This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record.
1. Trial at a Glance
DUALIS was a randomized, parallel, phase 3 trial evaluating dual therapy with boosted darunavir plus dolutegravir in HIV-Infection. The registry reports 269 enrolled participants and two treatment arms. The primary endpoint was the number and percentage of patients with fully suppressed HIV RNA below 50 cps/ml at week 48.
| Feature | DUALIS |
|---|---|
| Study acronym | DUALIS |
| ClinicalTrials.gov identifier | NCT02486133 |
| Phase | Phase 3 |
| Status | Completed |
| Condition | HIV-Infection |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | None |
| Primary purpose | Treatment |
| Enrollment | 269 |
| Number of arms | 2 |
| Lead sponsor | Technical University of Munich |
| Sponsor type | Other |
| Study start | 2015-07 |
| Primary completion | 2018-05-31 |
2. Clinical Question
The central clinical question was whether a dual-therapy strategy based on boosted darunavir plus dolutegravir could be evaluated against the trial's comparator arm with respect to virologic suppression at week 48 in people with HIV-Infection.
Population
Participants enrolled in the phase 3 DUALIS trial with the registry-listed condition of HIV-Infection.
Intervention
The trial title identifies dual therapy with boosted darunavir plus dolutegravir as the treatment strategy under evaluation.
Comparator
The registry describes two randomized arms, but the available record does not specify the arm-level comparator regimen in the information reported here.
Primary question
What proportion of patients have fully suppressed HIV RNA below 50 cps/ml at week 48, and how does that proportion compare between the randomized arms?
3. Trial Design
Interventions listed in the registry
The registry lists six drug interventions: Prezista, Norvir, Tivicay, Truvada, Kivexa, and Descovy. The available record identifies the overall dual-therapy strategy in the study title but does not provide enough arm-level detail here to assign each listed drug to a particular randomized arm.
| Registry-listed intervention | Type |
|---|---|
| Prezista | Drug |
| Norvir | Drug |
| Tivicay | Drug |
| Truvada | Drug |
| Kivexa | Drug |
| Descovy | Drug |
4. Trial Timeline
Study start
The DUALIS study began in July 2015.
Randomized treatment comparison
The registry describes a randomized, parallel, unmasked phase 3 treatment study with two arms.
Primary endpoint assessment
The primary endpoint assesses HIV RNA suppression at week 48 using NAT diagnostic testing.
Primary completion
The registry lists May 31, 2018 as the primary completion date.
5. Primary Endpoint
| Endpoint | Registry definition | Time frame |
|---|---|---|
| HIV RNA suppression | Number (%) of patients with fully suppressed HIV RNA < 50 cps/ml at week 48 | 48 weeks |
The registry specifies that HIV RNA suppression below 50 cps/ml was to be assessed at week 48 using a NAT diagnostic. The endpoint is therefore a binary outcome at a prespecified time point: each participant is classified according to whether the HIV RNA threshold has been met at week 48.
Primary statistical quantity
Patients with fully suppressed HIV RNA < 50 cps/ml at week 48
The primary endpoint is expressed as both a count and a percentage, allowing the result to be understood as an absolute proportion within each randomized treatment group.
6. Planned Analysis
The ClinicalTrials.gov record does not contain posted statistical analyses or numerical outcome results for the primary endpoint. The registry does specify what is to be measured: the number and percentage of patients with fully suppressed HIV RNA below 50 cps/ml at week 48.
How this endpoint is typically analyzed
For a randomized two-arm trial with a binary endpoint, the fundamental analysis is a comparison of the proportion meeting the response criterion in each treatment group. The observed proportion is calculated as the number of participants classified as suppressed divided by the number of participants included in the relevant analysis population.
where x is the number of participants with HIV RNA < 50 cps/ml at week 48 and n is the corresponding number of participants in the analysis population.
A treatment comparison can then be expressed using an absolute difference in proportions, a risk ratio, an odds ratio, or another prespecified binary-outcome estimand. Confidence intervals quantify uncertainty around the selected effect measure. A formal hypothesis test can also be used to assess evidence against a prespecified null hypothesis.
The registry does not specify which of these comparative measures or tests was prespecified for DUALIS. Consequently, they should be understood here as standard statistical approaches for this endpoint type rather than as documented DUALIS analysis methods.
Why the percentage is clinically informative
A percentage directly answers a clinically intuitive question: what fraction of participants achieved the virologic suppression threshold by the specified week? Unlike a relative effect measure, the percentage retains the absolute scale of the outcome.
For example, if two hypothetical groups had suppression proportions of 80% and 70%, the absolute difference would be 10 percentage points. A relative measure would express the same comparison differently. These measures are mathematically related but communicate different aspects of the treatment contrast.
7. Statistical Methodology
Binary endpoint analysis
The week-48 endpoint is naturally represented as a binary outcome. Each participant has one of two endpoint classifications: fully suppressed HIV RNA below 50 cps/ml at week 48, or not meeting that threshold according to the endpoint definition.
An absolute risk difference compares the observed suppression proportions between the two randomized groups. A positive value means the first group's observed proportion is higher; a negative value means it is lower.
Risk ratio
A risk ratio compares the probability of the endpoint between treatment groups. A ratio of 1 indicates equal proportions. Values above 1 indicate a higher observed probability in the numerator group, while values below 1 indicate a lower observed probability.
The risk ratio is a relative measure and should be interpreted alongside the underlying proportions. The same relative ratio can correspond to very different absolute differences depending on baseline event rates.
Odds ratio
An odds ratio compares the odds of suppression rather than the probabilities themselves. It can be useful in binary-outcome modeling, but it is not interchangeable with a risk ratio, particularly when the endpoint is common.
Confidence intervals
A confidence interval around a treatment-effect estimate describes statistical uncertainty associated with the estimate under the specified analytical framework. A narrow interval indicates greater precision than a wide interval, all else being equal.
The interval should not be interpreted as a range containing the true individual response probability for each participant. It quantifies uncertainty about a population-level parameter estimated from the trial.
Hypothesis testing
For a binary endpoint, a hypothesis test could compare the observed suppression proportions under a null hypothesis of no difference between treatment groups. Depending on the prespecified analysis plan and sample sizes, large-sample methods or exact methods may be appropriate.
The p-value would address the compatibility of the observed data with the specified null hypothesis. It would not measure the size or clinical importance of the treatment effect.
8. Statistical Methods Explained
Why is the primary endpoint a percentage?
The registry defines the primary endpoint as the number and percentage of patients with fully suppressed HIV RNA below 50 cps/ml at week 48. This produces a directly interpretable binary outcome: whether each participant reached the prespecified virologic threshold at the specified time.
Why is week 48 important?
The endpoint has a fixed 48-week time frame. Fixing the assessment time means the treatment groups are compared on the same temporal scale rather than allowing different follow-up durations to define the primary outcome.
Why use a randomized comparison?
Randomization creates the framework for comparing the two treatment strategies while reducing systematic differences in baseline characteristics between groups in expectation. It is the treatment assignment, rather than post-treatment choice, that defines the primary comparison.
What does a risk ratio tell us?
A risk ratio describes how the probability of virologic suppression in one group compares with the probability in another. A value of 1 means equal probabilities, while a value above or below 1 represents a relative difference. It does not by itself tell us the absolute number of additional patients achieving suppression.
Why report an absolute difference as well?
An absolute difference is often easier to translate into the number of additional patients achieving the endpoint. If one group has a suppression probability of 0.80 and another has 0.70, the absolute difference is 0.10, or 10 percentage points. This is distinct from saying that one probability is 14.3% higher relative to the other.
Why does a p-value not measure effect size?
A p-value describes evidence against a specified null hypothesis under the statistical model and testing procedure. It depends on both the observed difference and the amount of information available. A small p-value can accompany a modest effect in a large study, while a clinically meaningful effect may have substantial statistical uncertainty in a smaller study.
9. Analysis Population and Randomization
The registry identifies DUALIS as randomized, with 269 participants enrolled and two parallel arms. Randomization is central to the interpretation of a treatment comparison because it defines the groups before subsequent outcomes are observed.
| Design feature | Registry information | Statistical implication |
|---|---|---|
| Allocation | Randomized | Supports comparison based on treatment assignment. |
| Design model | Parallel | Participants are assigned to one of two concurrent treatment arms rather than receiving treatments sequentially within the same participant. |
| Masking | None | The study is unmasked; participants and investigators are not described as blinded. |
| Number of arms | 2 | The primary endpoint can be compared between two randomized groups. |
| Enrollment | 269 | The overall randomized-study population is reported as 269 enrolled participants. |
Why randomization does not automatically eliminate every source of uncertainty
Randomization protects the treatment comparison against many forms of systematic allocation bias, but it does not guarantee that the observed groups will be identical in every characteristic. Random variation remains, which is why confidence intervals and other measures of statistical uncertainty are important.
In addition, the primary endpoint is assessed at a specific follow-up time. Missing or unavailable week-48 assessments can affect the effective analysis population and the resulting estimate. The registry information available here does not specify a missing-data or imputation strategy.
10. The Role of NAT Diagnostic Testing
The registry specifies that HIV RNA suppression at week 48 will be assessed using NAT diagnostic testing. The endpoint is therefore based on an objective laboratory measurement rather than a subjective clinical rating.
Measurement threshold
The endpoint uses the explicit threshold of HIV RNA < 50 cps/ml.
Assessment time
The primary endpoint is assessed at week 48.
Endpoint format
The registry specifies both a number and percentage of patients meeting the suppression criterion.
Diagnostic method
NAT diagnostic testing is specified for assessment of HIV RNA suppression.
For statistical interpretation, the laboratory threshold creates a clinically meaningful binary classification from a quantitative measurement. That distinction matters: the primary endpoint does not ask for the mean or median HIV RNA concentration. It asks how many participants are below a defined threshold at the specified assessment time.
11. What a Complete Binary Endpoint Result Would Contain
A complete statistical presentation of the DUALIS primary endpoint would normally show the suppression result separately for each randomized arm and then provide a prespecified treatment-effect comparison.
| Result component | Statistical purpose |
|---|---|
| Number suppressed | Shows the numerator contributing to the endpoint proportion. |
| Analysis population size | Shows the denominator used for the percentage. |
| Percentage suppressed | Provides the absolute observed probability of meeting the endpoint. |
| Between-group effect estimate | Quantifies the treatment contrast using a prespecified effect measure. |
| 95% confidence interval | Quantifies statistical uncertainty around the effect estimate. |
| P-value, if prespecified | Tests the relevant null hypothesis under the specified testing framework. |
These components should be interpreted together. A percentage alone describes each group but does not quantify the between-group treatment contrast. Conversely, a relative effect without the underlying percentages can obscure the absolute frequency of the endpoint.
For a binary virologic endpoint, the most transparent presentation is to show the observed number and percentage in each randomized group and then report an explicitly defined between-group estimand with its confidence interval. The estimand should be identified before interpreting the magnitude of the treatment difference.
12. Missing Data and Week-48 Classification
The primary endpoint is defined at week 48, so the statistical analysis requires a rule for participants whose week-48 HIV RNA result is unavailable or whose endpoint status cannot be classified at that time.
The registry information available for this analysis does not specify a missing-data or imputation method. This is an important methodological distinction because different rules for handling missing week-48 assessments can lead to different estimated suppression proportions.
Observed-data approach
Participants with an evaluable week-48 result are classified according to the stated HIV RNA threshold.
Prespecified imputation
A trial protocol or statistical analysis plan can define how missing endpoint assessments are handled. No such strategy is specified in the registry information here.
For a randomized trial, the handling of missing outcomes is part of the estimand and analysis framework rather than a purely administrative decision. The denominator matters because changing which participants are counted can change the reported percentage even when the number of observed suppressed participants is unchanged.
13. Non-Inferiority Considerations
The registry information available here identifies the primary endpoint but does not state a non-inferiority margin or identify the primary hypothesis as a non-inferiority hypothesis.
That distinction is important because a non-inferiority trial is not interpreted in the same way as a conventional superiority comparison. A non-inferiority analysis requires a prespecified clinically justified margin defining how much worse the new strategy could be while still being considered non-inferior.
The margin is part of the trial's design and cannot be chosen after observing the treatment results. No such margin is reported in the registry information used here.
Accordingly, the week-48 suppression endpoint should not be described as establishing superiority or non-inferiority without the corresponding prespecified hypothesis and statistical analysis.
14. Multiplicity and Interim Analysis
The available ClinicalTrials.gov record identifies one registered primary endpoint: the number and percentage of patients with fully suppressed HIV RNA below 50 cps/ml at week 48.
The registry information available here does not specify a multiplicity adjustment, hierarchical testing strategy, alpha-spending procedure, or interim efficacy analysis.
| Design topic | What the registry reports | Statistical interpretation |
|---|---|---|
| Primary endpoint | One registered primary endpoint | The primary statistical question is centered on week-48 HIV RNA suppression. |
| Multiplicity adjustment | Not specified | No particular multiplicity procedure can be attributed to the registry record. |
| Interim efficacy analysis | Not specified | No interim alpha-spending or early-stopping framework is reported here. |
| Endpoint hierarchy | Not specified beyond the registered primary endpoint | Secondary or exploratory hypothesis-testing procedures cannot be characterized from the registry information available here. |
15. Stratification and Covariate Adjustment
The registry information identifies the trial as randomized and parallel but does not specify randomization strata, baseline covariates for adjustment, or a multivariable statistical model for the primary endpoint.
For a binary endpoint, covariate adjustment can sometimes improve precision or account for prespecified stratification factors. However, adjustment should follow the statistical analysis plan rather than being selected after seeing the treatment results.
16. Safety Analysis
The available registry information identifies the interventions and overall trial design but does not provide serious adverse event counts by arm or other numerical safety results.
Safety and efficacy answer different statistical questions. The primary endpoint asks whether patients meet a defined virologic suppression threshold at week 48. Safety analysis would instead summarize adverse events, serious adverse events, discontinuations, laboratory abnormalities, or other prespecified safety outcomes according to the trial's safety analysis population and definitions.
| Safety component | Registry information available here |
|---|---|
| Serious adverse events by arm | Not reported |
| Adverse event rates by arm | Not reported |
| Safety analysis population | Not specified |
| Exposure-adjusted safety rates | Not specified |
Because the registry does not provide numerical safety results in the available record, there is no arm-level safety estimate to interpret alongside the primary virologic endpoint.
17. Why an Unmasked Design Matters
DUALIS is listed as having no masking. This is a descriptive design characteristic rather than a conclusion about the quality of the trial.
For an objective laboratory endpoint such as HIV RNA suppression, the direct measurement itself is less dependent on subjective investigator judgment than endpoints involving symptom ratings or clinical assessments. Nevertheless, the absence of masking can matter for treatment decisions, co-interventions, follow-up behavior, reporting of subjective outcomes, and other aspects of trial conduct.
Objective endpoint
The primary endpoint uses a laboratory-defined HIV RNA threshold measured at a specified time point.
Unmasked treatment
The registry identifies the trial as having no masking, so treatment assignment was not blinded.
The statistical analysis should therefore remain focused on the prespecified laboratory endpoint while recognizing that an unmasked design can have implications for other trial processes.
18. Interpreting the Primary Endpoint Correctly
“Fully suppressed HIV RNA < 50 cps/ml at week 48” is a threshold-based endpoint. It does not describe the exact HIV RNA value for each participant, the magnitude by which a participant is below the threshold, or the trajectory of HIV RNA measurements between baseline and week 48.
The percentage represents the fraction of the relevant analysis population that meets the suppression criterion at week 48. It is an absolute measure of the frequency of the endpoint within a treatment group.
The percentage is not itself a relative treatment effect, a risk ratio, an odds ratio, or a measure of statistical significance. Those quantities require a comparison between treatment groups and an associated statistical framework.
The same number of suppressed participants can produce different percentages if the analysis denominator changes. Consequently, the numerator and denominator should be reported together whenever a binary endpoint is presented.
19. Statistical Interpretation of a Future Treatment Contrast
If a complete analysis reports the suppression percentage for each randomized arm, the first question is the size of the absolute difference. The next question is whether the difference is expressed on a relative scale, such as a risk ratio or odds ratio, and whether the confidence interval supports a sufficiently precise interpretation.
For example, an absolute difference of 10 percentage points has a direct interpretation regardless of the baseline suppression rate. A risk ratio of 1.10 would communicate a different property: the suppression probability in one group is 10% higher relative to the other group. These should not be treated as interchangeable statements.
| Measure | Question answered |
|---|---|
| Suppression proportion | What fraction of participants achieved HIV RNA < 50 cps/ml? |
| Absolute difference | How many additional percentage points separate the groups? |
| Risk ratio | How many times as large is the suppression probability in one group relative to the other? |
| Odds ratio | How do the odds of suppression compare between groups? |
| Confidence interval | How precisely is the treatment contrast estimated? |
| P-value | How compatible are the data with the specified null hypothesis? |
Because no numerical primary endpoint results are posted in the registry, these concepts define how the endpoint would normally be interpreted rather than describing an observed DUALIS treatment effect.
20. Limitations
- No posted statistical analysis: the ClinicalTrials.gov record available here contains the registered primary endpoint but no posted statistical analysis or numerical outcome result.
- No arm-level enrollment counts: the record reports 269 participants and two arms but does not provide the arm-specific sample sizes in the available information.
- Comparator details are limited: the record lists six drug interventions but does not provide sufficient arm-level mapping here to specify the complete regimen assigned to each randomized arm.
- Missing-data strategy: the available registry information does not specify how participants without an evaluable week-48 HIV RNA measurement are handled.
- Analysis method: the registry does not specify the treatment-effect measure, confidence-interval method, hypothesis test, or modeling approach for the primary endpoint.
- No multiplicity framework reported: the available record does not describe multiplicity adjustment, endpoint hierarchy beyond the registered primary endpoint, or alpha spending.
- No interim analysis described: the available registry information does not specify interim efficacy monitoring or stopping boundaries.
- No safety results: numerical serious adverse event or adverse event results by treatment arm are not reported in the available record.
- Unmasked design: the registry identifies the trial as having no masking, which is relevant when considering potential sources of bias for subjective outcomes and trial conduct.
21. Why This Trial Matters Statistically
DUALIS is a useful teaching case because its primary endpoint illustrates a different statistical structure from a time-to-event endpoint. Rather than estimating survival curves or a hazard ratio, the registered primary outcome asks whether participants meet a laboratory-defined threshold at a fixed follow-up time.
| Concept | How it appears in DUALIS |
|---|---|
| Randomization | Participants were randomly allocated to two treatment arms. |
| Parallel design | The trial used a parallel design rather than a crossover structure. |
| Binary endpoint | The primary endpoint classifies patients according to HIV RNA suppression below 50 cps/ml at week 48. |
| Absolute proportion | The endpoint is reported as the number and percentage of patients meeting the suppression criterion. |
| Between-group comparison | The randomized design provides the framework for comparing the endpoint between treatment arms. |
| Confidence interval | A confidence interval would quantify uncertainty around a treatment-effect estimate. |
| Hypothesis testing | A prespecified statistical test could assess the evidence for a difference between arms. |
| Missing data | Week-48 endpoint assessment creates a need for a prespecified rule for participants without evaluable results. |
| Unmasked design | The registry specifies no masking, which can matter for subjective aspects of trial conduct. |
| Objective laboratory measurement | The primary endpoint is based on HIV RNA measured using NAT diagnostic testing. |
22. Clinical Interpretation vs Statistical Interpretation
Statistical interpretation
The registered primary endpoint is a binary week-48 virologic suppression outcome. Its natural statistical representation is a proportion within each randomized arm followed by a prespecified between-group comparison.
Clinical interpretation
The endpoint asks whether patients reach the specific HIV RNA threshold of < 50 cps/ml at week 48. Its clinical meaning therefore depends on both the threshold itself and the proportion of participants reaching it.
Keeping these two interpretations separate is important. Statistical significance, if demonstrated in a formal analysis, would address evidence for a treatment-group difference. Clinical interpretation would additionally consider the magnitude of the difference, the precision of the estimate, the analysis population, and the context of the treatment strategies.
23. A Practical Framework for Reading the DUALIS Endpoint
This sequence illustrates the statistical logic of the primary endpoint. First, the endpoint threshold is defined. Second, the assessment time is fixed. Third, patients meeting the threshold are counted. Fourth, the number is expressed as a percentage of the relevant analysis population. Finally, the randomized treatment groups can be compared using a prespecified effect measure and inferential procedure.
24. Sources
- ClinicalTrials.gov: DUALIS — NCT02486133.
Continue through Clinical Biostats
Explore clinical trial results, statistical tutorials, and analysis concepts used throughout randomized clinical research.
25. Record Summary
DUALIS is a completed randomized phase 3 parallel trial of dual therapy with boosted darunavir plus dolutegravir in HIV-Infection. The registry reports 269 enrolled participants and two treatment arms, with no masking. Its registered primary endpoint is the number and percentage of patients with fully suppressed HIV RNA < 50 cps/ml at week 48, assessed using NAT diagnostic testing.
Statistically, the endpoint is a fixed-time binary outcome. The central analytical task is therefore to estimate the proportion of participants meeting the virologic suppression threshold in each randomized arm and compare those proportions using a prespecified treatment-effect measure and inferential framework. Absolute differences, relative measures, confidence intervals, and p-values answer different questions and should not be treated as interchangeable.
The ClinicalTrials.gov record does not report a posted statistical analysis or numerical primary endpoint result. It also does not specify the treatment-effect measure, inferential test, missing-data strategy, multiplicity framework, interim-analysis procedure, or arm-level enrollment counts in the information available here. Those methodological distinctions are essential when interpreting any future numerical results associated with the trial.