Cervical Cancer Screening Intervention Among Korean American Women: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Cervical Cancer Screening Intervention Among Korean American Women (NCT02594826), a randomized na study. The primary outcome is Number of Women Who Receive a Pap Smear Test.

2. Clinical Question

3. Trial Design

The registry describes a randomized, parallel-group behavioral trial with no masking. Participants were assigned to the culturally appropriate intervention or the general health education control. Because masking was listed as none, the record does not describe masked assessment or masked participants. It does not report the method used to generate or conceal the allocation sequence, or whether allocation was stratified.

The stated enrollment is 744. The primary outcome entries list 290 participants in the culturally appropriate intervention group and 298 in the general health education control group. The secondary outcome entries list 341 and 335 participants, respectively. These are outcome-specific entries; the registry does not explain the differences in group counts between outcomes, nor does it say whether they reflect missing outcome data, analysis-population definitions, or another reason. The numbers should therefore be read in the context of each outcome rather than treated as interchangeable analysis denominators.

The trial is listed as completed, with phase recorded as NA. It began in 2009-02, and primary completion is recorded as 2013-09. The registry gives Cervical Carcinoma as the condition. Fox Chase Cancer Center is listed as sponsor. These administrative details describe the registry record, not the quality or completeness of the statistical reporting.

The intervention was delivered in a community and church-based context according to its description. The comparator offered general health and cancer education. The registry does not report further implementation details, such as attendance, intervention fidelity, exposure to other screening services, or whether participants received additional support outside the assigned program. Those details would help clarify the contrast between the groups, but cannot be inferred from the intervention labels alone.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: Following eligibility screening and informed consent, women completed a baseline questionnaire, which included items on screening history and prior medical history. For a small number of women, responses on the baseline questionnaire were deemed to be inconsistent with study eligibility. Therefore, those women were excluded from further analysis.

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 Number of Women Who Receive a Pap Smear Test, measured over 12 months.

The primary endpoint is the number of women in each group who received a Pap smear test within the stated 12-month timeframe. It is recorded as a participant count, so the outcome represents whether the screening event was reported for a participant, rather than a continuous measurement such as a score or a time-to-event outcome. The registry’s group-level counts provide the number with the event and the number represented in each group.

Screening uptake is a direct way to assess the trial’s stated objective of increasing cervical cancer screening. Comparing the groups’ event proportions allows the result to be expressed as an absolute risk difference, while relative measures such as a risk ratio or odds ratio describe the contrast on a different scale. Each scale answers a related but distinct statistical question. The absolute difference is often useful for understanding the size of the group separation in the observed data; relative measures describe how the event occurrence compares between groups.

The endpoint does not, by itself, describe the quality or result of the test, later follow-up, or health outcomes. The record reports the number receiving a Pap smear test, not whether a test detected disease or changed subsequent care. The exact method used to establish test receipt, including whether it was self-reported or confirmed from records, is not reported in the registry record. The timing is specified as 12 months, but further details about the assessment window are not reported.

7. Results Posted in the Registry

• Culturally Appropriate Intervention: 72.1% (209/290) versus 10.1% (30/298) on control; risk difference 62.0 percentage points (95% CI 55.8 to 68.2), risk ratio 7.16 (5.06–10.13); unadjusted two-proportion z-test p = <0.001.

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.

Knowledge About Cervical Cancer

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard errors.
• Result:
  • Culturally Appropriate Intervention: mean 6.65 versus 6.03 on control; difference 0.63 (95% CI 0.14 to 1.11), standardized effect d = 0.20; z-test p = 0.011.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Culturally Appropriate Intervention than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The registry lists one secondary outcome: knowledge about cervical cancer, measured in units on a scale at 12 months post-program. Its posted results indicate a positive mean contrast favoring the culturally appropriate intervention group, and the reported unadjusted comparison has p = 0.011. The sponsor’s linear regression analysis reports p = 0.005. These findings concern knowledge, not the receipt of screening itself, and offer a different kind of information from the primary endpoint.

Considered alongside the primary result, the secondary outcome is directionally consistent with an educational intervention being associated with a difference in a knowledge measure as well as a difference in screening uptake. That pattern does not establish that knowledge changed first, that knowledge mediated screening behavior, or that the screening contrast was caused by a change in knowledge. The registry does not report a mediation analysis or a causal pathway analysis. The two outcomes should therefore be treated as related questions rather than evidence of a demonstrated mechanism.

Unadjusted secondary results generally carry less weight than a prespecified primary result because examining multiple outcomes can increase the chance of finding an apparently notable result even when no true differences exist. This is a general statistical concern. The registry record does not report a multiplicity procedure, a hierarchy for testing outcomes, or a detailed analysis plan establishing how the secondary endpoint was to be interpreted. The secondary p-value should thus be read as unadjusted evidence for that comparison, rather than as a result that has been protected against all multiple-testing concerns.

The registry reports a mean and standard error for each group, alongside a mean-difference estimate and interval. The scale’s range and the meaning of a unit difference are not reported. Without that context, statistical evidence of a difference does not indicate whether its size is important in practice. Nor does a group-level mean comparison show how individual participants’ knowledge changed over time; the posted results do not provide participant-level change patterns.

The record lists two outcome measures and two analyses overall. It does not report whether other measured outcomes were omitted from the posted record. That missing context limits assessment of selective reporting and the overall multiplicity burden. The secondary result is therefore useful as supporting context, but it should not be treated as an independent confirmation of the primary result or as proof of the intervention’s mechanism.

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

• Knowledge About Cervical Cancer

9. Statistical Interpretation

10. Statistical Methods Explained

What does the risk difference describe?

For a binary outcome, each participant is classified as having or not having the event. The risk, or event proportion, is the number with the event divided by the number in the group. The risk difference subtracts the control-group risk from the intervention-group risk. A positive difference in this comparison means that the event proportion is higher in the intervention group. It expresses the contrast on an absolute scale, in the same probability units as the group risks.

The recalculated risk difference for the Pap smear endpoint is 0.62, with a 95% confidence interval from 0.558 to 0.682. The group risks posted for the comparison are 0.721 and 0.101. These estimates are based on the listed group counts and are unadjusted: they do not account for participant characteristics or other variables in a model. A risk difference is often straightforward to interpret, but its value depends on the baseline risk in the comparison group and the population being studied.

How do risk ratios and odds ratios differ?

A risk ratio compares the event risk in one group with the event risk in another. An odds ratio compares event odds, where odds are the event probability divided by the probability of no event. Risk and odds are close when an event is uncommon, but can differ substantially when it is common. Accordingly, an odds ratio should not be interpreted as if it were a risk ratio or as an absolute increase in probability.

The registry’s recalculated comparison includes a risk ratio of 7.159, with its interval from 5.061 to 10.13, and an odds ratio of 23.05, with its interval from 14.61 to 36.38. These are unadjusted recalculations from the posted summary counts. The registry’s sponsor analysis reports a mixed-model odds ratio of 35.8, with a 95% confidence interval from 11.1 to 114.9. The estimators are on an odds scale but are not necessarily the same analysis: one is calculated from group-level summaries, while the other is reported from a mixed-model analysis. The model’s exact specification is not reported in the registry record.

What does a confidence interval communicate?

A confidence interval gives a range of values compatible with the data and the statistical model used to construct it. It conveys uncertainty around an estimate; it is not a range that contains the true effect with certainty, and it does not account automatically for biases in design, measurement, or missing data. A narrow interval indicates greater statistical precision than a wide interval, under the assumptions used for the calculation. Intervals should be interpreted on the scale of the measure: a risk-difference interval concerns an absolute contrast, whereas an odds-ratio interval concerns a relative odds contrast.

In repeated use of the same method under its assumptions, confidence intervals at a stated level would cover the target parameter at that level over repeated studies. That general interpretation does not mean there is a 95% probability that the particular fixed effect lies inside this particular interval. The registry labels the sponsor’s interval as 95%. The recalculated comparisons also give lower and upper confidence limits; the registry’s comparison fields do not separately state the confidence level for those recalculated limits.

What does a p-value contribute?

A p-value describes how unusual the observed data, or data more extreme, would be if the null hypothesis and the analysis assumptions held. A small p-value can be evidence against that null model, but it does not measure the probability the null is true. Nor does it measure the probability that the finding arose “by chance,” establish clinical importance, or capture uncertainty about bias and applicability.

The unadjusted primary comparison reports p < 0.001, and the sponsor’s mixed-model analysis also reports p < 0.001. The secondary unadjusted comparison reports 0.011, while the sponsor’s linear regression analysis reports 0.005. The methods behind the sponsor analyses are named, but their detailed specifications are not reported. The p-values therefore should be read alongside the estimates, intervals, outcome definitions, and limitations of the registry information.

Why might a mixed model be used?

In general, a mixed model can represent outcome variation associated with groups or clusters, and may account for correlation among observations that are not statistically independent. For example, data collected within shared settings can sometimes be correlated. This is a general explanation of the method, not a claim about the model used in this trial. The registry labels the primary sponsor analysis “Mixed Models Analysis” but does not report its random effects, fixed effects, link function, covariates, or treatment of any clustering. It is therefore not possible to identify from the registry record which features the model accounted for.

The secondary sponsor analysis is listed as linear regression. In general, linear regression models a mean outcome as a function of predictors. For a continuous outcome such as a knowledge score, the model can estimate a difference in means, possibly accounting for variables included in the model. The registry does not state which predictors were included or whether the reported p-value came from an adjusted or unadjusted model. No specific adjustment should be assumed.

Why can a simple recalculation differ from a model analysis?

A simple recalculation uses the posted group summaries and compares the groups without adjustment. A model-based analysis can use a different outcome scale, incorporate additional variables, or account for correlation according to its specified structure. Differences in estimates or intervals may therefore reflect the estimand, modelling choices, or analysis population rather than a calculation error. The registry marks no discrepancy for this trial. That does not reveal the model details; it indicates that the posted record does not flag a discrepancy requiring the sponsor result to be displaced.

All statistical methods depend on assumptions. For binary outcomes, the group comparison assumes that the event counts and denominators appropriately represent the analyzed participants. Model-based results additionally depend on the model specification and the handling of missing information. The registry does not report enough detail to assess all these assumptions. The most defensible reading combines the posted counts, the recalculated unadjusted contrast, and the sponsor’s reported model result while keeping their distinct roles clear.

11. Confidence Intervals and Precision

The primary risk-difference estimate is 0.62, and its 95% confidence interval runs from 0.558 to 0.682. The interval lies above zero, which is consistent with a higher screening proportion in the intervention group in the analyzed data. Its limits indicate the range of absolute contrasts compatible with the recalculation’s statistical assumptions. They do not account for possible outcome misclassification, selection, or other sources of bias.

The recalculated relative measures have wider ranges on their respective scales: the risk ratio is 7.159, with limits 5.061 and 10.13, and the odds ratio is 23.05, with limits 14.61 and 36.38. These intervals are not directly comparable to the risk-difference interval because their scales differ. Relative measures can appear numerically large when the control event proportion is low; interpretation should return to the observed group risks and the absolute difference.

The sponsor’s mixed-model odds ratio is 35.8, with a 95% confidence interval from 11.1 to 114.9. This interval is broad on the odds-ratio scale, indicating considerable uncertainty about the precise magnitude of that model-based relative estimate, despite the reported p-value of <0.001. Statistical evidence for a contrast and precision about its exact size are separate considerations. The interval’s lower and upper limits are compatible with effects of substantially different magnitudes.

For knowledge, the mean difference is 0.627 units, with a confidence interval from 0.143 to 1.111. This interval is above zero, but the registry does not describe the scale’s range, interpretation, or a threshold for a meaningful difference. Its statistical precision therefore does not by itself establish the practical importance of the score contrast. The reported standard errors describe uncertainty in the group means, but further information about the scale and analysis would be needed for a fuller interpretation.

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 illustrations based on the observed effect, not a reconstruction of the trial’s original design. For these outcomes, the figures refer to participants per arm, rather than a total number of events. The stated assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, a normal approximation, and detection of the stated fraction of the observed effect. The original trial’s sample-size assumptions and justification are not reported in the registry record.

The 80% power column gives the illustrated participants per arm for a design that would have an 80% probability of detecting the specified effect under the stated assumptions, if that effect and those assumptions were correct. The 90% power column uses a higher target probability of detecting that same specified effect. These are design-planning quantities, not observed results. This is not observed power.

The rows represent planning for 100%, 75%, and 50% of the observed effect. A smaller effect is harder to distinguish from random variation, so the illustration requires more participants per arm as the target effect becomes smaller. This relationship is a general feature of statistical planning, not evidence that the trial was designed around any particular row. The displayed requirements also depend on the stated approximation and allocation assumptions; they should not be read as a recommendation for another study without considering its design, outcome definition, analysis, and target population.

The primary and secondary outcomes have different observed effect scales, and their planning illustrations should be read separately. They are not a report of the number actually analyzed, a measure of the trial’s achieved power, or a substitute for the confidence intervals around the posted estimates. The trial’s own design assumptions are not reported in the registry record.

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 = 0.101; treatment proportion p₁ = 0.721; average p̄ = 0.411.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6957 = 1.3636.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5402 = 0.4547.
4. n = (1.3636 + 0.4547)² / (0.6200)² = 8.6, rounded up to 9 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 <- 0.100671140939597                          # control proportion
p1 <- 0.720689655172414                          # 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(2594826)
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 = 9 (80% power), 11 (90% power)
 75% of observed effect: n per arm = 15 (80% power), 20 (90% power)
 50% of observed effect: n per arm = 30 (80% power), 40 (90% power)
power.prop.test: 8.6 
Simulated power at n = 9 per arm: 0.843 (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

For the primary endpoint, the registry reports the sponsor’s mixed-model odds ratio as 35.8, with a 95% confidence interval from 11.1 to 114.9 and p < 0.001. The simple recalculation from the posted group counts gives an unadjusted risk difference of 0.62, with a 95% confidence interval from 0.558 to 0.682 and p < 0.001. The recalculation also gives unadjusted relative measures, but those should not be treated as identical to the sponsor’s model-based estimate merely because both include an odds ratio.

The sponsor analysis and the recalculation answer related questions using different statistical approaches and, in the primary comparison, different reported scales. The registry does not describe the mixed model’s specification or covariates, so the precise source of any difference between the model-based and summary-data estimates cannot be established. The record does not flag a discrepancy. The sponsor’s result is the trial’s own reported analysis; the recalculation provides a transparent, unadjusted comparison from posted summaries, not a correction or replacement.

For the secondary knowledge endpoint, the sponsor reports a linear regression analysis with p = 0.005. The unadjusted recalculation reports p = 0.011 for the mean difference. The exact regression model and its included variables are not reported in the registry record. The two p-values should therefore be described as results from separate analyses, not as interchangeable estimates.

15. Multiplicity, Missing Data and Analysis Populations

The registry does not report a multiplicity adjustment for the primary and secondary comparisons, a testing hierarchy, or a family-wise error procedure. The secondary outcome is presented with an unadjusted p-value. When several outcomes or analyses are considered, unadjusted p-values do not account for the greater opportunity to observe a low value among the set. This is a general statistical issue, not evidence that the posted finding is false. The registry lists two outcome measures and two analyses, but whether these represent the complete set of outcomes considered is not reported in the registry record.

Missing-data methods are also not reported. The primary outcome entries list 290 participants in the intervention group and 298 in the control group; the secondary entries list 341 and 335. The record does not explain why these outcome-specific group counts differ, whether every randomized participant contributed to each analysis, or how unavailable outcome information was handled. It does not report imputation, a complete-case approach, sensitivity analyses, or assumptions about missingness. No particular method should be inferred.

The analysis populations are not fully characterized. The record does not say whether analyses followed an intention-to-treat principle, included only participants who received the intervention, or used another population definition. It also does not provide a participant flow account detailing post-randomization exclusions, withdrawals, or outcome availability. These omissions matter because a randomized comparison is most readily interpreted when the analyzed groups and handling of missing outcomes are clear.

Details of allocation concealment, outcome ascertainment, protocol adherence, and the statistical analysis plan are not reported in the registry record. The sponsor’s analysis methods are named, but model specifications are not given. These gaps limit independent assessment of risk of bias and reproducibility. They do not negate the posted estimates; they define what can and cannot be concluded from the registry information alone.

16. 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 0% reporting threshold.

The serious adverse event fields report zero serious adverse events in the culturally appropriate intervention group and zero in the general health education control group. The other adverse event fields also report zero in both groups. These are registry-recorded counts for the groups and categories shown. The death fields are null rather than numerical counts, so they do not state a death total.

These entries should be read as reported event data, not as proof that no adverse experiences occurred. The registry does not explain the adverse-event collection period, definitions, monitoring procedures, completeness of reporting, or whether events were actively solicited. A zero count in a posted field can reflect the recorded results under the trial’s reporting process; it does not establish that the intervention was risk-free or that all possible harms were assessed.

The trial interventions were behavioral education and navigation approaches, but the intervention type does not remove the need to interpret safety reporting cautiously. The registry’s serious-event and other-event entries provide limited information about what was recorded. Further details about event descriptions and attribution are not reported in the registry record.

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

The posted primary outcome data show a substantially higher observed proportion of Pap smear receipt in the culturally appropriate intervention group than in the general health education control group at the 12-month assessment. The recalculated absolute contrast is 0.62, with its confidence interval above zero, and the sponsor’s mixed-model analysis also reports p < 0.001. These statistics describe the recorded comparison and its uncertainty under the relevant analyses.

The result does not, by itself, establish that the intervention would produce the same contrast in every setting or population. The trial concerns underserved Korean American women, and the intervention was described as church-based and culturally appropriate. How well the findings apply to other groups, communities, or service settings is not reported. The record does not provide enough detail about participant characteristics, implementation, or recruitment to evaluate transferability fully.

The result also does not identify the contribution of each component. The intervention combined culturally appropriate education focused on cervical cancer with navigation assistance. The trial compared that package with general health education, rather than reporting separate comparisons for education and navigation. The observed contrast cannot therefore show which component, combination, or contextual feature accounts for the difference.

Screening receipt is a process outcome. It does not establish screening quality, diagnostic findings, subsequent treatment, or health outcomes. The knowledge endpoint adds information about a related educational measure, but the registry does not show that knowledge changes mediated screening behavior. The results do not answer those questions.

Finally, statistical significance is not the same as certainty about causation, precision, or practical importance. The confidence intervals quantify statistical uncertainty under their methods, while limitations such as missing-data handling, outcome ascertainment, model specification, and multiplicity remain. Those details are not reported in the registry record. The findings are best interpreted as a strong observed group contrast in the posted screening outcome, with the limits of the available record kept in view.

18. Why This Trial Matters Statistically

This trial illustrates how the scale used to express a binary outcome changes the interpretation. The primary endpoint can be described as a count, a group event proportion, an absolute risk difference, a risk ratio, or an odds ratio. The risk difference communicates the contrast in absolute probability units. Relative measures describe comparison on ratio scales, and an odds ratio is not the same as a risk ratio, especially when the event is common in one group. Reporting the group counts and more than one effect scale helps readers distinguish these questions.

The result also illustrates the difference between a summary-data recalculation and a sponsor’s model-based analysis. The unadjusted risk difference is directly tied to the posted group summaries. The sponsor’s mixed-model odds ratio comes from a named modelling approach, but the record does not report its specification. These results can be informative together without being treated as identical estimands. A clear statistical interpretation should state which analysis produced each estimate and avoid attributing model features that are not documented.

Precision matters alongside statistical significance. The primary interval excludes zero on the risk-difference scale, while the sponsor’s odds-ratio interval spans a broad range of relative estimates. This distinction shows why a small p-value alone is insufficient to communicate how well the effect’s magnitude is known. The secondary knowledge result also demonstrates that statistical evidence and practical meaning are separate: the scale’s interpretation is not reported.

Finally, the record highlights the importance of transparent reporting beyond point estimates. Outcome-specific group counts differ, but the reasons are not stated; missing-data methods, multiplicity handling, and detailed model specifications are also absent. These reporting details shape how confidently readers can reconstruct the analysis. The trial is therefore useful for understanding both the interpretation of randomized comparisons and the limits created when analysis details are not fully reported.

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.

The registry record gives limited detail about how randomization was implemented. Allocation is listed as randomized, but sequence generation, allocation concealment, and any stratification are not reported. Masking is listed as none. The record does not explain how the primary outcome was ascertained or whether outcome assessors were independent of intervention delivery. These omissions constrain assessment of potential bias.

Group counts differ between the primary and secondary outcome entries, and the reasons are not reported. The record does not give a participant flow account, define the analysis populations, or describe missing-data handling. It is therefore not possible to determine from the registry record how the listed denominators relate to all randomized participants or whether missing outcomes could affect the estimates.

The sponsor’s primary analysis is identified as mixed models, but model details are absent. The secondary analysis is identified as linear regression, also without a reported model specification. Covariates, clustering assumptions, and sensitivity analyses are not described. The registry also does not report how multiplicity was addressed or provide a detailed analysis plan. These limitations matter for interpreting the relationship between sponsor analyses and unadjusted recalculations.

The intervention combined culturally appropriate, church-based education with navigation assistance. The record does not provide implementation, attendance, fidelity, or exposure details, so the contribution of individual elements cannot be separated. It also does not report enough participant-characteristic information to assess how representative the analyzed groups were of the broader intended population. The outcome records screening receipt and knowledge, not downstream clinical outcomes. These limitations define the scope of conclusions supported by the posted information.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Cervical Cancer Screening Intervention Among Korean American Women — NCT02594826.
• Registry record last updated 2021-01-26; status last verified 2021-01.
• Results first posted 2016-10-13.
• 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 describes a completed randomized, parallel-group behavioral trial with no masking, comparing a culturally appropriate, church-based cervical cancer intervention with navigation assistance against general health education. The stated enrollment is 744, and the trial is listed for underserved Korean American women.

The primary outcome was the number receiving a Pap smear test at 12 months. The posted counts were 209 of 290 in the intervention group and 30 of 298 in the control group. The unadjusted recalculated risk difference is 0.62, with a 95% confidence interval from 0.558 to 0.682 and p < 0.001. The sponsor’s mixed-model analysis reports an odds ratio of 35.8, with a 95% confidence interval from 11.1 to 114.9 and p < 0.001.

The secondary outcome was knowledge about cervical cancer at 12 months post-program. The registry does not report detailed missing-data, multiplicity, or model-specification information. These omissions limit independent assessment of the analyses and the broader interpretation of the results.