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 Effects of Biofeedback in Patients With Acute Cerebral Infarction (NCT03516227), a randomized na study. The primary outcome is Changes From Baseline Psychological Distress (Anxiety and Depression) at 1 and 3 Months.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT03516227 |
| Phase | NA |
| Condition | Acute Ischemic Stroke, Biofeedback, Autonomic Dysfunction, Cognitive Function |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Single |
| Enrollment | 40 |
| Arms in the primary outcome | n/a |
| Lead sponsor | National Defense Medical Center, Taiwan |
| Results posted | Yes |
2. Clinical Question
Population
Patients with acute cerebral infarction, described in the registry summary as patients with acute ischemic stroke. The record reports 40 enrolled and says 35 participants completed all aspects of the study: 19 in the intervention group and 16 in the control group.
Intervention
Heart rate variability biofeedback combined with abdominal breathing training. Participants in the experimental group received four training sessions.
Comparator
Routine care, also described as usual care, in the control group.
Primary question
How did psychological distress scores for anxiety and depression change from baseline to the 1-month and 3-month visits in the intervention and control groups?
The registered question concerns changes in anxiety and depression measured with the Hospital Anxiety and Depression Scale (HADS). The endpoint description says participants rated their own distress. Anxiety and depression are represented by separate seven-item subscales, with higher scores indicating a worse outcome. The registry record does not give the actual group scores or an estimated between-group contrast at either follow-up.
This is a randomized, controlled comparison, but the available record does not specify the allocation ratio or explain the randomization procedure. It also does not identify who was masked in the single-blind design. Those details matter when assessing how the groups were formed and how expectations or assessment procedures might affect responses. The registry summary describes additional repeated assessments of heart rate variability, cognitive function, and activities of daily living, but the primary question is psychological distress over time. The record does not report numerical findings for that question, so it cannot establish the direction or precision of any difference between the groups.
3. Trial Design
The registry describes a randomized, controlled, single-blind trial with a parallel-group design. Participants were assigned to an experimental group receiving heart rate variability biofeedback or a control group receiving routine care. The experimental intervention combined biofeedback with abdominal breathing and consisted of four training sessions. The record does not describe the content or duration of routine care, the interval between training sessions, or whether other aspects of care differed between groups.
The study enrolled 40 participants and is listed as completed. The reported start date is 2015-11-09, and the primary completion date is 2016-06-26. The registry lists the phase as NA. The sponsor is National Defense Medical Center, Taiwan. The study concerns acute cerebral infarction, with related conditions in the record including acute ischemic stroke, autonomic dysfunction, cognitive function, and psychological distress.
Measurements were collected before the intervention and at 1-month and 3-month visits. The stated primary outcome was change from baseline in psychological distress, assessed through anxiety and depression scores on HADS. The study summary also names heart rate variability, the Mini-Mental Status Examination, and the Barthel Index for activities of daily living among the repeated measures. The record lists five outcome measures, but does not provide numerical outcome results or analysis records.
The registry identifies the study as single-blind but does not say whether participants, intervention staff, outcome assessors, or analysts were masked. It does not report the method used to generate or conceal the random allocation sequence, eligibility criteria in detail, or the planned sample-size justification. The record also does not describe protocol deviations or the reasons participants did not complete all aspects of the study. These omissions limit assessment of implementation and the analysis population, without changing the fact that the registered design was randomized and parallel-group.
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: Five patients did not receive the allocated group treatment protocol (two had arrhythmias after inclusion and three refused baseline measurements)
Overall Study
| Stage | HRVBF | Control |
|---|---|---|
| STARTED | 20 | 20 |
| COMPLETED | 19 | 16 |
| NOT COMPLETED | 1 | 4 |
Analysis population for the primary outcome, as defined in the registry: Thirty-five participants completed all aspects of the study (19 intervention, 16 control)
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 | HRVBF | Control | Total |
|---|---|---|---|---|
| Participants analysed | 19 | 16 | 35 | |
| Age, Continuous | years | 67.6 (SD 11.4) | 67.2 (SD 7.6) | 67.4 (SD 9.7) |
| Sex: Female, Male: Female | Participants | 9 | 8 | 17 |
| Sex: Female, Male: Male | Participants | 10 | 8 | 18 |
| Race and Ethnicity Not Collected | Participants | — | — | 0 |
| Region of Enrollment: Taiwan | participants | 19 | 16 | 35 |
6. Primary Endpoint
The registry lists the primary outcome as Changes From Baseline Psychological Distress (Anxiety and Depression) at 1 and 3 Months, measured over baseline, 1-month, and 3-month visits.
| Element | Registry entry |
|---|---|
| Result type | Mean |
| Measure of spread | Standard Deviation |
| Unit | score on a scale |
| Endpoint classification | Continuous Endpoint |
The primary endpoint concerns changes from baseline in psychological distress at the 1-month and 3-month visits. Distress was measured with the Hospital Anxiety and Depression Scale, a 14-item ordinal scale intended for participants to rate their own anxiety-related and depression-related distress. The scale has separate seven-item anxiety and depression components. Each component ranges from 0 to 21, with higher scores indicating a worse outcome. The record describes a score of 8 or higher out of 21 as the cut-off used in earlier studies to identify anxiety or depression.
Using separate anxiety and depression scores allows the two forms of distress to be considered individually rather than collapsed into a single undifferentiated measure. Repeated measurement makes it possible, in principle, to compare how scores change over follow-up in each group. A between-group comparison of change would address whether the pattern differs between those assigned to biofeedback and those assigned to routine care. The registry record does not say which specific contrast or statistical model was planned for that purpose.
HADS is a self-rating measure, so it captures participants’ reported symptoms rather than an independently observed physiological event. Its score is useful for quantifying distress, but a score alone does not establish a diagnosis or explain why a participant reports a particular level of distress. The record gives no group means, standard deviations, change estimates, or confidence intervals for this endpoint. Consequently, the magnitude, direction, and precision of the registered primary outcome cannot be assessed from the registry results.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The registry record identifies the primary outcome and the scheduled assessments, but it does not report numerical results for anxiety or depression at baseline, 1 month, or 3 months. It also lists no outcome analyses. There is therefore no posted estimate of how scores changed in either group, no reported between-group comparison, and no confidence interval or p-value for the primary question. The available information describes what the study set out to measure, not the result of that measurement.
The record states that 40 participants were enrolled and that 35 completed all aspects of the study, including 19 in the intervention group and 16 in the control group. It does not explain the status of other enrolled participants, whether their partial data were included, or whether completion differed by outcome or visit. The completion figures alone do not show the pattern of missing measurements or how missing data may have affected an estimate.
The registered primary outcome is based on a continuous score summarized as a mean with standard deviation. In general, a mean describes the group’s average score, while a standard deviation describes how much individual scores vary around that average. Neither measure by itself estimates the uncertainty in a treatment-group contrast. That assessment generally requires an estimate of the contrast and its uncertainty, such as a confidence interval. No such endpoint results are reported in the registry record.
The registry’s adverse-event fields report zero serious events, zero other events, and zero deaths for each group, while also recording zero participants at risk in those fields. Those entries should not be treated as evidence that no adverse events occurred among all enrolled participants: the record does not provide a usable participant denominator for these fields or explain how the entries were collected. The adverse-event entries also do not answer the primary psychological distress question.
A separate planning illustration is based on the observed effect and specifies a two-sided significance level of 0.05, equal allocation, and a normal approximation. Such a calculation concerns hypothetical future sample-size planning under stated assumptions; it is not a trial outcome or evidence of the certainty of a finding. Because numerical endpoint results are not reported, the clinical magnitude and statistical precision of the primary comparison remain unknown. No conclusion about whether scores improved, worsened, or differed between groups can be drawn from this registry record.
9. Statistical Methods Explained
The registry identifies the primary outcome as a continuous measure summarized by a mean and standard deviation. It does not report the statistical analysis used to compare the groups or the changes over time. The explanations below describe general statistical concepts relevant to this design; they should not be read as a description of methods confirmed to have been used in the trial.
What does a mean and standard deviation describe?
For a continuous score, the mean is the arithmetic average across participants with measurements. The standard deviation describes the spread of individual scores around that mean. Reporting both can help characterize a group’s score distribution, but they do not by themselves answer whether the groups differ. To interpret a treatment comparison, readers also need a defined contrast, such as a difference between groups at follow-up or a difference in change from baseline, together with an estimate of its uncertainty. The registry record reports the outcome’s summary format but does not publish the necessary numerical endpoint results.
Why does the repeated-measures design matter?
Participants were assessed before the intervention and at 1-month and 3-month visits. In general, repeated observations from the same participant are related: a participant’s later score may be associated with that participant’s earlier score. An analysis that recognizes this dependence can estimate changes over time and compare patterns between groups while accounting for within-participant correlation. Treating every measurement as if it came from a different person would ignore that relationship. The registry record does not identify the model or method used for repeated measurements, so it is not possible to assess whether this dependence was handled in the analysis.
The question can be framed in more than one way. A comparison at each follow-up asks whether group scores differ at that visit; a comparison of change asks whether the amount of change from baseline differs between groups. Adjusting for baseline, analyzing change scores, or estimating group-by-time patterns are distinct approaches and can answer related but not identical questions. The record does not specify which contrast was planned or report any estimates, so an interpretation should not assume a particular approach.
What does a confidence interval add?
In general, a confidence interval expresses the uncertainty around an estimated effect under the assumptions of the analysis. Its width reflects how precisely the effect has been estimated: a narrower interval indicates greater precision, while a wider interval indicates less precision. The interval also helps readers consider which effect sizes remain compatible with the data. A p-value alone, even if reported, would not convey the range of plausible effects or their clinical importance. No confidence interval or p-value for the primary endpoint is reported in the registry record.
Why can over-dispersion matter?
Over-dispersion is a general issue for outcomes where the observed variability is greater than a statistical model expects. When a model underestimates variability, its uncertainty estimates may be too small, making evidence appear more precise than it is. The primary outcome is described as a continuous score with standard deviation, not as a count outcome; the registry record does not report a model or any assessment of over-dispersion. This concept is therefore a general statistical consideration, not a finding about this trial.
What does the sample-size illustration represent?
The planning illustration uses a two-sided significance level of 0.05, equal allocation, and a normal approximation. It considers detecting a stated fraction of the observed effect. In general, sample-size calculations depend on the target effect, variability, significance threshold, desired power, allocation, and analysis assumptions. A planning estimate is conditional on those choices and is not a measure of the evidence produced by the completed trial. It is not observed power. The registry record does not report the trial’s own sample-size justification or enough numerical endpoint results to evaluate the observed effect and its precision.
More generally, randomized allocation aims to make comparison groups similar on average at the start of a study, but randomization does not guarantee balance in every characteristic, particularly in a limited sample. Baseline results and a transparent analysis would help readers assess the comparison. The registry record does not give baseline HADS scores or describe the statistical handling of baseline measurements. It also does not specify adjustment for other variables. These are unreported analysis details, not evidence that a particular method was or was not used.
10. Multiplicity, Missing Data and Analysis Populations
The registry lists five outcome measures and the primary endpoint includes anxiety and depression assessed over more than one follow-up visit. When several outcomes, time points, or contrasts are tested, the chance of at least one apparently notable result can increase. In general, a statistical analysis plan can address this multiplicity by identifying a primary comparison in advance, setting a hierarchy for secondary questions, or applying an adjustment. The registry record does not report a multiplicity strategy, a hierarchy of outcome analyses, or analysis results. It is not possible to determine how multiple comparisons were handled.
The record states that 40 participants were enrolled and 35 completed all aspects of the study: 19 in the intervention group and 16 in the control group. It does not report the reasons for non-completion, timing of withdrawals, whether participants missed particular visits, or how many HADS measurements were available at each time point. Completion of all study aspects is not necessarily the same as complete data for every individual outcome, and the record does not clarify this distinction.
Methods for missing data—such as excluding incomplete records, using available measurements, or applying an imputation approach—are not reported in the registry record. Nor does the record define whether the primary analysis included all randomized participants or only a subset. Without those details and participant-level outcome availability, the effect of missing data on the reported estimate cannot be assessed. The registry lists no analysis records for the outcomes, so there is no posted estimate whose handling of multiplicity or missingness can be evaluated.
11. What the Result Does, and Does Not, Mean
The registry record establishes that a randomized, controlled trial compared four sessions of heart rate variability biofeedback combined with abdominal breathing against routine care in patients with acute cerebral infarction. It identifies anxiety and depression scores on HADS as the primary outcome and specifies assessments at baseline, 1 month, and 3 months. It also gives enrollment and completion counts. These are design and conduct details; they do not show whether psychological distress changed or whether the groups differed.
No numerical primary outcome results or analyses are reported. It is therefore not possible to describe a benefit, lack of benefit, increase, decrease, or between-group difference based on this record. The direction and size of any estimate, and how precisely it was estimated, are unknown. A stated endpoint and a completed trial do not by themselves establish a result.
Even when numerical score results are available, statistical interpretation should separate the estimated effect from its uncertainty and from its practical meaning. A confidence interval can show the range of effects compatible with an analysis, while a p-value addresses a different statistical question. Neither alone determines whether a difference is important to participants. The record does not supply an estimate, interval, or p-value for the primary endpoint, so none of these aspects can be judged for this comparison.
The cut-off of 8 or higher on either HADS subscale is described as a threshold used in earlier studies to identify anxiety or depression. A group mean cannot establish how many individuals crossed that threshold, and a score at or above it is not, by itself, a diagnosis. The registry provides no counts of participants above the threshold. Similarly, the reported zero adverse-event entries should not be interpreted as proof of absence of harm, given the recorded zero at-risk denominators and the lack of an explanation for those entries.
The findings that can be stated are limited to the registered design, outcome definitions, follow-up schedule, and completion figures. The record does not permit conclusions about efficacy, clinical importance, safety, durability, or how the intervention compares with routine care on the primary outcome.
12. Why This Trial Matters Statistically
This trial raises several statistical issues that are common in behavioral and clinical research. The randomized parallel-group design creates a framework for comparing groups, while repeated measurements allow change over time to be studied. Interpreting such a design requires more than knowing the endpoint: the comparison must be clearly defined, the within-participant dependence addressed, and the uncertainty around the estimate reported. The registry record describes the outcome and visits but does not provide the analysis or numerical findings needed to evaluate those elements.
The endpoint also illustrates the distinction between a continuous score and a threshold-based interpretation. HADS scores can be summarized as means and standard deviations, while a specified cut-off can be used to identify participants whose scores meet a threshold. Those summaries answer different questions. Mean changes describe average score patterns; threshold counts describe how many participants fall on one side of a chosen boundary. The registry record does not report either type of result.
Completion is relevant to interpretation because outcome estimates may depend on which participants and visits contribute data. The record reports 40 enrolled and 35 completing all aspects, but does not describe missing measurements or the analysis population. A transparent account of participant flow, outcome availability, and missing-data handling would help readers assess whether the reported comparison represents the randomized groups.
Finally, numerical results are essential for separating a potentially meaningful effect from an imprecisely estimated one. A point estimate, confidence interval, and clear account of the primary contrast would allow readers to assess magnitude and uncertainty. The registry record lists no analyses and contains no numerical outcome results, so its contribution is chiefly to define the planned comparison and measurements rather than to quantify the observed effect.
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.
Several trial-specific limitations are apparent from the registry record. Although the study is identified as randomized and single-blind, the record does not describe the randomization method, allocation concealment, or which parties were masked. Those details are important for judging the potential for selection and measurement biases. The intervention is described as four biofeedback training sessions, but session content and timing are not detailed. The control group is described as receiving routine or usual care, without a fuller account of that care.
The registry reports 40 enrolled and 35 completing all aspects, with 19 intervention and 16 control participants among those completions. It does not explain reasons for non-completion or indicate how many measurements were available for each outcome and visit. The analysis population, handling of missing data, and use of all randomized participants are not reported. These gaps limit assessment of the relationship between enrollment, follow-up, and any potential outcome estimate.
The endpoint relies on participant self-ratings of psychological distress. Such scores provide a structured measure, but they may reflect subjective reporting and do not by themselves establish a diagnosis or a change in other aspects of health. The record does not report baseline or follow-up scores, between-group estimates, confidence intervals, or analysis methods. As a result, neither the size nor the precision of the primary comparison can be assessed.
The record lists five outcome measures and multiple assessment times but does not report how outcomes or comparisons were prioritized for analysis. It also does not provide an explanation of the zero adverse-event denominators. Finally, eligibility criteria and other details needed to characterize the study population are not reported in the registry record, which restricts assessment of how well the enrolled participants represent other patient groups.
14. Related Tutorials
15. Related Statistical Calculators
16. Sources
- ClinicalTrials.gov: Effects of Biofeedback in Patients With Acute Cerebral Infarction — NCT03516227.
- Registry record last updated 2019-10-11; status last verified 2019-09.
- Results first posted 2019-10-11.
- 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
NCT03516227 is a completed randomized, controlled, single-blind parallel-group trial sponsored by National Defense Medical Center, Taiwan. It enrolled 40 participants with acute cerebral infarction. The experimental group received four sessions of heart rate variability biofeedback combined with abdominal breathing; the control group received routine care. The reported study dates are 2015-11-09 for the start and 2016-06-26 for primary completion.
The primary outcome was change in anxiety and depression scores on HADS, assessed at baseline and at 1-month and 3-month visits. The registry also names repeated measures of heart rate variability, cognitive function, and activities of daily living. It reports that 35 participants completed all aspects, including 19 in the intervention group and 16 in the control group.
Numerical primary outcome results and analysis records are not reported in the registry record. The record does not specify the statistical model, analysis population, missing-data approach, or multiplicity strategy. It therefore describes the trial design and measures but does not support an assessment of the estimated primary effect or its precision.