Simplifying Treatment and Monitoring for HIV (STREAM HIV): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Simplifying Treatment and Monitoring for HIV (STREAM HIV) (NCT04341779), a randomized na study. The primary outcome is Number of Participants With Viral Load Suppression (<200 Copies/ml) and Retained in Care at 72 Weeks.

2. Clinical Question

The primary question concerns a combined clinical and care-engagement outcome. It is not simply a comparison of viral-load results: participants must meet the suppression definition and be retained in care to count toward the endpoint. The registry record also describes secondary questions about acceptability and cost-effectiveness, but does not post results for those outcomes in the record.

3. Trial Design

STREAM HIV is a randomized, parallel-group trial with no masking. The registry lists an enrollment of 539 and the study status as active, not recruiting. The recorded start date is 2021-02-04, and the primary completion date is 2025-06-30. The registry identifies University of Washington as the sponsor and lists the phase as NA.

The intervention arm received point-of-care urine adherence testing and point-of-care viral-load testing. Its detailed description specifies monthly urine tenofovir testing with adherence counselling during the first 5 months, plus point-of-care viral-load testing at Month 6 and 12. The standard-of-care arm had routine adherence counselling and laboratory-based viral-load testing. The registry describes the study aim as assessing clinical efficacy and cost-effectiveness of this integrated monitoring model among people initiating first-line tenofovir disoproxil fumarate-based ART in South Africa.

The posted primary outcome is the number of participants with viral-load suppression below 200 copies/mL and retention in care at 72 weeks after ART initiation. The outcome record gives 270 participants in the intervention arm and 268 in standard care for this comparison, with 186 and 167 participants meeting the endpoint, respectively. These are the outcome denominators and counts recorded for the comparison; the registry does not explain in the posted results why the endpoint denominators differ from one another or from enrollment.

The record includes two secondary outcomes: acceptability of point-of-care tenofovir and viral-load testing, and cost-effectiveness of routine point-of-care testing compared with standard-of-care monitoring. Results for these secondary outcomes are not reported in the registry record. Further protocol details, including the trial's original sample-size justification and full statistical analysis plan, 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.

Before assignment: The initial enrollment was planned for 540. However, one person was considered ineligible after enrollment. Therefore, this person was not enrolled per-protocol.

Overall Study

Analysis population for the primary outcome, as defined in the registry: Number of study participants who have data for the comparison of viral load suppression (\<200 copies/ml) and retention in care at 72 weeks after ART initiation.

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 With Viral Load Suppression (<200 Copies/ml) and Retained in Care at 72 Weeks, measured over 72 weeks after ART initiation.

The primary endpoint counts participants who both had viral-load suppression and were retained in care at 72 weeks after ART initiation. Suppression is defined as a viral load below 200 copies/mL, measured using a laboratory-based reference assay performed by the South African National Health Laboratory Services. Retention is included as part of the same endpoint, so the count represents people meeting both conditions rather than suppression alone.

This combined outcome is relevant to the trial's stated aim of improving HIV outcomes through adherence testing and viral-load monitoring. Viral-load suppression captures a laboratory measure of treatment response, while retention in care captures continued engagement with care. Combining them makes the endpoint a measure of success across both dimensions, but it also means the result cannot identify from the count alone which component explains a difference between arms. Separate component results are not reported in the registry record.

The posted result is a count and an arm-specific proportion summarized through a risk-difference comparison. The endpoint counts are 186 of 270 in the intervention arm and 167 of 268 in standard care. These figures describe the posted comparison; the record does not give further details about the handling of participants without endpoint data, the exact analysis population, or any component-specific endpoint analysis.

7. Results Posted in the Registry

• Intervention Arm: 68.9% (186/270) versus 62.3% (167/268) on control; risk difference 6.6 percentage points (95% CI -1.4 to 14.6), risk ratio 1.11 (0.98–1.25); unadjusted two-proportion z-test p = 0.108.

8. Statistical Interpretation

9. Statistical Methods Explained

What does a risk difference describe?

In a binary-outcome comparison, each participant either meets the endpoint or does not. A risk difference subtracts the endpoint proportion in one arm from the endpoint proportion in the other. A positive estimate in this comparison indicates a higher observed endpoint proportion in the intervention arm; a negative estimate would indicate a higher observed proportion in standard care. The posted point estimate is positive, matching the direction of the reported counts.

A risk difference is an absolute contrast, expressed on the outcome-proportion scale. It differs from a risk ratio, which compares proportions multiplicatively. The registry also gives relative effect summaries, but their detailed estimates and intervals are not needed to interpret the primary posted comparison. The risk-difference estimate and its interval are the most direct way to describe the absolute arm contrast without implying that an observed association establishes a treatment effect with certainty.

How should the confidence interval be read?

A confidence interval describes the range of effect values reasonably compatible with the data and the statistical procedure used. It captures sampling uncertainty under the method's assumptions; it is not a guarantee that the true effect lies within a particular range. A narrow interval indicates more precision than a wide interval, while an interval spanning the null value leaves uncertainty about whether the underlying difference is positive or negative.

For the posted risk difference, the interval crosses zero, the null value for an absolute difference. This is consistent with the unadjusted p-value of 0.108: the reported comparison does not meet the two-sided significance level of 0.05 used in the planning illustration. That does not show that there is no difference. It indicates that the data, analyzed as reported, do not sharply distinguish the observed contrast from no difference at that conventional threshold.

What does the unadjusted p-value tell us?

A p-value describes how unusual data at least as incompatible with a null hypothesis would be if that null hypothesis and the test's assumptions were true. It is not the probability that the null hypothesis is true, and it is not a measure of clinical importance. The reported p-value is unadjusted. The registry does not provide a full description of the test, any covariate adjustment, or a multiplicity procedure, so the p-value should not be interpreted as if those analyses had been reported.

The confidence interval and p-value provide related but distinct information. The interval presents a range of compatible effects and their precision; the p-value summarizes evidence against a null value under a specified test. Neither says whether the combined endpoint's difference is important to participants or whether the intervention is cost-effective. Those questions require information beyond this comparison.

Why is the endpoint binary?

The endpoint classifies participants according to whether they met both stated conditions: suppression below 200 copies/mL and retention in care at 72 weeks. The registry reports counts of participants who met it, rather than a continuous viral-load measurement or time-to-event outcome. For a binary outcome, arm-specific counts and denominators describe how frequently the endpoint was observed, while a risk difference expresses the absolute contrast between those proportions.

Because this endpoint combines two conditions, its interpretation is narrower than the separate concepts of suppression and retention. A participant must satisfy both to count as meeting the endpoint. The record does not post separate counts for each component, so their individual contributions cannot be disentangled from the combined result.

What can and cannot be inferred from randomization?

Random allocation is intended to make trial arms comparable on average at the point of assignment, supporting a causal comparison when follow-up, outcome measurement, and analysis are appropriately handled. The registry identifies allocation as randomized, and masking as none. It does not provide enough detail in the posted results to assess allocation procedures, concealment, adherence to assigned strategies, or handling of missing outcomes. Randomization is important design information, but it does not remove uncertainty from an estimate or answer questions about unreported analysis choices.

These methods explanations are general. The registry record does not report the original sample-size justification, a detailed analysis plan, interim analyses, or an adjusted primary analysis. Those items should not be inferred from the posted comparison or from planning illustrations based on the observed effect.

10. Confidence Intervals and Precision

The reported risk-difference interval crosses zero, extending from a small negative value to a positive value. This indicates limited precision for deciding whether the underlying difference is below or above no difference. The unadjusted p-value is 0.108, consistent with that uncertainty. The estimate's positive direction is informative about the observed comparison, but the interval shows that the data are also compatible with a negative difference.

Precision concerns the range of effects compatible with the data, not whether the observed estimate is important in practice. A confidence interval is not a probability statement about the true effect after observing the data. It is a range generated by a statistical procedure with stated assumptions. The registry does not report enough analysis detail to assess all assumptions or to compare alternative interval methods.

The outcome denominators are 270 in the intervention arm and 268 in standard care. These counts give the scale of the posted comparison, but the record does not explain missing endpoint data or the analysis population in detail. The denominator information alone cannot establish how participants without recorded endpoint outcomes were handled. Separate component-specific estimates and intervals are not reported in the registry record, so precision for viral suppression alone or retention alone cannot be assessed from this result.

Overall, the point estimate gives a direction, while the interval makes clear that the direction and magnitude remain uncertain. The posted result does not support a precise claim that the intervention strategy differs from standard care on the combined endpoint.

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 and not total events. They are illustrations computed from the observed effect, using a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. Each row considers detecting a stated fraction of the observed effect: 100%, 75%, or 50%. The trial's original sample-size assumptions and justification are not reported in the registry record.

The 80% power column describes the number of participants per arm that the planning calculation associates with an 80% chance of detecting the specified effect under its assumptions, if that effect is present. The 90% power column describes the corresponding planning target for a 90% chance. For planning at 100% of the observed effect, the figures are 819 participants per arm for 80% power and 1,095 participants per arm for 90% power. At 75% of the observed effect, they are 1,472 and 1,970 per arm. At 50% of the observed effect, they are 3,347 and 4,480 per arm.

Smaller effects require more participants for the same target power because they are harder to distinguish from random variation. These numbers are planning illustrations based on the observed result, not observed power. They do not establish that the trial was designed around these assumptions, and they should not be read as the trial's original target enrollment.

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 = 0.623; treatment proportion p₁ = 0.689; average p̄ = 0.656.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6718 = 1.3167.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6702 = 0.5640.
4. n = (1.3167 + 0.5640)² / (0.0658)² = 818.1, rounded up to 819 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 <- 0.623134328358209                          # control proportion
p1 <- 0.688888888888889                          # 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(4341779)
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 = 819 (80% power), 1,095 (90% power)
 75% of observed effect: n per arm = 1,472 (80% power), 1,970 (90% power)
 50% of observed effect: n per arm = 3,347 (80% power), 4,480 (90% power)
power.prop.test: 818.1 
Simulated power at n = 819 per arm: 0.807 (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 a primary endpoint and two secondary outcomes: acceptability of point-of-care tenofovir and viral-load testing, and cost-effectiveness compared with standard-of-care monitoring. The record also reports four outcome measures, but does not provide results for the secondary outcomes. It gives no analyses in the posted analysis count. A detailed statistical analysis plan, including how multiple endpoints or tests were to be handled, is not reported in the registry record.

Multiplicity matters when several outcomes, time points, subgroups, or statistical tests are considered. The more comparisons made, the greater the chance that at least one may appear notable by chance if each is assessed without an adjustment or prespecified hierarchy. The record does not say whether such a hierarchy or adjustment was planned, so it is not possible to determine how the unadjusted p-value relates to any broader testing strategy.

Missing-data methods are also not described. The primary endpoint comparison includes 270 participants in the intervention arm and 268 in standard care, while enrollment is listed as 539. The registry record does not explain the relationship between these figures, identify participants without endpoint data, or state whether missing outcomes were imputed, excluded, or handled through another method. No assumptions about missingness should be inferred from the counts alone.

The record does not specify the analysis population, such as whether all randomized participants were included according to assigned arm, nor does it report sensitivity analyses for missing outcomes. It also does not report an adjusted primary analysis or interim analyses. These omissions limit assessment of how robust the posted comparison is to analysis choices. They do not demonstrate that the trial lacked such procedures; they mean the procedures are not reported in 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.

The serious adverse event record lists zero serious adverse events in the intervention arm among 270 participants at risk and zero in standard care among 269 participants at risk. The record separately lists 5 deaths in each arm. These are distinct reported figures: the absence of recorded serious adverse events in these fields should not be treated as an absence of deaths.

The figures are counts recorded in the registry, not comparative estimates with confidence intervals. No serious adverse event terms are listed among the top terms. The registry does not provide descriptions of the deaths, their timing, causes, relationship to study procedures, or any adjudication details. It also does not explain why the safety denominators differ from the primary endpoint denominators.

These entries do not establish that the monitoring strategies caused or did not cause any event. They describe what is recorded for the two groups in the posted safety data. Without further detail on collection, definitions, follow-up, and attribution, the counts should be interpreted narrowly rather than as a complete safety comparison.

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

The result describes the proportion of participants meeting a combined endpoint of viral-load suppression below 200 copies/mL and retention in care at 72 weeks after ART initiation. The endpoint count was 186 of 270 in the intervention arm and 167 of 268 in standard care. The reported risk-difference estimate is positive, so the observed comparison is in the direction of more participants meeting the combined endpoint in the intervention arm.

The interval crosses no difference, and the unadjusted p-value is 0.108. The result therefore does not establish a clear difference between the arms, nor does it establish that the strategies are equivalent. It indicates uncertainty about the size and direction of the underlying contrast. The comparison does not show which component of the combined endpoint accounts for the observed pattern, because separate component results are not reported in the registry record.

The result does not answer whether the intervention is cost-effective or acceptable to participants and providers. Although those are listed as secondary outcomes, results for them are not reported in the registry record. Nor does this endpoint establish effects on outcomes not included in the posted comparison.

Randomization is a strength of the design, but the record does not describe key analytic details such as the handling of missing outcomes, the analysis population, or adjustments for other variables. The status is active, not recruiting, and the record gives a primary completion date of 2025-06-30; it does not clarify whether these results are final. The posted data should therefore be read as a registry-recorded comparison with the stated limitations, not as a complete account of every trial question or analysis.

16. Why This Trial Matters Statistically

This trial poses a statistical question about an integrated care-monitoring strategy, not a single laboratory test in isolation. Its primary outcome combines viral suppression with retention, so interpretation requires attention to both the binary endpoint and the care-engagement component. A favorable direction in a combined measure does not reveal which component contributes to the pattern unless each is also reported separately.

The randomized parallel-group design supports a direct comparison between the point-of-care approach and standard care. At the same time, the confidence interval crosses the no-difference value, making precision central to interpretation. A point estimate alone could suggest a clearer distinction than the full uncertainty supports; the interval and unadjusted p-value keep that uncertainty visible.

The planning illustrations also show why sample-size assumptions matter. Their participant targets vary with the fraction of the observed effect used for planning, and smaller effects require larger planned groups to reach the same power target. These illustrations cannot be substituted for the trial's own design rationale, which is not reported in the registry record.

Finally, the registry's limited reporting of missing-data handling, analysis populations, multiplicity, and secondary outcomes affects how comprehensively the result can be assessed. Distinguishing what the posted comparison estimates from what remains unreported is essential to statistical interpretation. The endpoint result is informative about the recorded combined outcome, but it cannot answer every clinical, operational, or economic question raised by the study.

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 result is for a combined endpoint, and the registry does not post separate results for viral-load suppression and retention in care. As a result, the endpoint comparison cannot identify which component contributes to the observed contrast. The record also does not provide detail about the number or characteristics of participants missing endpoint data, or how such data were handled.

The registry does not describe the analysis population, adjusted models, sensitivity analyses, interim analyses, or the detailed statistical analysis plan. The unadjusted result is therefore difficult to place in the context of other prespecified analyses or robustness checks. Multiplicity procedures are not reported, despite the presence of secondary outcomes. Results for acceptability and cost-effectiveness are not reported in the registry record.

The outcome denominators are 270 and 268, while enrollment is listed as 539. The record does not explain this difference or how the reported denominators were formed. Safety data use denominators of 270 and 269, and the reason for that distinction is also not reported. These figures should not be reconciled by assuming a particular participant flow or analysis rule.

The registry status is active, not recruiting, and the primary completion date is 2025-06-30. The record does not state whether the posted results are final or describe the completeness of follow-up. Safety entries list deaths but do not provide causes, timing, or attribution. These reporting limitations constrain interpretation of the posted comparison; they do not establish that unreported procedures or information do not exist elsewhere.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Simplifying Treatment and Monitoring for HIV (STREAM HIV) — NCT04341779.
• Registry record last updated 2026-02-19; status last verified 2026-01.
• Results first posted 2026-02-19.
• 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

STREAM HIV is a randomized, unmasked, parallel-group study of point-of-care urine tenofovir adherence and viral-load testing compared with standard-of-care monitoring among people initiating first-line tenofovir disoproxil fumarate-based ART in South Africa. The registry lists enrollment of 539, status as active, not recruiting, and primary completion on 2025-06-30.

The primary endpoint is the number of participants with viral-load suppression below 200 copies/mL and retention in care at 72 weeks after ART initiation. The posted counts are 186 of 270 in the intervention arm and 167 of 268 in standard care. The risk-difference estimate is positive, but its interval crosses no difference; the unadjusted p-value is 0.108. Safety entries list zero serious adverse events in each arm and 5 deaths in each arm.

Results for the listed secondary outcomes, the original sample-size justification, and detailed analysis procedures are not reported in the registry record.