Use of Telehealth In-home Messaging to Improve GI (Gastrointestinal) Endoscopy Completion Rates: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Use of Telehealth In-home Messaging to Improve GI (Gastrointestinal) Endoscopy Completion Rates (NCT00310362), a randomized na study. The primary outcome is Appointment Nonadherence-colonoscopy.

2. Clinical Question

3. Trial Design

This completed study, NCT00310362, was a randomized, parallel-group trial with single masking. The registry lists enrollment as 3610 and identifies the VA Office of Research and Development as the sponsor. The recorded start was 2007-07 and primary completion was 2009-01. The phase is listed as NA. The record does not identify who was masked, or describe the masking procedure, so the single-masking label should not be interpreted more specifically.

Participants were assigned to usual care, IVR3, or IVR7. Usual care involved a nurse telephone call 7 days before the appointment. IVR3 sent an automated reminder 3 days before the appointment, and IVR7 sent one 7 days before it; both IVR approaches also provided preparation education. Thus, the arms varied in the type and timing of reminder. The record does not describe whether other aspects of care were standardized across arms.

The primary measure was colonoscopy appointment nonadherence over 3 months. Three additional outcomes were posted: nonattendance for flexible sigmoidoscopy, preparation nonadherence for colonoscopy, and preparation nonadherence for flexible sigmoidoscopy. Each is recorded as a participant count. The posted outcome results contain arm-level counts and unadjusted pairwise comparisons against usual care. The registry record does not report a separate analysis plan, an analysis population, or the sponsor's inferential analysis. The registry lists zero analysis entries.

The trial is best understood as a comparison of reminder strategies in a VA endoscopy setting. Random allocation supports a comparative interpretation in principle, but interpretation also depends on how outcomes were observed, how missing outcomes were handled, and whether the analysis followed the assigned groups. Those details are not reported in the registry record.

4. Participant Flow and Accounting

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

Overall Study

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 Appointment Nonadherence-colonoscopy, measured over 3 months.

The primary endpoint is appointment nonadherence for colonoscopy, expressed as a count of participants. Its recorded timeframe is 3 months. The registry defines a nonadherent appointment as one that was canceled or not attended. This combines two ways an appointment may fail to proceed; the record does not separate cancellations from failures to attend.

Appointment nonadherence is directly relevant to a trial of reminders because the intervention is intended to influence whether participants complete the scheduled endoscopy visit. It is an operational outcome: it measures attendance behavior, rather than the quality of the examination, the detection of disease, or a later health outcome. The registry's description notes that delayed or missed endoscopy can affect care capacity and access, which gives appointment completion practical importance in the study setting.

The endpoint does not itself establish why an appointment was canceled or missed. Nor does it show whether a participant later completed an endoscopy at another time. Those distinctions are not reported in the registry record. The counts are presented by randomized arm, with comparisons against usual care. Since this is a binary participant-level event summarized as counts, the comparative estimates concern the chance of recorded nonattendance during the stated timeframe, not a time-to-event measure.

The definition supports a clear distinction between appointment nonadherence and preparation nonadherence, which appears as a separate secondary outcome. Those outcomes should not be treated as interchangeable: one concerns whether the appointment was canceled or missed, while the other concerns adherence to procedure preparation.

7. Results Posted in the Registry

• IVR3: 41.9% (333/794) versus 44.1% (348/790) on control; risk difference -2.1 percentage points (95% CI -7.0 to 2.8), risk ratio 0.95 (0.85–1.07); unadjusted two-proportion z-test p = 0.396.
• IVR7: 38.0% (303/797) versus 44.1% (348/790) on control; risk difference -6.0 percentage points (95% CI -10.9 to -1.2), risk ratio 0.86 (0.77–0.97); unadjusted two-proportion z-test p = 0.015.

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.

Nonattendance-flexible Sigmoidoscopy

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Interactive Voice Response 3 Days Prior: 41.1% (169/411) versus 37.9% (155/409) on control; risk difference 3.2 percentage points (95% CI -3.5 to 9.9), risk ratio 1.09 (0.92–1.29); unadjusted two-proportion z-test p = 0.345.
  • Interactive Voice Response 7 Days Prior: 40.1% (164/409) versus 37.9% (155/409) on control; risk difference 2.2 percentage points (95% CI -4.5 to 8.9), risk ratio 1.06 (0.89–1.26); unadjusted two-proportion z-test p = 0.519.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Interactive Voice Response 3 Days Prior and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Interactive Voice Response 7 Days Prior and the control. The interval gives the range of effects that remain compatible with the data.

Preparation Nonadherence-colonoscopy

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Interactive Voice Response 3 Days Prior: 44.0% (349/794) versus 44.1% (348/790) on control; risk difference -0.1 percentage points (95% CI -5.0 to 4.8), risk ratio 1.00 (0.89–1.12); unadjusted two-proportion z-test p = 0.969.
  • Interactive Voice Response 7 Days Prior: 40.0% (319/797) versus 44.1% (348/790) on control; risk difference -4.0 percentage points (95% CI -8.9 to 0.8), risk ratio 0.91 (0.81–1.02); unadjusted two-proportion z-test p = 0.104.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Interactive Voice Response 3 Days Prior and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Interactive Voice Response 7 Days Prior and the control. The interval gives the range of effects that remain compatible with the data.

Preparation Non-adherence-flexible Sigmoidoscopy

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Interactive Voice Response 3 Days Prior: 42.1% (173/411) versus 40.1% (164/409) on control; risk difference 2.0 percentage points (95% CI -4.7 to 8.7), risk ratio 1.05 (0.89–1.24); unadjusted two-proportion z-test p = 0.562.
  • Interactive Voice Response 7 Days Prior: 41.1% (168/409) versus 40.1% (164/409) on control; risk difference 1.0 percentage points (95% CI -5.8 to 7.7), risk ratio 1.02 (0.87–1.21); unadjusted two-proportion z-test p = 0.776.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Interactive Voice Response 3 Days Prior and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Interactive Voice Response 7 Days Prior and the control. The interval gives the range of effects that remain compatible with the data.

The three secondary outcomes cover flexible sigmoidoscopy nonattendance, colonoscopy preparation nonadherence, and flexible sigmoidoscopy preparation nonadherence. Their posted comparisons do not show a consistent pattern of clearly separated risks across both reminder schedules. The intervals for these comparisons include zero, and their unadjusted p-values do not provide compelling evidence of a difference on their own. The detailed endpoint estimates and counts are reported with the outcome results; they should be interpreted in their own clinical context rather than pooled into a single overall effect.

This overall pattern does not contradict the primary IVR7 comparison, because attendance at colonoscopy and preparation adherence are different outcomes, and flexible sigmoidoscopy is a different procedure context. A reminder could relate differently to making an appointment, attending it, or following preparation instructions. The registry record does not report an analysis testing whether effects differ across these outcomes, so one should not infer such a difference from the separate results.

Secondary findings carry less weight when multiple outcomes and comparisons are considered without a reported multiplicity adjustment. In general, examining many hypotheses increases the chance that at least one apparently noteworthy result occurs by chance. The record does not state which secondary outcomes or contrasts were prespecified as confirmatory, nor does it report a correction procedure. The secondary results are therefore best treated as supportive or exploratory evidence, not as independent confirmation of the primary comparison.

The registry lists zero analysis entries. It also does not report a sponsor analysis or a hierarchy that would distinguish confirmatory from exploratory endpoints. Accordingly, the secondary outcome pattern adds context about appointment and preparation behavior, but it cannot resolve the uncertainty about the primary result or establish a broad effect of IVR reminders across all endoscopy-related behaviors.

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

• Nonattendance-flexible Sigmoidoscopy
• Preparation Nonadherence-colonoscopy
• Preparation Non-adherence-flexible Sigmoidoscopy

9. Statistical Interpretation

10. Statistical Methods Explained

These results compare binary outcomes: each participant either had a recorded event, such as a canceled or missed appointment, or did not. Counts are shown alongside the number of participants in each arm. For the primary endpoint, the registry reports the counts 348.0 of 790.0 for usual care, 333.0 of 794.0 for IVR3, and 303.0 of 797.0 for IVR7. The pairwise comparisons use usual care as the reference. They are not direct comparisons between IVR3 and IVR7.

What does a risk difference describe?

In general, a risk difference expresses the difference between the event risks in two groups. A value below zero for an intervention-versus-usual-care contrast means the event risk is lower in the intervention group under the contrast's stated direction; a value above zero means it is higher. The risk difference is an absolute measure, rather than a relative measure. It can be useful for understanding how far apart the group event risks are, but the estimate should be read together with its confidence interval.

The primary comparison estimates are negative for both IVR schedules. The interval for IVR3 spans zero, whereas the interval for IVR7 is below zero and has upper endpoint -0.012. These features describe uncertainty and direction; they do not establish equivalence for IVR3, nor do they remove the need to consider multiplicity and missing information for IVR7.

How should a confidence interval be read?

In general, a confidence interval gives a range of values compatible with the data and the method used to calculate the interval. A wider interval indicates less precision than a narrower interval. When an interval for a risk difference includes zero, the data are compatible with no group difference as well as with differences in either direction, within the interval. When it excludes zero, the interval is inconsistent with a zero difference at the corresponding confidence level, assuming the calculation and analysis are appropriate. The registry does not state the confidence level in its outcome description, so it should not be inferred from the posted interval alone.

For IVR3, the primary risk-difference interval includes zero. For IVR7, the upper endpoint is -0.012, and the interval is entirely negative. The unadjusted p-values are 0.396 and 0.015, respectively. A p-value describes how incompatible the observed data are with a specified null model under the analysis assumptions; it is not the probability that the null hypothesis is true, nor does it measure the size or clinical importance of an effect.

What does “unadjusted” mean?

The comparison values are unadjusted recalculations from the posted summary counts. In general, an unadjusted comparison does not account for covariates or for the fact that multiple outcomes and contrasts may be examined. The registry record lists no analysis entries and does not report a sponsor analysis, covariate-adjustment strategy, or multiplicity procedure. The posted p-values should therefore be read as nominal comparisons rather than as results adjusted for a broader analysis plan. The discrepancy indicator is false, but that does not supply missing details about the trial's own analysis.

Randomization is intended to make assigned groups comparable on average, including for characteristics that may not have been measured. In practice, comparative interpretation also depends on adherence to assigned groups, complete outcome measurement, and the analysis population. The registry record does not report these details. No adjustment for baseline characteristics should be assumed.

Why do binary outcomes and event counts matter?

For a binary outcome, the number of participants with the event is interpreted in relation to the group denominator. Counts alone can be misleading if group sizes differ, so the posted risk comparisons use both counts and denominators. General statistical practice also recognizes that the estimated risk difference has sampling uncertainty, especially when event counts are limited or when the observed difference is small. The confidence interval communicates this uncertainty more directly than the point estimate alone.

The primary endpoint groups cancellation and nonattendance together. This makes it a combined outcome: separate patterns for cancellations and missed visits cannot be recovered from the posted result. Preparation nonadherence is measured separately and should not be folded into the primary endpoint. Likewise, flexible sigmoidoscopy outcomes concern a different procedure context and are not substitutes for the colonoscopy primary endpoint.

The record does not report a model specification, missing-data method, intention-to-treat approach, per-protocol analysis, or prespecified analysis plan. In general, missing outcome data can alter a comparison if the chance of missingness differs between groups or is related to the outcome. Without reporting on outcome completeness and the analysis population, the posted comparisons cannot establish how those issues were handled. The statistical interpretation is consequently limited to the arm counts and the unadjusted comparisons available in the registry record.

11. Confidence Intervals and Precision

Precision is about how tightly the data estimate a group difference, not whether the difference is important. The confidence interval for the primary IVR3 comparison includes zero, so it is compatible with no difference as well as with differences in either direction. Its posted endpoints are -0.070 and 0.028. Because these endpoints contain four decimal places, they are not repeated; the interval's relationship to zero is the key point for interpretation.

The IVR7 primary interval is entirely below zero, with endpoints -0.109 and -0.012. It therefore supports a negative risk difference under the stated comparison direction, while still allowing a range of possible magnitudes. The upper endpoint is relatively close to zero, and the interval is not a single precise value. The corresponding unadjusted p-value is 0.015. For IVR3, the p-value is 0.396. Neither a p-value nor an interval alone establishes the practical importance of a difference.

Confidence intervals also depend on the method used to calculate them. The registry record does not describe the interval method or confidence level. The recalculated intervals should therefore be read as statistical summaries of the posted counts, with that methodological detail unavailable. The posted arm counts are the basis for the comparisons; they do not reveal whether every randomized participant had a fully observed outcome.

For the secondary endpoints, the intervals also include zero across the posted comparisons. This pattern indicates substantial uncertainty about the direction and magnitude of those comparisons. It does not show that all strategies are identical. Distinguishing a small difference from no difference can require more information, but no particular required sample size can be inferred as a trial conclusion from these intervals. The planning illustrations address hypothetical future studies and are not a measure of how conclusive this completed trial was.

12. Planning a Similar Trial: Sample Size

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

The sample-size illustrations are planning calculations based on observed effects, not the trial's original design assumptions. The registry record does not report the original sample-size justification or design assumptions. For these illustrations, the stated assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The target effect is set at the stated fraction of the observed effect: 100%, 75%, or 50%.

For these binary outcomes, the figures are participants per arm, not total participants and not total events. The two power columns describe the planned chance of detecting the target effect under the stated assumptions if that effect is present: 80% power and 90% power. These are planning targets, not guarantees. The 90% column calls for more participants per arm than the 80% column for the same target effect, because a study with more information has a greater chance of detecting that effect under the calculation's assumptions.

Rows for smaller fractions of the observed effect require larger planned group sizes. In general, smaller differences are harder to distinguish from random variation, so a study designed to detect them needs more participants. The calculations are outcome- and comparison-specific; the observed effect and uncertainty differ across endpoints and contrasts. They should not be read as one universal sample-size recommendation for the whole research question.

These values are not observed power. Observed power is not a useful substitute for examining the effect estimate and its uncertainty after a study. The illustrations instead show how future planning changes with a chosen effect size and target power, under the stated simplified assumptions. They do not establish that the completed trial was designed to those targets or that its conduct followed these assumptions.

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 = 0.441; treatment proportion p₁ = 0.419; average p̄ = 0.430.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.7001 = 1.3722.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.7000 = 0.5891.
4. n = (1.3722 + 0.5891)² / (-0.0211)² = 8,631.7, rounded up to 8,632 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 <- 0.440506329113924                          # control proportion
p1 <- 0.419395465994962                          # 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(310362)
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 = 8,632 (80% power), 11,555 (90% power)
 75% of observed effect: n per arm = 15,369 (80% power), 20,575 (90% power)
 50% of observed effect: n per arm = 34,631 (80% power), 46,361 (90% power)
power.prop.test: 8631.7 
Simulated power at n = 8,632 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.

14. Multiplicity, Missing Data and Analysis Populations

The registry record does not report whether the trial used a multiplicity strategy. The posted outcomes include one primary measure and three secondary measures, with two IVR-versus-usual-care comparisons for each. In general, multiple outcomes and comparisons create multiple opportunities to observe a low p-value by chance, even if no true differences exist. A prespecified hierarchy, adjusted testing procedure, or clear distinction between confirmatory and exploratory analyses can help interpret that issue. None is reported, so the p-value of 0.015 for the IVR7 primary comparison is an unadjusted result and should be treated as nominal.

The record also does not report an analysis plan, interim analyses, stopping rules, or a sample-size justification. These details should not be inferred from the posted results. The registry lists zero analysis entries, so it is not possible to compare the outcome results with a documented sponsor analysis or to establish which analyses were intended before the data were examined.

Missing-data handling is not reported in the registry record. The available counts and denominators do not describe whether any outcomes were missing, how missingness was defined, or whether missing participants were included in a particular analysis population. In general, missing outcomes can bias a comparison if the probability of being observed is related to treatment assignment or to the outcome itself. The registry record does not provide enough information to assess that possibility.

The analysis population is also not reported. It is not stated whether participants were analyzed according to randomized assignment, according to the reminder received, or under another rule. The record does not describe crossovers, deviations from assigned reminders, or exclusions from analysis. These omissions matter because randomization protects comparisons most directly when the analysis preserves the assigned groups and outcome data are handled transparently. They limit the strength of any causal interpretation of the posted unadjusted comparisons.

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

The posted results address whether two IVR reminder schedules were associated with colonoscopy appointment nonadherence compared with usual care involving a nurse call. The IVR7 primary comparison has a negative risk-difference interval that excludes zero, with an unadjusted p-value of 0.015. The IVR3 interval includes zero, and its unadjusted p-value is 0.396. These summaries indicate different levels of statistical evidence in the posted comparisons, not proof that one schedule is effective and the other is ineffective.

The result does not establish equivalence between IVR3 and usual care. A non-significant comparison is not evidence that the approaches have the same effect; an equivalence or noninferiority conclusion requires a suitable design, prespecified margin, and analysis, none of which is reported in the registry record. Nor does the record report a direct statistical comparison between IVR3 and IVR7. The difference between their separate p-values cannot itself establish that the schedules differ.

The outcome is appointment nonadherence, combining canceled and unattended colonoscopy appointments. It does not measure cancer detection, cancer incidence, procedure quality, or health outcomes after endoscopy. The registry's background describes concerns about endoscopy completion, clinic wait times, capacity, and costs, but the posted endpoint does not quantify those consequences. It also does not reveal the reasons for cancellation or nonattendance, or whether participants later completed a procedure.

The results should be considered alongside the missing analysis details. The registry does not report a multiplicity procedure, missing-data method, analysis population, or sponsor analysis. These omissions mean the observed comparisons cannot be interpreted as a complete account of the trial's planned statistical evaluation. The secondary outcomes do not provide a consistent, independently confirmed pattern across attendance and preparation measures.

Accordingly, the most supportable interpretation is that the registry's unadjusted primary comparison for IVR7 points toward lower colonoscopy nonattendance than usual care, while the IVR3 comparison remains uncertain. That is a description of the posted statistical evidence, not medical advice or a conclusion that either reminder approach should be adopted.

16. Why This Trial Matters Statistically

This trial is statistically interesting because it compares a new communication method with an active form of usual care. The nurse call included a reminder and an opportunity to ask questions, while IVR posted a reminder and preparation education on a different schedule. The relevant contrast is therefore not simply contact versus no contact. It is a comparison of specific reminder strategies against an existing clinical practice.

The design also separates reminder timing. IVR3 and IVR7 are distinct interventions, and the primary comparisons do not point to the same strength of evidence. Keeping those schedules separate avoids concealing variation behind a pooled estimate that the registry does not report. At the same time, the lack of a direct comparison between the schedules means the results do not establish which timing is preferable.

Endpoint definition matters statistically and practically. By combining canceled appointments and nonattendance, the primary measure offers a straightforward count of appointments that did not proceed as planned, but it cannot show which type of nonadherence contributed to the result. The separate preparation outcomes extend the question beyond attendance, while also increasing the number of results that require cautious interpretation.

The record illustrates why effect estimates, uncertainty, and analysis context need to be read together. The IVR7 comparison has an interval below zero, while the IVR3 comparison includes zero; neither observation should be detached from the unadjusted analysis, unreported multiplicity handling, and unavailable information about missing data and analysis populations. The planning illustrations further show how sample-size requirements depend on the target effect and power, but they cannot replace information about the trial's original design rationale.

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.

Important trial-specific details are absent from the registry record. It does not give eligibility criteria, participant characteristics, recruitment procedures, or the setting beyond its VA context. The meaning of single masking is not specified, and the record does not identify who was masked. These omissions limit assessment of how representative the participants were and how the procedures may have affected outcome recording.

The primary outcome groups canceled appointments with failure to attend and does not present those components separately. Reasons for nonattendance, rescheduling, and later completion are not reported. The record also does not state whether the endpoint was obtained from scheduling records, participant reports, or another source. These gaps restrict interpretation of what the event count represents in practice.

Outcome completeness, exclusions, adherence to assigned reminder schedules, and crossovers are not described. The analysis population and missing-data method are also not reported. As a result, the posted denominators cannot establish how outcome availability or deviations from assigned groups were handled.

The registry provides no reported analysis plan, sponsor inferential analysis, or multiplicity procedure. It does not report the confidence-interval method or confidence level. Secondary outcomes are available, but their role in a prespecified testing hierarchy is unknown. The record lists zero analysis entries.

Finally, the registry lists zero participants at risk and zero affected for serious and other adverse events in each adverse-event group, while deaths are null. Those entries do not provide detailed information about adverse-event collection or ascertainment. They should not be interpreted as a comprehensive safety assessment. These limitations are specific to the information recorded and do not establish that the corresponding trial procedures did not occur.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Use of Telehealth In-home Messaging to Improve GI (Gastrointestinal) Endoscopy Completion Rates — NCT00310362.
• Registry record last updated 2020-06-30; status last verified 2020-06.
• Results first posted 2014-09-25.
• 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

NCT00310362 was a completed, randomized, parallel-group, single-masked study sponsored by the VA Office of Research and Development. Enrollment is listed as 3610; the study start was 2007-07 and primary completion was 2009-01. The registry compares usual care with IVR reminders 3 days and 7 days before endoscopy appointments.

The primary outcome was colonoscopy appointment nonadherence over 3 months, defined as cancellation or nonattendance. The unadjusted IVR7 comparison has a risk-difference interval below zero and p = 0.015; the IVR3 interval includes zero and p = 0.396. Three secondary outcomes address flexible sigmoidoscopy attendance and preparation adherence, and their comparisons are also posted.

The registry record does not report the trial's original sample-size assumptions, analysis plan, missing-data method, or analysis population. Planning illustrations use a two-sided significance level of 0.05, equal allocation, and a normal approximation, and express requirements as participants per arm. Those calculations are planning aids based on observed effects, not observed power.