Efficacy and Safety of a New Formulation of Chondroitin Sulfate and Glucosamine Sulfate to Treat Knee Osteoarthritis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Efficacy and Safety of a New Formulation of Chondroitin Sulfate and Glucosamine Sulfate to Treat Knee Osteoarthritis (NCT01893905), a randomized phase 3 study. The primary outcome is Change in Pain According to VAS (0-100 mm).

2. Clinical Question

3. Trial Design

This completed Phase III trial used randomized allocation and a parallel-group design. The registry reports quadruple masking, but does not identify which participant, care, research, or assessment roles were masked. The sponsor was Tedec-Meiji Farma, S.A. The study began in 2013-06, and primary completion was recorded as 2014-08. Total enrollment was 158.

Participants were assigned to the chondroitin sulfate and glucosamine sulfate combination or placebo. The active arm description specifies chondroitin sulfate 1200mg and glucosamine sulfate 1500mg, taken orally once a day for 24 weeks. The placebo arm was also administered once a day for 24 weeks. The registry lists 80 participants in the combination arm and 78 in the placebo arm for the primary outcome.

The primary outcome was change in pain according to the VAS (0-100 mm), with a 24-week timeframe. The registry description also says change was calculated at week 4 and week 12, as well as week 24, but the posted primary result concerns the 24-week timeframe. The record lists one outcome measure and one analysis. Details of the recruitment process, eligibility criteria, allocation sequence, concealment procedures, and the trial’s original sample-size rationale 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 Change in Pain According to VAS (0-100 mm), measured over 24 weeks.

The primary endpoint measures change in pain from baseline using a visual analogue scale. A participant indicates pain by marking a position along a continuous 0-100 mm line between two endpoints. The registry describes negative change as a decrease in pain. For the primary result, the outcome is summarized as the mean change within each randomized arm at 24 weeks, with standard deviations also posted.

A continuous scale can retain gradations in participants’ reported pain rather than reducing responses to a small set of categories. Comparing average change between randomized groups can describe how their mean changes differ over the stated follow-up. That group comparison does not show how much participants’ responses varied beyond what the standard deviations indicate, nor does it describe every individual’s course.

The registry describes pain change at week 4, week 12, and week 24, while identifying 24 weeks as the primary outcome timeframe. The record does not report a threshold for a clinically important change, a responder definition, or additional endpoint details such as the precise procedure for handling missing measurements. Those omissions limit interpretation of what size of change would be considered important and how the endpoint was implemented.

7. Results Posted in the Registry

• CS+SG: mean -11.86 versus -20.57 on control; difference 8.71 (95% CI 7.96 to 9.46), standardized effect d = 3.61; z-test p = <0.001.

8. Statistical Interpretation

9. Statistical Methods Explained

What does the mean difference describe?

The endpoint is continuous, and the registry reports a mean change for each arm. A mean difference compares those two arm means, with the recalculation defined as CS+SG minus placebo. The posted arm means are -11.86 and -20.57, and the recalculated difference is 8.71 units. Since negative change represents a decrease in pain, a positive difference in this direction corresponds to the combination arm having a less negative mean change than the placebo arm. This describes the direction of the group contrast; it is not a comparison of each participant against their own treatment alternative.

How should the p-values be read?

In general, a p-value describes how compatible observed data are with a specified statistical null model, under the assumptions of the analysis. It is not the probability that the null hypothesis is true, and it does not measure clinical importance. The unadjusted recalculation gives p < 0.001. The registry separately reports p < 0.031 for the sponsor’s Pocock approach. The record does not provide enough detail to reproduce that sponsor analysis or to establish precisely how it relates to any interim monitoring.

What does the standard deviation add?

Standard deviation describes the spread of individual values around an arm’s mean. The posted standard deviations are 2.42 for CS+SG and 2.41 for placebo. They are not standard errors and are not confidence intervals. General statistical interpretation: a standard deviation characterizes variation among observations, whereas a standard error characterizes uncertainty in an estimated statistic. The registry lists 80 and 78 participants for the respective primary outcome arms, but it does not provide individual-level observations or additional distribution summaries.

What is recalculated?

The recalculated mean difference and its interval are simple unadjusted calculations from the posted summary statistics. They are not presented as the sponsor’s analysis. The recalculated estimate is 8.71, with a confidence interval from 7.957 to 9.463; the unadjusted p-value is p < 0.001. The sponsor’s separate reported analysis is identified as using a Pocock approach and gives p < 0.031, without a posted estimate or interval. The registry does not flag a discrepancy. Since the sponsor analysis is not fully described, neither its precise model nor the relationship between its p-value and the unadjusted calculation can be reconstructed from the record.

Why can design and analysis details matter?

General statistical explanation: random allocation is intended to make treatment groups comparable on average at assignment, while masking can reduce some forms of influence on care or assessment. The record states randomized allocation, parallel design, and quadruple masking, but does not specify implementation details or who was masked. General statistical explanation: the analysis population, outcome handling, and missing-data method can affect the interpretation of a comparison, especially when some measurements are unavailable. These details are not reported in the registry record.

The registry identifies the endpoint as a mean of a continuous measure and lists one analysis. It does not report the model used for the primary endpoint, covariate adjustment, a pre-specified analysis plan, or a complete description of how the sponsor’s Pocock approach was applied. The record also does not report whether there were interim analyses, how multiplicity was handled, or whether the analysis followed an intention-to-treat or another population definition. These absences do not establish that such procedures were not used; they mean they cannot be assessed from the registry record.

Why might over-dispersion matter?

General statistical explanation: over-dispersion occurs when observed variation is greater than a statistical model expects. If a model does not account for that extra variation, its uncertainty estimates may be too small, making intervals appear narrower and tests more decisive than warranted. The registry describes a mean and standard deviation for each arm, but does not report an over-dispersion assessment or a model that would allow one to evaluate it. The posted summary statistics alone are insufficient to make that assessment.

10. Confidence Intervals and Precision

A confidence interval gives a range of values compatible with the observed data under the assumptions of the calculation. It communicates uncertainty in an estimated contrast; it is not a range that contains a specified proportion of individual participants’ outcomes, and it does not assign a probability to the fixed true effect after the data have been observed.

For the unadjusted mean difference of 8.71 units, the recalculated interval extends from 7.957 to 9.463. Its positive direction matches the contrast defined as CS+SG minus placebo. The interval is relatively close around the point estimate, but precision should be understood in relation to the assumptions behind this unadjusted calculation and the completeness of the posted data. The interval is not identified as the sponsor’s interval; the sponsor analysis has no estimate or confidence interval posted in the registry record.

The arm standard deviations are 2.42 and 2.41. These describe observed spread, while the confidence interval addresses uncertainty in the mean difference. Neither statistic indicates whether the size of the contrast matters to patients. The registry record does not report a threshold for clinically important pain change, a distribution of individual changes, or the extent and handling of missing endpoint data. Those missing details limit conclusions about clinical meaning and the robustness of the precision estimate.

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 assumptions. The assumptions listed are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each row uses a stated fraction of the observed effect as the effect to detect. The figures are participants per arm, not total events. The record’s planning figures apply to this continuous-outcome comparison.

For planning around 100% of the observed effect, the table gives 2 participants per arm for 80% power and 2 participants per arm for 90% power. For 75% of the observed effect, it gives 3 participants per arm for 80% power and 3 participants per arm for 90% power. For 50% of the observed effect, it gives 5 participants per arm for 80% power and 7 participants per arm for 90% power. These are planning illustrations under the stated assumptions, not estimates of what was actually needed for the completed trial.

In general, power is the probability that a study will detect an effect of a specified size under the assumed model and design. The 80% and 90% columns show the per-arm planning counts associated with those target probabilities. Smaller effects need more participants to detect, all else equal; the rows illustrate that relationship. These calculations are not observed power. The trial’s own original sample-size justification and design 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 = 8.710; pooled standard deviation σ = 2.415.
2. n = 2 × (1.96 + 0.842)² × 2.415² / 8.710² = 1.2, rounded up to 2 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 <- 8.71                      # observed difference in means
sigma <- 2.41506927755388                     # 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(1893905)
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 = 2 (80% power), 2 (90% power)
 75% of observed effect: n per arm = 3 (80% power), 3 (90% power)
 50% of observed effect: n per arm = 5 (80% power), 7 (90% power)
power.t.test: 2.6 
Simulated power at n = 2 per arm: 0.494 (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. Analyses Reported by the Sponsor

The registry reports a sponsor analysis using a Pocock approach and gives p < 0.031. No sponsor estimate or confidence interval is posted, and the registry does not explain the approach or the analysis details needed to reproduce it. The sponsor result therefore offers a reported significance test but not a fully described estimate of the size or precision of the contrast.

The unadjusted recalculation from the posted arm summaries gives a mean difference of 8.71, a confidence interval from 7.957 to 9.463, and p < 0.001. These calculations are not the sponsor’s analysis and should not be treated as equivalent. The registry does not flag a discrepancy, but the limited sponsor reporting means the two p-values cannot be reconciled from the record. The sponsor analysis remains the distinct analysis reported by the registry; the recalculation adds a simple unadjusted comparison rather than replacing it.

14. Multiplicity, Missing Data and Analysis Populations

Multiplicity concerns arise when several hypotheses or outcomes are tested: the chance of at least one apparently notable result can increase as the number of tests grows. General statistical explanation: researchers may address this through a pre-specified testing hierarchy or an adjustment to the testing procedure. The registry lists one outcome measure and one analysis, with zero secondary outcomes posted. It does not provide a complete protocol or statistical analysis plan that would establish whether other outcomes or tests were planned, or how multiplicity was handled.

The sponsor analysis is described as using a Pocock approach, but the registry does not report whether interim analyses occurred, how many there were, or the boundaries or procedures used. It is therefore not possible to determine from the record how the approach relates to any repeated testing. The sponsor-reported p-value should be read as reported, without inferring unreported interim-monitoring details.

The record does not identify the analysis population, such as whether all randomized participants were included, nor does it state how missing endpoint measurements were handled. The primary result lists 80 participants in the CS+SG arm and 78 in the placebo arm, but the registry does not explain whether these counts represent randomized participants, participants with available endpoint measurements, or another analysis set. It also does not report a missing-data method, imputation procedure, or sensitivity analysis.

These reporting gaps affect how confidently a reader can assess the analysis’s implementation and robustness. They do not establish that the trial lacked a statistical plan, missing-data procedures, or multiplicity controls. They mean that these features cannot be evaluated using the registry record. The narrowest supported reading is the posted 24-week mean-change comparison, together with the limited information on the sponsor’s analysis.

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: breast cancer (1.3%), pulmonary embolism (1.3%), breast cancer metastatic (1.2%), diverticulitis (1.2%).

The registry reports serious adverse events in 2 of 80 participants at risk in the CS+SG arm and 2 of 78 participants at risk in the placebo arm. These are counts of affected participants as reported, not a comparison adjusted for follow-up or other factors. The record does not provide further detail about event timing, severity, relatedness to treatment, or statistical comparison. Deaths are not reported in the registry record.

The listed serious adverse event terms include Breast cancer (1.3%), Pulmonary embolism (1.3%), Breast cancer metastatic (1.2%), and Diverticulitis (1.2%). The registry does not report the denominators or further details for these term-level percentages. Other adverse events affected 20 of 80 participants in the combination arm and 18 of 78 in the placebo arm. These figures describe the posted counts; they do not establish why an event occurred or support conclusions about comparative safety on their own.

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

The posted data show that mean pain change at 24 weeks was negative in both arms, indicating decreases from baseline under the registry’s definition. The mean was -11.86 in the CS+SG arm and -20.57 in the placebo arm. The unadjusted mean difference, defined as CS+SG minus placebo, was 8.71 units. The recalculated confidence interval runs from 7.957 to 9.463, and the unadjusted p-value is p < 0.001. The direction of the contrast reflects a less negative mean change in the combination arm.

These summaries do not establish whether the difference is important to patients. No threshold for a clinically important change is reported in the registry record. A group mean also does not show how many participants improved, worsened, or had little change. The record does not post individual-level results, responder counts, or a distribution of changes that would answer those questions.

The statistical comparison is also not a complete account of the trial’s evidence. The sponsor separately reports a Pocock-approach p-value of p < 0.031, without an estimate or interval. The registry does not provide enough detail to reproduce that analysis or explain its relationship to the unadjusted calculation. Missing information about the analysis population, missing outcome data, and the original statistical plan further limits evaluation.

The results should therefore be interpreted as posted group-level pain-change statistics at a specified time point, not as a complete characterization of benefit, harm, or individual response. The serious adverse event counts are descriptive, with limited accompanying detail. Neither the efficacy statistics nor the adverse-event counts alone establish a clinical recommendation or a treatment effect beyond the comparison and methods documented in the registry record.

17. Why This Trial Matters Statistically

This trial illustrates why the direction and definition of a mean difference matter. Since negative VAS change represents a decrease in pain, the positive contrast defined as CS+SG minus placebo does not mean that pain increased in the combination arm. It reflects that the mean change there was less negative than the placebo mean. Reading the sign without the endpoint’s coding would obscure the actual direction of the posted group means.

It also illustrates the distinction between a sponsor-reported analysis and a simple unadjusted recalculation. The registry gives a Pocock-approach p-value, while the recalculation gives an estimate, interval, and unadjusted p-value from the arm summaries. Those outputs answer related but not necessarily identical statistical questions, and the record does not provide enough methodological detail to reconcile them fully.

Finally, a statistically precise contrast is not automatically a clinically important one. The registry gives no threshold for meaningful change and does not post individual response distributions or secondary outcomes. The trial’s randomized, masked, parallel design is relevant context, but incomplete reporting on the analysis population, missing measurements, and statistical plan constrains assessment of the result’s robustness. These distinctions are central to interpreting registry results responsibly.

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.

The registry record does not provide detailed eligibility criteria, baseline participant characteristics, or a description of how randomization and masking were implemented. Although the design is recorded as randomized, parallel, and quadruple masked, the roles included in masking are not identified. The original sample-size justification and planned assumptions are also not reported in the registry record.

For the primary outcome, the record gives arm means, standard deviations, and participant counts, but does not explain the analysis population or how missing measurements were handled. It does not report a complete statistical analysis plan, covariate adjustment, or enough information to reproduce the sponsor’s Pocock approach. There are no secondary outcomes posted. The record also does not state whether interim analyses took place or how multiplicity was handled.

The endpoint is based on a subjective pain scale, and the record does not provide a clinically important difference threshold or individual-level response summaries. The listed serious adverse-event terms lack reported denominators and further context, and deaths are not reported in the registry record. These limitations concern what can be assessed from the registry record; they should not be taken as proof that the trial did not collect or analyze information that is not posted there.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Efficacy and Safety of a New Formulation of Chondroitin Sulfate and Glucosamine Sulfate to Treat Knee Osteoarthritis — NCT01893905.
• Registry record last updated 2017-08-04; status last verified 2017-04.
• Results first posted 2017-08-04.
• 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

NCT01893905 is a completed randomized, quadruple-masked, parallel-group Phase III trial sponsored by Tedec-Meiji Farma, S.A. It enrolled 158 participants with knee osteoarthritis and compared a once-daily chondroitin sulfate and glucosamine sulfate combination with placebo over 24 weeks.

The primary outcome was change in VAS pain at 24 weeks. Posted mean changes were -11.86 for CS+SG and -20.57 for placebo. The unadjusted recalculation gives a mean difference of 8.71, with a confidence interval from 7.957 to 9.463 and p < 0.001. The registry also reports a sponsor Pocock-approach result of p < 0.031, without an estimate or interval.

Important details—including the analysis population, missing-data handling, original sample-size rationale, and full statistical plan—are not reported in the registry record.