Minimally Invasive Surgery and rtPA for Intracerebral Hemorrhage Evacuation: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Minimally Invasive Surgery and rtPA for Intracerebral Hemorrhage Evacuation (NCT00224770), a randomized phase 2 study. The primary outcome is Safety Outcome Number 1: Rate of Mortality.

2. Clinical Question

3. Trial Design

The registry identifies NCT00224770 as a phase 2, randomized, parallel-group trial with no masking. Its listed condition is intracerebral hemorrhage, and enrollment is 141. The record gives a start date of 2005-08, a primary completion date of 2012-08, and a completed status. These dates and the completed status describe the registry entry; they do not establish when each analysis was conducted or whether the record was updated later.

The listed management groups are Medical Management, MISTIE Surgical Management, and ICES Surgical Management. Medical Management is described as standard-of-care medical management according to American Heart Association guidelines. MISTIE combines minimally invasive surgery and clot lysis with rt-PA. ICES uses intraoperative stereotactic CT-guided endoscopic surgery for mechanical removal, with best medical care and no rt-PA, according to the arm description. The record’s brief summary says the ICES cohort is analyzed separately and describes MISTIE intention-to-treat subjects as the hypothesized test group.

The primary outcome is the rate of mortality, measured as the percentage of participants who died during the first 30 days after randomization. The displayed outcome groups contain 42 participants in Medical Management, 54 in MISTIE Surgical Management, and 14 in ICES Surgical Management. The registry also lists ordinal Modified Rankin Scale outcomes at Day 180 and Day 365, clot-size reduction by the end of treatment, and postoperative clot-size reduction among secondary outcomes.

The registry record does not report the original sample-size justification, allocation ratio, details of sequence generation, or the statistical analysis plan. It identifies the trial as randomized but does not describe those implementation details in the information recorded for this analysis.

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: Each study center was required to demonstrate proficiency in the technical aspects of enrollment, stabilization, surgery, and drug administration. This proficiency was demonstrated on at least one pilot patient prior to randomization of the first patient in the investigational cohort of 110 randomized patients.

Overall Study

Reasons for not completing this period:

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: The decision was made to exclude the pilot patients from the analysis...

6. Primary Endpoint

The registry lists the primary outcome as Safety Outcome Number 1: Rate of Mortality, measured over 30 days from randomization.

The primary endpoint is the percentage of participants who died during the first 30 days after randomization. It is a binary outcome: each participant either died within that specified period or did not. The registry labels it a safety outcome and reports the result as a percentage of participants, with a 90% confidence interval for each management group.

Mortality is used to summarize an important safety event over a defined follow-up period. A fixed period makes the outcome interpretable as a proportion of each group, rather than as a time-to-event estimate. The record’s definition anchors the window to randomization, so the outcome concerns deaths in the first 30 days after that point.

The endpoint does not, by itself, describe the timing of deaths within the window, their causes, or later mortality. Those details are not reported in the registry record for this primary outcome. Nor does this endpoint measure functional recovery or the degree of clot reduction; the registry lists functional and clot-size outcomes separately as secondary outcomes. The mortality percentages therefore address a specific safety question and should not be treated as a complete account of all benefits or harms.

7. Results Posted in the Registry

• MISTIE Surgical Management: 14.8% (8/54) versus 9.5% (4/42) on control; risk difference 5.3 percentage points (95% CI -7.7 to 18.3), risk ratio 1.56 (0.50–4.82); unadjusted two-proportion z-test p = 0.437.
• ICES Surgical Management: 7.1% (1/14) versus 9.5% (4/42) on control; risk difference -2.4 percentage points (95% CI -18.5 to 13.8), risk ratio 0.75 (0.09–6.16); unadjusted two-proportion z-test p = 0.787.

8. Statistical Interpretation

9. Statistical Methods Explained

For a binary endpoint such as 30-day mortality, the observed proportion is the number of participants with the event divided by the number in the group. The registry reports the resulting group percentages and 90% confidence intervals. A confidence interval describes a range of values reasonably compatible with the observed data under the method used. It does not mean that a particular percentage of participants had outcomes within the interval, and it does not guarantee that the interval contains the true value.

What does a risk difference describe?

A risk difference compares two event proportions by subtracting one group’s proportion from the other’s. It is an absolute contrast, rather than a relative measure. A value of zero corresponds to equal proportions. A positive value indicates a larger proportion in the group named first in the contrast as expressed by the calculation; a negative value indicates a smaller proportion. The recalculated contrasts are stated as proportions: 0.053 for MISTIE versus Medical Management and -0.024 for ICES versus Medical Management. The posted group percentages and these contrasts should be read together: the contrasts express the size and direction of the observed separation, while the confidence intervals show uncertainty around it.

Why use Fisher Exact?

The registry reports Fisher Exact analyses for the comparisons. In general, Fisher’s exact test evaluates evidence against equal event proportions using the observed counts in the comparison groups. It can be useful when event counts are limited. The registry records p-values of 0.324 for the Medical Management and MISTIE comparison and 0.633 for Medical Management and ICES. A p-value describes how unusual data at least as incompatible with the null hypothesis would be if that hypothesis and the test assumptions held. It is not the probability that the null hypothesis is true, and it does not measure the probability that a treatment caused the observed result.

Why can recalculations differ from the registry analysis?

A recalculation from posted summary statistics is an unadjusted calculation. It uses the available group-level counts and does not incorporate participant-level covariates or other adjustments that might be part of a fuller analysis. The registry’s own analysis is the sponsor-reported result and takes precedence when the two differ. In this record, the discrepancy flag is false, but the reported analyses and recalculated comparisons still have different p-values and interval information. The registry record does not explain the reason for those differences or provide enough detail to reconstruct every analysis choice. The sponsor’s analysis should therefore be distinguished from the simple recalculation rather than treated as interchangeable with it.

What does over-dispersion mean in general?

Over-dispersion is a general statistical issue that arises when observed variability is greater than a model assumes. In that situation, a model that treats outcomes as less variable can produce confidence intervals that are too narrow or tests that overstate precision. The mortality outcome is reported as a binary percentage, and the registry information does not identify an over-dispersion model or report an assessment of over-dispersion. This general concept should not be read as a claim that over-dispersion was present in this trial.

How should the intervals and p-values be combined?

Intervals and p-values answer related but distinct questions. An interval communicates a range of effect sizes compatible with the data under a specified procedure. A p-value summarizes evidence against a null value under a test. Neither replaces the observed percentages, and neither alone indicates whether a difference is important. Here, the recalculated intervals extend to values on both sides of no difference for both comparisons. The reported p-values likewise do not provide evidence that clearly distinguishes the groups, while the broad intervals leave substantial uncertainty about the magnitude and direction of any underlying difference.

The registry record does not report whether there were interim analyses, the original statistical plan, the handling of missing outcome data, or a full description of analysis populations for each outcome. Those omissions limit how completely the reported analyses can be evaluated. The methods visible in the record support a careful descriptive reading of the posted estimates, not assumptions about unreported procedures.

10. Confidence Intervals and Precision

The group-level mortality estimates are accompanied by 90% confidence intervals. Medical Management has an estimate of 9.5% and an interval from 3.3% to 20.5%. MISTIE Surgical Management has an estimate of 14.8% and an interval from 7.6% to 25.1%. ICES Surgical Management has an estimate of 7.1% and an interval from 0.4% to 29.7%. These ranges are wide relative to their estimates, and the ICES interval is especially broad. The group sizes—42, 54, and 14—provide context for that uncertainty.

The unadjusted comparison intervals also span values below and above no difference: -0.077 to 0.183 for MISTIE versus Medical Management, and -0.185 to 0.138 for ICES versus Medical Management. Thus, the data do not locate the underlying contrasts precisely. The estimates point in different directions, but the uncertainty ranges allow a range of possible effects in either direction.

In general, confidence intervals tend to be wider when event counts or sample sizes are limited, because the data contain less information about the underlying proportion. This is a general explanation of precision, not a claim about the trial’s design calculations. A wide interval should not be interpreted as evidence of no effect; it indicates that the estimate is not precise enough to rule out materially different possibilities. The registry does not report the original sample-size justification or the confidence-interval method for every posted analysis, so the precision assessment should remain tied to the reported estimates and intervals.

11. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The sample-size figures are planning illustrations based on the observed effect, not the trial’s original design calculation. For this binary outcome, the figures are participants per arm, not total events. The planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation; the effect to detect is the stated fraction of the observed effect. The trial’s own original design assumptions and sample-size justification are not reported in the registry record.

The 80% power column describes the number of participants per arm that the planning calculation estimates would be needed to have an 80% chance of detecting the specified effect under those assumptions, if that effect were the true one. The 90% power column uses a higher target chance of detecting it and therefore generally calls for more participants. These are planning quantities, not a guarantee that a study with that size will produce a statistically significant result.

For MISTIE compared with Medical Management, the planning illustration gives 599 participants per arm for 80% power and 801 for 90% power when planning for 100% of the observed effect. For 75% of the observed effect, it gives 1,014 and 1,357; for 50%, it gives 2,168 and 2,902. For ICES compared with Medical Management, the corresponding figures are 2,115 and 2,830 for 100% of the observed effect, 3,881 and 5,196 for 75%, and 9,006 and 12,056 for 50%.

Smaller effects require more participants to distinguish them from random variation under the same planning assumptions. These figures are not observed power; they are prospective illustrations derived from the observed contrasts and stated assumptions. They should not be read as evidence that the completed trial had a particular probability of finding an effect.

12. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.095; treatment proportion p₁ = 0.148; average p̄ = 0.122.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.4624 = 0.9062.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.4608 = 0.3878.
4. n = (0.9062 + 0.3878)² / (0.0529)² = 598.2, rounded up to 599 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
p0 <- 0.0952380952380952                          # control proportion
p1 <- 0.148148148148148                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(224770)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 599 (80% power), 801 (90% power)
 75% of observed effect: n per arm = 1,014 (80% power), 1,357 (90% power)
 50% of observed effect: n per arm = 2,168 (80% power), 2,902 (90% power)
power.prop.test: 598.2 
Simulated power at n = 599 per arm: 0.800 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

13. Analyses Reported by the Sponsor

The registry’s sponsor-reported comparisons use Fisher Exact methods. For Medical Management versus MISTIE, the posted risk difference is 5.3, the 95% confidence interval has an upper bound of 16.2, and the p-value is 0.324. For Medical Management versus ICES, the posted risk difference is -2.4, the 95% confidence interval has an upper bound of 11.2, and the p-value is 0.633. The record does not post lower bounds for these two intervals.

The simple unadjusted recalculations from the posted summary counts give risk differences of 0.053 for MISTIE versus Medical Management and -0.024 for ICES versus Medical Management, with p-values of 0.437 and 0.787, respectively. These values are not a replacement for the sponsor’s own analysis. The sponsor-reported result takes precedence when interpreting the trial’s posted analysis; recalculations are a separate, summary-statistic-based check. The registry does not explain why the analysis outputs differ in their p-values or interval presentation, nor does it provide enough detail to reconstruct every sponsor analysis choice.

Both sets of results leave uncertainty about the direction and magnitude of the group contrasts. The posted p-values do not establish equivalence, and the incomplete interval bounds in the sponsor-reported results prevent a full reading of those intervals from the registry entry alone.

14. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary mortality outcome and four secondary outcomes: ordinal Modified Rankin Scale at Day 180, ordinal Modified Rankin Scale at Day 365, clot-size reduction by the end of treatment, and postoperative clot-size reduction. It records nine outcome measures and 17 analyses. With multiple outcomes or comparisons, the chance of obtaining at least one apparently notable result by chance can increase unless the analysis plan addresses multiplicity. This is a general statistical concern; the registry record does not report whether or how multiplicity was addressed.

The record does not report interim analyses or a statistical plan specifying a testing hierarchy, adjustment method, or prespecified handling of multiple comparisons. It also does not report whether outcome analyses were adjusted for baseline characteristics. Those details should not be inferred from the existence of the reported p-values or from the trial’s randomized designation.

The brief summary describes MISTIE intention-to-treat subjects as the hypothesized test group and says the ICES cohort is to be analyzed separately. The outcome record lists separate group sizes for Medical Management, MISTIE, and ICES, but does not provide a complete account of the analysis population for each posted endpoint. It does not report the numbers of participants with missing 30-day mortality data, how missing data were handled, or whether all randomized participants were included in each comparison. The registry record also does not describe exclusions or deviations from assigned management for the primary outcome.

These omissions matter because analysis populations, missingness, and multiplicity can affect how results should be interpreted. They do not demonstrate that a particular problem occurred. They mean that the available record does not support a detailed assessment of those aspects of the trial’s analysis. Accordingly, the mortality comparisons are best read as the reported estimates and tests, with the scope and uncertainty described, rather than as a complete audit of the trial’s statistical conduct.

15. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

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: respiratory failure (11.9%), intracranial hypertension (9.5%), ischemia cerebrovascular (9.5%), pneumonitits (9.3%), intracranial hemorrhage: tissue, enlargement (7.4%).

The registry reports serious adverse events in 23 of 42 participants in Medical Management, 28 of 54 in MISTIE Surgical Management, and 6 of 14 in ICES Surgical Management. It reports other adverse events in 24 of 42, 41 of 54, and 8 of 14 participants, respectively. These are group counts of participants affected, not event totals, and they should be kept distinct from the primary outcome of death within 30 days.

The listed leading serious adverse-event terms include Respiratory failure at 11.9%, Intracranial hypertension at 9.5%, Ischemia cerebrovascular at 9.5%, Pneumonitits at 9.3%, Intracranial hemorrhage: Tissue, Enlargement at 7.4%, and Cardiac arrest at 7.1%. The registry lists these percentages but does not provide their denominators or explain whether a participant could contribute to more than one term in the information summarized for these terms.

These figures describe adverse-event reporting in the registry and do not, by themselves, establish causation or permit a complete comparison of event severity, duration, or attribution. The record’s adverse-event death counts are not reported in the registry record. The event listings should therefore be interpreted alongside, but not substituted for, the separately defined 30-day mortality endpoint.

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

The posted results describe 30-day mortality proportions in three management groups and show that the estimates are uncertain. The observed rates are 9.5% for Medical Management, 14.8% for MISTIE Surgical Management, and 7.1% for ICES Surgical Management. The recalculated comparisons are compatible with effects in either direction, and the sponsor-reported p-values do not establish a clear difference. The results therefore do not provide a precise estimate of how the underlying mortality proportions compare.

That interpretation is not the same as concluding that the approaches have equal safety. Failure to demonstrate a clear difference is not proof of equivalence or absence of an effect. The intervals are broad, and the ICES group includes 14 participants for this endpoint. In general, limited information can leave potentially important differences unresolved; that is a statement about uncertainty, not a claim that any particular difference exists.

The outcome is limited to death during the first 30 days after randomization. It does not answer questions about mortality after that period, functional outcomes, or clot reduction. Although the registry lists secondary outcomes involving Modified Rankin Scale at Day 180 and Day 365 and clot-size reduction, their results are not included in the mortality details. The 30-day endpoint also cannot identify causes of death from the information recorded.

The registry’s brief summary frames the trial’s purpose as assessing safety and preliminary effectiveness of minimally invasive surgery with rt-PA and says the ICES cohort is analyzed separately. The mortality data alone are not a complete evaluation of safety or effectiveness, and they do not support medical advice. Their statistical meaning is narrower: observed group percentages, uncertain comparisons, and limitations in the analysis details recorded in the registry.

17. Why This Trial Matters Statistically

This trial illustrates why reporting an event percentage without its uncertainty can invite overconfident interpretation. The group estimates give a clear description of what was observed, but the confidence intervals and group sizes show how much uncertainty remains. The difference between a point estimate and a well-supported conclusion is especially important when interpreting safety outcomes with limited event information.

It also shows why absolute contrasts, group rates, and test results should be read together. The risk differences convey the direction and size of the observed separations, while the intervals communicate that the underlying contrasts are not precisely determined. P-values alone would not reveal the range of plausible effects or establish that management groups are equivalent.

The separate ICES cohort is a further interpretive consideration. The registry summary says it is to be analyzed separately, and its posted outcome group is smaller than the other listed groups. That context helps explain why its estimate is particularly uncertain, without implying that the cohort’s result is uninformative or that its outcome differs from the other groups.

Finally, the record demonstrates the value of distinguishing sponsor-reported analyses from simple recalculations, and of identifying what the registry does not describe. The lack of reported detail about original design assumptions, interim analyses, missing data, and multiplicity does not prove a flaw. It marks the boundary of what can be concluded from the posted information. For biostatisticians and clinical researchers, that boundary is part of the statistical interpretation.

18. 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.
• Participant counts were reconstructed from posted percentages and arm sizes, so small rounding differences are possible.

The posted group sizes for the primary outcome are 42 for Medical Management, 54 for MISTIE Surgical Management, and 14 for ICES Surgical Management. The small ICES group is reflected in its wide 90% confidence interval, from 0.4% to 29.7%, and makes its mortality estimate imprecise. The uncertainty is not confined to ICES: the intervals for Medical Management and MISTIE also span a broad range of percentages.

The registry’s primary endpoint is restricted to mortality during the first 30 days after randomization. The record does not provide cause-specific mortality, later mortality results, or a detailed description of how deaths were ascertained. The adverse-event information supplies group counts and selected event terms, but not enough detail in this record to assess attribution, duration, recurrence, or overlap among reported event categories.

The trial includes three listed management groups, while the brief summary says the ICES cohort is to be analyzed separately and identifies MISTIE intention-to-treat subjects as the hypothesized test group. The registry does not provide a full account of the population included in every analysis or the handling of missing outcome data. The analysis plan, original sample-size justification, allocation procedures, and interim-analysis information are also not reported in the registry record.

The sponsor-reported comparison intervals have upper bounds but no posted lower bounds. The registry also does not explain the differences between the sponsor-reported p-values and the unadjusted recalculations. These limits mean the displayed results support descriptive and cautious comparative interpretation, but not a complete reconstruction of the trial’s statistical methods or a definitive conclusion about comparative safety.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Minimally Invasive Surgery and rtPA for Intracerebral Hemorrhage Evacuation — NCT00224770.
• Registry record last updated 2015-06-18; status last verified 2015-06.
• Results first posted 2015-06-18.
• 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.

22. Record Summary

NCT00224770 is a completed phase 2 randomized parallel-group trial with no masking, listed for intracerebral hemorrhage. Enrollment is recorded as 141. The registry lists Medical Management, MISTIE Surgical Management, and ICES Surgical Management. Its primary safety outcome is the percentage of participants who died during the first 30 days after randomization.

Posted mortality rates are 9.5% for Medical Management, 14.8% for MISTIE, and 7.1% for ICES. The recalculated unadjusted contrasts have broad intervals that include values on either side of no difference. The registry reports Fisher Exact comparisons with p-values of 0.324 and 0.633; those analyses do not establish a clear group difference. The ICES estimate is particularly imprecise given its outcome group size of 14.

The record lists secondary functional and clot-size outcomes, along with adverse-event information. It does not report the original sample-size justification, interim analyses, full missing-data handling, or a complete statistical analysis plan.