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 The Thai Surgical Intensive Care Study (Thai-SICU Study) (NCT01354197), a registered phase not stated study. The primary outcome is Overall Mortality.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT01354197 |
| Phase | Phase not stated |
| Condition | Complications of Surgical and Medical Care: General Terms, Adverse Reaction to Drug, Sequelae of External Causes of Morbidity and Mortality |
| Allocation | Not Stated |
| Design model | Not Stated |
| Masking | Not Stated |
| Enrollment | 4652 |
| Arms in the primary outcome | n/a |
| Lead sponsor | Royal College of Anesthesiologists of Thailand |
| Results posted | Yes |
2. Clinical Question
Population
Patients admitted to surgical intensive care units at participating Thai university hospitals during the cohort period. The registry describes nine participating university-based surgical intensive care units and lists an enrollment of 4652. Further eligibility criteria and a more detailed definition of the population are not reported in the registry record.
Intervention
There was no assigned intervention. The study observed patients admitted to surgical intensive care units as part of a cohort. The registry describes the arm as all intensive care unit admissions during the cohort time, rather than as a treatment group.
Comparator
No randomized or otherwise assigned comparator is described. The survivor and non-survivor groups in the adverse-event information are defined by survival at day 28, so they are outcome groups, not like-for-like control arms.
Primary question
What number of patients did not survive up to 28 days after surgical intensive care unit admission, and what adverse events occurred in these units? The registry lists overall mortality as the primary outcome and intensive care unit readmission as a secondary outcome. It does not report the numerical result for the primary outcome.
The study is descriptive rather than a test of whether one care strategy is better than another. Its central clinical-statistical question concerns outcomes after admission in the participating units. Because there is no assigned intervention or suitable control group, the registry information cannot isolate the effect of a particular treatment or compare a treatment with usual care. The survivor categories can support descriptions of how adverse-event counts were recorded across outcome groups, but they cannot establish that an event caused survival status or that survival status caused an event.
The record names complications of surgical and medical care, adverse drug reactions, sequelae of external causes of morbidity and mortality, and critical illness among the conditions of interest. These broad terms help describe the study context, but do not establish detailed diagnostic criteria or the clinical reasons for each admission. The registry's description therefore frames a broad observational question about mortality and adverse events in surgical intensive care, rather than a narrowly defined question about a single disease or intervention.
3. Trial Design
The Thai-SICU Study is registered as a completed, multicenter observational study. Its official title describes a study of outcomes and adverse events in surgical intensive care units at Thai university hospitals. The record identifies the Royal College of Anesthesiologists of Thailand as sponsor. The listed study start is 2011-04, and the listed primary completion is 2012-11.
The brief study description states that the study observed 28 days mortality and the occurrence of adverse events in nine participating university-based surgical intensive care units. The registry lists enrollment of 4652. It describes a single arm consisting of all intensive care unit admissions to surgical intensive care during the cohort time. No treatment assignment, randomization, masking, or comparator arm is described. Allocation is listed as not stated; the record also cautions that the arms are not a like-for-like control comparison.
The primary outcome is overall mortality, expressed as the number of participants who did not survive up to 28 days after surgical intensive care unit admission. Readmission to intensive care is listed as a secondary outcome. The registry indicates two outcome measures, but does not provide a numerical result for the primary mortality measure or a numerical readmission result in the information recorded for this study.
The adverse-event figures are organized by whether patients survived at day 28. This is a descriptive grouping based on an observed outcome, not a randomized design feature. The registry presents serious and other adverse-event counts for the survivor and non-survivor groups. Details of how events were identified, whether they were counted once per person or repeatedly, and the exact observation rules are not reported in the registry record.
The record does not describe an intended sample-size calculation, a recruitment target rationale, a comparison hypothesis, or a prespecified statistical analysis plan. These omissions matter when interpreting how the study was designed and what inferences its results can support. In particular, the observational structure supports description of the participating units and recorded outcomes, but does not by itself support causal conclusions about care practices or interventions.
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 | All ICU Admission Patients |
|---|---|
| STARTED | 4652 |
| COMPLETED | 4652 |
| NOT COMPLETED | 0 |
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 | Survive at Day 28 | Non-survive at Day 28 | Total |
|---|---|---|---|---|
| Participants analysed | 4010 | 642 | 4652 | |
| Age, Continuous | years | 61.4 (SD 17.3) | 63.7 (SD 17.4) | 61.7 (SD 17.3) |
| Sex: Female, Male: Female | Participants | 1692 | 231 | 1923 |
| Sex: Female, Male: Male | Participants | 2318 | 411 | 2729 |
| Race/Ethnicity, Customized: Thai | participants | 4010 | 642 | 4652 |
| Region of Enrollment: Thailand | participants | 4010 | 642 | 4652 |
6. Primary Endpoint
The registry lists the primary outcome as Overall Mortality, measured over 28 days.
| Element | Registry entry |
|---|---|
| Result type | Number |
| Measure of spread | not applicable |
| Unit | participants |
| Endpoint classification | Binary Endpoint |
The primary endpoint is overall mortality, measured as the number of participants who did not survive up to 28 days after admission to a surgical intensive care unit. The timeframe begins with the admission described in the registry. The endpoint is a binary outcome at the participant level: a patient either survived through the stated follow-up period or did not. The registry's measure is expressed as a count of participants, not as a time-to-event measure or as a continuous clinical measurement.
Mortality is used in clinical research because it is an important outcome that can be defined without relying on a subjective judgment of symptom improvement. In general, a fixed follow-up period also makes the timing of outcome assessment explicit. These general properties do not resolve whether follow-up was complete, how deaths were ascertained, or whether the registry's operational definition included particular circumstances; those details are not reported in the registry record.
The record identifies the measure and its timeframe but does not give the observed number of deaths or a corresponding proportion. It also leaves the population field for the primary outcome blank. Therefore, the endpoint can be described conceptually, but the registry information does not establish the numerical mortality result or the exact analysis population used for that outcome.
Readmission to intensive care is listed as a secondary outcome, but its measurement definition and result are not reported in the registry record. The primary mortality endpoint should not be conflated with the adverse-event counts, which are presented separately by day-28 survival status.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The registry establishes the study's purpose and broad design: observe 28-day mortality and adverse events among patients admitted to surgical intensive care units in Thai university hospitals. It records enrollment of 4652 and describes participation by nine university-based units. These registry entries identify the scope of the cohort, but they do not provide a numerical estimate of the primary mortality outcome. Although overall mortality is named as the primary measure, the count of participants who did not survive through 28 days is not reported in the record.
The adverse-event information is more numerical, but requires careful reading. For patients who survived at day 28, the registry lists 4010 at risk for serious adverse events and 1070 affected; for other adverse events it lists 4010 at risk and 253 affected. For patients who did not survive at day 28, it lists 642 at risk for serious adverse events and 519 affected; for other adverse events it lists 642 at risk and 90 affected. These are counts, not adjusted treatment effects or causal comparisons. The record does not explain whether the affected counts represent people with at least one event, event occurrences, or another counting convention.
The two outcome groups were defined by survival status. That makes the grouping useful for describing the recorded adverse-event information in relation to eventual status, but it does not make the groups comparable as if they had been assigned different treatments. Differences in illness severity, timing, duration of exposure to intensive care, and other factors could affect both event recording and survival; the registry does not provide analyses that address such factors. No inference that an event caused death, or that avoiding an event would change survival, follows from these counts alone.
The listed percentages for serious adverse-event terms are Sepsis at 46.7%, Acute kidney injury at 46.0%, Cardiac arrest at 32.6%, Acute lung injury/ Acute respiratory distress syndrome at 18.5%, New arrhythmia at 16.0%, and Upper gastrointestinal hemorrhage at 3.4%. The denominator and calculation basis for these term percentages are not reported in the registry record. They should therefore be read as reported descriptive percentages, not assumed to represent risk among all admissions or a comparison between groups.
No inferential analyses are listed. The record does not report confidence intervals, p-values, adjusted estimates, or a statistical model. Consequently, the precision of the primary mortality result and the uncertainty around the adverse-event figures cannot be assessed from the registry information. The figures describe recorded counts; they do not establish whether observed differences are statistically distinguishable from chance or clinically attributable to a particular exposure.
The study's observational setting can provide a record of outcomes in the participating units, but generalization beyond those settings requires information about how the units and patients were selected and how representative they were. Those details are not reported. The appropriate interpretation is therefore descriptive and bounded by the participating units, the registry's definitions, and the information it reports.
9. Statistical Methods Explained
The registry describes an observational cohort rather than a randomized trial. In a randomized comparison, assignment to an intervention is used to make groups comparable, on average, with respect to both measured and unmeasured factors. This study instead describes admissions to surgical intensive care and organizes some adverse-event information by whether patients survived at day 28. Because survival status is an outcome, not an assigned exposure, the groups cannot be interpreted as treatment and control arms. This distinction is central to interpreting any apparent contrast between their recorded event counts.
What does a binary outcome mean?
A binary outcome has two possible states for each participant. For the primary endpoint, those states are survival or non-survival through the specified 28 days after admission. A count of participants who experienced the event is a natural summary for such an endpoint. A count alone, however, does not communicate uncertainty or provide a comparison unless the relevant denominator, analysis population, and comparison are also defined. The record names the endpoint and gives its timeframe, but does not report its observed count or the population used for the primary analysis.
The adverse-event counts include both an at-risk count and an affected count for each outcome group and event category. In general, a count of affected participants can be interpreted as an incidence proportion only when the unit counted, the eligible population, and the observation period are clear. Here the registry does not state whether an affected count means people with at least one event, the total number of event episodes, or another convention. Nor does it specify whether participants could contribute to multiple event terms. The counts should therefore be presented as recorded, without converting them into newly calculated rates.
What would uncertainty measures add?
A confidence interval is a general statistical tool for expressing uncertainty around an estimate. It describes a range of values that are compatible with the data and a specified method, under the assumptions of that method. A narrow interval generally indicates greater precision than a wide interval, but an interval does not remove bias from outcome misclassification, missing follow-up, confounding, or an unrepresentative sample. The registry information does not include confidence intervals for the primary outcome or the adverse-event figures, so their statistical precision cannot be evaluated from this record.
A p-value is a measure tied to a statistical model and a stated null hypothesis. It is not the probability that the null hypothesis is true, and it does not measure the size or clinical importance of an effect. The registry lists zero analyses and does not report p-values or hypotheses tested. There is therefore no basis in the record for describing a statistically significant result or for treating the absence of a reported test as evidence of no difference.
How should group contrasts be read?
Descriptive counts can be useful for understanding what was recorded in a clinical setting. A contrast between groups becomes more difficult to interpret when group membership depends on an outcome that occurs after admission. The survivor and non-survivor groups may differ in baseline condition, illness course, opportunity to experience or record an adverse event, and length of time under observation. The registry does not report adjusted analyses addressing these issues. A simple difference in counts or proportions, even if calculated elsewhere, would not by itself estimate a causal effect.
For clustered or multicenter data, observations from patients treated in the same unit may be more alike than observations from different units. General statistical methods can account for such dependence when the design and data support it. Ignoring dependence can make uncertainty estimates too optimistic. The record identifies nine participating units but does not report whether analyses accounted for clustering, differences between hospitals, or unit-level variation. No particular modeling approach should be inferred.
Over-dispersion is a general issue that can arise when outcome variation exceeds what a statistical model assumes. If a model assumes less variation than the data contain, standard errors may be too small and intervals too narrow. Whether over-dispersion is relevant depends on the outcome type and analysis. Since no analysis model is reported, it is not possible to determine whether this issue arose or how it was handled in this study.
What does the planning illustration mean?
The sample-size calculations associated with the posted material are planning illustrations based on an observed effect, not the study's own design justification. The stated assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; the target effect is a stated fraction of the observed effect. The record does not report the actual numerical effect used for these calculations or the resulting planning counts in the information available for interpretation. These calculations do not describe observed power and should not be mistaken for evidence that the observational cohort had a particular probability of detecting an effect.
More generally, a sample-size calculation depends on assumptions about the effect to detect, variability or event frequency, significance level, allocation, and analysis method. Changing those assumptions changes the planning result. A calculation based on an observed effect is especially dependent on that estimate and does not replace a prespecified rationale. The registry does not report a protocol-based sample-size justification, so the planning illustration cannot be attributed to the study's original design.
What can be concluded from the methods record?
The methods information supports a description of a multicenter observational cohort, its named primary endpoint, and the way adverse-event figures are grouped. It does not establish a complete analysis strategy. Definitions of the analysis population, handling of missing data, adjustment for confounding, clustering methods, multiplicity procedures, and uncertainty estimation are not reported in the registry record. Statistical interpretation must remain limited to the definitions and counts explicitly recorded.
10. Multiplicity, Missing Data and Analysis Populations
The registry lists two outcome measures: overall mortality as the primary outcome and readmission to intensive care as a secondary outcome. It does not provide an analysis plan stating how these outcomes were prioritized statistically, whether multiple hypotheses were tested, or whether any adjustment for multiplicity was planned. Multiplicity refers generally to the increased chance of at least one false-positive finding when multiple statistical tests are conducted. Whether an adjustment is appropriate depends on the study questions and analysis plan; no such plan is reported.
The registry lists zero analyses. It does not report a numerical primary mortality result, a readmission result, confidence intervals, or p-values. The absence of reported analysis results means that statistical significance and precision cannot be evaluated from the registry record. It should not be interpreted as evidence that an outcome had no association or that no analysis occurred outside the information recorded.
Missing-data methods are not reported. The record does not say whether follow-up was complete, how unavailable outcome information was handled, or whether any patients were excluded from an analysis. The primary outcome population field is blank, so the exact analysis population for mortality is not specified. The enrollment figure describes the study, but does not establish the denominator for every endpoint.
The registry also does not report whether the analysis used all eligible admissions, how repeated admissions were handled, or whether clustering by unit was considered. These details affect interpretation of both estimates and uncertainty. Without them, it is not possible to determine whether analyses followed an intention-to-treat-like principle, a complete-case approach, or another population definition. No such approach should be assumed for this observational study.
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 |
|---|---|---|---|
| Survive at Day 28 | 1070 | 4010 | 26.7% |
| Non-survive at Day 28 | 519 | 642 | 80.8% |
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 |
|---|---|
| Survive at Day 28 | 253 of 4010 |
| Non-survive at Day 28 | 90 of 642 |
Most frequent serious adverse event terms
| Event term | Organ system | Survive at Day 28 | Non-survive at Day 28 |
|---|---|---|---|
| Sepsis | Infections and infestations | 607 of 4010 | 300 of 642 |
| Acute kidney injury | Renal and urinary disorders | 491 of 4010 | 295 of 642 |
| New arrhythmia | Cardiac disorders | 184 of 4010 | 103 of 642 |
| Acute lung injury/ Acute respiratory distress syndrome | Respiratory, thoracic and mediastinal disorders | 119 of 4010 | 119 of 642 |
| Cardiac arrest | Cardiac disorders | 17 of 4010 | 209 of 642 |
| Intra-abdominal hypertension | Gastrointestinal disorders | 54 of 4010 | 17 of 642 |
| Myocardial infarction | Cardiac disorders | 47 of 4010 | 19 of 642 |
| Upper gastrointestinal hemorrhage | Gastrointestinal disorders | 33 of 4010 | 22 of 642 |
| Seizure | Nervous system disorders | 26 of 4010 | 19 of 642 |
| New stroke | Nervous system disorders | 9 of 4010 | 2 of 642 |
Most frequent non-serious adverse event terms
| Event term | Organ system | Survive at Day 28 | Non-survive at Day 28 |
|---|---|---|---|
| Delirium | Nervous system disorders | 117 of 4010 | 46 of 642 |
| Reintubation within 72 hours | Respiratory, thoracic and mediastinal disorders | 112 of 4010 | 29 of 642 |
| Unplanned extubation | Respiratory, thoracic and mediastinal disorders | 41 of 4010 | 16 of 642 |
| Pneumothorax | Respiratory, thoracic and mediastinal disorders | 14 of 4010 | 11 of 642 |
| Symptomatic deep venous thrombosis | Vascular disorders | 10 of 4010 | 4 of 642 |
| Self reported medication error | Surgical and medical procedures | 7 of 4010 | 3 of 642 |
| Pulmonary aspiration | Respiratory, thoracic and mediastinal disorders | 2 of 4010 | 1 of 642 |
Serious events with the highest arm-level rate: sepsis (46.7%), acute kidney injury (46.0%), cardiac arrest (32.6%), acute lung injury/ acute respiratory distress syndrome (18.5%), new arrhythmia (16.0%).
The registry gives serious and other adverse-event counts separately for patients who survived and did not survive at day 28. Among survivors, the serious-event figures are 4010 at risk and 1070 affected, while the other-event figures are 4010 at risk and 253 affected. Among non-survivors, the corresponding serious-event figures are 642 at risk and 519 affected, and the other-event figures are 642 at risk and 90 affected. These should be read as the counts recorded for the outcome-defined groups, not as a comparison of assigned treatments.
The registry also lists serious adverse-event terms with percentages: Sepsis, 46.7%; Acute kidney injury, 46.0%; Cardiac arrest, 32.6%; Acute lung injury/ Acute respiratory distress syndrome, 18.5%; New arrhythmia, 16.0%; and Upper gastrointestinal hemorrhage, 3.4%. The basis for these percentages, including their denominator and whether terms can overlap, is not reported in the registry record. They cannot be assumed to represent the proportion of all enrolled patients with each event.
In general, adverse-event interpretation depends on clear definitions, ascertainment methods, timing, and whether events are counted per person or per occurrence. Those details are not reported. Since the groups are based on survival status, these figures do not show that an event caused non-survival or that an event was more likely under a particular treatment. They describe recorded safety information, with important uncertainty about its exact measurement basis.
12. What the Result Does, and Does Not, Mean
The study describes mortality and adverse events in patients admitted to surgical intensive care units at participating Thai university hospitals. Its primary endpoint is the number of participants who did not survive up to 28 days after admission. The registry also records adverse-event counts by survival status and lists intensive care unit readmission as a secondary outcome. These are clinically relevant outcomes for describing the experience of patients in the participating settings.
The record does not report a numerical result for the primary mortality endpoint or for readmission. It therefore cannot support a numerical statement about the observed 28-day mortality outcome, nor can it establish the precision of such an estimate. The adverse-event counts and listed percentages provide descriptive information, but their counting definitions and percentage denominators are not reported. They should not be read as complete estimates of risk across all admissions.
This is not a randomized treatment comparison. No intervention or like-for-like control group is described, and the adverse-event groups are defined by survival at day 28. Consequently, the record does not show that a treatment, practice, or adverse event caused a particular outcome. It also does not establish that any care strategy works or does not work. Associations, if calculated from more complete data, would still require careful handling of confounding, timing, and differences among patients and units.
The record's description is most informative as a statement of study scope, outcome definitions, and the categories of adverse events recorded. It identifies a multicenter cohort with enrollment of 4652 and nine participating university-based surgical intensive care units. It does not establish representativeness beyond those units, completeness of follow-up, or consistency of outcome ascertainment across centers.
These limits do not make the descriptive information meaningless. They define what it can support: a contextual account of outcomes recorded in the study setting, rather than a causal or comparative treatment conclusion. Stronger interpretation would require the outcome results, clearer definitions and denominators, and a documented analysis plan. Such information is not reported in the registry record.
13. Why This Trial Matters Statistically
Multicenter observational studies can describe outcomes in real clinical settings where patients are not assigned to treatments for research purposes. The Thai-SICU Study's focus on surgical intensive care brings mortality and adverse events into the same broad clinical context. Its record identifies nine participating university-based units, a 28-day mortality question, and adverse-event categories, including serious events. This makes the study relevant as a description of the questions being tracked in these settings.
Statistically, the study also illustrates why outcome definitions and group construction matter. A mortality endpoint defined over a stated period is conceptually clear, but a count without a reported analysis population or numerical result does not provide a complete estimate. Adverse-event counts grouped by eventual survival can inform description, yet they do not form a causal comparison. The same numbers can be easy to overinterpret if denominators, event-counting rules, and the timing of events are not clear.
The registry record also demonstrates the value of distinguishing descriptive reporting from inferential analysis. It lists enrollment and event information, but records zero analyses and does not report uncertainty estimates or an analysis plan. Without those details, readers cannot judge statistical precision, account for differences among units, or determine how missing data were handled. That distinction is important in clinical research: a substantial observational cohort does not automatically produce a precise or unbiased estimate for every endpoint.
The study therefore matters statistically as an example of both the potential and the limits of registry-based outcome reporting. It identifies clinically important outcomes and a multicenter setting, while leaving key analytic details unreported. Careful reading preserves the value of the descriptive record without claiming more than it establishes.
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.
Several limitations specific to this study follow from the information recorded. The primary endpoint is defined as the number of participants who did not survive up to 28 days after admission, but the numerical mortality result is not reported. The primary outcome population field is blank, and the record does not state how deaths were ascertained or how complete follow-up was. These omissions prevent assessment of the endpoint estimate and its precision.
The adverse-event information lacks a reported operational definition. The record does not explain how serious and other events were identified, whether counts refer to people or event episodes, how recurrent events were treated, or whether patients could appear under multiple terms. It also does not state the denominators underlying the listed serious-event percentages. As a result, the percentages cannot safely be interpreted as risks for the full cohort or for a defined subgroup.
The survivor and non-survivor categories are based on the outcome of interest. They are not intervention and control groups, and the record does not report adjusted analyses to account for patient characteristics, illness severity, exposure time, or differences among units. The study's observational design therefore limits causal interpretation. The registry identifies nine participating university-based units, but does not report how those centers were selected or whether they represent other settings.
Other important design and analysis details are not reported in the registry record, including eligibility criteria, handling of repeated admissions, missing-data methods, clustering adjustments, and a prespecified statistical plan. The secondary outcome of intensive care unit readmission is named but lacks a reported result and detailed definition. These limitations should be considered alongside the completed status and enrollment information, which alone do not resolve the uncertainty about measurement or analysis.
15. Related Tutorials
16. Related Statistical Calculators
17. Sources
- ClinicalTrials.gov: The Thai Surgical Intensive Care Study (Thai-SICU Study) — NCT01354197.
- Registry record last updated 2017-03-27; status last verified 2017-02.
- Results first posted 2017-03-27.
- 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
The Thai Surgical Intensive Care Study, also titled the Multicenter Study of Outcome and Adverse Events in Surgical Intensive Care Unit of Thai University Hospital, is registered as completed. The Royal College of Anesthesiologists of Thailand is listed as sponsor. The study start is 2011-04 and primary completion is 2012-11. Enrollment is listed as 4652.
The study observed 28 days mortality and adverse events in nine participating university-based surgical intensive care units. The primary outcome is overall mortality, defined as the number of participants who did not survive up to 28 days after admission. Intensive care unit readmission is listed as a secondary outcome. The record also provides serious and other adverse-event counts grouped by survival status, and lists serious-event terms with percentages.
The record does not report the numerical primary mortality result, the readmission result, or any analyses. It does not report key details about analysis populations, missing data, event ascertainment, or statistical methods. The available information supports a bounded descriptive account, not a causal treatment comparison.