This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. The registry record identifies the primary endpoint and reports results, but the ClinicalTrials.gov record contains no formal statistical-analysis entries.
1. Trial at a Glance
SECOND-LINE was a randomized, parallel-group, unmasked phase 4 treatment trial in HIV infection. It enrolled 558 participants to compare two second-line combination antiretroviral therapy options after virological failure of a standard NNRTI + 2N(t)RTI first-line regimen.
| Feature | SECOND-LINE |
|---|---|
| Trial name | SECOND-LINE |
| NCT identifier | NCT00931463 |
| Therapeutic area | Infectious Disease |
| Condition | HIV Infections |
| Phase | Phase 4 |
| Status | COMPLETED |
| Allocation | RANDOMIZED |
| Design model | PARALLEL |
| Masking | NONE |
| Primary purpose | TREATMENT |
| Enrollment | 558 |
| Lead sponsor | Kirby Institute |
| Sponsor type | OTHER_GOV |
| Trial dates | Start: 2009-09; Primary completion: 2012-09 |
| Results posted | Yes |
| Outcome measures posted | 5 |
2. Clinical Question
The trial addresses a second-line treatment question in participants with HIV infection following virological failure of a standard NNRTI + 2N(t)RTI first-line regimen. The registered primary endpoint asks whether participants achieve plasma HIV RNA below a prespecified threshold 48 weeks after randomization.
Population
Participants with HIV infection following virological failure of a standard NNRTI + 2N(t)RTI first-line regimen.
Intervention option
Ritonavir-boosted lopinavir with 2N(t)RTI, as identified in the ClinicalTrials.gov record.
Comparator option
Ritonavir-boosted lopinavir with raltegravir, as identified in the ClinicalTrials.gov record.
Primary question
What proportion of participants have plasma HIV RNA < 200 copies/mL 48 weeks after randomization?
3. Trial Design
Ritonavir-boosted lopinavir + 2N(t)RTI
- Ritonavir-boosted lopinavir
- 2N(t)RTI
- Second-line combination antiretroviral therapy
Ritonavir-boosted lopinavir + raltegravir
- Ritonavir-boosted lopinavir
- Raltegravir
- Second-line combination antiretroviral therapy
4. Primary Endpoint
| Endpoint | Registry definition | Time frame | Endpoint type |
|---|---|---|---|
| Participants With Plasma HIV RNA < 200 Copies/mL 48 Weeks After Randomization | Participants with plasma HIV RNA < 200 copies/mL | 48 weeks following randomization | Binary |
This is a binary endpoint: each participant is classified according to whether the prespecified HIV RNA threshold is achieved at the specified follow-up time. That structure is important because the natural statistical comparison is a comparison of proportions rather than a time-to-event comparison.
The analysis therefore focuses on the proportion of participants meeting the binary criterion in each randomized group and on the uncertainty around the between-group difference or ratio.
5. Statistical Methodology
Comparison of binary proportions
For a randomized two-arm trial with a binary primary endpoint, a standard analysis begins with the proportion of participants meeting the endpoint in each treatment group. The treatment effect can then be summarized using an absolute risk difference, a relative risk, or an odds ratio, with a confidence interval for the selected effect measure.
Here, p1 and p0 represent the observed proportions meeting the endpoint in the two randomized groups.
Hypothesis testing
A formal two-group analysis of a binary endpoint can use a Pearson chi-square test when its assumptions are appropriate. Fisher's exact test is an alternative when expected cell counts are small. A confidence interval for the treatment effect should accompany any hypothesis test because the p-value alone does not describe the magnitude or precision of the treatment difference.
Analysis population
Randomization creates the basis for an intention-to-treat comparison, in which participants are analyzed according to their randomized assignment. The ClinicalTrials.gov record does not specify the precise analysis population, handling of participants with missing 48-week HIV RNA measurements, or the prespecified statistical analysis plan. Those details should therefore not be inferred from the registry summary.
Missing binary outcomes
For a binary endpoint assessed at a fixed follow-up time, missing outcome measurements require an explicit analysis rule. Depending on the prespecified protocol and estimand, possible approaches can include defining missing outcomes according to a conservative failure rule, analyzing observed outcomes only, or using a model-based method. The ClinicalTrials.gov record does not identify which approach was prespecified for this trial.
6. Planned Analysis
The ClinicalTrials.gov record states that results have been posted and that the primary endpoint has results posted for it. However, the ClinicalTrials.gov record contains no formal statistical analyses. Consequently, no formal treatment-effect estimate, confidence interval, or p-value for the primary endpoint is available in the ClinicalTrials.gov record.
What would normally be analysed?
For the registered endpoint, the analysis would normally compare the proportion of participants with plasma HIV RNA < 200 copies/mL at 48 weeks in the two randomized treatment groups. A complete statistical presentation would ordinarily report the numerator and denominator in each arm, the resulting proportions, an effect measure such as a risk difference or relative risk, a confidence interval, and a p-value if a formal hypothesis test was prespecified.
| Element | What the ClinicalTrials.gov record establishes | What is not reported |
|---|---|---|
| Primary endpoint | Plasma HIV RNA < 200 copies/mL 48 weeks after randomization | — |
| Endpoint type | Binary | — |
| Results posted | Yes | — |
| Formal statistical analyses | None reported | Treatment-effect estimate, confidence interval, p-value |
| Arm-level primary endpoint counts | Not reported in the ClinicalTrials.gov record provided | Numerators and denominators |
7. Serious Adverse Events
The ClinicalTrials.gov record includes serious adverse-event counts by arm. These figures describe affected participants among the stated at-risk populations. They do not constitute a formal statistical comparison because no comparative estimate or p-value is reported.
| Group | Participants with serious adverse events | At risk |
|---|---|---|
| Ritonavir-boosted lopinavir + 2N(t)RTI | 23 | 271 |
| Ritonavir-boosted lopinavir + raltegravir | 24 | 270 |
The reported counts are close in absolute number, but the denominators are also slightly different. A proper comparative safety analysis would therefore use the arm-specific proportions rather than comparing the raw counts alone. The ClinicalTrials.gov record does not provide a confidence interval, p-value, risk ratio, risk difference, or odds ratio for serious adverse events.
8. Statistical Methods Explained
Why is the primary endpoint analysed as binary data?
The registry defines the primary endpoint as whether participants have plasma HIV RNA below 200 copies/mL at a fixed point, 48 weeks following randomization. Each participant therefore has an outcome that can be represented as yes or no. Binary statistical methods are designed for this type of endpoint.
What does an absolute risk difference tell us?
The risk difference subtracts the proportion meeting the endpoint in one treatment group from the corresponding proportion in the other group. It expresses the comparison on an absolute probability scale. For example, a difference of 0.10 would correspond to a 10-percentage-point difference, although no such value is reported for SECOND-LINE in the ClinicalTrials.gov record.
What does a relative risk tell us?
The relative risk compares the probability of meeting the endpoint in one group with the probability in the other. A relative risk of 1 indicates equal observed probabilities, values above 1 indicate a higher probability in the numerator group, and values below 1 indicate a lower probability. The ClinicalTrials.gov record does not report a relative risk for the primary endpoint.
Why should a confidence interval accompany a p-value?
A p-value addresses the compatibility of the observed data with a specified null hypothesis under the statistical model. It does not quantify the size of the treatment effect. A confidence interval adds information about the precision of the estimated effect and therefore gives a more informative description of the uncertainty surrounding the comparison.
Why does randomization matter?
Randomization establishes the treatment assignment mechanism before outcomes are observed. In a properly conducted randomized comparison, baseline differences are attributable to chance rather than systematic assignment. This is why the randomized treatment groups provide the foundation for causal comparison of the assigned strategies.
Why can the serious-adverse-event counts not be compared directly?
Raw counts do not account for the number of participants at risk in each group. Here, the registry-reported denominators are 271 and 270. A statistical comparison should therefore begin with the corresponding proportions and then quantify uncertainty around the difference or ratio.
9. Interpreting the Primary Endpoint Correctly
The primary endpoint asks whether a participant has plasma HIV RNA < 200 copies/mL at 48 weeks following randomization. It is a fixed-time binary endpoint, not a measure of the time required to reach that threshold and not a time-to-event endpoint.
If the arm-level proportions and a formal comparison were available, the estimated effect would describe the difference in the probability of meeting the specified virologic criterion at 48 weeks. It would not, by itself, describe every aspect of viral-load trajectories over the preceding 48 weeks.
A confidence interval around a risk difference, relative risk, or odds ratio would quantify statistical uncertainty under the chosen analysis framework. It would not describe the range of individual patient responses.
A p-value would not measure the magnitude of the treatment effect, the clinical importance of the result, or the probability that one treatment is effective. Those questions require examination of the effect estimate, its uncertainty, the endpoint definition, and the clinical context.
10. Binary Endpoint Analysis in More Detail
A useful way to represent the primary analysis is with a two-by-two table. The rows correspond to randomized treatment groups and the columns correspond to whether the participant meets the 48-week virologic endpoint.
| HIV RNA < 200 copies/mL | Does not meet endpoint | Total | |
|---|---|---|---|
| Ritonavir-boosted lopinavir + 2N(t)RTI | a | b | a + b |
| Ritonavir-boosted lopinavir + raltegravir | c | d | c + d |
The observed response proportions would be calculated as a/(a+b) and c/(c+d). The ClinicalTrials.gov record does not provide a, b, c, or d for the primary endpoint, so the numerical treatment effect cannot be reconstructed without introducing information outside the allowed data.
Risk difference
The risk difference is directly interpretable in percentage-point terms after multiplying by 100. Its confidence interval would quantify uncertainty around the estimated difference.
Relative risk
The relative risk compares the probabilities of meeting the endpoint. It is often particularly intuitive when the endpoint is a desirable binary outcome.
Odds ratio
The odds ratio compares odds rather than probabilities. It should not be interpreted as though it were automatically equivalent to a risk ratio, particularly when the endpoint is common.
11. Analysis of Serious Adverse Events
The serious-adverse-event data can be placed on the same binary framework. For the two groups, the ClinicalTrials.gov record is 23 affected participants among 271 at risk and 24 affected participants among 270 at risk.
| Group | Affected | At risk | Outcome structure |
|---|---|---|---|
| Ritonavir-boosted lopinavir + 2N(t)RTI | 23 | 271 | 23 affected; remaining participants not identified in the ClinicalTrials.gov record as having the serious AE |
| Ritonavir-boosted lopinavir + raltegravir | 24 | 270 | 24 affected; remaining participants not identified in the ClinicalTrials.gov record as having the serious AE |
For safety interpretation, it is important to distinguish a count of affected participants from a count of adverse-event episodes. The ClinicalTrials.gov record specifically describe affected participants divided by participants at risk. They do not state the number of individual serious-adverse-event episodes or their timing.
12. Trial Timeline
Trial start
The registry profile identifies September 2009 as the study start.
Randomized parallel-group study
The trial was conducted as a randomized, parallel-group, unmasked treatment study.
Primary endpoint time point
The registered primary endpoint is assessed 48 weeks following randomization.
Primary completion
The registry profile identifies September 2012 as the primary completion date.
Registry status
the ClinicalTrials.gov record lists the study status as COMPLETED and indicates that results have been posted.
13. Design Features Supported by the Registry Data
| Design feature | What the ClinicalTrials.gov record shows | Statistical implication |
|---|---|---|
| Randomization | RANDOMIZED | Provides the basis for a randomized comparison of assigned treatment groups. |
| Parallel design | PARALLEL | Participants remain in one randomized treatment pathway rather than being assigned sequentially to multiple treatments. |
| Masking | NONE | The trial is unmasked; outcome assessment and treatment knowledge can therefore be relevant when interpreting endpoints and potential biases. |
| Primary purpose | TREATMENT | The trial evaluates treatment strategies rather than prevention, diagnostic, or other primary purposes. |
| Phase | Phase 4 | The registry classifies the trial as a phase 4 study. |
| Arms | 2 | The design has two treatment groups. |
Design topics not specified in the ClinicalTrials.gov record
the ClinicalTrials.gov record does not report a non-inferiority margin, crossover design, factorial design, Bayesian analysis, interim-analysis strategy, multiplicity adjustment, or specific missing-data imputation method. These features are therefore not assigned to SECOND-LINE on this page.
14. Limitations
- No formal statistical-analysis entries reported: the ClinicalTrials.gov record contains no formal statistical analyses, so numerical treatment effects, confidence intervals, and p-values for the primary endpoint cannot be reported from the provided record.
- Primary endpoint counts are not reported: although the registry profile states that results are posted for the primary endpoint, the ClinicalTrials.gov record does not give the arm-level numbers achieving plasma HIV RNA < 200 copies/mL at 48 weeks.
- Fixed-time endpoint: the primary endpoint is measured at 48 weeks, so it does not by itself describe the complete longitudinal pattern of virologic response.
- Unmasked design: the trial is recorded as having no masking. Knowledge of assigned treatment can be relevant to the interpretation of outcomes and study conduct, depending on how outcomes were assessed.
- Missing-data handling is unspecified: the ClinicalTrials.gov record does not identify how participants without a 48-week primary endpoint measurement were handled in the formal analysis.
- Safety comparison is incomplete: serious-adverse-event counts and at-risk denominators are reported, but no formal comparative safety analysis is provided.
- Outcome detail is limited: the registry-reported profile states that five outcome measures were posted but provides only the registered primary endpoint in the detailed endpoint ClinicalTrials.gov record.
15. Why This Trial Matters Statistically
SECOND-LINE provides a useful teaching example of a randomized clinical-trial question with a binary primary endpoint. Unlike a time-to-event trial, the primary statistical question is centered on whether participants satisfy a prespecified virologic criterion at a particular follow-up time.
| Concept | How it appears in SECOND-LINE |
|---|---|
| Randomization | Participants were assigned in a randomized two-arm design. |
| Parallel groups | The registry identifies a PARALLEL design model. |
| Binary endpoint | The primary endpoint is whether plasma HIV RNA is < 200 copies/mL 48 weeks after randomization. |
| Absolute effect | A risk difference would express the treatment comparison in percentage-point terms. |
| Relative effect | A relative risk or odds ratio could summarize the proportional difference between groups. |
| Confidence intervals | A complete statistical analysis would quantify uncertainty around the estimated treatment effect. |
| Hypothesis testing | A chi-square or Fisher's exact framework can be used for a two-group binary comparison when appropriate. |
| Missing data | A fixed-time binary endpoint requires a prespecified rule for participants without an evaluable 48-week outcome. |
| Safety analysis | Serious adverse events are reported as affected participants among participants at risk. |
| Unmasked design | The registry identifies the trial as having no masking. |
16. Clinical Biostats Interpretation of the Available Results
The registry profile confirms that results have been posted for the primary binary endpoint, but it does not provide the arm-level primary endpoint counts or a formal statistical-analysis entry. Therefore, the magnitude, direction, confidence interval, and p-value of the primary treatment comparison cannot be stated from the ClinicalTrials.gov record without importing information from outside the defined source set.
The ClinicalTrials.gov record reports 23 serious-adverse-event-affected participants among 271 at risk in the ritonavir-boosted lopinavir + 2N(t)RTI group and 24 among 270 at risk in the ritonavir-boosted lopinavir + raltegravir group. These are descriptive counts with denominators, not a formal hypothesis test.
The available registry summary does not establish whether the two strategies differed statistically on the primary virologic endpoint or whether their serious-adverse-event rates differed statistically. A valid conclusion about either comparison requires the underlying arm-level outcome data and the prespecified analysis framework.
17. A Practical Analysis Workflow
If the complete arm-level primary endpoint data were available, a transparent analysis would proceed in a small number of steps.
The final presentation should retain the distinction between the descriptive endpoint rate, the treatment-effect estimate, the confidence interval, and the hypothesis-test result. This prevents the common mistake of treating a p-value as though it were the treatment effect itself.
18. Related Statistical Concepts
Learn more about the methods used in this trial:
19. Related Statistical Calculators
20. Sources
- ClinicalTrials.gov: NCT00931463.
- PubMed: PMID 33399309.
- PubMed: PMID 25723472.
- PubMed: PMID 24204757.
- PubMed: PMID 23769235.
Continue through the Clinical Biostats statistical tutorials
Explore the statistical concepts that underlie randomized comparisons, binary endpoints, treatment-effect measures, confidence intervals, and clinical-trial analysis.
21. Record Summary
SECOND-LINE is a completed randomized phase 4 trial in HIV infection evaluating two options for second-line combination antiretroviral therapy following virological failure of a standard NNRTI + 2N(t)RTI first-line regimen. The ClinicalTrials.gov record identifies a two-arm parallel design, enrollment of 558 participants, and a binary primary endpoint defined as participants with plasma HIV RNA < 200 copies/mL 48 weeks following randomization.
The principal statistical teaching point is the distinction between a registered endpoint and a reported formal analysis. The ClinicalTrials.gov record states that results were posted, but they contain no statistical-analysis entries and no arm-level primary endpoint counts. A statistically complete assessment therefore requires the underlying outcome data before a treatment effect, confidence interval, or p-value can responsibly be stated.
The ClinicalTrials.gov record provides descriptive serious-adverse-event counts: 23 of 271 at risk in the ritonavir-boosted lopinavir + 2N(t)RTI group and 24 of 270 at risk in the ritonavir-boosted lopinavir + raltegravir group. These figures can be used to understand the structure of the safety comparison, but the ClinicalTrials.gov record does not contain a formal comparative analysis.