Assessing the Incidence of Postoperative Delirium Following Aortic Valve Replacement: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Assessing the Incidence of Postoperative Delirium Following Aortic Valve Replacement (NCT03950440), a registered phase not stated study. The primary outcome is Number of Participants With Postoperative Delirium.

2. Clinical Question

The study’s stated aim is to identify the incidence of postoperative delirium in the first five postoperative days among patients undergoing different aortic valve replacement techniques. Its primary outcome is a participant count, rather than a reported comparison of time to delirium or a modeled risk contrast. The registry record describes the primary analysis population as propensity weighted, but does not report the resulting outcome counts or an incidence estimate.

The study also planned to examine delirium onset, duration, and severity, along with adverse events, activities of daily living, and lengths of stay. These outcomes broaden the clinical questions beyond whether delirium was identified: they concern when it occurred, how long it lasted, and how patients fared during later follow-up. The record does not report results for those outcomes.

3. Trial Design

NCT03950440, titled “Assessing the Incidence of Postoperative Delirium Following Aortic Valve Replacement,” is recorded as an observational study sponsored by Universitaire Ziekenhuizen KU Leuven. The study began on 2018-09-23, had a primary completion date of 2020-01-20, and is marked completed. Enrollment is listed as 250. These registry entries describe the study record, not an independently verified account of recruitment or conduct.

The study concerns patients undergoing different aortic valve replacement techniques. The recorded groups are TAVI Patients (n=84), Mini-sternotomy (n=82), and Full-sternotomy (n=84). The record explicitly characterizes this as not a randomized comparison; allocation is not stated, and the arms are not a like-for-like control comparison. It therefore does not support interpreting raw differences between groups as effects caused by the procedure.

The primary outcome is the number of participants with postoperative delirium. The registry describes delirium assessment using the 3D-CAM, CAM-ICU, and nurse reports, combined with interviews with caregivers and the patient. The stated timeframe is through the fifth postoperative day in hospital, described as an average of 5 days. If delirium occurred, follow-up continued until resolution or discharge, whichever came first, up to 19 days.

Planned secondary outcomes include onset, duration, and severity of delirium; adverse events assessed using VARC-2 composite endpoints at 30 days and 6 months; instrumental activities of daily living at the 6-month interview; and intensive care unit and hospital length of stay. Six-month telephone follow-up for activities of daily living is described in the study summary. Results for these outcomes 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.

Before assignment: Between September 2018 and January 2020, 250 patients were included: 84 patients in the TAVI group and 166 in the SAVR group. During the screening, 43 patients were non eligible (age \<70 years, additional valve surgery, already participated or were intubated) and 41 patients were excluded (did not speak Dutch, baseline delirium, declined to participated).

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Propensity weighted analysis

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.

Population: Baseline characteristics without propensity weighting.

6. Primary Endpoint

The registry lists the primary outcome as Number of Participants With Postoperative Delirium, measured over Until the fifth postoperative day in hospital (an average of 5 days). In case of POD, followup was continued until resolution of POD or discharge (whichever comes first) up to 19 days..

The primary endpoint counts participants identified as having postoperative delirium. It is a binary event outcome at the participant level: a person either meets the study’s delirium identification approach during the specified follow-up or does not. The record names the 3D-CAM, CAM-ICU, and nurse reports combined with interviews with caregivers and the patient as the assessment sources.

The planned observation window is through the fifth postoperative day in hospital, described as an average of 5 days. For participants with delirium, follow-up was to continue until resolution or discharge, whichever came first, up to 19 days. This extension means that the stated endpoint window is not simply the same fixed duration for every participant who develops delirium.

Counting affected participants is useful for describing how often an event was identified in the study population. It can also support comparisons if group definitions, follow-up, ascertainment, and analysis are clearly specified. In general, however, a count alone does not convey the timing, duration, or severity of each event. The registry lists those as separate secondary outcomes.

The record describes the primary population as a propensity weighted analysis, but does not report its delirium count, an incidence estimate, uncertainty around an estimate, or details of the weighting procedure. The primary endpoint result is therefore not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry describes a participant-count endpoint and a propensity weighted analysis population. It does not report a full statistical analysis plan, the primary outcome counts, or an effect estimate. The statistical concepts below are general explanations of how such methods are commonly interpreted; they are not claims about unreported procedures in this study.

What does a participant count represent?

A count records how many participants met an event definition during the stated observation period. For this study, the event is postoperative delirium assessed through the named assessment methods. Counts can be informative, but they depend on both the event definition and the opportunity to observe it. Differences in follow-up length, assessment frequency, or ascertainment can influence whether an event is identified. The record describes a usual assessment period through the fifth postoperative day and extended follow-up for cases of delirium, but it does not report further operational details such as the assessment schedule or rules for resolving conflicting assessment sources.

A count is not by itself an incidence rate or a comparative effect estimate. An incidence estimate requires a clear denominator and a definition of the period at risk; a rate also requires person-time. A group comparison may use a difference, ratio, or another measure, but whichever measure is selected should be stated with its uncertainty. The primary outcome results needed to interpret such quantities are not reported in the registry record.

What does propensity weighting do?

In general, propensity methods model the probability of receiving a particular exposure or treatment given measured characteristics. Weighting can then create an analysis population in which measured characteristics are more balanced across groups, depending on the method and its assumptions. This is not the same as random assignment. It cannot automatically balance factors that were not measured or were measured poorly, and extreme weights can make estimates unstable. Good interpretation requires knowing the variables included, the weighting target, diagnostics for balance, and the variance method used to calculate uncertainty.

The registry labels the primary population as propensity weighted but does not report those methodological details. It also does not state the allocation process. As a result, the degree to which weighting addressed differences among groups cannot be assessed from the registry record. Even a well-conducted weighting analysis would require assumptions about measured and unmeasured confounding for a causal interpretation.

Why does uncertainty matter?

In general, a confidence interval describes a range of estimates compatible with the data and the method under repeated sampling assumptions. Its width reflects statistical precision: wider intervals indicate less precision, while narrower intervals indicate greater precision. A p-value, when used, assesses how compatible the data are with a specified null hypothesis under the model; it is not a measure of clinical importance or proof that a causal effect exists. Neither a confidence interval nor a p-value for the primary delirium outcome is reported in the registry record.

For observational group comparisons, uncertainty calculations also depend on the analysis method. Weighting can change the effective contribution of participants, so a variance procedure should account for the weighting structure. The registry record does not report the variance estimator or the model used for the primary endpoint. It is not possible to infer those choices from the label “propensity weighted.”

What is the role of planning calculations?

In general, sample-size planning asks how many participants may be needed to detect a chosen effect under stated assumptions. A planning illustration is not an observed result and does not show the probability that a completed study detected an effect. The planning assumptions described for the illustration are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation, with the effect to detect expressed as a fraction of the observed effect. The illustration is not the trial’s own design. Its calculated sample-size values are not reported in the registry record, and the record does not report the trial’s sample-size justification.

Those assumptions should not be confused with the design of this observational study. The registry records three groups and states that allocation was not stated. A two-arm equal-allocation planning illustration does not turn those groups into a randomized comparison, nor does it provide evidence about the primary result. The record also does not report an interim analysis plan or the statistical plan used for the completed study.

Why can event counts be difficult to compare?

In general, raw event counts combine the number of participants observed with how frequently events occur. Counts should be read with their corresponding denominators, follow-up definitions, and ascertainment methods. A larger count does not necessarily mean a greater underlying risk when group sizes or observation differ. For adverse events, standardized definitions and consistent follow-up are also important. The safety figures recorded for this study are descriptive, and the information available does not include a comparative analysis that adjusts for the observational design.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("TAVI Patients (n=84)", "Mini-sternotomy (n=82)", "Full-sternotomy (n=84)")
n      <- c(84, 82, 84)
events <- c(13, 41, 45)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  TAVI Patients (n=84)  84  13  15.5  8.5  25.0  9.3  24.7  9.3  24.7
ARM  Mini-sternotomy (n=82)  82  41  50.0  38.7  61.3  39.4  60.6  39.4  60.6
ARM  Full-sternotomy (n=84)  84  45  53.6  42.4  64.5  43.0  63.8  43.0  63.9

11. Multiplicity, Missing Data and Analysis Populations

The study lists multiple outcomes: the primary delirium count, delirium onset, duration and severity, adverse-event composite endpoints at 30 days and 6 months, instrumental activities of daily living at the 6-month interview, and intensive care unit and hospital length of stay. When several outcomes or comparisons are tested, the chance of at least one apparently unusual result can rise. In general, multiplicity procedures can control this issue, or a study can distinguish a primary confirmatory outcome from exploratory outcomes. The registry record does not report a multiplicity strategy or a statistical analysis plan explaining how the listed outcomes were handled.

The record gives a primary endpoint and identifies secondary outcomes, but does not report whether secondary analyses were prespecified in detail, whether subgroup comparisons were planned, or how multiple comparisons were interpreted. The number of outcome measures recorded is 10. That count alone does not show how many statistical tests were conducted or whether any adjustment was applied.

Missing-data methods are not reported in the registry record. The description of planned follow-up does not establish whether all participants completed each assessment, how incomplete delirium assessments were classified, or how missing follow-up data were handled. In general, missingness can affect an estimate if the likelihood of missing data is related to participants’ characteristics or outcomes. A complete-case analysis, imputation, or other approach would need to be described before its assumptions could be assessed.

The registry describes the primary analysis population as propensity weighted, but does not specify which participants were included in that population or how exclusions and missing covariates were addressed. It does not report the analysis population for each secondary outcome. Interim analyses and stopping rules are also not reported in the registry record. These omissions limit assessment of how the planned outcomes were analyzed and how robust any estimates might be.

12. 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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: valve dysfunction (15.5%), nyha 3/44 (13.1%), rehospitalization (11.9%), all stroke (10.7%).

The registry’s adverse-event record lists serious and other adverse-event counts by group. Serious adverse events affected 36 of 84 TAVI patients, 21 of 82 mini-sternotomy patients, and 17 of 84 full-sternotomy patients. Other adverse events affected 28 TAVI patients, 18 mini-sternotomy patients, and 18 full-sternotomy patients. The listed deaths were 8, 2, and 1, respectively. These are raw counts, not adjusted estimates or tests of differences.

The serious adverse-event terms listed include Valve dysfunction at 15.5%, NYHA 3/44 at 13.1%, Rehospitalization at 11.9%, and All stroke at 10.7%. The registry record does not provide further detail in the information available about the denominators or definitions underlying these percentages, so they should be read as listed rather than reinterpreted.

In general, adverse-event counts need context about definitions, follow-up, and the number of participants assessed. The group sizes differ, and the study is observational rather than a randomized comparison; allocation is not stated, and the arms are not a like-for-like control comparison. The counts therefore describe recorded events but do not establish that a technique caused a difference in safety. A statistical comparison and its uncertainty are not reported in the registry record.

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

The registry record establishes that this completed observational study set out to identify postoperative delirium after different aortic valve replacement techniques. It specifies the delirium assessment sources, the intended observation period, and additional outcomes relating to delirium, adverse events, function, and length of stay. These details help define the questions the study was designed to address.

However, the primary outcome counts and incidence estimate are not reported in the registry record. There is therefore no reported numerical result with which to describe the observed frequency of delirium, its precision, or differences among groups. The record also does not report a primary-outcome confidence interval or p-value. No conclusion about the direction or magnitude of a group difference can be drawn from the primary outcome information recorded.

The adverse-event counts are descriptive. They do not, on their own, show that one technique caused more or fewer events. Because the study is observational and the groups are not described as like-for-like controls, participant characteristics and other differences could contribute to raw contrasts. The propensity weighted analysis label does not resolve that interpretive issue without details of the weighting method, covariates, diagnostics, and uncertainty calculations.

In general, an observational study can provide useful information about event frequency and clinical follow-up in the population it observes. Its findings require careful interpretation, particularly when comparing groups that were not randomized. Here, the record does not include the primary results or enough analysis detail to evaluate the strength of any comparative claims.

The registry’s planned outcomes should also be distinguished from observed results. Delirium onset, duration and severity; VARC-2 adverse-event composites; activities of daily living at 6 months; and lengths of stay are listed as outcomes, but their results are not reported in the registry record. The most defensible interpretation is therefore about the study’s aims and recorded design, not about whether delirium or other outcomes were more or less common with any particular technique.

14. Why This Trial Matters Statistically

This study raises a clinically relevant statistical problem: how to describe postoperative delirium when patients receive different aortic valve replacement techniques and are assessed through multiple information sources. The endpoint depends on identifying delirium, so the assessment approach and follow-up window are integral to interpreting any event count. The registry specifies use of the 3D-CAM, CAM-ICU, and nurse reports combined with interviews, and describes follow-up through the fifth postoperative day with continued follow-up for cases until resolution or discharge, up to 19 days.

The design also illustrates why observational comparisons require careful framing. The groups are not randomized, allocation is not stated, and the registry explicitly says the arms are not a like-for-like control comparison. Statistical adjustment, including propensity weighting, can address some measured differences under particular assumptions, but it does not make an observational study equivalent to a randomized trial. Clear reporting of the target population, variables used, balance diagnostics, and uncertainty is important for judging what adjusted comparisons mean.

Finally, the registry makes the distinction between a study question and an interpretable result especially important. It records the primary endpoint and several secondary outcomes, but the primary delirium counts and results for the secondary outcomes are not reported in the registry record. Without those results, the study can be understood as a planned assessment of delirium and follow-up outcomes, but its estimates and precision cannot be assessed. This is a reminder that transparent reporting of outcome data and analysis methods is essential for statistical interpretation.

15. 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.

A central limitation of the record is that it does not report the primary delirium count, an incidence estimate, or uncertainty around that estimate. The same lack of results applies to the listed secondary outcomes. Consequently, the registry record does not allow an assessment of the size or precision of the study’s primary finding.

The comparison is observational, allocation is not stated, and the groups are explicitly not a like-for-like control comparison. Patient characteristics and clinical decisions that influenced the choice of technique could differ across groups. Although the primary population is described as propensity weighted, the registry record does not report the variables included, the weighting method, balance diagnostics, or the approach to uncertainty. The extent to which measured differences were addressed cannot be judged, and unmeasured confounding remains a general concern for observational comparisons.

The delirium assessment combines named screening approaches with nurse reports and interviews. The record does not report additional details about assessment frequency, assessor training, adjudication, or how disagreements among information sources were resolved. Follow-up is described through the fifth postoperative day, with longer follow-up for cases up to 19 days, but further detail on completeness of assessment and missing observations is not reported in the registry record.

The record does not report a statistical analysis plan, missing-data methods, multiplicity handling, interim analyses, or a sample-size justification. It also does not provide enough information to evaluate the denominators and definitions behind the listed serious adverse-event term percentages. These gaps limit interpretation of both the primary endpoint and the descriptive safety figures.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Assessing the Incidence of Postoperative Delirium Following Aortic Valve Replacement — NCT03950440.
• Registry record last updated 2023-02-16; status last verified 2022-05.
• Results first posted 2023-02-16.
• 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.

19. Record Summary

NCT03950440 is a completed observational study sponsored by Universitaire Ziekenhuizen KU Leuven, with enrollment listed as 250. Its stated aim was to identify postoperative delirium after different aortic valve replacement techniques. The primary endpoint was the number of participants with postoperative delirium, assessed using the 3D-CAM, CAM-ICU, and nurse reports combined with interviews.

The registry describes assessment through the fifth postoperative day and continued follow-up for delirium until resolution or discharge, up to 19 days. Recorded groups are TAVI, mini-sternotomy, and full-sternotomy. The comparison is not randomized, allocation is not stated, and the arms are not a like-for-like control comparison. The primary analysis population is described as propensity weighted.

Secondary outcomes include delirium characteristics, adverse-event composites, activities of daily living at 6 months, and lengths of stay. The registry record does not report the primary delirium results, the secondary outcome results, or the details needed to evaluate the weighting and statistical analysis.