it.
The distinction between accuracy and agreement is also important. Comparing a clinician’s rating with a neurophysiologist’s assessment can support a correctness measure, while a kappa coefficient describes agreement under a chance-adjusted framework. The measures answer related but non-identical questions. Reporting both can help distinguish matching assessments from agreement relative to a reference, provided the underlying data and calculation details are available.
Domain-specific reporting may also matter. A single overall percentage could conceal variation among symmetry, sedation level, seizure activity, and artefact assessments. Conversely, separate domain results raise questions about multiplicity and how uncertainty should be interpreted across several outcomes. The registry does not report the analyses or results needed to examine these issues, making the record a useful illustration of why clear outcome definitions and analysis reporting matter.
Finally, the study’s before-and-after framing makes design details central. Without reported allocation, a comparison group, or an analysis plan, changes cannot be attributed to training with confidence. The registry’s caution that the arms are not a like-for-like control comparison further discourages a treatment-effect interpretation. Statistical interpretation depends not only on a result, but also on whether the design supports the claim being made.
13. 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 main limitation for interpreting this trial record is the absence of numerical results for the primary and secondary outcomes. The registry does not report the percentage of correct evaluations, domain-specific estimates, the kappa coefficient, or uncertainty measures. As a result, performance and agreement cannot be quantified from the record.
Important details about the measurement process are also not reported. The record does not specify the correctness rule, the number of evaluations, how evaluations were linked to recordings or clinicians, or how the neurophysiologist reference was established. It does not describe assessor masking or explain how disagreements were resolved. These details affect what an accuracy percentage would represent and how much confidence to place in it.
The structure of the observations is unclear. The number of clinicians, recordings, and assessments is not reported, nor is it clear whether clinicians assessed repeated tracings or whether patients contributed multiple recordings. If observations were clustered, the chosen analysis would need to account for that dependence to estimate uncertainty appropriately. The registry does not report whether this issue was addressed.
The record does not provide a statistical analysis plan, sample-size justification, missing-data procedures, analysis populations, or a multiplicity strategy. It also does not establish a randomized, like-for-like control comparison. These omissions limit causal interpretation of the before-and-after framing and make it difficult to assess the intended scope of inference. The adverse-event entries list zero counts with zero at-risk counts, so they do not support a substantive safety assessment.
14. Related Tutorials
15. Related Statistical Calculators
16. Sources
- ClinicalTrials.gov: Continuous Quantified EEG in NeuroIntensive Care — NCT02901262.
- Registry record last updated 2024-03-22; status last verified 2023-09.
- Results first posted 2024-03-22.
- 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.
17. Record Summary
The registry record describes a completed study of continuous and quantitative EEG interpretation in neuro-intensive care. It lists enrollment of 30, a start date of 2016-01, a primary completion date of 2017-01, and Azienda Ospedaliera San Gerardo di Monza as sponsor.
The stated purpose is to assess staff interpretation of symmetry, sedation level, seizure activity, and artefact before and after in-house training. The primary outcome is the percentage of correct evaluations by neuro-intensivists after training compared with neurophysiologist evaluations, with a timeframe of 2 months. Agreement after training is listed as a secondary outcome.
The registry does not report numerical primary or secondary results, confidence intervals, or analysis details. It lists 0 analyses and notes that allocation is not stated and the arms are not a like-for-like control comparison. The record establishes the study question and completion status, but does not provide enough outcome information to judge observed accuracy, agreement, or precision.