Ending the HIV Epidemic Through Point-of-Care Technologies (EHPOC): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Ending the HIV Epidemic Through Point-of-Care Technologies (EHPOC) (NCT04793750), a randomized na study. The primary outcome is Participants Linked Either to Care (PrEP or ART).

2. Clinical Question

3. Trial Design

The registry record describes a randomized, parallel-group study with no masking. It lists an enrollment of 224 and a status of completed. The study title is Ending the HIV Epidemic Through Point-of-Care Technologies (EHPOC), and the official title describes performance evaluation of novel point-of-care HIV tests in Baltimore. The phase is recorded as NA, consistent with the record's focus on diagnostic testing rather than a drug phase.

Participants were assigned to POC HIV VL Testing or SOC HIV Testing. In the POC arm, participants received standard-of-care tests—the DPP HIV-Syphilis Test System and OraQuick—plus an HIV point-of-care viral-load test. The SOC arm received routine standard-of-care HIV testing. The listed interventions are Cepheid GeneXpert HIV-1 Qual POC HIV VL test, DPP HIV-Syphilis test system, and OraQuick. Further implementation details, including the randomization procedure and whether assignment was stratified, are not reported in the registry record.

The primary outcome was linkage to care during a 12 Weeks period. The record defines linkage as at least one in-person or telehealth interaction with a clinical team about HIV ART or PrEP within the study follow-up period. The arm counts were 51 of 113 in the POC HIV VL Testing arm and 50 of 111 in the SOC HIV Testing arm.

Two secondary outcomes are listed: time to linkage to either PrEP or ART and participants reporting condom-less sex. Results for those outcomes are not reported in the registry record. The record identifies Johns Hopkins University as sponsor. More detailed information about recruitment settings, eligibility, follow-up completeness, and protocol procedures is 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 Participants Linked Either to Care (PrEP or ART), measured over 12 Weeks.

The primary endpoint counts participants linked either to care through PrEP or ART during 12 Weeks of follow-up. A participant met the endpoint if they had at least one in-person or telehealth interaction with a clinical team about HIV ART or PrEP during the study follow-up period. The posted result is a count of participants meeting that definition, rather than a measure of viral load, the time until linkage, or the number of clinical contacts.

Linkage is relevant to the study's stated focus because the registry rationale concerns how testing technologies may connect people with HIV treatment or prevention services. A point-of-care test could potentially make results available within a clinical visit, while the endpoint asks whether a clinical interaction about ART or PrEP occurred during follow-up. The endpoint therefore captures a service-engagement outcome, not the diagnostic accuracy of a test itself. Test performance measures are not detailed in the posted primary result.

The primary outcome counts were 51 participants in POC HIV VL Testing and 50 in SOC HIV Testing. The registry describes the endpoint's timeframe as 12 Weeks. It does not report whether linkage was independently verified, how incomplete follow-up was handled, or whether a participant's interaction was classified by blinded reviewers. Those details would help clarify ascertainment and are not reported in the registry record.

7. Results Posted in the Registry

• POC HIV VL Testing: 45.1% (51/113) versus 45.0% (50/111) on control; risk difference 0.1 percentage points (95% CI -12.9 to 13.1), risk ratio 1.00 (0.75–1.34); unadjusted two-proportion z-test p = 0.989.

8. Statistical Interpretation

9. Statistical Methods Explained

What does randomization contribute?

Randomization assigns participants to study arms by chance. In general, this approach is intended to make the groups comparable, on average, with respect to both measured and unmeasured characteristics at the time of assignment. That makes a between-arm comparison more informative about the intervention than a simple comparison of groups formed by clinical choice. Randomization does not guarantee identical groups, particularly in a study of limited size, and it does not correct problems arising after assignment, such as incomplete outcome ascertainment.

The registry record identifies the design as randomized and parallel, with no masking. Parallel means the two groups receive their assigned strategies over the study period rather than crossing over between strategies as part of the design. No masking means participants or study personnel were not described as blinded. The precise allocation procedure, concealment process, and any safeguards around outcome assessment are not reported in the registry record.

What is a risk difference?

General: a risk difference compares the proportion of participants with an outcome in one group with the proportion in another. It is expressed as an absolute difference between outcome probabilities. A value of zero corresponds to equal observed risks; positive or negative values indicate the direction of the contrast according to the comparison's stated ordering. In this record, the posted comparison is a risk difference for linkage, and its estimate is very close to zero.

A risk difference is often direct to interpret because it describes an absolute contrast. It is sensitive, however, to the outcome definition and follow-up period. Here, the outcome is having at least one clinical interaction about ART or PrEP within 12 Weeks. It should not be interpreted as a comparison of time to linkage or of test accuracy.

How should the p-value be understood?

General: a p-value measures how unusual the observed data, or data more extreme, would be under a specified null model and its assumptions. A large p-value does not demonstrate that there is no effect, and it is not the probability that a treatment or testing strategy is ineffective. The posted unadjusted p-value is 0.989. The registry record does not report the full test specification, whether the analysis was prespecified, or whether other analyses were adjusted for covariates.

Why use a confidence interval?

General: a confidence interval presents a range of values compatible with the data and the method used to construct it. It communicates uncertainty about an estimate more directly than a point estimate alone. An interval containing zero for a difference is consistent with no difference under the interval's statistical procedure, but it can also include nonzero differences. The interval's width matters: a wide interval allows a broader range of possible effects than a narrow one.

For the posted risk difference, the interval includes values in both directions and is broad relative to an estimate close to zero. This means the point estimate should not be treated as an exact or definitive effect size. The record does not specify whether the interval was calculated using a particular small-sample correction or other method, so further details should not be assumed.

What is over-dispersion?

General: over-dispersion arises when observed variability is greater than a statistical model assumes. When present and unaccounted for, it can make standard errors too small and intervals too narrow, overstating precision. It is especially relevant in some count and clustered-data settings. The posted primary outcome is a binary participant-level event count, and the record does not report a model assessing over-dispersion or clustering. There is therefore no basis in the registry record to claim that over-dispersion was present or that a particular correction was used.

What analysis details are available?

The registry record reports arm-level outcome counts and an unadjusted comparison. It does not report a detailed statistical analysis plan, a covariate-adjusted model, a missing-data method, a prespecified interim analysis, or the exact analysis population. Those details can affect how results are interpreted and reproduced. The posted summary statistics support a basic description of the observed comparison, but they do not establish which additional analytic decisions were made.

10. Confidence Intervals and Precision

The reported risk-difference estimate is very close to zero, while its confidence interval spans negative and positive values. This combination is important: the point estimate describes the center of the comparison, but the interval conveys the range of uncertainty around it. Because the interval includes zero, the posted data are compatible with no difference under the interval's method. Its extension in both directions also means that the data do not precisely determine the direction of a possible difference.

General: confidence intervals are not guarantees that the true value lies within a particular observed interval. They are generated by a statistical method that, under its assumptions, has a stated long-run coverage property. Interpretation depends on the method, the design, and how the data were collected. The record does not give enough detail to assess every assumption behind the interval.

The observed event counts were 51 of 113 in the POC HIV VL Testing arm and 50 of 111 in the SOC HIV Testing arm. These counts are close, but closeness of counts does not itself establish precision or equivalence. Precision depends on the information in the full comparison, including the number of participants and the outcome frequency, as reflected in the interval. The record does not report a prespecified equivalence margin or a non-inferiority margin.

The reported p-value of 0.989 is consistent with little evidence against a null comparison under the unadjusted test. It should be considered alongside the interval and the endpoint definition, not as a standalone conclusion. A confidence interval can help readers distinguish a narrowly estimated near-zero difference from an uncertain estimate that happens to be near zero. In this case, the interval's range cautions against treating the central estimate as a definitive statement about the size or direction of a true effect.

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 participants or total events. They illustrate sample sizes under the stated assumptions, using the observed effect as a starting point and then considering planning scenarios for 100% of the observed effect, 75% of the observed effect, and 50% of the observed effect. For each scenario, the table gives separate estimates for planning for 80% power and 90% power. The curve should be read as a visual representation of how the planning requirement changes as the effect to detect changes.

The calculation assumptions listed in the registry record 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. For the full observed effect, the figures are 5,053,414 participants per arm for 80% power and 6,765,087 participants per arm for 90% power. For planning for 75% of the observed effect, the figures are 8,983,456 participants per arm for 80% power and 12,026,298 participants per arm for 90% power. For planning for 50% of the observed effect, the figures are 20,211,895 participants per arm for 80% power and 27,057,992 participants per arm for 90% power.

In general, a smaller effect is harder to distinguish from random variation, so it requires more participants under otherwise similar assumptions. The large planning estimates reflect the very small observed risk-difference estimate used in these illustrations; they are not the original trial's design calculations. The trial's own original sample-size assumptions are not reported in the registry record. These figures are planning illustrations, 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 = 0.450; treatment proportion p₁ = 0.451; average p̄ = 0.451.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.7037 = 1.3792.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.7037 = 0.5922.
4. n = (1.3792 + 0.5922)² / (0.0009)² = 5,053,413.3, rounded up to 5,053,414 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 <- 0.45045045045045                          # control proportion
p1 <- 0.451327433628319                          # 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(4793750)
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 = 5,053,414 (80% power), 6,765,087 (90% power)
 75% of observed effect: n per arm = 8,983,456 (80% power), 12,026,298 (90% power)
 50% of observed effect: n per arm = 20,211,895 (80% power), 27,057,992 (90% power)
power.prop.test: 5053413 
Simulated power at n = 5,053,414 per arm: 0.798 (20000 replicates)

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

13. Multiplicity, Missing Data and Analysis Populations

The registry record lists three outcome measures, including the primary linkage outcome and two secondary outcomes: time to linkage to either PrEP or ART, and participants reporting condom-less sex. The number of secondary outcomes posted is two, but results for those secondary outcomes are not reported in the registry record. The record also indicates no posted analyses. The primary comparison has an unadjusted p-value, but a broader analysis plan is not described.

Multiplicity arises when multiple outcomes, comparisons, or analyses are tested. General: the more tests that are performed, the greater the chance that at least one result may appear statistically notable by chance alone. Researchers can address this through prespecification, hierarchical testing, or adjustment procedures, among other approaches. Whether any such strategy was planned or used for this study is not reported in the registry record. No multiplicity adjustment should therefore be assumed.

Missing data can also affect interpretation. If participants do not complete follow-up or if linkage status is unknown for some participants, the observed counts may not represent all randomized participants in the same way. The registry record gives arm sizes and outcome counts, but does not report the number with missing endpoint data, the reason for any missing data, or how missing outcomes were handled. It does not describe imputation or sensitivity analyses.

The analysis population is not reported in the registry record. It is not clear whether the posted comparison included all randomized participants, only participants with observed outcome data, or another population. The record also does not state how participants were handled if they did not follow their assigned testing strategy. These omissions limit assessment of how the primary estimate relates to the randomized design.

These are reporting gaps, not evidence that a particular procedure was or was not performed. Without details on multiplicity, missingness, and analysis populations, the primary comparison can be described as posted, but its full analytic context cannot be reconstructed from the registry record.

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.

Most frequent non-serious adverse event terms

The serious adverse event figures report zero serious adverse events in both arms: the POC HIV VL Testing arm had 0 among 113 participants at risk, and the SOC HIV Testing arm had 0 among 111 participants at risk. Deaths were also recorded as 0 in each arm. These entries describe the counts recorded in the registry, not proof that no unrecorded events occurred.

The record also lists other adverse events: 16 among 113 participants at risk in the POC HIV VL Testing arm and 25 among 111 participants at risk in the SOC HIV Testing arm. The registry entry does not provide the event terms or further detail about severity, timing, relatedness, or whether individual participants could have had more than one event. No top serious adverse event terms are listed.

General: adverse-event counts are descriptive safety information. Comparisons need care when event definitions, ascertainment, follow-up, and exposure differ or are not fully described. The registry record does not provide enough detail to determine the clinical significance or cause of the recorded events. These figures should not be interpreted as evidence that either testing strategy caused or prevented adverse events.

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

The posted primary result is a comparison of linkage counts during 12 Weeks, using a definition based on at least one clinical interaction about HIV ART or PrEP. The POC HIV VL Testing arm had 51 participants meeting the endpoint among 113, and the SOC HIV Testing arm had 50 among 111. The risk-difference estimate is very close to zero, and the unadjusted p-value is 0.989. The confidence interval spans values in both directions, so the estimate remains uncertain.

These data do not establish equivalence between the strategies. Equivalence requires a prespecified margin and an analysis designed to assess whether differences are small enough to fall within that margin; neither is reported in the registry record. Likewise, the result does not establish that the testing approach improves or fails to improve linkage. It describes the observed endpoint comparison and its statistical uncertainty.

The primary endpoint is service linkage, not diagnostic accuracy. The trial title and rationale concern point-of-care HIV testing, and the record lists testing technologies, but the posted primary result does not report sensitivity, specificity, or other performance measures for those tests. Nor does this result establish effects on viral suppression, HIV transmission, PrEP use, or time to ART initiation. Two secondary outcomes are listed, but their results are not reported in the registry record.

The record also does not report key details needed to assess the robustness of the result, including missing endpoint data, the analysis population, adjusted analyses, and a complete statistical plan. Consequently, the appropriate reading is bounded: the posted counts are similar, the estimated contrast is near zero, and uncertainty remains about the range of differences compatible with the data. Broader clinical conclusions are not supported by this primary comparison alone.

16. Why This Trial Matters Statistically

This trial addresses a statistical question that is easy to blur with a broader clinical question: whether adding point-of-care HIV viral-load testing to standard-of-care tests changes linkage to HIV care or prevention, as defined by a clinical interaction about ART or PrEP. Keeping the endpoint definition in view prevents a result about linkage from being mistaken for a result about test accuracy or downstream health outcomes.

The record also illustrates why estimates and uncertainty should be read together. The primary counts are close, and the estimated risk difference is very close to zero, but the confidence interval extends in both directions. A near-zero point estimate alone might sound definitive; the interval makes clear that the comparison is not that exact. The unadjusted p-value adds information about the test against a null comparison, but it does not resolve uncertainty about clinical importance or prove equivalence.

Planning calculations based on an observed effect can be instructive because they show how sample-size requirements change when the effect of interest is smaller. In this case, the observed effect used for the illustrations is very small, and the resulting participant requirements are correspondingly large. Such calculations are not a substitute for a prospectively specified design justification, which is not reported in the registry record.

Finally, transparent reporting of missing data, analysis populations, multiplicity, and endpoint ascertainment would help readers assess how robust the estimate is. Their absence from the registry record does not prove that these issues were ignored; it limits what can be concluded from the posted results. Statistically careful interpretation separates the observed data from assumptions and from wider claims about testing and clinical outcomes.

17. Limitations

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

The primary comparison is reported as an unadjusted result, and further details about the statistical model and test are not reported in the registry record. The record does not describe the randomization procedure, allocation concealment, or how the absence of masking may have affected outcome assessment. These details matter when evaluating the possibility of selection or ascertainment effects.

The primary outcome depends on identifying at least one in-person or telehealth interaction about ART or PrEP during 12 Weeks. The record does not report how interactions were verified, whether outcome assessors were independent, or whether definitions were applied consistently across arms. Follow-up completeness and missing endpoint data are also not reported in the registry record.

Although the registry reports enrollment of 224, the arm sizes in the posted primary comparison are 113 and 111. The record does not explain the relationship between enrollment and endpoint availability or identify the analysis population. It also does not report participant characteristics that would help assess the study population's representativeness.

The primary result concerns linkage, rather than test performance or downstream outcomes. Results for the two listed secondary outcomes are not reported in the registry record. The record does not provide an original sample-size justification, a complete statistical analysis plan, multiplicity procedures, interim-analysis details, or prespecified criteria for interpreting clinically important differences. It also does not report an equivalence or non-inferiority margin.

Finally, the reported adverse-event summary lacks event terms and details about seriousness, relatedness, timing, and ascertainment. These reporting limitations restrict assessment of safety and the broader interpretation of the comparison. They should be distinguished from evidence that any particular outcome or analysis was absent from the study itself.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Ending the HIV Epidemic Through Point-of-Care Technologies (EHPOC) — NCT04793750.
• Registry record last updated 2026-04-23; status last verified 2026-04.
• Results first posted 2026-04-22.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

21. Record Summary

The registry record describes NCT04793750, Ending the HIV Epidemic Through Point-of-Care Technologies (EHPOC), sponsored by Johns Hopkins University. It is a completed, randomized, parallel-group study with no masking and enrollment of 224. The study compares POC HIV VL Testing, which adds an HIV point-of-care viral-load test to standard-of-care tests, with routine SOC HIV Testing.

The primary endpoint is linkage to care, defined as at least one in-person or telehealth interaction with a clinical team about HIV ART or PrEP during 12 Weeks. The posted counts were 51 of 113 in the POC arm and 50 of 111 in the SOC arm. The comparison's risk-difference estimate is very close to zero; its confidence interval includes values in both directions, and the unadjusted p-value is 0.989.

Two secondary outcomes are listed, but their results are not reported in the registry record. Missing-data procedures, analysis populations, and the trial's original sample-size assumptions are also not reported in the registry record.