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
The Partners Demonstration Project was a phase 4, single-group study evaluating PrEP and ART implementation strategies for HIV-1 serodiscordant couples. The registry describes a prevention-focused study with enrollment of 1013 participants.
| Feature | Partners Demonstration Project |
|---|---|
| NCT ID | NCT02775929 |
| Title | Partners Demonstration Project of PrEP and ART |
| Condition | HIV Infection |
| Design | Single-group, open-label |
| Primary purpose | Prevention |
| Interventions | FTC-TDF PrEP and ART |
| Sponsor | University of Washington |
2. Clinical Question
Population
High-risk HIV-1 serodiscordant couples defined using a validated risk scoring tool.
Intervention
FTC-TDF PrEP for HIV-1 uninfected partners and ART for HIV-1 infected partners.
Comparator
No comparator arm was used. The registry describes a single-group design.
Primary question
Can PrEP and ART strategies be implemented with measurable recruitment, uptake, adherence, preference, safety, and feasibility outcomes?
3. Trial Design
Allocation
NA
Masking
None
Number of arms
1
Study period
Start date: 2012-11. Primary completion date: 2016-06.
4. Endpoints
The registry lists multiple primary endpoints assessed over 24 months, covering recruitment, implementation, adherence, preferences, safety, and infant outcomes.
| Endpoint category | Registry definition |
|---|---|
| Recruitment feasibility | Screen-to-eligible ratio and proportion of eligible couples deciding to enroll. |
| Cost-effectiveness | Incremental cost-effectiveness ratio (ICER) per HIV related infection, death, and DALY averted. |
| PrEP initiation and adherence | Initiation, missed doses, detectable and quantifiable PrEP levels, and PrEP holds. |
| ART initiation and adherence | ART initiation and suppressed plasma HIV-1 RNA levels. |
| Preferences and behaviors | Choices, concerns, sexual frequency, condom use frequency, and relationship outcomes. |
| Pregnancy and infant outcomes | HIV infection, congenital abnormalities, serious adverse events, and infant growth measures. |
5. Planned Analysis
No results are posted on ClinicalTrials.gov for this record. The registry describes measurements that were planned over 24 months, but does not report completed statistical estimates.
For implementation outcomes such as recruitment ratios, enrollment proportions, uptake, and adherence measures, descriptive statistics would typically summarize counts and proportions. For cost-effectiveness outcomes, the registry specifies estimation of ICERs comparing PrEP with current practice per incident HIV-1 case, HIV related death, and DALY averted.
For infant growth outcomes, the registry describes linear models using available infant growth z-scores over study duration, with change over time represented by regression against study month.
6. Statistical Methodology
Descriptive analysis of implementation outcomes
Many endpoints in this study are feasibility and implementation measures. These are commonly summarized using proportions, counts, and confidence intervals when appropriate. Interpretation focuses on observed implementation patterns rather than randomized treatment comparisons.
Cost-effectiveness analysis
The registry states that the cost-effectiveness analysis will follow World Health Organization guidelines and estimate incremental cost-effectiveness ratios. An ICER compares the additional cost of an intervention with the additional health outcome achieved.
ICER = (Cost of intervention − Cost of comparator) / (Effect of intervention − Effect of comparator)
Longitudinal infant growth models
The registry describes infant length, weight, and head circumference measured as age- and gender-specific z-scores using World Health Organization growth curves. The slope of change over time is estimated using available measurements and regression against study month.
7. Statistical Methods Explained
Why is there no treatment-effect hazard ratio?
The study uses a single-group design. Without a randomized comparator arm, a treatment-effect hazard ratio comparing two assigned groups is not part of the registry-described analysis framework.
How are feasibility outcomes interpreted?
Measures such as screen-to-eligible ratio and enrollment decisions describe whether recruitment strategies reach the intended population and whether eligible couples choose participation.
Why are adherence measures important statistically?
PrEP adherence endpoints combine behavioral reporting and biological measurements. Together, these provide complementary information about medication use patterns.
What does an ICER represent?
An incremental cost-effectiveness ratio summarizes additional cost relative to additional health benefit. It is interpreted in the context of the healthcare decision framework being applied.
Why use longitudinal regression for infant growth?
Repeated measurements from the same infant contain information about growth trajectories. A longitudinal model accounts for change over time rather than relying on a single measurement.
8. Limitations
- The registry describes a single-group design, limiting direct causal comparisons with an untreated or alternative-treatment group.
- No posted statistical analyses or numerical results are available in the ClinicalTrials.gov record.
- Multiple endpoints were planned across implementation, adherence, behavioral, pregnancy, and infant outcomes.
- Interpretation of feasibility outcomes differs from interpretation of randomized efficacy trials.
9. Why This Trial Matters Statistically
This study illustrates an important category of clinical research: implementation-focused prevention studies. Statistical analysis is not limited to comparing clinical events between randomized groups. Recruitment efficiency, medication uptake, adherence, cost-effectiveness, and longitudinal outcomes can each require different analytical frameworks.
The Partners Demonstration Project also demonstrates how clinical research questions determine statistical design. A prevention implementation question may prioritize feasibility metrics, behavioral measurements, and health-economic modeling rather than a conventional superiority endpoint.