Strategy to Prevent the Onset of Clinically-Apparent Rheumatoid Arthritis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Strategy to Prevent the Onset of Clinically-Apparent Rheumatoid Arthritis (NCT02603146), a randomized phase 2 study. The primary outcome is Number of Participants Who Developed Clinically-Apparent Rheumatoid Arthritis (CL - RA) From Treatment Initiation to Month 36 By Treatment Arm.

2. Clinical Question

The central statistical question is whether the recorded frequency of clinically-apparent rheumatoid arthritis differed between the hydroxychloroquine and placebo groups during the specified follow-up. This is a prevention question in a population selected for an elevated autoantibody, not a study of treatment for established rheumatoid arthritis. The posted primary endpoint is a participant count, and the analysis population is described as modified intent-to-treat: randomized participants who received at least one dose of assigned study drug and met entry criteria.

The registry record describes the purpose as assessing safety and effectiveness for preventing future onset of rheumatoid arthritis. The posted results show event counts and an analysis comparing arms, but they do not establish what would happen in people outside the described population. The primary endpoint definition also includes a specific clinical classification threshold or qualifying joint findings with an erosion on imaging; it is not simply a positive antibody test. The distinction matters because elevated anti-CCP3 describes the recruitment group, while clinically-apparent rheumatoid arthritis is the outcome being counted.

3. Trial Design

This was a phase 2, randomized, double-masked, parallel-group study sponsored by the National Institute of Allergy and Infectious Diseases. The registry record lists an enrollment of 144 and a status of terminated. It records a start date of 2016-04-27 and a primary completion date of 2022-11-01. The reason for termination is not reported in the registry record.

Participants were assigned to hydroxychloroquine or placebo. The hydroxychloroquine arm description specifies 200-400 mg daily, based on ideal body weight, for 12 months; the placebo arm is described as taking placebo pills on the same daily schedule for 12 months. The primary outcome timeframe extends from baseline to Month 36, so the posted outcome horizon is longer than the stated treatment period. The registry record does not explain the follow-up procedures after the 12-month treatment period.

Recruitment was intended to identify healthy participants with elevated anti-CCP3 levels. The listed strategies included prescreening first-degree relatives of patients with rheumatoid arthritis, health-fair attendees, and people with elevated anti-CCP3 identified in rheumatology clinics in the absence of inflammatory arthritis. These routes may identify people with different backgrounds, but the registry record does not report recruitment totals or outcomes by recruitment route.

The posted primary arm counts are 69 participants in the hydroxychloroquine group and 73 in the placebo group. The registry record describes the primary analysis population as modified intent-to-treat, including randomized participants who received at least one dose and met entry criteria. Reasons that the posted arm counts differ from the enrollment figure, or details of exclusions from the analysis population, are not reported in the registry record. The original design assumptions, including the planned sample size justification, 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:

Analysis population for the primary outcome, as defined in the registry: The modified-intent-to-treat population includes all randomized participants who received at least one dose of assigned study drug and met entry criteria.

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 intent-to-treat population includes all randomized participants.

6. Primary Endpoint

The registry lists the primary outcome as Number of Participants Who Developed Clinically-Apparent Rheumatoid Arthritis (CL - RA) From Treatment Initiation to Month 36 By Treatment Arm, measured over Baseline to Month 36.

The primary endpoint counts participants who developed clinically-apparent rheumatoid arthritis from treatment initiation to Month 36. The posted measure is a count of participants, reported separately by randomized arm. This makes the endpoint a binary event outcome at the participant level: for the endpoint, each participant either met the stated event definition during the timeframe or did not.

The registry definition follows the 2010 American College of Rheumatology (ACR)/European League Against Rheumatism (EULAR) Classification Criteria. Clinically-apparent rheumatoid arthritis is defined either by a score of at least 6 indicating definite rheumatoid arthritis, or by a joint examination consistent with inflammatory arthritis together with at least 1 erosion confirmed by x-ray imaging of the hands, wrists, and feet. The endpoint therefore combines a classification score route and a clinical-and-imaging route.

This endpoint is used because the study asks about preventing the onset of clinically apparent disease in people selected for elevated anti-CCP3. Counting the defined clinical outcome is more directly aligned with that question than counting antibody positivity alone. It is also distinct from the secondary inflammatory arthritis outcomes, which are separately listed in the record and do not necessarily require the primary endpoint definition. The timeframe is important: these results describe events through Month 36, while the assigned treatment is described as lasting 12 months.

7. Results Posted in the Registry

• Hydroxychloroquine: 29.0% (20/69) versus 32.9% (24/73) on control; risk difference -3.9 percentage points (95% CI -19.1 to 11.3), risk ratio 0.88 (0.54–1.44); unadjusted two-proportion z-test p = 0.616.

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 Who Developed Clinically-Apparent Rheumatoid Arthritis (CL-RA) From Treatment Initiation to Month 12 By Treatment Arm

• 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:
  • Hydroxychloroquine: 15.9% (11/69) versus 17.8% (13/73) on control; risk difference -1.9 percentage points (95% CI -14.2 to 10.4), risk ratio 0.90 (0.43–1.86); unadjusted two-proportion z-test p = 0.767.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Hydroxychloroquine and the control. The interval gives the range of effects that remain compatible with the data.

Number of Participants Who Developed Inflammatory Arthritis (IA) From Treatment Initiation to Month 12.

• 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:
  • Hydroxychloroquine: 17.4% (12/69) versus 17.8% (13/73) on control; risk difference -0.4 percentage points (95% CI -12.9 to 12.1), risk ratio 0.98 (0.48–1.99); unadjusted two-proportion z-test p = 0.948.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Hydroxychloroquine and the control. The interval gives the range of effects that remain compatible with the data.

Number of Participants Who Developed Inflammatory Arthritis (IA) From Treatment Initiation to Month 36

• 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:
  • Hydroxychloroquine: 30.4% (21/69) versus 37.0% (27/73) on control; risk difference -6.6 percentage points (95% CI -22.1 to 9.0), risk ratio 0.82 (0.52–1.31); unadjusted two-proportion z-test p = 0.409.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Hydroxychloroquine and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes broaden the picture beyond the primary Month 36 endpoint. They include clinically-apparent rheumatoid arthritis at an earlier timepoint, inflammatory arthritis over two timeframes, time to clinically-apparent rheumatoid arthritis, and self-reported joint symptoms. The posted results cover 14 secondary outcomes. These measures address related but distinct aspects of disease onset and symptoms, so they should not be treated as repeated versions of the primary endpoint.

Across the posted binary secondary outcomes with arm counts, the observed arm comparisons do not establish a clear difference. Their uncertainty is substantial, and the record does not give a multiplicity-adjusted interpretation. The secondary data can therefore be used as context for the primary result, but they do not overturn or independently settle it. In particular, related endpoints can share participants and clinical features; agreement across them is not the same as multiple independent confirmations.

In general, when many endpoints are examined, the chance of at least one apparently unusual finding can increase even if there is no underlying effect. Unless a testing hierarchy or multiplicity adjustment is specified, individual secondary p-values or apparent patterns should be interpreted cautiously. The registry record does not report a correction procedure or state which secondary outcomes were designated as confirmatory. Their relative priority and inferential status are therefore not clear from the posted information.

These outcomes may still be useful for generating questions about timing, inflammatory arthritis, or symptom reporting for future research. But the secondary findings are best regarded as supportive or exploratory descriptions rather than proof of a prevention effect. The primary endpoint remains the main posted test of the stated study question, and its confidence interval indicates considerable uncertainty.

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

• Number of Participants Who Developed Clinically-Apparent Rheumatoid Arthritis (CL-RA) From Treatment Initiation to Month 12 By Treatment Arm
• Number of Participants Who Developed Inflammatory Arthritis (IA) From Treatment Initiation to Month 12.
• Time to Development of Clinically-Apparent Rheumatoid Arthritis (CL - RA) By Treatment Arm
• Number of Participants Who Developed Inflammatory Arthritis (IA) From Treatment Initiation to Month 36
• Number of Participant Self-Reported Painful Joints By Treatment Arm
• Number of Participant Self-Reported Stiff Joints By Treatment Arm
• Number of Participant Self-Reported Swollen Joints By Treatment Arm
• Number of Participant Self-Reported Painful Joints in the Hands, Wrists and Feet By Treatment Arm

9. Statistical Interpretation

10. Statistical Methods Explained

The outcome is a participant count, and the main comparison concerns the chance of experiencing the defined event in each randomized group. In general, a risk difference subtracts the event probability in one group from the event probability in the other. A negative risk difference in the posted comparison is oriented toward fewer events in the hydroxychloroquine group than in the placebo group. Its value is an estimate, not a direct statement about what would happen to every participant.

What does a confidence interval express?

In general, a confidence interval describes the uncertainty around an estimated effect under the assumptions of the analysis. A narrow interval indicates greater precision than a wide interval. When a risk-difference interval includes zero, the data are compatible with no between-group difference as well as with values on either side of zero. In this trial, the sponsor's interval for the primary risk difference extends from -0.336 to 0.220. Its span across zero is why the posted comparison does not sharply distinguish the direction of the difference.

How should a p-value be read?

In general, a p-value measures how unusual the observed data, or data more extreme, would be under a specified null model and its assumptions. It is not the probability that the null hypothesis is true, and it does not measure the size or importance of an effect. The sponsor reports p = 0.522 for its primary Chi-squared comparison. The simple unadjusted re-calculation reports p = 0.616. They arise from different summaries and should not be substituted for one another or read as proof that the arms are identical.

Why can the posted analyses differ?

The arm counts are direct event totals, while the sponsor also reports cumulative probabilities and a risk-difference comparison. The registry record does not identify the method used for the cumulative probability estimates, so the reader cannot fully reconstruct how they were obtained from the record alone. A simple calculation from counts is not necessarily equivalent to an analysis that uses a time-to-event or cumulative-probability approach, different analysis rules, or other information. The record does not state which such features, if any, were used. The sponsor's reported analysis should therefore be read as its own posted result, and the re-calculation as a separate, unadjusted summary.

What does the analysis population mean?

The primary analysis population is described as modified intent-to-treat. It includes randomized participants who received at least one dose of assigned study drug and met entry criteria. This differs from a definition that necessarily includes every randomized person regardless of treatment received or eligibility assessment. The record does not report the number excluded from this population, reasons for any exclusions, or the handling of missing endpoint information. Those details matter because estimates can depend on who is included and how incomplete follow-up is treated.

What is the role of the secondary outcomes?

The record lists secondary measures including clinically-apparent rheumatoid arthritis at Month 12, inflammatory arthritis at Month 12 and Month 36, time to clinically-apparent rheumatoid arthritis, and participant-reported joint symptoms. These outcomes can add context about timing and related manifestations, but they do not replace the prespecified primary question. The record does not state whether the secondary analyses were adjusted for the number of comparisons. In general, examining many outcomes raises the chance that some results may look notable by chance, so unadjusted secondary findings are better treated as exploratory unless the analysis plan establishes otherwise.

Finally, the trial record does not report its original statistical analysis plan in enough detail to establish the planned model, handling of missing data, interim monitoring, or multiplicity procedure. It is therefore appropriate to describe the analyses that are posted, distinguish sponsor results from simple recalculations, and avoid inferring unreported methods.

11. Confidence Intervals and Precision

The main posted comparison has substantial uncertainty. The sponsor's risk-difference estimate is -0.058, and its 95% confidence interval runs from -0.336 to 0.220. The interval includes zero, so the posted estimate is compatible with no difference as well as differences in either direction. The interval is broad relative to the estimate, which means the data do not pin down the size or direction of the contrast precisely.

The sponsor's arm-specific cumulative probability estimates also have 95% confidence intervals: 0.213 to 0.459 for hydroxychloroquine and 0.268 to 0.519 for placebo. These intervals overlap. Overlap alone is not a formal test of the difference between groups; the reported comparison is the relevant posted test for that question. The record does not report the method used to derive the cumulative probability intervals.

In general, confidence intervals summarize statistical uncertainty under an analysis and its assumptions. They do not account automatically for every possible limitation, such as missing data, outcome misclassification, or selection into the analysis population. Here, the modified-intent-to-treat definition is stated, but the record does not describe missing-data handling or provide enough information to assess its effect. Precision should therefore be considered alongside the population and analysis details, not in isolation.

The confidence intervals should not be read as a range guaranteed to contain the true effect, nor as evidence that all values within the range are equally plausible. They show that this trial's estimate is not highly precise. This is an important distinction from concluding that there is no effect: the posted interval is consistent with a range of possibilities, and the trial record alone does not resolve them.

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 sample-size table and curve are planning illustrations based on the observed effect, not the trial's original design calculations. For the binary outcomes, the entries are participants per arm, not total events. The planning assumptions listed for these illustrations are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The effect to detect is set to the stated fraction of the observed effect. The trial's own original design assumptions are not reported in the registry record.

The 80% power column describes the number of participants per arm estimated to give an 80% chance of detecting the specified effect under the planning assumptions, if that effect and those assumptions were correct. The 90% power column makes the same type of projection for a 90% chance. These are design-planning quantities, not guarantees about the results of a future study. The table is not observed power.

Rows that use 100%, 75%, or 50% of the observed effect represent different assumed effect sizes for planning. Smaller assumed effects require more participants per arm to detect with the same target power, which is why the entries rise as the assumed fraction decreases. The observed effect in a small study can be uncertain, so these illustrations should not be treated as a precise forecast of the sample size needed to answer the research question.

The curve presents the same general planning relationship visually. Read it as a way to see how projected sample size changes with the effect assumption, not as an additional result from the trial. These calculations do not recover the original protocol's rationale, and the registry record does not report the justification for the enrolled sample size.

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 = 0.329; treatment proportion p₁ = 0.290; average p̄ = 0.309.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6537 = 1.2812.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6531 = 0.5496.
4. n = (1.2812 + 0.5496)² / (-0.0389)² = 2,213.7, rounded up to 2,214 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 <- 0.328767123287671                          # control proportion
p1 <- 0.289855072463768                          # 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(2603146)
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 = 2,214 (80% power), 2,963 (90% power)
 75% of observed effect: n per arm = 3,971 (80% power), 5,315 (90% power)
 50% of observed effect: n per arm = 9,009 (80% power), 12,059 (90% power)
power.prop.test: 2213.7 
Simulated power at n = 2,214 per arm: 0.804 (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 sponsor's primary result is a Chi-squared comparison of risk difference: -0.058, with a 95% confidence interval from -0.336 to 0.220 and p = 0.522. The record also reports cumulative probabilities of 0.336 for hydroxychloroquine and 0.394 for placebo, each with its own 95% confidence interval. The method for those cumulative probability estimates is not reported in the registry record.

The recalculation is a simple unadjusted comparison based on the posted group counts, and its p-value is 0.616. It is not a reproduction of every aspect of the sponsor's analysis. The sponsor's risk-difference estimate and the recalculated estimate are based on different reported summaries; the counts and the sponsor's cumulative probabilities also do not have to be interchangeable quantities. The registry record does not give sufficient methodological detail to explain fully how the sponsor's cumulative estimates were produced.

Accordingly, read the sponsor's posted analysis as the trial's reported inferential result and the recalculation as a transparent supplementary check from summary counts. The two p-values are both above conventional thresholds, but neither establishes equivalence or proves the absence of an effect. The sponsor's confidence interval is particularly useful for seeing the range of differences compatible with its analysis. No discrepancy is flagged in the posted record.

15. Multiplicity, Missing Data and Analysis Populations

The record lists 15 outcome measures and 26 analyses, including a primary outcome and multiple secondary outcomes. Testing many outcomes can increase the probability of obtaining at least one apparently notable result by chance. In general, multiplicity methods address this issue by controlling the overall error rate, using a prespecified hierarchy, or identifying which comparisons are confirmatory and which are exploratory. Whether any such strategy was planned or applied is not reported in the registry record.

The record also does not report a full statistical analysis plan, an order of testing, or a distinction between confirmatory and exploratory secondary analyses. Without that information, it is difficult to assess the inferential weight of individual secondary comparisons. The absence of a reported adjustment does not prove that no adjustment was used; it means the registry record does not document one.

The primary population is described as modified intent-to-treat: randomized participants who received at least one dose of assigned study drug and met entry criteria. The record does not report exclusions from this population, the number with missing primary endpoint data, or how missing outcomes were handled. It also does not explain whether participants were analyzed according to their assigned arm in every posted analysis. These omissions limit an independent assessment of possible bias from exclusions or incomplete follow-up.

The registry record does not report interim analyses, stopping rules, or the reason the study was terminated. It also does not report the original sample-size justification. These details would help readers understand whether the study's course or analysis plan changed and how the available sample size related to the intended question. They should not be inferred from the trial status or the posted results.

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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: myocardial infarction (1.4%), gastrointestinal haemorrhage (1.4%), autoimmune hepatitis (1.4%), foot fracture (1.4%), colon cancer (1.4%).

The safety summary reports serious adverse events among participants at risk: 5 affected participants in the hydroxychloroquine group, with 71 at risk, and 5 in the placebo group, with 73 at risk. It separately reports other adverse events affecting 25 participants in the hydroxychloroquine group and 31 in the placebo group. The record also lists 1 death in the hydroxychloroquine group and 0 in the placebo group. These are descriptive counts; the registry summary does not provide a causal attribution for individual events.

The listed serious-event terms are myocardial infarction, gastrointestinal haemorrhage, autoimmune hepatitis, foot fracture, colon cancer, and desmoid tumour, each shown as 1.4%. The record does not explain whether these terms overlap within participants, provide detail about timing, or state whether investigators considered events related to study treatment. Do not interpret the term list as establishing causation or comparing risks without additional analysis.

These figures should be read in the context of the reported groups at risk and the trial's limited size. A numerical difference in counts alone cannot determine whether an event pattern was caused by treatment. The safety summary is not a substitute for a detailed adverse-event report, and the registry record does not provide further event-level information.

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

The primary results show the number of participants meeting the registry's clinical definition of rheumatoid arthritis in each arm and a sponsor-reported comparison with a confidence interval spanning values on both sides of no difference. Statistically, the posted analysis does not resolve the direction or size of a difference with precision. That is a statement about what these data can establish, not a recommendation about treatment.

The trial asks a focused question in people with elevated anti-CCP3 who were recruited as healthy participants and did not have inflammatory arthritis at screening. Its findings do not automatically apply to people with established rheumatoid arthritis, people without elevated anti-CCP3, or populations recruited by other means. The registry record does not provide subgroup results by recruitment pathway or other participant characteristics.

The endpoint has a specific clinical definition and a Month 36 timeframe. It is not interchangeable with antibody levels, inflammatory arthritis alone, or self-reported joint symptoms. The treatment is described as lasting 12 months, while the primary endpoint is assessed through Month 36; the record does not detail follow-up procedures or the handling of incomplete follow-up.

A non-significant comparison does not show that the treatments are equivalent, and an estimate favoring one arm does not by itself demonstrate a causal effect. The confidence interval and analysis population are essential context. The reason for termination, missing-data methods, and original sample-size rationale are not reported in the registry record, further limiting the conclusions that can be drawn from the posted material.

Thus, the proper takeaway is descriptive and bounded: the record reports primary and secondary outcomes from this randomized comparison, with considerable uncertainty around the primary between-arm estimate. It neither establishes a definitive prevention effect nor proves that no effect exists. Broader conclusions would require evidence beyond these posted results.

18. Why This Trial Matters Statistically

This trial illustrates several statistical issues common in prevention research. First, the outcome is relatively specific: participants are counted when they meet a defined clinical classification or the alternative joint-and-imaging criterion. Clear endpoint definitions help readers distinguish the clinical event of interest from risk markers such as elevated anti-CCP3.

Second, effect estimates need to be read together with their uncertainty. The primary sponsor-reported risk difference is accompanied by a confidence interval that spans zero and extends in both directions. Looking only at a point estimate or p-value would hide how much uncertainty remains. The interval makes clear that the observed trial does not precisely determine the contrast between arms.

Third, the record shows why the analysis itself matters. The sponsor reports cumulative probabilities and a Chi-squared risk-difference comparison, while a simple unadjusted recalculation from group counts yields a different p-value. These summaries should be distinguished rather than collapsed into one answer. More complete reporting of methods would allow readers to understand and reproduce the sponsor's analysis.

Finally, multiple secondary outcomes, a modified-intent-to-treat population, and a terminated study status all make transparent reporting especially important. The registry record does not document multiplicity procedures, missing-data methods, the reason for termination, or the original sample-size rationale. Those gaps do not invalidate the posted figures, but they constrain interpretation. Statistically, the page is a reminder that a useful result includes not just counts and tests, but also the population, endpoint definition, analysis choices, and precision.

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 limitations specific to this record constrain interpretation. The sponsor reports cumulative probabilities for the primary endpoint, but the method used to produce those estimates is not reported in the registry record. The posted count-based recalculation is therefore a supplementary summary rather than a complete reconstruction of the sponsor's analysis. The sponsor's risk-difference result and simple recalculation should be kept distinct.

The primary analysis population is modified intent-to-treat, but the record does not report the number of participants excluded from that population or the reasons for exclusion. It also does not report how missing outcome information was handled. Although the study has a Month 36 endpoint, details on follow-up completion are not given. These omissions make it difficult to assess the influence of incomplete data.

The registry record lists multiple outcomes and analyses but does not report a multiplicity procedure or a hierarchy for secondary outcomes. This limits the confirmatory interpretation of secondary comparisons. The study status is terminated, but the reason is not reported; the original sample-size assumptions and justification are also not reported. It is not possible to infer from the record whether the termination altered planned analyses or follow-up.

The recruitment summary describes several routes to identifying participants with elevated anti-CCP3, but outcomes by recruitment route are not reported. The record also does not provide enough detail to assess representativeness beyond the stated eligibility context. Finally, the serious adverse-event summary lacks event-level timing and attribution. These limitations do not change the counts as posted, but they restrict what can be concluded from them.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Strategy to Prevent the Onset of Clinically-Apparent Rheumatoid Arthritis — NCT02603146.
• Registry record last updated 2025-06-12; status last verified 2025-05.
• Results first posted 2023-11-21.
• 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

The registry record describes a randomized, double-masked, parallel-group phase 2 study of hydroxychloroquine versus placebo for prevention of clinically-apparent rheumatoid arthritis in participants with elevated anti-CCP3. Hydroxychloroquine and placebo were described as daily treatments for 12 months. The primary outcome counts participants meeting the registry's clinical definition of rheumatoid arthritis through Month 36.

The posted primary comparison reports a risk difference of -0.058, with a 95% confidence interval from -0.336 to 0.220 and p = 0.522. The interval includes no difference and leaves substantial uncertainty about the direction and size of the contrast. Secondary outcomes and safety counts are also posted, but the record does not report multiplicity procedures, missing-data methods, the reason for termination, or the original sample-size justification.

These results describe the posted analysis in the stated population. They do not establish equivalence, prove a prevention effect, or settle the broader research question.