St. Joe's Invitation Effectiveness Study: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for St. Joe's Invitation Effectiveness Study (NCT04658888), a randomized na study. The primary outcome is Percentage of Cervical Cancer Screening.

2. Clinical Question

The question is about the effect of adding a self-sampling option to a screening reminder, rather than comparing a reminder with no contact. Both groups received a letter intended to notify them that screening was due. The contrast therefore concerns the modified letter and request-card option against the standard letter's recommendation to arrange screening with a primary care provider.

The record describes a randomized, parallel-group study with a planned allocation ratio of 2 to 1, with more recipients assigned to the standard-letter group. The posted group results are 40 participants with screening among 333 in the modified-letter group and 95 among 631 in the standard-letter group. The registry reports the estimated screening proportions as 0.120 and 0.151, respectively. These are descriptive results for the groups as recorded; the registry does not report further characteristics that would show how the groups differed at baseline.

The primary comparison is reported as a risk difference, with a separate unadjusted p-value. The record does not report the detailed analysis population or missing-data rules, so the meaning of the group denominators cannot be elaborated beyond the posted results. The clinical question is thus narrow: how screening completion was recorded over the specified follow-up after these two letter strategies.

3. Trial Design

The St. Joe's Invitation Effectiveness Study is listed as a randomized, parallel-group trial with no masking. Its registry enrollment is 964. The record lists the phase as NA and the status as completed. The study dates recorded are a start date of 2021-04-26 and a primary completion date of 2021-12-01.

Participants identified in the electronic medical record as being out of date with cervical cancer screening were assigned to one of two letter approaches. The standard-letter arm received a reminder that screening was overdue and a recommendation to contact a primary care provider to schedule an appointment. The modified-letter arm received an overdue-screening notification, information on self-sampling, and a card to return to request a kit. The intervention is also named as the Evalyn Self-Sampling Brush.

The study description specifies a 2-to-1 assignment ratio, with two assignments to the control arm for each assignment to the intervention arm. The posted outcome groups contain 631 participants in the standard-letter arm and 333 in the modified-letter arm. These are the reported arm sizes; no additional participant accounting should be inferred from them.

The primary outcome timeframe is three months after the letter is sent. The primary endpoint is described as comparing the percentage of screenings completed in the two groups. The study summary also names method of screening completion and use of the self-sampling kit as outcomes to be monitored over that period. It identifies overall screening rate across people mailed a letter as a secondary endpoint, and the registry lists one secondary outcome.

The record does not report the randomization sequence method, allocation concealment, a masking procedure, or further operational details of follow-up. No original statistical analysis plan or trial-specific sample-size assumptions are reported in the registry record. The planning illustration associated with the posted result uses stated assumptions for planning only; it should not be read as documentation of the trial's original design rationale.

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: No participants who qualified were excluded from the study.

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 Percentage of Cervical Cancer Screening, measured over Three months after the letter is sent..

The primary endpoint is the percentage of cervical cancer screenings completed in the standard-letter group compared with the percentage completed in the modified-letter group. Its timeframe is three months after the letter is sent. The registry reports the outcome as a count of participants with the event, which means that the posted group results describe how many participants in each arm were recorded as completing screening during the endpoint period.

This endpoint is used to address the study's central comparison: whether screening completion differed when the reminder letter included an option to request a self-sampling kit, compared with a standard reminder letter. Counting completion within a defined follow-up period makes the endpoint interpretable as a binary outcome for each participant: screening completion was or was not recorded for that period. The record does not provide more detail about how completion was verified, how screening methods were classified, or how records with incomplete follow-up were handled.

The posted results report 40 participants with screening among 333 in the modified-letter group and 95 among 631 in the standard-letter group. The estimated proportions are reported as 0.120 and 0.151, respectively. These group estimates are the basis for the reported risk difference of -0.030. The registry's broader study description also mentions screening method and self-sampling kit use, but the primary endpoint result alone does not provide a breakdown by screening method or kit use. The registry record does not report a separate numerical result for kit requests or method of completion.

7. Results Posted in the Registry

• Modified Letter With Self-sampling Request Card: 12.0% (40/333) versus 15.1% (95/631) on control; risk difference -3.0 percentage points (95% CI -7.5 to 1.4), risk ratio 0.80 (0.57–1.13); unadjusted two-proportion z-test p = 0.195.

8. Statistical Interpretation

9. Statistical Methods Explained

For a binary endpoint, each participant is classified according to whether the event was recorded during the specified follow-up. Here, the event is completion of cervical cancer screening within three months after the letter was sent. The registry reports counts of participants with the event and group denominators, and also provides estimated proportions. The group proportions are 0.120 for the modified-letter group and 0.151 for the standard-letter group.

What does the risk difference express?

A risk difference subtracts one group's event proportion from the other group's event proportion. The reported comparison is -0.030 for the modified-letter group relative to the standard-letter group. This is an absolute contrast on the proportion scale. Its negative direction reflects the lower estimated proportion in the modified-letter group. Unlike a ratio, a risk difference describes the separation between proportions in their original scale. It can be useful for understanding the size and direction of an absolute contrast, though interpretation still depends on the outcome definition and the study context.

The interval reported for the risk difference runs from -0.075 to 0.014. An interval gives a range of values compatible with the estimate under the method used to construct it. The registry does not state the confidence level or construction method, so those details cannot be specified. Because the interval includes zero, the reported precision does not rule out a zero difference or differences in either direction. This does not prove equivalence; establishing equivalence would require a prespecified margin and an analysis designed for that question, neither of which is reported in the registry record.

How do ratio measures differ?

The risk ratio compares the event proportions by division, while the odds ratio compares the odds of the event between groups. Both are relative measures, so their no-association value is 1. The reported risk ratio is 0.798, with an interval from 0.565 to 1.126. The reported odds ratio is 0.770, with an interval from 0.518 to 1.144. Both estimates are below 1, consistent in direction with the reported risk difference, but both intervals extend across 1. These measures use different scales and should not be treated as interchangeable. The odds ratio in particular is not itself a difference in event proportions.

For a binary outcome, over-dispersion can matter when observations vary more than a model assumes. General explanation: an analysis that assumes less variability than is actually present can understate uncertainty and make evidence appear more precise than it is. The registry does not report whether over-dispersion was assessed or whether a model that could account for it was used. This is relevant as a general statistical consideration, not a claim about the analysis in this trial.

What does the p-value add?

The unadjusted p-value is 0.195. In general, a p-value measures how compatible the observed data are with a specified null hypothesis under a particular statistical model and test procedure. It does not give the probability that an observed pattern occurred by chance, the probability that a treatment effect exists, or the size of an effect. A larger p-value does not show that groups are identical. Interpretation requires the estimate, its interval, the study design, and the analysis assumptions together.

The record labels the p-value as unadjusted. That label indicates that an adjustment for multiplicity is not reflected in this reported p-value, but the registry record does not explain which tests were considered together or whether any other multiplicity procedure was used. The registry lists zero analysis records. This does not establish that no analyses were conducted; it means that analysis details are not posted as registry analysis records.

What can randomization support?

Random assignment is intended to make groups comparable, on average, with respect to factors known and unknown at assignment. General explanation: when allocation is implemented as planned and outcome ascertainment and analysis are appropriate, randomized comparisons can support causal interpretation more strongly than nonrandomized comparisons. The registry lists the study as randomized and specifies a 2-to-1 allocation plan. It does not report the sequence-generation method, concealment of assignments, or enough analysis detail to assess implementation and analysis populations. The result should therefore be described as a randomized group comparison, while avoiding claims that exceed the posted record.

The two arms differ as a package of communication and access options. The modified letter includes information about self-sampling and a request card; the standard letter recommends contacting a primary care provider. A comparison of these arms does not separate the effect of the letter wording, the request card, access to a kit, or the device itself. That would require further design or outcome information not reported in the registry record.

Finally, the endpoint is bounded to the three-month period after mailing. The registry describes the outcome as screening completion, with the method of completion and kit use also of interest. The posted primary comparison gives no method-specific results. General explanation: a binary endpoint can summarize whether an event occurred within a period, but it does not preserve information about timing within that period or what happens later. The record does not report time-to-event analyses or longer-term follow-up results.

10. Confidence Intervals and Precision

The primary risk-difference estimate is -0.030, and its reported interval extends from -0.075 to 0.014. The interval is important because the point estimate alone does not show how uncertain the estimated contrast is. The interval spans zero, so the posted precision is consistent with no difference as well as with a contrast in either direction. Its lower and upper limits also show that values meaningfully different from the point estimate remain compatible with the reported result.

The registry does not state the confidence level or the method used to construct the interval. It is therefore appropriate to describe it as the reported interval rather than assign it a confidence level not stated in the record. The interval should not be interpreted as a range containing a fixed proportion of individual outcomes, nor as the probability that the true effect lies within those limits.

The ratio estimates provide another view of uncertainty. The risk ratio is 0.798, with an interval from 0.565 to 1.126. The odds ratio is 0.770, with an interval from 0.518 to 1.144. Each interval includes 1, the no-association value for a ratio. These ranges leave uncertainty about both the direction and magnitude of the relative contrast.

The unadjusted p-value of 0.195 is not a substitute for the interval. The p-value addresses compatibility with a null model under the test assumptions, whereas the interval communicates a range of effect estimates compatible with the method used. Together, the reported estimate, interval, and p-value indicate that the registry result is not highly precise about the size or direction of the difference. The record does not provide subgroup estimates, adjustment details, or further information that would narrow interpretation.

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 planning illustration is expressed as participants per arm, not as a total number of events. It uses equal allocation between the two arms, a two-sided significance level of 0.05, and a normal approximation. The effect to detect is set to a stated fraction of the observed effect: the illustration includes planning for 100% of the observed effect, planning for 75% of the observed effect, and planning for 50% of the observed effect. The trial's own original design assumptions are not reported in the registry record.

For 100% of the observed effect, the illustration gives 1,983 participants per arm for 80% power and 2,653 participants per arm for 90% power. For 75% of the observed effect, it gives 3,608 participants per arm for 80% power and 4,830 participants per arm for 90% power. For 50% of the observed effect, it gives 8,304 participants per arm for 80% power and 11,117 participants per arm for 90% power.

The 80% power column describes a planning scenario in which a study designed with the stated assumptions would have an 80% chance of detecting the specified effect if that effect were the true one and the assumptions held. The 90% power column uses the corresponding 90% target. These are planning targets, not guarantees that a future study will detect an effect. The values are per arm, so each describes the planned size of each comparison group, rather than a combined total.

Smaller effects require more participants to detect under otherwise unchanged assumptions. The curve should be read as the same general relationship between the effect assumed for planning and the required sample size; it is not evidence about the trial's achieved power or its original enrollment rationale. These calculations are planning illustrations based on the observed effect, not observed power.

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, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.151; treatment proportion p₁ = 0.120; average p̄ = 0.135.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.4838 = 0.9482.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.4833 = 0.4068.
4. n = (0.9482 + 0.4068)² / (-0.0304)² = 1,982.0, rounded up to 1,983 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 <- 0.150554675118859                          # control proportion
p1 <- 0.12012012012012                          # 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(4658888)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 1,983 (80% power), 2,653 (90% power)
 75% of observed effect: n per arm = 3,608 (80% power), 4,830 (90% power)
 50% of observed effect: n per arm = 8,304 (80% power), 11,117 (90% power)
power.prop.test: 1982 
Simulated power at n = 1,983 per arm: 0.798 (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.

13. Multiplicity, Missing Data and Analysis Populations

The registry lists three outcome measures and one posted secondary outcome. The study description identifies the primary endpoint as screening completion by group and describes method of completion and self-sampling kit use as outcomes of interest. Overall screening rate among people mailed a letter is named as a secondary endpoint. The registry does not report a full outcome hierarchy or the detailed analysis plan for all listed measures.

Multiplicity arises when several outcomes, comparisons, or analyses are examined. General explanation: considering many tests can increase the chance of finding at least one apparently unusual result even when no underlying difference exists. Statistical plans can address this through prespecified prioritization or adjustment methods. The registry labels the reported p-value as unadjusted, and no multiplicity adjustment procedure is reported in the registry record. The number of analysis records listed is zero; that does not establish that no analyses were carried out, only that analysis records are not posted.

The record does not report how missing outcomes were handled, whether participants with incomplete follow-up were excluded, or how missing screening status was classified. It also does not report whether the analysis followed an intention-to-treat, per-protocol, or other population definition. Therefore, the posted denominators and counts can be described as recorded, but the treatment of participants without ascertainable outcome status cannot be inferred.

Further details that are not reported in the registry record include the statistical model or test used for the primary comparison, any covariate adjustment, the handling of deviations from assigned letters, and the method for calculating the reported intervals. The record also does not provide a detailed plan for subgroup analyses or an explanation of whether interim analyses were planned or conducted. None of these omissions demonstrates that a method was or was not used; they mark limits on what can be concluded from the registry entry.

These uncertainties matter because analytic choices can affect estimates, standard errors, and interpretation. They do not erase the posted results, but they mean the results are best described at the level the record supports: observed group counts and reported effect estimates for the stated three-month endpoint, with analysis details not fully documented.

14. Safety Results

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

Deaths

Other (non-serious) adverse events

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

The registry's adverse-event entries report zero serious adverse events and zero other adverse events in each arm. The entries list 333 participants at risk for these outcomes in the modified-letter group and 631 in the standard-letter group. Deaths are also reported as zero in each group, and no serious adverse-event terms are listed.

These entries describe what is recorded in the registry under its adverse-event fields. They do not establish that no event of any kind occurred, and they should not be read as a comprehensive assessment of safety beyond the events and reporting represented in the record. The registry does not report further definitions, ascertainment procedures, or details about how adverse events were collected and classified.

The intervention is a letter-based invitation with an option to request a self-sampling kit, while the comparator is a standard letter. The adverse-event figures do not provide information about screening completion or kit use; those are separate outcomes. Nor do the zero entries explain whether all participants had the same observation period for adverse events. The record's primary follow-up timeframe is three months after the letter is sent, but additional safety-follow-up details are not reported in the registry record.

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

The posted result shows that the estimated screening proportion was 0.120 in the modified-letter group and 0.151 in the standard-letter group. The risk-difference estimate for the modified-letter group relative to the standard-letter group is -0.030, with a reported interval from -0.075 to 0.014. The unadjusted p-value is 0.195. These results describe the comparison recorded for screening completion within three months after mailing.

The interval includes zero, and the intervals for the risk ratio and odds ratio include 1. This means the posted estimates do not precisely distinguish a difference in either direction from no difference. It does not demonstrate that the two approaches are equivalent, and it does not establish that the modified-letter approach has no effect. A conclusion of equivalence would require a prespecified margin and an appropriate analysis, neither of which is reported.

The result also does not isolate the effect of self-sampling itself. The modified arm combines a different letter with information about a kit and a request-card option. The standard arm recommends arranging screening with a primary care provider. Without results on kit requests, kit use, or screening method, the primary comparison cannot show which elements were associated with screening completion.

Nor does the three-month outcome describe screening after that period, longer-term follow-up, or outcomes beyond the registry's stated endpoint. The record does not report demographic or clinical subgroup results, missing-data procedures, or detailed analysis methods. As a result, the interpretation should stay close to the posted groups, outcome definition, estimates, and uncertainty. The findings are evidence about this randomized comparison as recorded, not a general statement about all invitation strategies or all screening settings.

16. Why This Trial Matters Statistically

This study addresses a statistical question about an intervention that changes the route by which a screening invitation is acted upon. The comparator is not absence of outreach: both groups receive a letter. The central contrast is whether adding information and an option to request self-sampling is associated with a different recorded screening proportion than a standard reminder. That distinction helps define the estimand and prevents overinterpreting the comparison as the effect of outreach versus no outreach.

The endpoint is binary and time-bounded: screening completion within three months after the letter is sent. Reporting both group proportions and an absolute risk difference makes the scale of the comparison explicit, while the risk ratio and odds ratio offer relative summaries. The measures point in a similar direction, but their intervals include their respective no-association values. The estimates therefore illustrate why statistical interpretation should combine effect size and uncertainty rather than focus on a single p-value.

The design also highlights the importance of separating an intervention package from its components. A modified invitation can affect awareness, willingness to request a kit, and the route used to complete screening. The primary result does not distinguish those pathways. Method-specific and kit-use results would be needed to describe those aspects, and such numerical results are not reported in the registry record.

Finally, the planning illustration makes the uncertainty of planning from an observed effect visible. Its required sample sizes change with the effect size assumed, and it uses equal allocation even though the registry describes an unequal 2-to-1 assignment plan. This is a planning scenario, not a reconstruction of the trial's original design. Together, the design and result show why clear endpoint definitions, complete analysis reporting, and careful distinction between observed estimates and planning assumptions are important in clinical research.

17. Limitations

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

The registry record does not provide participant characteristics or baseline screening details, so it is not possible to describe the composition of the outcome groups or assess their comparability from posted baseline information. It does not report how screening completion was verified, whether all completion methods were captured consistently, or how records with uncertain screening status were treated.

Although the study description identifies kit use and method of completion as outcomes of interest, the posted primary comparison does not provide a numerical breakdown for those outcomes. The record also does not report the number of participants who requested a kit, received one, or used one. As a result, the primary result cannot distinguish the effect of offering self-sampling from the other elements of the modified letter.

The record lists a 2-to-1 allocation plan but does not describe the randomization sequence, allocation concealment, or whether implementation followed that plan. Masking is listed as none. Details about outcome ascertainment, analysis populations, protocol deviations, and missing data are not reported in the registry record. The statistical test, interval construction, adjustment strategy, and original sample-size rationale are also not reported.

The primary outcome is limited to three months after the letter is sent. The registry does not report later screening outcomes or longer-term follow-up. It lists one secondary outcome, but the record does not include a numerical result for that outcome in the posted results. These omissions do not determine what happened; they limit what can be concluded from the information recorded.

The reported comparison is between two complete letter strategies, so it cannot identify the separate contribution of each intervention component. In addition, the planning illustration uses equal allocation and hypothetical effect fractions, while the study description states a 2-to-1 allocation. The illustration is useful for understanding planning assumptions but is not evidence of the trial's original design choices or achieved power.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: St. Joe's Invitation Effectiveness Study — NCT04658888.
• Registry record last updated 2022-07-06; status last verified 2022-06.
• Results first posted 2022-07-06.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

21. Record Summary

The registry lists the St. Joe's Invitation Effectiveness Study as a completed, randomized, parallel-group trial with no masking. It records enrollment of 964, phase NA, a start date of 2021-04-26, and a primary completion date of 2021-12-01. The study concerns cervical cancer screening among female patients identified as out of date with screening at Penn State Health St. Joe's Residency Clinic.

The comparison is a modified reminder letter with self-sampling information and a request-card option versus a standard reminder letter. The primary endpoint is screening completion within three months after the letter is sent. The posted results report 40 participants with screening among 333 in the modified-letter group and 95 among 631 in the standard-letter group. The risk difference is -0.030, with a reported interval from -0.075 to 0.014; the unadjusted p-value is 0.195.

The registry lists one secondary outcome and zero analysis records. Detailed missing-data, analysis-population, and statistical-method information is not reported in the registry record. The planning illustration is based on the observed effect and stated planning assumptions, not the trial's original design rationale.