Effects of Magnesium on Individuals Undergoing Coronary Artery Bypass Graft Surgery: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Effects of Magnesium on Individuals Undergoing Coronary Artery Bypass Graft Surgery (NCT00041392), a randomized phase 2 study. The primary outcome is Cognitive Function.

2. Clinical Question

3. Trial Design

The registry record describes a completed phase 2 trial, with randomized allocation, triple masking and a parallel-group model. The study was sponsored by Duke University. Its official title was PeriOperative Interventional Neuroprotection Trial (POINT), and its stated purpose was to examine the effects of supplemental magnesium on neurocognitive function among people undergoing CABG.

The record lists enrollment of 389. The intervention arms were magnesium, described as 100 mg/kg, and 0.9% saline, described as 100 mg/kg. The posted outcome results include 198 participants in the magnesium arm and 191 in the saline arm. These are the group counts attached to the posted outcome, not a substitute for information about every stage of participant flow.

The primary completion date is recorded as 2009-05, and the study start as 2001-12. The status is completed. The registry record does not report the details of how the random sequence was generated, how allocation was concealed, or which people were masked. “Triple” is the recorded masking category; the identities of the masked groups are not reported in the registry record.

The study used a parallel comparison: participants were assigned to one of the two intervention arms rather than being described as receiving both interventions in sequence. The posted primary outcome is a continuous cognitive change score measured between baseline and 6 weeks. Further details about treatment delivery, follow-up procedures, adherence, and the analysis population 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.

Overall Study

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.

6. Primary Endpoint

The registry lists the primary outcome as Cognitive Function, measured over Measured at baseline and 6 weeks.

The primary endpoint is cognitive function, expressed as a continuous change score from baseline to 6 weeks. The record describes a cognitive index built from four domains: verbal memory; abstraction and visuo-spatial orientation, described as executive function; visual memory; and attention and concentration. Factor analysis of baseline cognitive test scores was used to select a four-factor solution intended to characterize cognitive function while minimizing redundancy among measures.

The index at baseline was calculated as the mean of the four preoperative domain scores. Its scale was set to have a mean of zero and standard deviation of 0.5. On that scale, a positive score is above the mean, a negative score is below the mean, and 0.5 represents 1 standard deviation above the mean. The outcome change score is the 6-week index minus the baseline index. Thus, a positive change represents improvement on the index and a negative change represents decline.

A continuous score retains information about the magnitude and direction of change rather than reducing participants to categories. Comparing group averages can therefore describe the average separation between the randomized arms on this measure. It does not, by itself, describe the range of individual responses or establish how many participants improved or declined. Details about the cognitive tests, their scoring procedures beyond the stated index construction, and the handling of incomplete assessments are not reported in the registry record.

7. Results Posted in the Registry

• 0.9 % Saline: mean 0.12 versus 0.07 on control; difference 0.05 (95% CI -0.01 to 0.11), standardized effect d = 0.18; z-test p = 0.078.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is continuous, so its summary uses group means and standard deviations. A mean describes the average score in a group; a standard deviation describes how spread out individual scores are around that group average. These summaries answer different questions: the mean locates the group average, while the standard deviation gives information about variation among participants. The posted standard deviation is 0.28 in each arm.

What does the mean difference represent?

The outcome is calculated by subtracting the baseline cognitive index from the 6-week index. The comparison reports the mean difference as 0.05, with saline relative to magnesium. The two underlying group means are 0.07 for magnesium and 0.12 for saline. A positive difference in this stated direction means the saline group mean is higher on the change-score scale. Since positive change denotes improvement in the index, the direction of the estimate favors a higher average change in saline for this particular outcome. Direction alone does not determine whether the difference is important or reliable.

A group-level mean difference is not an individual treatment effect. It does not show that all participants changed in the same direction or that each participant in one group had a particular response. The standard deviations offer context about participant-level variability, but the registry record does not provide a distribution plot or counts of participants with positive, zero, or negative change.

How should the p-value be understood?

The unadjusted p-value is 0.078. In general, a p-value measures the compatibility of the observed data, or more extreme data, with a specified null model. It is not a measure of the probability that the null hypothesis is true. It is also not the probability that chance alone caused the results. Interpretation depends on the model and analysis choices, and those details beyond the posted unadjusted result are not reported in the registry record.

The record’s sample-size planning assumptions include a two-sided significance level of 0.05, but that value is stated for the planning illustration. It should not be taken as evidence that the trial’s own analysis plan used that threshold. The original analysis plan and its significance criteria are not reported in the registry record. The p-value should therefore be read as the posted unadjusted result, not as a complete account of the trial’s prespecified inferential approach.

Why does the interval matter?

The reported interval for the mean difference runs from -0.006 to 0.106. In general, a confidence interval describes uncertainty around an estimate under a statistical procedure and its assumptions. Its endpoints show a range of values compatible with the data under that procedure. The interval includes zero, which corresponds to no mean difference. It also includes values on either side of zero. The confidence level and method used to construct this interval are not reported in the registry record, so those details cannot be specified.

Precision is distinct from direction. An estimate can point in a particular direction while still having substantial uncertainty. The interval should be considered alongside the point estimate rather than replaced by a binary reading of the p-value. Neither the interval nor the p-value establishes clinical importance; the record does not report a clinically meaningful threshold for this index.

What does the randomized comparison contribute?

Random assignment is intended, in general, to make treatment groups comparable at the point of assignment with respect to measured and unmeasured characteristics. That design can support causal comparisons when allocation, follow-up, and analysis are conducted appropriately. However, the registry record does not report sequence-generation details, concealment procedures, participant flow, adherence, or the handling of missing outcomes. The posted results alone cannot establish how these aspects were managed.

The comparison is parallel-group, and the recorded masking category is triple. Masking can, in general, reduce the influence of expectations on care, assessment, or reporting, depending on who is masked and how. The roles covered by triple masking are not reported in the registry record, so its practical scope cannot be assessed from this information.

What do the planning calculations represent?

The sample-size illustrations use the observed mean difference as their reference effect and consider stated fractions of that effect. They are planning examples, not evidence about how many participants the original trial was designed to recruit. The record specifies a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation for these illustrations. It does not report the assumptions used for the trial’s original design or its sample-size justification.

These planning counts are not observed power. In general, power is a design-stage probability of detecting a specified effect under specified assumptions. Once a study has been conducted, calculating power from the observed effect does not add information beyond the observed estimate and its uncertainty. The illustrations are best interpreted as showing how the indicated participant requirements vary when the effect to detect is set to different fractions of the observed effect.

10. Confidence Intervals and Precision

The mean difference is 0.05, reported as 0.9% saline relative to magnesium. Its interval runs from -0.006 to 0.106. The lower endpoint is slightly below zero, while the upper endpoint is above zero. Thus the interval includes no difference and values in either direction. It also includes positive differences larger than the point estimate. The interval gives a sense of uncertainty that is not captured by the point estimate alone.

A confidence interval is generally interpreted as a range generated by a procedure that, under repeated use and its assumptions, has a stated coverage property. The confidence level and construction method are not reported in the registry record. It would therefore be inappropriate to label this interval with a particular confidence level. The record also does not describe whether the interval is adjusted for any additional analysis or multiplicity.

The unadjusted p-value is 0.078, but a p-value and an interval convey different aspects of the evidence. The p-value measures compatibility with a null model under a test; the interval displays a range of estimates consistent with the interval procedure. Neither turns the result into a definitive yes-or-no conclusion. The estimate’s direction and the interval’s width should be considered together.

Precision is limited to what is reported for this endpoint. The group standard deviations are 0.28 for magnesium and 0.28 for saline, while the posted group counts are 198 and 191. The registry record does not report the number with missing cognitive measurements, the confidence level, the interval method, or a precision target for the original design. Consequently, the interval can be described as posted, but its construction and coverage cannot be independently characterized from the registry record.

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 planning figures are participants per arm, not total events. They are illustrations based on the observed effect, not the trial’s own recruitment target or original design calculation. The registry record specifies the planning assumptions: a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The target effect is set to the stated fraction of the observed effect. The assumptions used for the trial’s original design are not reported in the registry record.

The 80% power column gives the illustrated number of participants per arm for a design with an 80% chance of detecting the selected effect under those assumptions. The 90% power column gives the corresponding number per arm for a 90% chance. These are planning concepts, not guarantees about results in a future study. The displayed figures use the observed effect as their starting point and also consider smaller fractions of it.

Smaller effects require more participants to detect under otherwise unchanged assumptions. The planning values therefore rise when the target is reduced from 100% of the observed effect to 75% or 50% of that effect. They should be read as conditional illustrations: changing the effect size or any of the planning assumptions can change the required participant count.

This is not observed power. The figures do not describe the probability that the completed trial successfully detected its true effect, and they do not replace the observed estimate, interval, or p-value. The trial’s own sample-size justification and original assumptions are not reported in the registry record.

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, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = 0.050; pooled standard deviation σ = 0.280.
2. n = 2 × (1.96 + 0.842)² × 0.280² / 0.050² = 492.3, rounded up to 493 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
delta <- 0.05                      # observed difference in means
sigma <- 0.28                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^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 (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(41392)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 493 (80% power), 660 (90% power)
 75% of observed effect: n per arm = 876 (80% power), 1,172 (90% power)
 50% of observed effect: n per arm = 1,970 (80% power), 2,637 (90% power)
power.t.test: 493.2 
Simulated power at n = 493 per arm: 0.800 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-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. Multiplicity, Missing Data and Analysis Populations

The registry record lists one outcome measure and no posted secondary outcomes. It also reports 0 analyses. The posted primary outcome therefore has an unadjusted mean comparison, but the record does not describe a full analysis plan or any procedures for handling multiple outcomes, repeated testing, or subgroup comparisons. The fact that one primary measure is listed does not establish what analyses were prespecified or performed outside the posted results.

Multiplicity arises when several statistical tests or outcomes are considered: the chance of at least one apparently unusual result can increase as the number of tests grows. In general, adjustment procedures can address this issue, but whether such procedures were planned or used in this trial is not reported in the registry record. The displayed p-value is explicitly unadjusted and should not be interpreted as multiplicity-adjusted evidence.

Missing data are another important uncertainty. The endpoint requires cognitive measurements at baseline and 6 weeks, but the registry record does not report how many participants had measurements at both times, how incomplete assessments were handled, or whether any imputation or sensitivity analyses were used. The outcome group counts alone do not establish the amount or pattern of missing data.

The analysis population is not described. The registry record does not say whether the comparison followed an intention-to-treat approach, included only participants with observed measurements, or used another population definition. It also does not report exclusions after randomization or reasons for them. These details matter because estimates can be affected when the analyzed group differs from the randomized group, especially if missingness or exclusion is related to outcomes or assigned treatment.

Accordingly, the posted unadjusted result can be described, but the record does not permit a complete assessment of multiplicity control, missing-data assumptions, or analysis-population choices. These are reporting gaps, not evidence that a particular approach was or was not used.

14. 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: pleural effusion (4.2%), bleeding (4.0%), respiratory failure (4.0%), stroke (3.1%), pacemaker placement (2.0%).

The safety record reports serious adverse events affecting 37 of 198 participants in the magnesium group and 32 of 191 participants in the 0.9% saline group. Other adverse events are reported for 137 of 198 participants in the magnesium group and 109 of 191 participants in the saline group. These are the posted counts and denominators; no additional comparison or rate calculation is needed to read them.

The serious-event terms listed most often in the record include pleural effusion at 4.2%, bleeding at 4.0%, respiratory failure at 4.0%, stroke at 3.1%, pacemaker placement at 2.0%, and hypotension at 2.0%. The record does not specify in this summary how these term percentages were calculated or provide arm-specific counts for each listed term. They should not be treated as comparative estimates between groups.

Deaths are not reported in the registry record; a missing death count should not be interpreted as zero. The available adverse-event summaries do not establish whether an event was caused by an intervention. Details about event definitions, timing, severity grading, and the method used to collect or adjudicate events are not reported in the registry record.

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

The posted results describe average cognitive change from baseline to 6 weeks in a randomized comparison of magnesium and 0.9% saline among people undergoing CABG. The magnesium mean change was 0.07, and the saline mean change was 0.12. The mean difference, stated as saline relative to magnesium, was 0.05. Since positive change indicates improvement on the cognitive index, the observed group averages were positive, and the saline average was higher on this score.

The estimate is uncertain. Its reported interval, -0.006 to 0.106, includes zero and extends to values in either direction. The unadjusted p-value is 0.078. These statistics do not establish a definitive difference or its absence. They describe the result and its uncertainty under the analyses represented by the posted figures. The confidence level and interval method are not reported in the registry record.

The result does not show the pattern of individual outcomes. Group means cannot tell whether most participants improved, whether changes varied substantially, or whether a smaller subset drove the average. The registry record does not report counts by direction of change or a clinical threshold for meaningful change. The index is a composite measure based on four cognitive domains, and its stated score construction gives a way to express average change rather than a direct account of every cognitive function.

The outcome is reported at 6 weeks; this does not establish longer-term cognitive outcomes. The posted record lists no secondary outcome results, and it does not report missing-data methods, analysis-population definitions, or an adjusted analysis. The findings should therefore be understood as the posted comparison for the specified endpoint, not as a complete account of all potential outcomes or analyses.

These statistical results are not medical advice and do not establish whether a treatment should be used. They do not prove that magnesium improves or worsens cognition, nor that saline is superior. They quantify the observed group comparison and the uncertainty reported for it, while leaving important clinical and analytical details unreported.

16. Why This Trial Matters Statistically

This trial illustrates how a clinical question can be expressed as a continuous outcome rather than a binary event. The cognitive index combines information from four domains, and its change score preserves direction and magnitude. A mean comparison can then describe the average shift between randomized groups. This is useful statistically, but the index’s scale and construction are essential to interpretation: positive and negative values have meanings defined by the registry record, and the mean difference is not a count of participants who improved.

The example also shows why a point estimate should not be read alone. The posted difference is 0.05, while its interval extends from -0.006 to 0.106 and the unadjusted p-value is 0.078. The estimate’s direction is apparent, but the interval includes zero and a range of possible magnitudes. Distinguishing effect size from precision helps prevent a single test result from being treated as a complete summary of evidence.

Another statistical issue is the separation between results and planning. The sample-size illustrations use the observed effect and stated assumptions, while the original trial’s design assumptions are not reported in the registry record. Such illustrations can show how required sample sizes depend on the target effect and desired power, but they cannot be used as evidence that the completed trial had a particular probability of finding an effect.

Finally, interpretation depends on reporting beyond the headline comparison. Missing outcomes, analysis-population definitions, multiplicity procedures, interval construction, and masking roles are not reported in the registry record. Identifying these gaps is part of statistical interpretation: it distinguishes what the posted figures directly support from what cannot be assessed. The trial is therefore a useful case for reading a continuous endpoint alongside its design, uncertainty, and reporting limits.

17. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The available registry information leaves several trial-specific questions unresolved. It does not report detailed eligibility criteria, the randomization procedure, allocation concealment, or the roles included in triple masking. These details would help assess how participants entered the trial and what protections against bias were in place. The record’s design labels alone do not answer those questions.

The primary endpoint is a derived cognitive index. The record explains the four domains and baseline construction, but it does not provide further psychometric information, the underlying test list, or a clinically meaningful threshold for the change score. Consequently, statistical differences on this index cannot be translated into a specific functional or patient-important change from the registry information alone.

Only the 6-week outcome is described in the posted primary result. The registry record does not report longer follow-up for this endpoint, participant-level distributions, or the number of participants with complete measurements at both time points. It also does not describe missing-data handling, post-randomization exclusions, adherence, or the analysis population. These omissions limit assessment of whether the posted comparison reflects all randomized participants and how sensitive it may be to incomplete outcome data.

The mean-difference interval is reported without its confidence level or construction method. The p-value is unadjusted, and the analysis plan is not reported. No secondary results are posted, and the record gives no information about prespecified subgroup or sensitivity analyses. Safety information includes group-level adverse-event counts and a list of serious-event terms, but deaths are not reported in the registry record and term-specific group comparisons are not available in the stated results.

These limitations concern what can be assessed from the registry record; they do not establish that a particular method was absent from the trial. They should be kept distinct from the observed group means and comparison, which remain the posted results for the described endpoint.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Effects of Magnesium on Individuals Undergoing Coronary Artery Bypass Graft Surgery — NCT00041392.
• Registry record last updated 2013-07-30; status last verified 2012-04.
• Results first posted 2012-10-30.
• 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.

21. Record Summary

The registry record describes a completed phase 2, randomized, triple-masked, parallel-group trial sponsored by Duke University. It compares magnesium, described as 100 mg/kg, with 0.9% saline, also described as 100 mg/kg, among people undergoing CABG. Enrollment is recorded as 389, with outcome counts of 198 in the magnesium arm and 191 in the saline arm.

The primary endpoint is continuous cognitive change from baseline to 6 weeks. The posted means are 0.07 for magnesium and 0.12 for saline; the mean difference is 0.05 for saline relative to magnesium. The reported interval is -0.006 to 0.106, and the unadjusted p-value is 0.078. The interval’s confidence level is not reported in the registry record.

The registry record lists no secondary outcome results and reports 0 analyses. Important details about missing data, analysis populations, multiplicity, and the trial’s original sample-size assumptions are not reported in the registry record. The safety results include serious and other adverse-event counts, while deaths are not reported.