Effectiveness of Direct-to-Patient Outreach on Colorectal Cancer Screening Within a Low Income and Diverse Population: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Effectiveness of Direct-to-Patient Outreach on Colorectal Cancer Screening Within a Low Income and Diverse Population (NCT01385579), a randomized na study. The primary outcome is Completion of a Colorectal Cancer Screening.

2. Clinical Question

In statistical terms, the central comparison is between the proportion of participants with a recorded screening event in the care manager outreach arm and the proportion in usual care during the defined follow-up period. The question concerns completion recorded in the health record, rather than a direct measurement of screening knowledge, intention, or access. It also does not ask whether any one screening method is superior to another: completion of any of the listed guideline-approved forms qualifies for the endpoint.

Random allocation makes the comparison an experimental one, although the registry record does not report the allocation sequence or implementation details. The registry record also does not report a more detailed eligibility definition, baseline screening status, or subgroup questions. The findings therefore address the enrolled trial population and the defined record-based endpoint; they do not, by themselves, describe every setting or population in which outreach might be used.

3. Trial Design

The registry record describes a randomized, parallel-group trial with no masking. The study was listed as completed, with enrollment of 202. The recorded start was 2010-01 and primary completion was 2010-06. The sponsor was Northwestern University. The phase entry is NA. The registry record does not provide further detail about the sequence generation, concealment of allocation, or the process used to assign participants.

Participants were assigned to usual care or care manager outreach. Usual care permitted providers to refer patients for colorectal cancer screening according to their ordinary health center protocol and practice; it did not include additional preventive care manager outreach. The intervention arm received a mailed letter, screening education, an FOBT kit, and directions for completing and returning the kit. Thus, the contrast is between a bundled outreach program and the existing referral practice, not between receipt and nonreceipt of all screening-related care.

The primary endpoint was assessed within 4 months of initiation of outreach, with the registry record specifying a deadline of June 30, 2010. Completion was defined by documentation in the electronic health record of colonoscopy, sigmoidoscopy, or FOBT. The outcome is binary at the participant level: the record indicates whether qualifying screening was documented within the stated period.

The registry record does not report the trial’s original sample-size assumptions, the planned allocation ratio, or detailed statistical procedures. It also does not report an analysis population definition, handling of missing records, or whether outcome ascertainment was conducted by assessors unaware of assignment. Because masking is listed as none, assignment was not masked according to the registry entry; the record does not explain whether any particular outcome-assessment steps were taken to reduce knowledge of assignment.

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

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 Completion of a Colorectal Cancer Screening, measured over within 4 months of the initiation of outreach (by June 30, 2010).

The primary endpoint is completion of a colorectal cancer screening, counted as participants with documentation of a guideline-approved screening test in the electronic health record. Eligible tests were colonoscopy, sigmoidoscopy, or FOBT. Documentation had to fall within 4 months of the start of outreach, by June 30, 2010. The reported group values are counts of participants meeting that definition, rather than a continuous measure such as a score or time until screening.

This endpoint is relevant to the trial’s stated question because the intervention was intended to prompt patients to complete screening. A record-based completion measure captures whether a qualifying test was documented during the follow-up window. It also gives a clear event definition that can be compared across the randomized groups. The registry record does not specify whether screening completed outside the health system could be captured, or how incomplete or unavailable electronic records were treated.

The endpoint combines three screening methods into one outcome. It therefore reflects documented completion of any listed method, not the use of a particular test or the quality of the screening process. The record does not report results separately by method. Nor does it report a clinical outcome such as cancer detection or mortality. The primary endpoint should consequently be interpreted as a short-term screening-completion measure, as defined by the registry record, rather than as a measure of longer-term health outcomes.

7. Results Posted in the Registry

• Care Manager Outreach: 29.8% (31/104) versus 5.1% (5/98) on control; risk difference 24.7 percentage points (95% CI 14.9 to 34.5), risk ratio 5.84 (2.37–14.42); unadjusted two-proportion z-test p = <0.001.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is binary: each participant either had a qualifying screening documented within the follow-up window or did not. The reported arm values are event counts, and the arm sizes give the denominators for the comparison. For a binary endpoint, the event probability in each arm can be summarized as a risk, and the difference between those risks is a risk difference. The registry record reports the comparison as a risk difference, alongside relative measures and an unadjusted p-value.

What does the risk difference express?

As general statistical context, a risk difference subtracts the event risk in one group from the event risk in the other. A value above zero for the stated comparison means a higher observed event risk in the intervention arm. The measure is expressed on the probability scale and is an absolute contrast, unlike a ratio. The registry record reports a positive risk-difference estimate and an interval that remains above zero. Its exact estimate and interval are not repeated because the recorded values have more than three decimal places.

An absolute measure can be useful because it directly describes the scale of the difference in event probability. Its interpretation still depends on the endpoint definition, follow-up period, and population being studied. Here, the event is record-documented completion of any of three listed screening methods within the specified period, rather than completion of a single test or a later clinical outcome.

How do relative measures differ?

As general statistical context, a risk ratio divides the event risk in one group by the event risk in the other. A value above one indicates a higher observed risk in the numerator group. The reported risk ratio is 5.842, with an interval from 2.367 to 14.42. An odds ratio compares the odds of an event, where odds are the event probability divided by the probability of no event. The reported odds ratio is 7.899, with an interval from 2.926 to 21.32. Both reported intervals lie above the null value of one, but the measures use different scales and should not be read as the same quantity.

The odds ratio can appear farther from one than the risk ratio when the event is not rare. For this reason, the risk ratio may often be more directly intuitive when discussing event probabilities. Neither relative measure conveys the absolute difference on its own; the risk-difference framework answers a distinct question. The registry record reports all these comparison summaries, and they should be considered together with the event counts and the operational endpoint definition.

What does the p-value tell us?

As general statistical context, a p-value is calculated under a model that specifies no group difference and other assumptions relevant to the test. It describes how often results at least as incompatible with that model might arise from random variation if the model were correct. It is not the probability that the trial result occurred by chance, is not a measure of effect magnitude, and does not prove that every assumption is satisfied. The posted unadjusted p-value is < 0.001. “Unadjusted” indicates that the reported value is not described as adjusted for multiple comparisons; the registry record does not report additional details about the test procedure.

Randomization is intended, in general, to make assignment independent of participant characteristics on average. It supports a comparison between assigned groups, but it does not guarantee identical characteristics in a particular trial. The registry record does not provide baseline characteristics or details of allocation implementation, so the extent of balance cannot be evaluated from the record. Likewise, the comparison is based on the outcome as documented; random assignment does not remove possible limitations in outcome capture.

What assumptions matter for this endpoint?

As general statistical context, a binary-outcome comparison relies on correctly identifying events and denominators and on an appropriate analysis for the design. The record defines a qualifying event and reports the arm counts and sample sizes. It does not report how missing or inaccessible records were handled, whether all randomized participants were analyzed in their assigned groups, or which exact inferential procedure generated the reported comparison. Those details are not reported in the registry record and should not be inferred from the displayed estimates alone.

The groups were assigned in a parallel design, so participants belong to one assigned arm rather than crossing between conditions in a sequence. The record lists no masking. In general, knowledge of assignment can affect behavior or documentation, particularly when the endpoint depends on health-record documentation and the intervention itself changes contact with patients. This observation is a general methodological consideration, not evidence that such bias occurred in this trial.

Finally, a statistically clear difference in the defined endpoint does not establish a difference in outcomes that were not measured or reported. The endpoint captures short-term, documented screening completion. It does not separately identify method-specific completion, long-term screening, cancer detection, or mortality. The registry record lists no secondary posted outcomes and no analyses, so additional outcome claims or analyses should not be attributed to the trial record.

10. Confidence Intervals and Precision

The registry record reports a confidence interval for the risk difference, as well as intervals for the risk ratio and odds ratio. Confidence intervals are statistical ranges constructed by a method that, under repeated use and its assumptions, would cover the underlying parameter at the stated confidence level. The confidence level itself is not reported in the registry record. These intervals describe sampling uncertainty under the analysis model; they do not account automatically for measurement problems, missing outcomes, selection, or departures from model assumptions.

For the risk difference, the reported interval lies entirely above zero. For the risk ratio and odds ratio, the reported intervals lie entirely above one, their respective no-difference values. This means the reported intervals exclude the corresponding null values. The risk ratio interval ranges from 2.367 to 14.42, and the odds ratio interval ranges from 2.926 to 21.32. The registry record’s risk-difference values are not repeated because they have more than three decimal places.

The relative-effect intervals span a broad range of values, so their point estimates should not be mistaken for highly precise estimates of the magnitude. The interval for an effect can be statistically separated from its null value while still including materially different effect sizes. The event counts are 5 in usual care and 31 in outreach; the amount of information about a binary outcome depends in part on how many events are observed. No extra precision should be inferred from the small p-value alone.

Confidence intervals also do not describe how likely the true effect is to fall at any particular point within the interval. They are not a guarantee that the effect in another population or setting will fall within the same range. The registry record does not report sensitivity analyses or alternate models that would show how robust the estimates are to different assumptions.

11. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The sample-size figures are planning illustrations based on the observed effect, rather than a record of the original trial’s design calculations. They are participants per arm, not total participants across arms or total events. The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The effect targeted in each scenario is a stated fraction of the observed effect.

The 80% power column gives the number of participants per arm estimated to provide an 80% probability of detecting the specified effect under those planning assumptions, if that effect and the assumptions were correct. The 90% power column uses a higher target probability of detecting that effect. Power is a design property calculated for a specified effect and analysis framework; it is not a guarantee that a future study will detect an effect.

The planning entries are 36 participants per arm for 80% power and 48 for 90% power when planning for 100% of the observed effect. For 75% of the observed effect, the corresponding figures are 56 and 74 participants per arm. For 50% of the observed effect, they are 102 and 136 participants per arm. Smaller effects need more participants for the same target power under the stated assumptions, because a smaller difference is harder to distinguish from random variation.

These figures are illustrative, not observed power. The registry record does not report the original sample-size justification or its own planning assumptions, so the illustrations should not be attributed to the trial’s design team. They also rely on the observed effect as a reference; an observed estimate can be an uncertain basis for planning another study.

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 = 0.051; treatment proportion p₁ = 0.298; average p̄ = 0.175.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.5368 = 1.0521.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5076 = 0.4272.
4. n = (1.0521 + 0.4272)² / (0.2471)² = 35.9, rounded up to 36 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 <- 0.0510204081632653                          # control proportion
p1 <- 0.298076923076923                          # 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(1385579)
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 = 36 (80% power), 48 (90% power)
 75% of observed effect: n per arm = 56 (80% power), 74 (90% power)
 50% of observed effect: n per arm = 102 (80% power), 136 (90% power)
power.prop.test: 35.9 
Simulated power at n = 36 per arm: 0.842 (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 record lists one outcome measure and no posted secondary outcomes. It lists zero analyses, while also reporting a primary comparison and its summary statistics. The record does not describe a prespecified statistical analysis plan, any family of hypotheses, or a multiplicity adjustment. It is therefore not possible to determine from the record whether the reported comparison was the only planned test or whether additional analyses were conducted but not posted.

As general statistical context, testing multiple outcomes or comparisons can increase the chance of at least one apparently unusual result if each test is evaluated without accounting for the wider set of tests. Adjustment methods can address that issue in some settings, but the record does not report whether any adjustment was planned or used. The posted p-value is described as unadjusted, so it should be read as the reported test result rather than as evidence of a multiplicity-adjusted conclusion.

Missing-data procedures are not reported in the registry record. The record gives arm sizes and the counts meeting the endpoint, but it does not say whether every assigned participant had complete outcome ascertainment, whether missing electronic records were treated as no screening, or whether any imputation or sensitivity analysis was performed. A denominator and an event count do not, by themselves, explain the handling of missingness.

The analysis population is also not reported in the registry record. In particular, the record does not state whether the comparison follows an intention-to-treat approach, whether participants were excluded after randomization, or how deviations from the assigned intervention were treated. These omissions limit evaluation of how the reported comparison relates to the randomized assignment. No assumptions about the handling of missing data, protocol deviations, or analysis populations should be inferred.

14. 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 1% reporting threshold.

The registry record reports zero serious adverse events among 98 participants at risk in usual care and zero among 104 at risk in care manager outreach. It also reports zero other adverse events in each arm, with the same at-risk counts. These are the adverse-event figures recorded for the groups; they do not establish that no events of any kind occurred outside the registry’s definitions or reporting process.

The death entries are null, which means death counts are not reported in the registry record. A null entry should not be interpreted as zero deaths. The record lists no serious adverse-event terms, but it does not provide definitions, collection procedures, follow-up details specific to safety, or an explanation of how missing safety information was handled.

As general statistical context, an observed count of zero does not prove that the underlying event risk is zero. The amount of information about uncommon events can be limited, particularly when event counts are absent. These figures should therefore be read as reported counts within the registry record, not as a comprehensive assessment of safety or as a basis for medical advice.

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

The main statistical result is a difference in documented screening completion between the randomized arms during the defined follow-up period. The counts were 5 in usual care and 31 in care manager outreach, and the reported unadjusted p-value is < 0.001. The risk-difference interval lies above zero; the relative-effect intervals also lie above their null value of one. These summaries describe the posted comparison for the registry-defined endpoint.

The outcome is a record of completion, not a direct measure of whether a participant received or understood the mailed materials. The registry definition accepts colonoscopy, sigmoidoscopy, or FOBT, and no method-specific results are reported. Nor are longer-term outcomes reported. The result should not be extended to cancer diagnosis, cancer prevention, mortality, or the quality of screening, none of which is the primary endpoint described in the record.

The intervention was a bundle of contact, education, a mailed FOBT kit, and return instructions. The comparison cannot identify which element contributed to the observed difference. It also compares the bundle with usual care that could include provider referral, rather than with an absence of screening-related care. The registry record does not report intervention adherence or detailed usual-care activity.

The trial was randomized, which is an important design feature for comparing assigned groups, but the available record does not document all procedures needed to judge implementation, missingness, or analysis choices. The outcome depends on electronic health record documentation, and the record does not explain capture of screening performed elsewhere. These issues qualify interpretation without showing that a particular bias occurred.

Accordingly, the posted data support a narrow description: in this trial, more participants assigned to care manager outreach had qualifying screening documented within the specified period than participants assigned to usual care. The result does not establish the same effect in other populations, health systems, or follow-up periods, and it does not prescribe what any individual should do.

16. Why This Trial Matters Statistically

This trial is statistically notable because it evaluates a practical outreach strategy using a randomized comparison and a concrete binary endpoint. The endpoint is not a broad self-reported intention or attitude: it is documentation of a qualifying screening test in the electronic health record within a defined period. That makes the outcome directly aligned with the study’s stated question about screening completion, while also making record capture an important part of interpretation.

The trial compares additional direct-to-patient contact with usual care rather than comparing screening with no screening. This is a meaningful distinction for estimating the incremental association of outreach in a setting where providers may already refer patients. It also means the result applies to a bundled program; the separate effects of a letter, educational material, a kit, and instructions are not identified.

For biostatistical reading, the trial illustrates why event counts, absolute contrasts, relative measures, and uncertainty should be considered together. The reported risk difference describes an absolute comparison; the risk ratio and odds ratio give relative comparisons on different scales. Their confidence intervals communicate uncertainty beyond the point estimates, while the p-value addresses compatibility with a no-difference model rather than effect importance.

The registry record also illustrates the limits of interpreting a result from incomplete reporting. Missing-data handling, analysis populations, original design assumptions, and detailed analysis procedures are not reported in the registry record. The randomized design is valuable, but evaluating how the analysis preserved that design requires information beyond the group counts and summary comparison.

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.

Several trial-specific limitations follow from what the registry record does and does not describe. The primary outcome relies on electronic health record documentation. The record does not say whether screening conducted outside the health system could be captured, or how incomplete records were treated. Thus, the recorded endpoint may not fully describe all screening activity, although the registry entry does not establish that under-recording occurred.

The intervention is a bundle, including a letter, educational information, an FOBT kit, and instructions. Because the comparison evaluates the package, it does not identify the effect of each component separately. The record also does not report whether the mailed materials were received, whether participants returned kits, or the extent of engagement with outreach. Such implementation details are not reported in the registry record.

The record lists no masking. It does not explain whether outcome review or data extraction was conducted without knowledge of assignment. Since the outcome is documented completion and the intervention includes direct contact, knowledge of assignment is a methodological consideration, but the registry record does not establish that it altered outcome recording.

The population is described broadly as low income and racially and ethnically diverse, but detailed eligibility criteria and participant characteristics are not reported. This limits assessment of how closely the enrolled population resembles other settings. The record also does not report subgroup results, method-specific screening results, or secondary outcomes.

Finally, the registry record does not report the original sample-size justification, missing-data methods, analysis population, or detailed statistical plan. These gaps limit evaluation of the analysis and its robustness. The follow-up is defined as within 4 months, so the record does not address persistence of screening behavior or longer-term clinical outcomes.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Effectiveness of Direct-to-Patient Outreach on Colorectal Cancer Screening Within a Low Income and Diverse Population — NCT01385579.
• Registry record last updated 2014-01-06; status last verified 2013-11.
• Results first posted 2014-01-06.
• 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

NCT01385579 is a completed randomized, parallel-group trial sponsored by Northwestern University. It compared usual care with care manager outreach among a low-income and racially and ethnically diverse population. The intervention involved a mailed letter, educational information, an FOBT kit, and instructions. The registry record lists no masking.

The primary endpoint was documented completion of colonoscopy, sigmoidoscopy, or FOBT within 4 months of outreach initiation, by June 30, 2010. The posted counts were 5 of 98 in usual care and 31 of 104 in outreach. The reported unadjusted p-value was < 0.001, with a positive risk-difference comparison and relative-effect intervals above one.

The record reports zero serious and zero other adverse events in both arms; deaths are not reported. It lists no secondary posted outcomes and does not report details on missing-data handling, analysis populations, or the original design assumptions. Interpretation is therefore limited to the reported short-term, record-documented screening outcome and comparison.