Safety and Efficacy of COBI-boosted Atazanavir Versus Ritonavir-boosted Atazanavir Each Administered With Emtricitabine/Tenofovir Disoproxil Fumarate in HIV-1 Infected, Antiretroviral Treatment-Naive Adults: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Safety and Efficacy of COBI-boosted Atazanavir Versus Ritonavir-boosted Atazanavir Each Administered With Emtricitabine/Tenofovir Disoproxil Fumarate in HIV-1 Infected, Antiretroviral Treatment-Naive Adults (NCT01108510), a randomized phase 3 study. The primary outcome is Percentage of Participants With HIV-1 RNA < 50 Copies/mL at Week 48.

2. Clinical Question

The question concerns a comparison between two complete treatment regimens, not a comparison of cobicistat and ritonavir given alone. Both groups received atazanavir and the same emtricitabine/tenofovir disoproxil fumarate background combination. The reported primary outcome is a binary endpoint: each participant is classified according to whether the viral-load threshold was met at the specified assessment, and the group result is expressed as a percentage.

The posted Week 48 percentages were 85.2% for atazanavir plus cobicistat and 87.4% for atazanavir plus ritonavir. The registry also identifies a sponsor analysis as non-inferiority or equivalence analysis, but the non-inferiority margin and the decision rule are not reported in the registry record. Therefore, the endpoint percentages and their uncertainty can be described, but the registry information alone does not establish the formal criterion used to declare non-inferiority.

3. Trial Design

This was a Phase 3, randomized, parallel-group study with triple masking. The official title describes it as double-blind; the registry’s masking field specifies triple masking. Enrollment was 698, and the study status is completed. The registry lists a start in 2010-04 and primary completion in 2011-11.

Participants were assigned to one of two once-daily regimens: atazanavir plus cobicistat and emtricitabine/tenofovir disoproxil fumarate, or atazanavir plus ritonavir and emtricitabine/tenofovir disoproxil fumarate. Placebos for the respective boosting agents were included in the arm descriptions. The registry summary states that allocation was planned in a 1:1 ratio and stratified by HIV-1 RNA level at screening, with categories at or below 100,000 copies/mL and above 100,000 copies/mL.

The primary analysis population comprised participants who were randomized and received at least one dose. The posted group counts for the Week 48 outcome were 344 and 348. These analysis counts total the registry’s stated enrollment, but the record does not report a more detailed account of screening, exclusions, or participant flow. The endpoint assessed the percentage with HIV-1 RNA below 50 copies/mL at Week 48 using the snapshot algorithm, which uses viral load at the specified time point within an allowed window together with study-drug discontinuation status.

The registry lists additional viral-load outcomes at Week 96, Week 144, and Week 192, and CD4 cell-count changes at Weeks 48, 96, 144, and 192. Numeric results are posted for the viral-load outcomes, while numeric results for the listed CD4 outcomes are not reported in the registry record. Details such as the trial’s original sample-size justification, allocation concealment procedures, and prespecified 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: 867 participants were screened.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Intent-to-Treat (ITT) Analysis Set: participants who were randomized and received at least one dose of study drug

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.

Population: Safety Analysis Set: participants who were randomized and received at least on dose of study drug

6. Primary Endpoint

The registry lists the primary outcome as Percentage of Participants With HIV-1 RNA < 50 Copies/mL at Week 48, measured over Week 48.

The primary endpoint was the percentage of participants with HIV-1 RNA below 50 copies/mL at Week 48. It is a binary outcome summarized at group level: a participant either meets the viral-load threshold under the endpoint definition or does not. The primary analysis population was the Intent-to-Treat Analysis Set, defined in the registry as randomized participants who received at least one dose of study drug.

The endpoint used the snapshot algorithm. According to the registry description, response status depends on the viral load at the prespecified time point within an allowed time window and on study-drug discontinuation status. This makes the endpoint more than a reading of a single laboratory value in isolation: discontinuation status is part of the classification rule. The precise details of the allowed window and the handling of every possible discontinuation circumstance are not reported in the registry record.

As a general statistical point, a threshold endpoint can be clinically and operationally straightforward to interpret, but it discards information about how far a participant’s viral load lies above or below the threshold. The group percentages describe the proportion meeting the defined criterion; they do not describe the full distribution of viral-load values or, by themselves, the reasons participants did not meet the criterion. The registry reports 85.2% for the cobicistat regimen and 87.4% for the ritonavir regimen at Week 48.

7. Results Posted in the Registry

• ATV+RTV+FTC/TDF: 87.4% (304/348) versus 85.2% (293/344) on control; risk difference 2.2 percentage points (95% CI -2.9 to 7.3), risk ratio 1.03 (0.97–1.09); unadjusted two-proportion z-test p = 0.404.

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.

Percentage of Participants With HIV-1 RNA < 50 Copies/mL at Week 96

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • ATV+RTV+FTC/TDF: 79.3% (276/348) versus 77.9% (268/344) on control; risk difference 1.4 percentage points (95% CI -4.7 to 7.5), risk ratio 1.02 (0.94–1.10); unadjusted two-proportion z-test p = 0.653.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between ATV+RTV+FTC/TDF and the control. The interval gives the range of effects that remain compatible with the data.

Percentage of Participants With HIV-1 RNA < 50 Copies/mL at Week 144

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • ATV+RTV+FTC/TDF: 74.1% (258/348) versus 72.1% (248/344) on control; risk difference 2.0 percentage points (95% CI -4.6 to 8.7), risk ratio 1.03 (0.94–1.13); unadjusted two-proportion z-test p = 0.544.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between ATV+RTV+FTC/TDF and the control. The interval gives the range of effects that remain compatible with the data.

Percentage of Participants With HIV-1 RNA < 50 Copies/mL at Week 192

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • ATV+RTV+FTC/TDF: 79.7% (55/69) versus 71.6% (53/74) on control; risk difference 8.1 percentage points (95% CI -5.9 to 22.1), risk ratio 1.11 (0.92–1.34); unadjusted two-proportion z-test p = 0.261.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between ATV+RTV+FTC/TDF and the control. The interval gives the range of effects that remain compatible with the data.

The posted secondary viral-load percentages follow a broadly similar pattern across the earlier scheduled assessments: the ritonavir group’s percentage is higher than the cobicistat group’s at Weeks 96 and 144. At Week 192, the posted percentages also favor the ritonavir group numerically, but the analysis counts are much smaller than at the earlier assessments. The recalculated intervals for each of these comparisons include zero. These observations do not replace the primary Week 48 result and should not be treated as independent confirmation of a treatment difference.

The Week 96, Week 144, and Week 192 outcomes use the same threshold of HIV-1 RNA below 50 copies/mL. Their posted group results are follow-up outcomes at distinct time points, not repeated measurements that can automatically be combined into one overall conclusion. The Week 192 analysis counts are 74 and 69; the registry does not explain the reasons for those counts in the information. Numeric results for the listed CD4 cell-count change outcomes are not reported in the registry record.

As a general statistical point, multiple secondary outcomes and time points create multiple opportunities to observe an apparently notable result by chance. Unless the testing strategy accounts for multiplicity, individual secondary p-values can have less evidential weight than a single prespecified primary comparison. The registry does not report a multiplicity adjustment or a complete hierarchy of testing. The secondary results are therefore best read as supporting descriptive information, with attention to their confidence intervals and analysis populations, rather than as a set of separate definitive tests.

The overall pattern is compatible with the primary result in the limited sense that the posted comparisons do not show a clear separation established by their recalculated intervals. It does not prove that the regimens have identical effects at every time point. Nor does the sequence of percentages establish why outcomes changed over time, because the registry record does not provide participant-level trajectories or detailed reasons for endpoint classification.

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

• Percentage of Participants With HIV-1 RNA < 50 Copies/mL at Week 96
• Percentage of Participants With HIV-1 RNA < 50 Copies/mL at Week 144
• Percentage of Participants With HIV-1 RNA < 50 Copies/mL at Week 192
• Change From Baseline in CD4 Cell Count at Week 48
• Change From Baseline in CD4 Cell Count at Week 96
• Change From Baseline in CD4 Cell Count at Week 144
• Change From Baseline in CD4 Cell Count at Week 192

9. Statistical Interpretation

10. Statistical Methods Explained

What does the risk difference represent?

For a binary endpoint, the risk difference compares the proportions meeting the endpoint in two groups. In the registry’s recalculation, the comparison is atazanavir plus ritonavir relative to atazanavir plus cobicistat. The Week 48 estimate is 0.022. The corresponding interval ranges from -0.029 to 0.073. These values are reported on the risk-difference scale used for the recalculation; they should not be silently converted to another scale or confused with the sponsor’s reported difference in percentages.

A risk difference is an absolute comparison of proportions. Unlike a relative measure, it does not express one group’s proportion as a multiple of the other group’s proportion. The posted percentages offer a direct description of each group, while the estimated risk difference and its interval describe the between-group contrast and its uncertainty.

What does the p-value tell the reader?

The unadjusted p-value for the recalculated Week 48 comparison is 0.404. In general, a p-value is calculated under a null hypothesis and a statistical model; it measures how compatible the observed data are with that null model. It is not the probability that the null hypothesis is true, nor is it a measure of clinical importance. A value that does not cross a chosen significance threshold does not prove there is no difference, and it cannot establish equivalence without an appropriate design and margin.

The registry does not report an unadjusted p-value for the sponsor’s analysis. The reported sponsor result is a difference in percentages with a confidence interval. The registry’s non-inferiority or equivalence label does not supply the missing margin or the complete decision rule.

Why use a confidence interval?

A confidence interval gives a range of values compatible with the data under the method used, with a stated confidence level. It communicates both the estimated contrast and its precision. The recalculated Week 48 interval extends across zero. This means the data do not distinguish a positive contrast from a negative contrast at the level represented by that interval. The interval also shows that the uncertainty is not limited to values very close to the point estimate.

Confidence intervals do not guarantee that the true value lies within a particular interval, and their interpretation depends on the analysis method and assumptions. The sponsor’s reported interval has a stated confidence level of 95.2%; the posted secondary sponsor intervals are stated as 95%. The recalculated interval and sponsor interval are not interchangeable because they arise from different reported analyses and are presented with different contrast conventions.

Why does the snapshot definition matter?

The primary endpoint uses the snapshot algorithm, which combines the viral-load result at the prespecified time point within an allowed window with study-drug discontinuation status. In general, a composite classification rule can affect who counts as a responder and how missing or unavailable endpoint information is handled. The registry description identifies the broad components, but the detailed rules for each discontinuation and missing-data circumstance are not reported in the registry record. Interpretation should therefore remain tied to the stated algorithm rather than treating the percentages as an unqualified measure of viral load at any time.

How should the group counts be understood?

The Week 48 outcome is reported for 344 participants in the cobicistat group and 348 in the ritonavir group. The primary population includes randomized participants who received at least one dose. A binary percentage summarizes the number meeting the endpoint relative to the analysis group, but does not display participant-level patterns. The registry gives the event counts used for recalculation as 293 and 304, respectively. Those counts and group denominators underpin the posted percentages and contrast; no further participant-level data are reported.

What does masking and randomization contribute?

As a general principle, random allocation is intended to reduce systematic differences between treatment groups at baseline, while masking can limit some forms of bias in care and outcome assessment. The registry identifies the study as randomized and triple masked, and its official title describes a double-blind study. The summary also describes 1:1 assignment stratified by screening HIV-1 RNA category. The registry does not report detailed implementation procedures, so the design labels should not be taken as a substitute for a full account of allocation concealment or masking success.

What cannot be inferred from the posted test?

The recalculated test is unadjusted. It does not incorporate a reported multiplicity procedure, and it should not be described as the trial’s complete prespecified statistical plan. The registry’s original sample-size assumptions, interim analyses, and full analysis specifications are not reported in the registry record. General statistical interpretation can explain the meaning of estimates and intervals, but it cannot fill in those unreported protocol details.

11. Confidence Intervals and Precision

The Week 48 recalculated risk difference is 0.022, with a confidence interval from -0.029 to 0.073. The interval includes zero and extends to values in both directions. This is a reminder that the point estimate alone is not a complete account of the result: plausible values under the interval method include contrasts on either side of no difference. The interval is not narrow enough to support a claim of exact equality.

The sponsor reports a difference in percentages of -2.2, with a 95.2% confidence interval from -7.4 to 3.0. This reported contrast is presented for the groups in the order cobicistat then ritonavir, unlike the orientation of the recalculated risk difference. The registry’s non-inferiority or equivalence label does not provide the margin needed to judge the interval against a formal non-inferiority criterion. The sponsor’s full statistical method is not reported in the registry record.

As a general explanation, a confidence interval reflects sampling uncertainty under the method used, not all possible sources of bias or uncertainty. Its width depends on factors including the amount of information and the variability of the data. Here, the interval should be read alongside the group counts and the endpoint definition. The primary comparison is based on 344 and 348 participants, while later outcomes have different analysis counts. In particular, Week 192 has 74 and 69 participants, so its estimates have less participant information than the earlier posted viral-load comparisons.

Neither the confidence interval nor the p-value alone determines clinical importance. They describe statistical uncertainty in a specified contrast. The registry does not report a threshold for a clinically important difference, so statistical precision cannot be translated into a judgment about the importance of a particular contrast.

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 illustrations are for participants per arm, not a total across both arms and not total events. They use a two-sided significance level of 0.05, equal allocation, and a normal approximation. The effect to detect is set to the stated fraction of the observed effect. The trial’s own original design assumptions and sample-size justification are not reported in the registry record.

For the primary Week 48 comparison, planning for 100% of the observed effect gives 3,906 participants per arm for 80% power and 5,229 participants per arm for 90% power. Planning for 75% of the observed effect gives 7,060 participants per arm for 80% power and 9,451 participants per arm for 90% power. Planning for 50% of the observed effect gives 16,144 participants per arm for 80% power and 21,612 participants per arm for 90% power.

In plain terms, 80% power is the probability, under the specified assumptions and if the chosen effect is real, of detecting that effect using the stated test. The 90% column represents a higher target probability of detection under those same assumptions. The corresponding required group sizes are larger. A smaller effect is harder to distinguish from random variation, so planning to detect a smaller fraction of the observed effect requires more participants per arm.

These figures are hypothetical planning illustrations based on the observed effect, not the enrollment target used to design this trial. They are not observed power and do not tell whether the completed study was adequately designed. The planning relationship is read in the same way across the curve: assumptions about smaller effects correspond to larger required participant counts, while the power target and statistical assumptions also shape the planned size.

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 = 0.852; treatment proportion p₁ = 0.874; average p̄ = 0.863.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.4868 = 0.9541.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.4865 = 0.4095.
4. n = (0.9541 + 0.4095)² / (0.0218)² = 3,905.6, rounded up to 3,906 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 <- 0.851744186046512                          # control proportion
p1 <- 0.873563218390805                          # 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(1108510)
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 = 3,906 (80% power), 5,229 (90% power)
 75% of observed effect: n per arm = 7,060 (80% power), 9,451 (90% power)
 50% of observed effect: n per arm = 16,144 (80% power), 21,612 (90% power)
power.prop.test: 3905.6 
Simulated power at n = 3,906 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.

14. Analyses Reported by the Sponsor

The registry reports the sponsor’s primary analysis as a difference in percentages of -2.2, with a 95.2% confidence interval from -7.4 to 3.0. It labels this analysis as non-inferiority or equivalence. The registry does not report a p-value for this analysis, its method, the non-inferiority margin, or the full decision rule. Without the margin and rule, the label alone is not enough to independently verify a formal non-inferiority conclusion.

The registry also gives an unadjusted recalculation: risk difference 0.022, confidence interval from -0.029 to 0.073, and p-value 0.404. This recalculation compares the ritonavir regimen with the cobicistat regimen, while the sponsor’s reported group order is cobicistat followed by ritonavir. The signs therefore reflect different contrast directions and should not be compared as if they use the same orientation. The analyses also use different reported forms.

The discrepancy flag is false. The recalculation is a simple unadjusted calculation from posted summary statistics; it is not a replacement for the sponsor’s own analysis. The registry’s reported sponsor result should be read as the trial’s reported analysis, while the recalculation offers a separate view of the posted data and its uncertainty. No unreported adjustment or analysis method should be inferred.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists one primary outcome and several secondary outcomes, including viral-load response at Weeks 96, 144, and 192 and CD4 cell-count change at Weeks 48, 96, 144, and 192. When several outcomes or time points are tested, the chance of at least one apparently unusual result can increase if each is judged using an unadjusted threshold. As a general statistical principle, multiplicity procedures can manage this issue, but the registry record does not report whether such a procedure was planned or used.

The primary endpoint description specifies the snapshot algorithm and the Intent-to-Treat Analysis Set. It states that the endpoint uses viral load within an allowed time window together with study-drug discontinuation status. The registry does not report the full rules for all missing measurements, withdrawals, discontinuations, or other circumstances that may affect classification. Those details matter because the definition determines how participants without a straightforward Week 48 result contribute to the endpoint.

The record also does not report a detailed analysis plan, interim analyses, the trial’s original sample-size justification, or the non-inferiority margin. These omissions prevent a complete assessment of the planned testing hierarchy, handling of missing data, and formal non-inferiority decision rule. They should not be filled by assumptions based on common practice.

The sponsor analysis is labeled non-inferiority or equivalence, while the registry’s independent recalculation is unadjusted. These should be kept distinct: an unadjusted difference test does not answer every question addressed by a non-inferiority framework. The posted data support description of the observed percentages and their reported uncertainty, but the record does not contain all information needed to reconstruct the trial’s full confirmatory statistical strategy.

16. Safety Results

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

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: subcutaneous abscess (1.2%), renal failure acute (1.1%), chest pain (0.9%), pyrexia (0.9%), herpes zoster (0.9%).

The registry reports serious adverse events for 64 of 344 participants in the atazanavir-plus-cobicistat group and 50 of 348 participants in the atazanavir-plus-ritonavir group. These are counts of participants recorded as affected among the listed groups. They do not by themselves describe event severity beyond the registry’s serious-event classification, timing, duration, or whether an event was considered related to study treatment.

The record also gives other adverse-event counts of 311 of 344 and 312 of 348, respectively. It does not provide additional detail in these figures about event types or their clinical significance. Death counts are not reported in the registry record.

The registry lists subcutaneous abscess at 1.2%, renal failure acute at 1.1%, and chest pain, pyrexia, herpes zoster, and depression each at 0.9% among the top serious-event terms. The listed terms are not accompanied by arm-specific counts in the information reported. These summaries should not be read as a complete safety profile or as evidence about causality. Comparing safety requires considering event definitions, exposure, follow-up, and the full clinical context, details that are not all reported in the registry record.

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

The posted results describe a randomized comparison of two atazanavir-based regimens in antiretroviral treatment-naive adults with HIV-1. At Week 48, 85.2% of the cobicistat group and 87.4% of the ritonavir group met the endpoint definition of HIV-1 RNA below 50 copies/mL. The recalculated unadjusted risk difference has an interval that includes zero. The sponsor’s reported analysis gives a difference in percentages of -2.2 with a 95.2% confidence interval from -7.4 to 3.0.

These findings do not establish that the regimens are exactly equivalent, and a non-significant unadjusted comparison is not proof of no difference. A formal non-inferiority conclusion depends on the margin and decision rule, neither of which is reported in the registry record. The sponsor’s analysis is identified as non-inferiority or equivalence, but the available record does not provide every detail needed to evaluate that framework independently.

The endpoint is a threshold-based classification at a specified time point, using the snapshot algorithm. It does not show the full distribution of viral-load measurements or explain why individual participants did not meet the threshold. The secondary viral-load results provide further time-point summaries, but those estimates are uncertain, and the Week 192 counts are smaller. The numeric results for the listed CD4 outcomes are not reported in the registry record.

These results also do not determine what an individual should do, establish outcomes in populations outside the enrolled study population, or resolve questions not measured by the posted endpoints. They are statistical summaries of the registry’s reported study outcomes. Any broader clinical interpretation requires the complete trial report and relevant evidence beyond the information recorded.

18. Why This Trial Matters Statistically

This trial illustrates why the treatment contrast, outcome definition, and analysis population all matter in interpreting a clinical study. The two groups shared atazanavir and emtricitabine/tenofovir disoproxil fumarate; the boosting agent differed. The result therefore addresses the comparison of complete regimens as assigned, rather than the isolated effect of either boosting drug.

It also illustrates the distinction between a difference test and a non-inferiority question. A conventional unadjusted comparison asks whether the data provide evidence of a difference under its null model. Non-inferiority asks whether the data exclude a prespecified degree of unacceptable disadvantage. The latter requires a margin and an analysis rule. The registry identifies the sponsor’s analysis as non-inferiority or equivalence but does not report the margin, making the distinction especially important.

Finally, the study shows why uncertainty and follow-up counts belong alongside percentages. Similar-looking response percentages are not proof of identical effects, and later estimates based on fewer participants may be less precise. Secondary outcomes across several visits also raise multiplicity questions. Reading the estimates, intervals, endpoint rules, and analysis limitations together gives a more disciplined account than relying on a single percentage or p-value.

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.
• Participant counts were reconstructed from posted percentages and arm sizes, so small rounding differences are possible.

The registry record does not report the non-inferiority margin, the sponsor analysis method, or the full decision rule, so the formal non-inferiority interpretation cannot be independently reconstructed. It also does not provide the trial’s original sample-size justification, detailed protocol assumptions, or full statistical analysis plan.

The primary endpoint depends on the snapshot algorithm, but detailed rules for its allowed time window and handling of specific missing-data or discontinuation circumstances are not reported. Participant flow and reasons for the smaller Week 192 analysis counts are not given. Numeric results for the listed CD4 cell-count outcomes are also not reported in the registry record.

Multiplicity handling across outcomes and time points is not reported. The recalculated comparisons are unadjusted and should not be mistaken for the complete sponsor analysis. The registry provides aggregate counts and percentages, not participant-level data, so it does not permit examination of outcome trajectories, baseline balance, or subgroup patterns beyond the described stratification. Safety summaries lack detailed event timing, treatment attribution, and arm-specific counts for the listed serious-event terms. These are limits of the information reported in the registry record, not proof that the underlying study lacked such procedures or data.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Safety and Efficacy of COBI-boosted Atazanavir Versus Ritonavir-boosted Atazanavir Each Administered With Emtricitabine/Tenofovir Disoproxil Fumarate in HIV-1 Infected, Antiretroviral Treatment-Naive Adults — NCT01108510.
• Registry record last updated 2016-05-23; status last verified 2016-04.
• Results first posted 2014-10-28.
• 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

NCT01108510 is a completed Phase 3 randomized, triple-masked, parallel-group study sponsored by Gilead Sciences. It enrolled 698 antiretroviral treatment-naive adults with HIV-1 and compared once-daily atazanavir plus cobicistat and emtricitabine/tenofovir disoproxil fumarate with atazanavir plus ritonavir and the same background combination.

The primary endpoint was the percentage with HIV-1 RNA below 50 copies/mL at Week 48 using the snapshot algorithm. The posted percentages were 85.2% and 87.4%. The registry reports a sponsor non-inferiority or equivalence analysis and an unadjusted recalculation; key details of the formal non-inferiority framework are not reported in the registry record. Secondary viral-load outcomes are posted at Weeks 96, 144, and 192, while numeric CD4 results are not reported in the registry record.