Effect of Varied Outreach Methods on Appointment Scheduling and Appointment Completion Using MyChart: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Effect of Varied Outreach Methods on Appointment Scheduling and Appointment Completion Using MyChart (NCT04994691), a randomized na study. The primary outcome is Number of Participants Who Completed Well Child Check Visit.

2. Clinical Question

The clinical question concerns outreach and use of scheduled preventive care. The trial record describes reminders directed to children overdue for a WCC, rather than a comparison of clinical treatments. The primary outcome is based on documentation in the electronic health record, so the result describes recorded visit completion under that definition. Appointment scheduling was a secondary outcome, and receipt of a first COVID-19 vaccination was also listed as a secondary outcome. These outcomes address different steps or events and should not be treated as interchangeable with the primary measure.

The record identifies the trial as randomized, parallel, and triple masked. It does not report which roles were masked, how assignment was generated or concealed, or whether participants and families were aware of their message assignment. Those details are not reported in the registry record. The result therefore speaks to the assigned outreach strategies and the recorded outcomes, while interpretation of implementation and participant awareness depends on information the record does not give.

3. Trial Design

The registry record describes a randomized, parallel trial with triple masking and an enrollment of 945. The study is marked completed, and its phase is listed as not applicable. Children were eligible as described in the objective if they were age 6-17 years, due for a WCC visit, and lacked a visit scheduled in the next 45 days. The record names three outreach conditions: no message, a standard MyChart message, and a tailored MyChart message.

The no-message arm received no message. The standard-message arm received 2 MyChart reminders stating that the child was overdue for a well-child check. The tailored-message arm received 2 such reminders with the date of the last WCC and the child’s age added. This is a comparison of message strategies, not a comparison of different clinical services. The record does not report how long families had access to schedule after receiving a reminder beyond the outcome timeframe.

The primary outcome was assessed within 8 weeks of intervention, or within 8 weeks of randomization for the control group. It counted participants with a completed WCC, defined by the presence of well-visit billing-code documentation in the electronic health record. The secondary scheduling outcome used the same general timeframe. The vaccination outcome was assessed within 8 weeks of intervention or control-group randomization, with its wording referring to receipt of a first COVID-19 vaccination.

The record lists the start date as 2021-07-30 and primary completion as 2021-10-07. It identifies Children’s Hospital Medical Center, Cincinnati as sponsor. Randomization procedures, masking roles, the original sample-size justification, interim analyses, and the prespecified statistical analysis plan are not reported in the registry record. The record also does not describe participant flow or explain the different denominators posted for the vaccination outcome. Those gaps limit assessment of how the planned design was carried out and how the posted analyses relate to all randomized participants.

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 Number of Participants Who Completed Well Child Check Visit, measured over Within 8 weeks of the date of intervention or the date of randomization for the control group.

The primary endpoint is the number of participants who completed a well-child check visit within 8 weeks of the intervention date, or within 8 weeks of randomization for the no-message control group. A participant counted as having completed the visit if a well-visit billing code was documented in the electronic health record. The posted counts are 40 of 315 in the no-message arm, 76 of 315 in the standard-message arm, and 61 of 315 in the tailored-message arm.

This endpoint is used because it captures whether an overdue preventive-care visit was completed during a defined follow-up window. The electronic health record billing-code definition makes the event operational: it is based on documentation rather than participant recall. In general, a clearly defined endpoint helps make comparisons between assigned groups interpretable. However, the measure reflects recorded completion under the registry’s definition. It does not describe the content or quality of the visit, whether a family attempted to schedule but could not attend, or any health outcome after the visit.

The outcome is binary at the participant level: each person either has a documented completion event during the specified period or does not. The reported result is therefore a count of participants with an event, accompanied by the arm denominator. A count alone does not convey uncertainty about the underlying event rate; comparisons and confidence intervals help describe the observed contrast and its precision. The endpoint timeframe also matters: it supports an interpretation about recorded WCC completion during that window, not about later visits or longer-term care patterns.

7. Results Posted in the Registry

• No Message: 12.7% (40/315) versus 24.1% (76/315) on control; risk difference -11.4 percentage points (95% CI -17.4 to -5.4), risk ratio 0.53 (0.37–0.75); unadjusted two-proportion z-test p = <0.001.
• Tailored Message: 19.4% (61/315) versus 24.1% (76/315) on control; risk difference -4.8 percentage points (95% CI -11.2 to 1.7), risk ratio 0.80 (0.60–1.08); unadjusted two-proportion z-test p = 0.147.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Number of Participants Who Scheduled an Appointment for a Well Child Check Visit.

• 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:
  • No Message: 9.5% (30/315) versus 18.4% (58/315) on control; risk difference -8.9 percentage points (95% CI -14.3 to -3.5), risk ratio 0.52 (0.34–0.78); unadjusted two-proportion z-test p = 0.001.
  • Tailored Message: 14.9% (47/315) versus 18.4% (58/315) on control; risk difference -3.5 percentage points (95% CI -9.3 to 2.3), risk ratio 0.81 (0.57–1.15); unadjusted two-proportion z-test p = 0.240.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically lower for No Message than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Tailored Message and the control. The interval gives the range of effects that remain compatible with the data.

Number of Participants Who Received Their First COVID-19 Vaccination

• 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:
  • No Message: 3.7% (3/82) versus 16.7% (14/84) on control; risk difference -13.0 percentage points (95% CI -22.0 to -4.1), risk ratio 0.22 (0.07–0.74); unadjusted two-proportion z-test p = 0.006.
  • Tailored Message: 4.8% (4/83) versus 16.7% (14/84) on control; risk difference -11.8 percentage points (95% CI -21.1 to -2.6), risk ratio 0.29 (0.10–0.84); unadjusted two-proportion z-test p = 0.014.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically lower for No Message than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically lower for Tailored Message than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The registry lists appointment scheduling for a WCC and receipt of a first COVID-19 vaccination as secondary outcomes, both assessed within 8 weeks under the stated intervention or control timing. The posted results show an overall pattern in which the standard-message arm has more recorded events than the no-message arm for these secondary outcomes. The tailored-message arm lies between those arms in the scheduling outcome. The vaccination outcome has different posted denominators across arms, and the reason is not reported in the registry record.

These secondary outcomes provide context about whether outreach may relate to steps that precede or accompany preventive care, but they do not replace the primary endpoint. Scheduling is not the same as completion: a scheduled visit may not be attended, while a completed visit is captured by the primary endpoint’s billing-code definition. Vaccination is a separate event and does not establish WCC completion. Any relationship between these measures is not reported, and the record does not show whether a participant’s secondary event was associated with a primary event.

The posted secondary analyses include unadjusted recalculations and, for scheduling, sponsor-reported logistic-regression results. The vaccination outcome has no sponsor-reported analyses listed. Results from secondary outcomes generally carry less weight than a prespecified primary outcome because they are additional questions and can be more vulnerable to chance findings when many comparisons are examined. The registry does not report whether a multiplicity adjustment was applied or how these outcomes were prioritized in the analysis plan. Thus, secondary patterns are useful as exploratory context, not as independent confirmation of the primary result.

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

• Number of Participants Who Scheduled an Appointment for a Well Child Check Visit.
• Number of Participants Who Received Their First COVID-19 Vaccination

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is a binary participant-level event: a WCC completion documented by a well-visit billing code within the stated timeframe. The posted arm counts and denominators allow an unadjusted comparison of event proportions. General statistical explanation: an event proportion expresses how often the defined event occurred within an arm, while a contrast between arms describes how those observed proportions differ. The recalculated summaries are explicitly unadjusted and use the posted summary counts.

What does a risk difference describe?

A risk difference subtracts one arm’s event proportion from the other arm’s event proportion. Its sign depends on the order of subtraction, so the group order must be kept clear. For the no-message versus standard-message recalculation, the estimate is -0.114: the no-message proportion is lower in the observed counts. For tailored message versus standard message, the estimate is -0.048. A risk difference is expressed on the absolute-proportion scale, which makes the size of the difference interpretable in relation to the event itself. It is not an adjusted model estimate.

How does a risk ratio differ?

A risk ratio compares event proportions by division rather than subtraction. The registry reports risk ratios from logistic regression for the primary group contrasts. A ratio is relative to the chosen reference and its direction depends on which arm is placed in the numerator. The record lists the group pairs and the risk-ratio values, but interpretation should retain that ordering rather than assuming an orientation from the value alone. A ratio and a risk difference answer related but distinct questions: one is relative, while the other is absolute.

What does logistic regression contribute?

Logistic regression is a general method for modeling a binary outcome. The registry identifies logistic regression as the method for its reported primary analyses and labels the reported parameter as a risk ratio. The record does not describe covariates, model specification, adjustment variables, handling of clustering, or how the reported ratio was obtained from the model. Those features are not reported in the registry record, so the posted method label does not permit a fuller reconstruction of the analysis.

Odds ratios also appear in the recalculated summaries. An odds ratio compares the odds of an event between arms; odds are not the same quantity as event proportions. For a common event, the odds ratio can differ meaningfully in magnitude from a risk ratio. General explanation: choosing a measure affects how an effect is expressed, not the underlying observed counts. The displayed risk differences, risk ratios, and odds ratios should therefore not be treated as interchangeable numerical versions of one quantity.

How should confidence intervals and p-values be read?

A confidence interval describes the range of estimates compatible with the data and the method under repeated-sampling assumptions. A narrower interval indicates greater precision for the corresponding estimate, while a wider interval indicates more uncertainty. For the recalculated tailored-versus-standard primary contrast, the interval runs from -0.112 to 0.017 and includes zero. General explanation: when a risk-difference interval includes zero, the data are compatible with no absolute difference under that calculation. This does not prove the arms are equivalent or that the true difference is exactly zero.

A p-value summarizes how unusual data at least as inconsistent with a specified null hypothesis would be under that hypothesis and the analysis assumptions. It is not the probability that the null hypothesis is true, nor a measure of clinical importance. The recalculated p-values are unadjusted. The registry reports p = 0.000 for one primary contrast; this is the registry’s displayed value, not a claim that a probability is literally zero. The record does not state whether its analyses account for multiple comparisons.

What is the distinction between the posted and recalculated analyses?

The recalculated estimates use the posted group counts and denominators as simple, unadjusted summaries. The registry’s own primary analyses use logistic regression and report risk ratios with confidence intervals and p-values. These approaches can differ because they may use different measures, model structure, or analytic choices. The registry record does not explain the observed discrepancy. Because that discrepancy is flagged, the sponsor’s own analysis takes precedence for interpretation of the trial’s inferential findings; the recalculations are descriptive checks rather than replacements.

The registry lists a randomized parallel design, but the method used to generate or conceal assignment is not reported. It also does not report the original analysis plan, missing-data approach, analysis population, or interim-analysis procedures. These are not details that can be inferred from the posted counts or from the fact that randomization and triple masking are listed. General statistical principles can clarify what the reported measures mean, but they cannot fill in unreported trial procedures.

11. Confidence Intervals and Precision

Confidence intervals convey uncertainty around an estimated contrast. The registry reports 95% confidence intervals for its logistic-regression risk ratios, while the recalculated risk differences also have intervals. General explanation: under the relevant statistical method and assumptions, a confidence interval is designed to capture the underlying parameter at its stated rate over repeated studies. It is not a statement that there is a 95% probability that this particular interval contains the fixed true value.

For the unadjusted primary comparison of no message with standard message, the recalculated risk difference is -0.114 and its interval extends from -0.174 to -0.054. The interval stays below zero, consistent with a lower observed completion proportion in the no-message arm under the specified subtraction order. For tailored message versus standard message, the estimate is -0.048 and the interval extends from -0.112 to 0.017. Because the interval crosses zero, the recalculated data allow a range of contrasts that includes no absolute difference. The interval also allows a difference in either direction; the point estimate alone should not be treated as conclusive.

The sponsor’s model-based estimates have their own intervals: the registry reports 1.92 (95% CI 1.38 to 2.60) for the no-message and standard-message comparison, 1.52 (95% CI 1.06 to 2.13) for no message and tailored message, and 1.26 (95% CI 0.94 to 1.66) for standard message and tailored message. These are risk-ratio intervals, not risk-difference intervals, and the group ordering in the record matters. The interval for the standard-message and tailored-message comparison includes 1, the null value for a ratio. This is consistent with its reported p = 0.124, without implying equivalence.

Precision is not the same as importance. A confidence interval can be statistically informative without establishing whether a difference matters to families or clinical practice. The record does not provide a prespecified threshold for a meaningful difference. The discrepancy between recalculated and sponsor analyses also means that the interval from one approach should not be substituted for the interval from the other. Both uncertainty and analysis provenance matter when interpreting these results.

12. Planning a Similar Trial: Sample Size

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

The planning figures are participants per arm for the binary outcomes in this trial, not total events. They illustrate the number of participants that would be needed under stated assumptions to detect a selected fraction of the observed effect. For example, rows describe planning for 100%, 75%, or 50% of the observed effect; the columns describe participants per arm for 80% power and participants per arm for 90% power.

General explanation: power is the probability that a specified analysis will detect an effect of a particular size when that effect is present, under the planning assumptions. The 80% and 90% columns therefore represent different target probabilities of detecting the chosen effect. The higher target generally requires a larger planned sample. Smaller effects are harder to distinguish from random variation, so planning for a smaller fraction of the observed effect requires more participants. The figures should be read as planning illustrations, not as a claim about what the trial was originally designed to detect.

The stated calculation assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation; the effect to detect is the stated fraction of the observed effect. The record notes that the planning illustration is computed from the observed effect rather than from the trial’s own design. The trial’s original sample-size assumptions and justification are not reported in the registry record. These planning figures are not observed power.

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 = 0.241; treatment proportion p₁ = 0.127; average p̄ = 0.184.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.5481 = 1.0743.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5421 = 0.4563.
4. n = (1.0743 + 0.4563)² / (-0.1143)² = 179.4, rounded up to 180 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 <- 0.241269841269841                          # control proportion
p1 <- 0.126984126984127                          # 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(4994691)
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 = 180 (80% power), 240 (90% power)
 75% of observed effect: n per arm = 339 (80% power), 453 (90% power)
 50% of observed effect: n per arm = 804 (80% power), 1,076 (90% power)
power.prop.test: 179.4 
Simulated power at n = 180 per arm: 0.804 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Analyses Reported by the Sponsor

The registry reports sponsor analyses using logistic regression, with risk ratios, 95% confidence intervals, and p-values. For the no-message and standard-message groups, the reported risk ratio is 1.92 (95% CI 1.38 to 2.60; p = 0.000). For no message and tailored message, it is 1.52 (95% CI 1.06 to 2.13; p = 0.023). For standard message and tailored message, it is 1.26 (95% CI 0.94 to 1.66; p = 0.124). The contrasts are identified by the group pairs and ordering in the registry record.

The recalculated results are simple unadjusted contrasts from the posted counts and denominators, expressed as risk differences, with corresponding unadjusted uncertainty measures. These are not the same analytic quantity or method as the registry’s logistic-regression risk ratios. The primary recalculation for no message versus standard message has an interval below zero, while the tailored-versus-standard recalculation has an interval that includes zero. The existence of a discrepancy between the recalculated summaries and sponsor analyses is flagged in the record, but its reason is not reported.

Because the discrepancy is true, the sponsor’s own analysis takes precedence when interpreting the trial’s reported inferential results. The recalculations can help readers see what the posted raw counts imply under a simple unadjusted comparison, but they should not be used to replace the sponsor’s model-based results or to claim that the registry analysis has been reproduced. The record does not provide sufficient detail to resolve the discrepancy independently.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary outcome and secondary outcomes, and it reports several arm comparisons. When multiple outcomes or comparisons are tested, the chance of at least one apparently unusual result can increase. General statistical explanation: multiplicity procedures can control that risk by adjusting the analysis or by defining a testing hierarchy in advance. The registry record does not report whether any such procedure was used, whether comparisons were prespecified in a hierarchy, or how the secondary analyses were handled. The recalculated p-values are unadjusted.

The record also does not report the original statistical analysis plan, the covariates or model details for the logistic regressions, interim analyses, or a rationale for the original enrollment target. These omissions do not invalidate the posted results, but they limit assessment of whether the displayed analyses were the complete prespecified set and whether the reported inferential measures reflect any adjustments. The planning illustrations based on observed effects are not a substitute for the trial’s original design assumptions.

Missing-data handling and analysis populations are not reported in the registry record. The primary outcome denominators are posted as 315 in each arm, but the record does not describe exclusions, missing outcome status, participant flow after assignment, or whether the analyses followed an intention-to-treat principle. The vaccination outcome has posted denominators of 82, 84, and 83 by arm; the reason those denominators differ from the primary outcome denominators is not reported. It would be inappropriate to infer whether this reflects eligibility, missing data, a subset, or another analysis decision.

The registry identifies the study as randomized and triple masked, but the methods used to implement allocation and masking are not reported. It does not specify which participants, clinicians, outcome assessors, or analysts were masked. These gaps matter because they affect how confidently readers can assess selection, measurement, and analysis. The record supports interpretation of the posted estimates and their stated methods, while leaving these design and analysis details unresolved.

16. Safety Results

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

Deaths

Other (non-serious) adverse events

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

The serious adverse event entries report 0 affected participants in each of the no-message, standard-message, and tailored-message arms. The other adverse event entries also report 0 affected participants in each arm, and deaths are recorded as 0 in each arm. The number at risk is listed as 315 for each arm for both serious and other adverse events.

These figures should be read as the adverse events recorded in the registry, not as proof that no event of any kind occurred. The registry does not list serious adverse event terms, and it does not provide further detail on how events were collected, defined, or monitored. A zero posted count can mean no event was recorded under the registry’s reporting process; it does not by itself establish the absence of all harms or guarantee safety. The record also does not describe the duration or completeness of safety surveillance beyond the outcome information posted.

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

The primary results describe documented WCC completion during the trial’s stated follow-up window among children assigned to the outreach conditions. The posted counts are 40 of 315 with no message, 76 of 315 with a standard message, and 61 of 315 with a tailored message. The count pattern is consistent with more recorded completions in the standard-message arm than in the no-message arm, while the tailored-message count falls between them. The recalculated unadjusted contrast between tailored and standard messages is less precise and its confidence interval includes zero.

The registry’s own logistic-regression results are the sponsor’s reported inferential analyses, and the discrepancy flag means those analyses take precedence over the recalculations. The reported risk-ratio results support distinctions between the listed arm pairs according to their stated confidence intervals and p-values, with the standard-message and tailored-message comparison less conclusive than the comparisons involving no message. This statistical description should not be translated into a claim that a message strategy works or does not work for every family.

The endpoint is a billing-code-documented visit, not a direct measure of the reasons a family did or did not schedule or attend. The results do not establish why the arms differed, whether message content caused the observed pattern, or whether reminders changed longer-term care. The record does not report outcomes beyond the defined follow-up period, family experience, visit quality, or subsequent health. It also does not provide a threshold for what would count as a meaningful difference.

For the same reasons, the secondary outcomes require separate interpretation. Scheduling does not guarantee completion, and vaccination is not a substitute measure for WCC attendance. The secondary findings do not establish a mechanism linking reminders, scheduling, completion, or vaccination. Missing information about multiplicity, analysis populations, and the original statistical plan further limits claims based on multiple comparisons. The most defensible meaning is narrow: the record reports outcome counts and model-based comparisons for assigned outreach strategies over specified periods, with uncertainty and analytic limitations that should remain part of any interpretation.

18. Why This Trial Matters Statistically

This trial illustrates why an outcome definition and its timeframe are central to interpreting a clinical study. A WCC completion recorded by a billing code is a concrete event, but it is not identical to intention to schedule, actual attendance without a code, quality of care, or a later health outcome. Separating scheduling from completion prevents readers from treating a step toward care as equivalent to a completed visit.

It also highlights the importance of distinguishing absolute and relative effect measures. The recalculated risk differences describe absolute contrasts from posted counts, while the registry reports risk ratios from logistic regression. These measures are related but not interchangeable, and the direction of a contrast depends on group ordering. A reader who checks the arm labels, reference group, and measure can avoid mistakenly assigning a result to the wrong comparison.

The record also demonstrates why reported analyses and transparent recalculations should be kept distinct. A simple recalculation can make the relation between counts and an unadjusted contrast easier to understand. It cannot reproduce a model-based analysis when model details are missing. When the registry flags a discrepancy, the sponsor’s analysis takes precedence for the trial’s inferential conclusions, while the recalculation remains a descriptive comparison of posted summaries.

Finally, the trial shows the interpretive value of uncertainty and documentation. Confidence intervals help distinguish a relatively clear contrast from one that remains compatible with a range of possibilities. Missing information about multiplicity, missing data, masking roles, and the original sample-size rationale limits what can be concluded beyond the reported estimates. These are not merely technical details: they determine how confidently findings can be compared, generalized, or used to motivate further study.

19. Limitations

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

The primary outcome depends on well-visit billing-code documentation in the electronic health record. The record does not say whether all completed visits would necessarily generate the specified code or how coding differences were handled. This limits interpretation to recorded completion under the stated definition rather than every possible instance of care.

The population is specific: children age 6-17 years who were due for a WCC and had no visit scheduled in the next 45 days. Results may not describe children who already had appointments, children outside this age range, or other populations. The record does not report further eligibility details or how participants were identified and approached.

The tailored message added the child’s age and date of last WCC to reminders. The record does not isolate the contribution of each added detail, describe message delivery or engagement, or report whether recipients read the messages. Thus, the arm comparison concerns the assigned message package, not a separately tested effect of each message element.

The record does not explain the discrepancy between the unadjusted recalculations and sponsor analyses, nor does it provide enough model detail to reproduce the sponsor estimates. It also does not report multiplicity handling, missing-data procedures, participant flow, analysis populations, masking roles, or the original sample-size justification. The different denominators posted for vaccination are unexplained. These limitations constrain assessment of robustness and generalizability without changing the fact that the registry reports the stated results.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Effect of Varied Outreach Methods on Appointment Scheduling and Appointment Completion Using MyChart — NCT04994691.
• Registry record last updated 2024-08-28; status last verified 2024-08.
• Results first posted 2024-08-28.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

23. Record Summary

The registry record for NCT04994691 describes a completed randomized, parallel, triple-masked study of MyChart outreach to children age 6-17 years who were due for a WCC and lacked a scheduled visit in the next 45 days. The arms were no message, two standard reminders, and two tailored reminders that added the date of the last WCC and the child’s age. Enrollment is listed as 945.

The primary outcome was a WCC documented by a well-visit billing code within 8 weeks. Posted counts were 40 of 315 for no message, 76 of 315 for standard message, and 61 of 315 for tailored message. The registry reports logistic-regression risk ratios for primary arm comparisons. Because a discrepancy with simple unadjusted recalculations is flagged, the sponsor’s reported analyses take precedence for inferential interpretation.

Secondary outcomes were WCC appointment scheduling and receipt of a first COVID-19 vaccination. The record does not report multiplicity handling, missing-data procedures, analysis populations, the original sample-size rationale, or the explanation for differing vaccination denominators. Adverse-event entries list zero affected participants in each arm.