Convalescent Plasma to Stem Coronavirus (CSSC-001): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Convalescent Plasma to Stem Coronavirus (CSSC-001) (NCT04323800), a randomized phase 2 study. The primary outcome is Efficacy of Treatment at Day 28 as Assessed by Number of Participants Who Develop SARS-Cov-2 Infection.

2. Clinical Question

3. Trial Design

CSSC-001 was a Phase 2, randomized, triple-masked, parallel-group study. It compared high-titer anti-SARS-CoV-2 plasma with SARS-CoV-2 non-immune plasma among adults exposed to COVID-19. The registry lists enrollment of 180 and marks the study completed. The start date is 2020-06-10, and the primary completion date is 2021-04-22. The registry does not report the criteria used to define exposure, the detailed eligibility rules, the randomization procedure, or the methods used to maintain masking.

The primary outcome was the number of participants who developed SARS-CoV-2 infection by Day 28. Infection was defined in the outcome description as symptoms compatible with infection and/or positive molecular testing, regardless of disease severity. The posted primary outcome counts were 12 participants in the high-titer group and 13 in the non-immune group. The corresponding group sizes were 81 and 87. These are the reported outcome-analysis group figures; the registry does not report in this record why they differ from the listed enrollment.

The registry lists severe disease as a secondary outcome with a timeframe of up to 28 days. It also reports adverse-event counts and terms. The exact analysis populations, handling of participants who did not complete follow-up, and details of any other outcome measures are not reported in the registry record. The registry lists 4 outcome measures and 3 analyses, but the available posted results identify one primary endpoint and one secondary outcome. The original trial protocol and its detailed statistical plan 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

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.

6. Primary Endpoint

The registry lists the primary outcome as Efficacy of Treatment at Day 28 as Assessed by Number of Participants Who Develop SARS-Cov-2 Infection, measured over Day 28.

The primary endpoint counted participants who developed SARS-CoV-2 infection by Day 28. The registry’s description includes symptoms compatible with infection and/or positive molecular testing and specifies that infection was counted regardless of severity. This makes the endpoint a participant-level binary outcome: each participant either met the stated infection criteria during the timeframe or did not.

In general, a cumulative-incidence endpoint can describe whether an event occurred over a defined follow-up period. It is useful when the trial question concerns the proportion of participants experiencing an event within a clinically specified timeframe, rather than the event’s timing or duration. Here, the stated timeframe was Day 28, and the endpoint addressed infection after exposure. It therefore differs from an endpoint focused only on severe illness: an infection could qualify regardless of severity.

The registry does not report a more detailed case-adjudication process, the relative role of symptoms and molecular testing, or how incomplete testing or follow-up was handled. Those details matter when interpreting how consistently the endpoint could be ascertained across groups. The result should be understood as a comparison of the posted participant counts under the registry’s stated endpoint definition, not as evidence about outcomes beyond the recorded timeframe.

7. Results Posted in the Registry

• SARS-CoV-2 Non-immune Plasma: 14.9% (13/87) versus 14.8% (12/81) on control; risk difference 0.1 percentage points (95% CI -10.6 to 10.9), risk ratio 1.01 (0.49–2.08); unadjusted two-proportion z-test p = 0.981.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Number of Participants With Severe Disease

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • SARS-CoV-2 Non-immune Plasma: 2.3% (2/87) versus 0.0% (0/81) on control; risk difference 2.3 percentage points (95% CI -0.9 to 5.4); unadjusted two-proportion z-test p = 0.170.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between SARS-CoV-2 Non-immune Plasma and the control. The interval gives the range of effects that remain compatible with the data.

The registry posts one secondary outcome: severe disease through up to 28 days. It is related to the primary endpoint because severe disease occurs within the broader context of infection, but it asks a narrower question than whether participants developed infection regardless of severity. The secondary result has few events and its uncertainty interval includes zero. It therefore does not provide a precise, independent resolution of the primary question.

The overall pattern should be considered in light of the small amount of event information and the distinction between the endpoints. A secondary outcome can add context, but a result based on sparse events is especially sensitive to the outcomes of a small number of participants. The registry does not report a multiplicity adjustment or a prespecified hierarchy for interpreting secondary outcomes. Without such information, unadjusted secondary results carry less weight because examining multiple outcomes can increase the chance of apparently notable findings arising by chance.

The available secondary result should not be treated as proof that the primary comparison is confirmed or contradicted. It addresses a different outcome and does not overcome the uncertainty in the primary estimate. Further secondary-outcome definitions, planned analyses, and analysis populations are not reported in the registry record.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Number of Participants With Severe Disease

9. Statistical Interpretation

10. Statistical Methods Explained

For the primary endpoint, the registry reports participant counts in each group. The recalculated comparison uses those posted summary counts to estimate an unadjusted difference in risk. It does not include covariate adjustment or additional participant-level information. The registry’s own analysis is reported separately and uses a Restricted Mean Survival Test statistic with a risk-difference parameter.

What does a risk difference express?

A risk difference compares the proportion of participants with the event in one group with the proportion in the other. A value near zero describes similar observed proportions, while a positive or negative value indicates which group had the larger observed proportion according to the comparison’s ordering. In general, risk differences are expressed on the absolute probability scale, which can make the size of an observed contrast easier to interpret than a relative measure. The recalculated estimate is close to zero, but its uncertainty is substantial relative to that estimate.

What does a confidence interval contribute?

A confidence interval gives a range of estimates compatible with the data under the method used to construct it. In general, it reflects both the observed contrast and the amount of information available. A wide interval signals limited precision. When an interval for a risk difference includes zero, the data are compatible with no difference as well as with differences in either direction. That does not show that the true effect is exactly zero; it shows that the data do not narrowly distinguish among those possibilities.

How should the p-value be read?

The recalculated p-value is unadjusted. In general, a p-value is calculated under a null hypothesis and a statistical model, and describes how often results at least as extreme as the observed result would occur if that null model were correct. It is not a measure of the probability that a treatment effect exists, nor does it indicate whether an effect would be important. The registry’s sponsor analysis has a separate p-value and method. The record does not report sufficient detail to reconstruct that analysis or identify why it differs from the simple recalculation.

Why can a binary endpoint be uncertain?

For a binary endpoint, the information in the comparison depends on the number of participants with and without the event, as well as the group sizes. When event counts are limited, estimates can vary substantially from one set of participants to another. In general, this produces wider uncertainty intervals and makes the estimate less able to rule out effects in either direction. A small p-value or a large p-value cannot substitute for examining the estimate and its confidence interval.

What does over-dispersion mean, and does it apply?

In general, over-dispersion occurs when variation in the data is greater than a chosen statistical model expects. It can make standard errors too small if the extra variation is ignored. The primary endpoint is reported as a participant count for a binary event, and the registry does not report an over-dispersion adjustment or describe an over-dispersed model. No claim about over-dispersion is therefore warranted for this result.

How does the registry analysis differ in form?

The registry names a Restricted Mean Survival Test statistic and reports a risk difference, but does not report a confidence interval for that result. In general, survival methods can account for the timing of events and censoring, while a simple comparison of event counts does not use event timing. The registry record does not report the event-time data, censoring details, or test implementation needed to explain exactly how the named method was applied to this endpoint. The reported sponsor result and the recalculated unadjusted comparison should therefore be kept distinct rather than treated as interchangeable analyses.

The recalculation is descriptive and depends only on the posted group counts. It cannot recover features that require participant-level records, such as adjustment for baseline factors, event timing, missing outcome status, or protocol-defined analysis populations. Those details are not reported in the registry record.

11. Confidence Intervals and Precision

The recalculated risk-difference confidence interval extends from a negative value to a positive value, so it includes zero. Its width indicates that the estimate is not precise enough to exclude differences in either direction. This is important because the point estimate is close to zero: the estimate alone may appear reassuringly small, but the interval shows that the data do not tightly constrain the range of compatible effects.

In general, a confidence interval describes uncertainty in an estimated effect under the method used to calculate it. It is not a range containing a fixed proportion of individual outcomes, nor does it assign a probability to the true effect. The interval’s interpretation depends on the calculation and assumptions behind it. The registry does not report enough detail to assess every assumption used by the sponsor analysis, whose reported result has no confidence limits.

The primary endpoint counts were 12 and 13 in groups of 81 and 87. A comparison based on these event counts has limited information for distinguishing small or moderate differences. This is not a claim that the groups are the same; it is a reason to treat the estimated contrast cautiously. The unadjusted p-value of 0.981 is consistent with the uncertainty reflected by the interval, but the p-value does not express precision by itself.

For the secondary outcome, the registry posts a risk-difference interval whose bounds lie on opposite sides of zero. This likewise indicates uncertainty rather than a settled group difference. The sponsor analysis does not have a posted interval, so its precision cannot be assessed from confidence limits in the registry record. More detailed population, missing-data, and analysis information is not reported in the registry record.

12. 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 participants or total events. The listed assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect used for planning is a stated fraction of the observed effect. These are planning illustrations calculated from the observed effect, not the trial’s own original design assumptions; those original assumptions are not reported in the registry record.

For planning for 100% of the observed effect, the figures are 1,218,887 participants per arm for 80% power and 1,631,743 participants per arm for 90% power. For planning for 75% of the observed effect, they are 2,164,992 participants per arm for 80% power and 2,898,309 participants per arm for 90% power. For planning for 50% of the observed effect, they are 4,866,913 participants per arm for 80% power and 6,515,415 participants per arm for 90% power.

In plain terms, the 80% power column gives the modeled participants per arm needed to have an 80% chance of detecting the specified effect under the stated assumptions; the 90% column uses a higher target chance. Smaller effects need more participants because they are harder to distinguish from random variation. Since the observed effect is close to zero, these illustrations produce very large planning figures. This is not observed power: it is a forward-looking calculation based on the observed effect and stated assumptions, not a measure of what the completed study achieved. The registry does not report the original sample-size justification or design assumptions.

13. 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 = 0.148; treatment proportion p₁ = 0.149; average p̄ = 0.149.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.5033 = 0.9864.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5033 = 0.4236.
4. n = (0.9864 + 0.4236)² / (0.0013)² = 1,218,886.7, rounded up to 1,218,887 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 <- 0.148148148148148                          # control proportion
p1 <- 0.149425287356322                          # 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(4323800)
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 = 1,218,887 (80% power), 1,631,743 (90% power)
 75% of observed effect: n per arm = 2,164,992 (80% power), 2,898,309 (90% power)
 50% of observed effect: n per arm = 4,866,913 (80% power), 6,515,415 (90% power)
power.prop.test: 1218887 
Simulated power at n = 1,218,887 per arm: 0.799 (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.

14. Analyses Reported by the Sponsor

The registry’s sponsor analysis reports a risk difference of 0.01, with p = 0.42, using a Restricted Mean Survival Test statistic. It does not post confidence limits for this result. The recalculated comparison is an unadjusted analysis of the posted group counts and has p = 0.981, with a confidence interval that crosses zero. The two results should be read as separate analyses rather than combined into a single estimate.

The registry record does not report enough detail about the sponsor’s test implementation, analysis population, or handling of follow-up to explain the difference between the sponsor result and the recalculation. The sponsor result is the trial’s reported analysis; the recalculation is a simple comparison based on posted summary statistics. Neither p-value alone establishes the presence or absence of an effect. The estimate and its precision, the method used, and the information not reported in the registry record all matter to interpretation.

15. Multiplicity, Missing Data and Analysis Populations

The registry record does not report a multiplicity-control procedure, a testing hierarchy, or a prespecified strategy for deciding which outcome should carry confirmatory weight. The record lists 4 outcome measures and 3 analyses, while the posted results identify one primary outcome and one secondary outcome. Without the complete protocol and statistical analysis plan, it is not possible to determine whether other endpoints or analyses were planned, how they were prioritized, or whether any adjustment was intended.

In general, testing several outcomes or conducting several analyses can raise the chance of at least one result appearing notable by chance. Multiplicity procedures can address this issue, but the appropriate approach depends on the trial’s planned questions and analysis strategy. The unadjusted secondary result should therefore be interpreted cautiously; the registry does not report whether it was adjusted for other tests.

The registry also does not report the analysis populations used for the posted results, how participants with missing outcome status were handled, or whether any imputation or sensitivity analysis was performed. These omissions matter because the primary endpoint depends on identifying infection during follow-up. In general, missing status can affect a group comparison if it is handled differently across groups or is related to the chance of having an event.

Randomization supports a fair comparison when the assigned groups are analyzed according to an appropriate, prespecified plan, but the registry record alone does not establish which analysis population was used. Details of participant flow, protocol deviations, outcome ascertainment, and missing-data assumptions are not reported in the registry record. Conclusions should remain limited to the posted summaries and the sponsor analysis as recorded.

16. Safety Results

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

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: grade 3 or 4 adverse events (14.9%), severe transfusion reaction (1.1%).

The registry lists serious adverse-event counts of 4 in the high-titer plasma group and 14 in the non-immune plasma group. It lists other adverse-event counts of 24 and 44, respectively, and deaths of 0 in each group. The record also lists Grade 3 or 4 adverse events at 14.9% and severe transfusion reaction at 1.1% among the top serious-event terms.

These are descriptive posted safety results. The registry record does not report the detailed definitions, event-level descriptions, attribution assessments, or analysis methods needed to determine whether the group counts reflect a treatment-related difference. Counts alone do not establish causation, and the listed serious-event terms should not be assumed to represent all events or to share the same denominator unless the record specifies that information. Follow-up details for safety outcomes are not reported in the registry record. The figures should be read as recorded safety observations, not as a standalone comparison of risk.

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

The results describe a randomized comparison of high-titer anti-SARS-CoV-2 plasma with non-immune plasma among adults exposed to COVID-19. For the primary endpoint, the posted counts are close, and the unadjusted risk-difference estimate is close to zero. However, the confidence interval spans values on both sides of zero. The data therefore do not precisely determine the direction or size of a group difference in Day 28 infection.

This does not establish equivalence. Demonstrating equivalence requires a defined margin and an analysis designed to assess whether the full uncertainty interval lies within that margin. Such a margin and an equivalence analysis are not reported in the registry record. Nor does a large p-value show that no effect exists. The posted analysis is better described as inconclusive about the size and direction of a difference, given its uncertainty.

The primary endpoint was infection regardless of severity, while severe disease was a separate secondary outcome with sparse events. The secondary result cannot resolve the uncertainty in the primary comparison. The sponsor’s reported analysis is distinct from the simple recalculation and lacks posted confidence limits. Details about the sponsor method are not reported in the registry record, so the reason for the different p-values cannot be established from these results alone.

The findings describe the recorded outcomes within the registry’s stated timeframe and populations. They do not establish outcomes beyond Day 28, effects in people outside the study population, or effects under different treatment procedures. Additional protocol, follow-up, missing-data, and analysis details are not reported in the registry record. The statistical interpretation should therefore stay focused on what the posted data show: the observed primary counts are similar, while the estimate remains too imprecise to rule out differences in either direction.

18. Why This Trial Matters Statistically

This trial illustrates why an observed contrast and its precision should be read together. The primary event counts are close, but the confidence interval extends across zero and leaves meaningful uncertainty about the group difference. A small estimated contrast does not, by itself, show that groups are equivalent or that a clinically relevant difference has been ruled out.

It also demonstrates the importance of distinguishing a registry-reported analysis from an unadjusted recalculation based on summary counts. The sponsor reports a Restricted Mean Survival Test statistic and a risk difference, while the recalculation uses posted group counts. Their different p-values do not mean that one can be substituted for the other. Method details are needed to understand how each analysis was conducted and what assumptions it used.

The secondary outcome and safety counts provide additional context, but sparse outcomes and limited methodological detail constrain how much can be inferred from them. Multiplicity, analysis populations, and missing-data handling are not reported in the registry record, so those aspects cannot be evaluated. Taken together, the trial is a useful example of how design labels, outcome definitions, point estimates, intervals, and reporting gaps each shape statistical interpretation. The central lesson is to avoid turning an imprecise comparison into a categorical conclusion.

19. 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 affect interpretation. The primary outcome is defined using symptoms compatible with infection and/or positive molecular testing, but the registry does not report how cases were adjudicated, how testing was scheduled, or how participants without complete testing or follow-up were handled. Differences in ascertainment could affect a comparison, but the registry record does not provide enough detail to evaluate that possibility.

The registry lists enrollment of 180, while the posted primary outcome groups contain 81 and 87 participants. The record does not explain the relationship between listed enrollment and the outcome-analysis group sizes or provide participant flow and reasons for any exclusions. The analysis populations and missing-data procedures are not reported in the registry record.

The primary outcome is limited to Day 28, and the registry does not report longer-term outcomes for this comparison. The secondary severe-disease outcome has sparse events, which limits its precision. The sponsor analysis has no posted confidence interval, and its method is named without sufficient detail to reproduce or fully assess it. The registry also does not report the original sample-size rationale, multiplicity plan, or full statistical analysis plan. These reporting gaps make it difficult to assess design assumptions and analysis choices beyond the summaries recorded.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Convalescent Plasma to Stem Coronavirus (CSSC-001) — NCT04323800.
• Registry record last updated 2022-04-26; status last verified 2022-04.
• Results first posted 2022-04-22.
• 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.

23. Record Summary

CSSC-001 was a completed, randomized, triple-masked, parallel Phase 2 study comparing high-titer anti-SARS-CoV-2 plasma with SARS-CoV-2 non-immune plasma in adults exposed to COVID-19. The registry lists enrollment of 180, a start date of 2020-06-10, and primary completion on 2021-04-22.

The primary endpoint was development of SARS-CoV-2 infection by Day 28, regardless of disease severity. The posted group counts were 12 and 13, with group sizes of 81 and 87. The unadjusted recalculation is close to zero but imprecise, while the sponsor reports a separate analysis with p = 0.42. Severe disease was the posted secondary outcome. Additional details on analysis populations, missing data, multiplicity, and the sponsor’s method are not reported in the registry record.