SCALE-UP Utah: Community-Academic Partnership to Address COVID-19 Testing Among Utah Community Health Centers: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for SCALE-UP Utah: Community-Academic Partnership to Address COVID-19 Testing Among Utah Community Health Centers (NCT04939532), a randomized na study. The primary outcome is Reach-TM Engage.

2. Clinical Question

3. Trial Design

NCT04939532 is a completed randomized, parallel-design study sponsored by the University of Utah. The registry lists single masking; it does not specify which people were masked. The phase is listed as NA. Enrollment is recorded as 44729. The study started on 2021-03-29, and primary completion was recorded on 2022-06-17.

The two arms were Text-Messaging (TM) and Text-Messaging Plus Patient Navigation (TM+PN). Both used text messaging as a population health management strategy. TM was described as using electronic health record data to identify patients at high risk for infection or severe disease, then contacting them about testing needs. TM+PN added patient navigation, including real-time community health worker assistance to address barriers and provide motivation. The registry summary also describes testing logistics as part of the messaging approach.

The study’s long-term objective was to increase the reach, acceptance, uptake, and sustainability of COVID-19 screening and testing among Utah Community Health Center patients. Its summary describes a rapid-cycle research approach: interventions could be tested with small numbers of clinics or patients over short timeframes, evaluated, adapted and retested, or extended to additional clinics or patients. The registry record does not explain how that approach was incorporated into the randomization or the primary analysis.

The primary outcome was Reach-TM Engage, described as the proportion of patients responding to the text-messaging intervention, with a 12 months timeframe. The posted results give arm-specific participant counts and denominators. The registry record does not report the randomization unit, allocation procedures, masking details, or the trial’s original sample-size assumptions. The overall enrollment entry and the outcome-specific arm denominators describe different record fields; the primary result should be interpreted using the denominators posted for that outcome.

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

Analysis population for the primary outcome, as defined in the registry: Proportion of patients that respond to the Text Messaging intervention

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 Reach-TM Engage, measured over 12 months.

Reach-TM Engage is the primary outcome. The registry describes it as the proportion of patients who respond to the Text Messaging intervention, with a 12 months timeframe. The results are recorded as participant counts: 2594 participants in the TM arm and 2401 in the TM+PN arm. The corresponding outcome denominators are 22140 and 22580.

Engagement is a relevant intermediate measure for a program that aims to reach patients by text. A response indicates that a patient engaged with the messaging, but it does not by itself establish whether the patient completed testing. The registry lists testing uptake among the secondary outcomes, including total patients tested as indicated in the health record and uptake among eligible patients. Those outcomes address different steps in the pathway from outreach to testing.

The primary endpoint’s stated proportion and the posted participant counts should be read together: the count is the number of participants with the recorded response, and the denominator is the corresponding arm population for this outcome. The registry record does not provide further operational details about what qualified as a response, how response status was ascertained, or how missing response information was handled. Those details would help clarify precisely what the endpoint captures and how it relates to actual testing behavior.

7. Results Posted in the Registry

• Text-Messaging Plus Patient Navigation (TM+PN): 10.6% (2401/22584) versus 11.7% (2594/22145) on control; risk difference -1.1 percentage points (95% CI -1.7 to -0.5), risk ratio 0.91 (0.86–0.96); unadjusted two-proportion z-test p = <0.001.

8. Statistical Interpretation

9. Statistical Methods Explained

The posted comparison concerns a binary outcome: each participant either had a recorded response to the text-messaging intervention or did not. For a binary endpoint, the arm-specific event count and denominator describe the observed response frequency. Here, the counts are 2594 of 22140 for TM and 2401 of 22580 for TM+PN. The registry labels the outcome as a participant count and describes the underlying measure as a proportion.

What does a risk difference express?

A risk difference subtracts the event probability in one group from the event probability in the reference group, according to the comparison’s stated ordering. It is an absolute contrast: zero corresponds to equal observed probabilities, while a negative estimate in this record indicates a lower probability in TM+PN than in TM. The posted risk difference is -0.011. Its confidence interval extends from -0.017 to -0.005. Because the entire interval is below zero, the interval is inconsistent with an exact zero difference at the confidence level used for that interval, subject to the analysis assumptions.

Risk differences are often useful because they retain the probability scale. They can be related to the size of the observed contrast without converting it into a relative measure. Their interpretation depends on the endpoint definition and the populations included in each denominator. The registry record does not report how missing endpoint information was handled, so the posted denominators should not be assumed to represent a particular missing-data strategy.

How do risk ratios and odds ratios differ?

A risk ratio divides one group’s event probability by the other group’s probability. A value of one corresponds to equal probabilities. The posted risk ratio is 0.908, with a confidence interval from 0.861 to 0.956. An odds ratio compares event odds rather than probabilities; its posted estimate is 0.897, with a confidence interval from 0.845 to 0.951. A value of one is also the equality point for an odds ratio. Since probabilities and odds are different quantities, a risk ratio and odds ratio are not interchangeable, even when they describe the same two groups and endpoint.

These measures express the same basic arm comparison in different ways. They do not identify the mechanism behind the observed difference. In particular, the primary endpoint is engagement with messaging. The estimates do not establish whether navigation changed testing behavior, whether any change occurred through motivation or logistics, or whether another factor explains the recorded response contrast.

What does the p-value contribute?

The reported unadjusted p-value is p < 0.001. In general, a p-value is calculated under a specified statistical model and a null hypothesis, often a hypothesis of no difference between groups. It describes how unusual the observed data, or more extreme data, would be if that null model and its assumptions were correct. It does not provide the probability that the null is true, and it does not quantify the clinical or operational importance of the result.

The word unadjusted matters because the registry lists multiple outcomes, but does not report a multiplicity adjustment for this comparison. Whether adjustment was planned or used in another analysis is not reported in the registry record. The posted p-value should therefore be read as the unadjusted value attached to this result, not as evidence that a broader family of statistical tests was controlled.

What can and cannot be assessed from the record?

The registry identifies the study as randomized, single-masked, and parallel. General statistical reasoning treats random allocation as a way to make groups comparable on average, but the record does not give the allocation unit or procedure, or identify who was masked. It also does not report the statistical model, any covariate adjustment, clustering treatment, or analysis population. Those details matter when evaluating how the confidence interval and p-value were obtained, especially if allocation or implementation involved clinics or repeated rapid-cycle activity.

The registry lists no analyses in its analysis count. That does not establish that no analysis was conducted; it means that analysis entries are not reported in the registry record. The posted comparison still supplies an estimate, confidence interval, and p-value, but the full analytic specification cannot be reconstructed from those entries alone.

In general, over-dispersion occurs when observed variability is greater than a statistical model assumes. When outcomes are clustered, for example by clinic, treating all observations as independent may understate uncertainty if the analysis does not account for that structure. This is a general statistical concern, not a claim that over-dispersion occurred in this trial. The registry record does not report whether clustering or over-dispersion was assessed or modeled.

10. Confidence Intervals and Precision

The risk-difference estimate is -0.011, and its 95% confidence interval runs from -0.017 to -0.005. A confidence interval gives a range of values compatible with the observed data under the method’s assumptions. It helps distinguish the estimated direction of the contrast from the uncertainty around its size. Here the interval lies entirely below zero, the no-difference value for a risk difference.

The relative measures also have intervals: the risk ratio is 0.908, with a 95% confidence interval from 0.861 to 0.956; the odds ratio is 0.897, with a 95% confidence interval from 0.845 to 0.951. For both ratios, one is the equality value. The intervals do not include one. These intervals describe uncertainty in the corresponding relative estimates and should not be confused with the interval for the absolute risk difference.

Precision is not the same as importance. A confidence interval can be relatively informative about the direction while still allowing a range of possible magnitudes. The endpoint also matters: these estimates concern recorded response to messaging, not the secondary testing-uptake outcomes. The record does not report whether the intervals account for clustering, repeated implementation cycles, or other design features.

Confidence intervals are not guarantees that the true effect lies within the displayed limits. Their usual interpretation depends on the statistical method and its assumptions. In general, repeated use of the same interval procedure under those assumptions would produce intervals containing the true value at the stated confidence level. The registry record does not give enough methodological detail to assess all assumptions behind these particular intervals.

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 planning figures are participants per arm, not total events. The illustrations use two-sided significance level 0.05, equal allocation between the two arms, and a normal approximation. Each row uses the stated fraction of the observed effect as the effect to detect. These are planning illustrations based on the observed result, not the trial’s original design assumptions; the original assumptions are not reported in the registry record.

For planning at 100% of the observed effect, the table gives 13,299 participants per arm for 80% power and 17,803 participants per arm for 90% power. At 75% of the observed effect, it gives 23,894 participants per arm for 80% power and 31,986 participants per arm for 90% power. At 50% of the observed effect, it gives 54,322 participants per arm for 80% power and 72,721 participants per arm for 90% power.

In plain terms, the 80% power column indicates the per-arm planning size for a design with an 80% chance of detecting the specified effect under the listed assumptions, if that effect is real. The 90% column describes the corresponding planning size for a 90% chance. Smaller effects require more participants because they are harder to distinguish from random variation. The curve should be read as a visual expression of the relationship between the effect being targeted and the required sample size, not as an additional trial result. This is not observed power.

12. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.117; treatment proportion p₁ = 0.106; average p̄ = 0.112.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.4455 = 0.8732.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.4455 = 0.3749.
4. n = (0.8732 + 0.3749)² / (-0.0108)² = 13,298.9, rounded up to 13,299 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 <- 0.117137051253105                          # control proportion
p1 <- 0.106314204746723                          # 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(4939532)
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 = 13,299 (80% power), 17,803 (90% power)
 75% of observed effect: n per arm = 23,894 (80% power), 31,986 (90% power)
 50% of observed effect: n per arm = 54,322 (80% power), 72,721 (90% power)
power.prop.test: 13298.9 
Simulated power at n = 13,299 per arm: 0.803 (20000 replicates)

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

13. Multiplicity, Missing Data and Analysis Populations

The registry lists five outcome measures, including the primary outcome and four secondary outcomes. The secondary measures concern total patients tested as indicated in health records, navigation reach and engagement measures, and uptake among eligible patients. Multiple outcomes can create multiple opportunities for statistical findings, so a prespecified approach to multiplicity can help readers understand how results were interpreted across outcomes.

A multiplicity strategy might specify a hierarchy of outcomes, divide a statistical error rate across tests, or identify selected analyses as exploratory. Which approach, if any, was used for this study is not reported in the registry record. The primary comparison is accompanied by an unadjusted p-value, but the record does not report a multiplicity-adjusted value or a broader testing plan. The posted p-value should therefore be considered in the context of the listed outcome set rather than treated as a complete account of all statistical evidence.

The registry’s analysis count is zero, and no analysis entries are posted. This does not show that analyses were not performed; the analysis plan and the methods used for each outcome are not reported in the registry record. In particular, the record does not specify the analysis population, whether the analysis followed assigned groups, how protocol deviations were handled, or whether any covariates were included.

Missing-data methods are also not reported in the registry record. For the primary response outcome, the record gives counts and denominators but does not explain how participants with unknown or unrecorded response status were classified. It does not report imputation, sensitivity analyses, or follow-up completeness. The extent and implications of missing outcome data therefore cannot be assessed from the posted information.

The study describes a rapid-cycle approach involving repeated evaluation and possible adaptation of intervention messages. The registry record does not state how repeated cycles were reflected in the statistical analysis, whether results were pooled across cycles, or whether any interim looks at outcome data were planned. The interim analysis plan is not reported in the registry record. These omissions limit assessment of how the reported uncertainty and p-value relate to all analyses conducted.

14. Safety Results

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

Deaths

Other (non-serious) adverse events

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

The serious adverse event entries report zero events in each arm. The registry gives 22145 participants at risk and zero serious adverse events for TM, and 22584 at risk and zero serious adverse events for TM+PN. The deaths field is also zero in both arms.

The other adverse event entries likewise report zero events in both arms: 22145 participants at risk and zero other events for TM, and 22584 at risk and zero other events for TM+PN. These are the posted registry counts. They should be read as reported event figures, not as proof that no adverse events occurred outside the recorded definitions, follow-up, or reporting process.

The adverse-event denominators differ from the primary outcome denominators, which are 22140 for TM and 22580 for TM+PN. Each denominator belongs to its respective registry outcome entry and should not be substituted for the other. The registry record does not provide further details about event definitions, ascertainment, duration of safety follow-up, or whether events were actively solicited. Zero recorded events therefore provide limited information about the completeness or sensitivity of safety monitoring.

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

The posted primary result indicates a lower observed response proportion in TM+PN than in TM alone for Reach-TM Engage. The estimated risk difference is -0.011, with a 95% confidence interval from -0.017 to -0.005, and the unadjusted p-value is p < 0.001. The relative estimates point in the same direction: the risk ratio is 0.908 and the odds ratio is 0.897. These findings describe the recorded response outcome and its estimated between-arm contrast.

They do not establish that adding navigation reduced COVID-19 testing. Response to a message and completion of a test are distinct steps. Testing uptake appears among the secondary outcomes, but its results are not included in the posted primary comparison. The result also does not establish why response differed, or whether the navigation component changed barriers, motivation, or logistical support in the intended way.

Randomization is an important design feature for comparing intervention groups, but interpretation depends on how assignment and analysis were carried out. The registry record does not report the allocation unit, detailed analysis methods, missing-data handling, or analysis population. Those details affect how confidently the posted comparison can be interpreted as an effect of assignment to the intervention arms.

The confidence interval provides information about precision for this endpoint, but it does not resolve questions about outcome definition, generalizability, or the broader testing objective. The study concerns Utah Community Health Centers and interventions based on patient outreach in that setting. The registry record does not provide enough participant or clinic characteristics to determine how closely the results might apply to other settings or patient populations.

The findings are therefore best read as a statistical comparison of recorded messaging response between the two randomized arms, with uncertainty expressed by the posted interval. They do not answer every question about COVID-19 testing uptake, long-term sustainability, or the implementation of community health worker navigation.

16. Why This Trial Matters Statistically

This trial raises a useful statistical distinction between engagement and uptake. A text response is a measurable step in an outreach pathway, while testing recorded in a health record is a separate outcome. Keeping those endpoints distinct prevents an engagement result from being read as though it directly demonstrated a change in testing behavior.

The comparison also illustrates why the effect scale matters. The risk difference expresses an absolute contrast, while the risk ratio and odds ratio express relative contrasts. Each has a different interpretation, even when all are calculated from the same arm data. Reporting more than one scale can help readers understand the result, provided the measures are not mistaken for separate findings.

Precision and statistical evidence are related but not identical. The confidence interval describes a range of estimates compatible with the data under the analysis assumptions, while the p-value addresses compatibility with a specified null model. Neither alone describes practical importance. In this study, the endpoint and its operational definition remain central to interpreting both measures.

The registry’s rapid-cycle description also makes the analysis plan particularly relevant. General statistical methods may need to account for repeated cycles, clinic-level implementation, or adaptation over time, depending on how the study was conducted. The registry record does not report how those features were handled. Recognizing that limitation is part of interpreting the reported estimates responsibly, rather than assuming a particular model or analysis strategy.

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 details needed to evaluate the primary estimate are not reported in the registry record. The randomization unit and allocation procedure are unspecified, as are the identity of the masked parties and the statistical model used for the primary comparison. The record does not say whether analysis accounted for clinic-level clustering or for the rapid-cycle implementation described in the study summary.

The primary endpoint is described as the proportion responding to text messaging, but further operational detail is not reported in the registry record. The response criteria, ascertainment process, and treatment of unknown response status are not specified. This makes it difficult to judge whether response was measured consistently across arms or how incomplete information may have affected the estimate.

The record lists four secondary outcomes but does not post their results in the information summarized for this comparison. It also does not report a prespecified multiplicity approach, analysis populations, missing-data methods, or interim analysis plan. These gaps limit assessment of the full outcome pattern and the context for the unadjusted p-value.

Generalizability is also difficult to assess. The study concerns Utah Community Health Centers, but detailed participant and clinic characteristics are not reported in the registry record. The description of the interventions explains their broad components, but further details about implementation, eligibility, and exposure to messaging or navigation are not available. These limitations do not negate the posted estimate; they define what can be concluded from it.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: SCALE-UP Utah: Community-Academic Partnership to Address COVID-19 Testing Among Utah Community Health Centers — NCT04939532.
• Registry record last updated 2026-02-04; status last verified 2025-12.
• Results first posted 2026-02-04.
• 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

NCT04939532, SCALE-UP Utah, was a completed randomized, single-masked, parallel study sponsored by the University of Utah. It compared Text-Messaging with Text-Messaging Plus Patient Navigation among patients in Utah Community Health Centers. Enrollment is recorded as 44729. The study started on 2021-03-29, and primary completion was recorded on 2022-06-17.

The primary outcome was Reach-TM Engage, described as response to the text-messaging intervention over 12 months. The posted response proportions were 0.117 for TM and 0.106 for TM+PN. The risk difference was -0.011, with a 95% confidence interval from -0.017 to -0.005, and the unadjusted p-value was p < 0.001. The result concerns messaging response; the registry record does not report the trial’s original sample-size assumptions, analysis plan, or missing-data methods.