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
VA NEPHRON-D was a randomized, double-blind, parallel phase 3 treatment trial in patients with kidney disease, nephropathy, and type 2 diabetes. The principal comparison evaluated combination therapy with an ARB and ACEI against ARB monotherapy using a time-to-event composite renal endpoint.
| Feature | VA NEPHRON-D |
|---|---|
| Trial name | VA NEPHRON-D: Diabetes iN Nephropathy Study |
| Phase | Phase 3 |
| Conditions | Kidney Disease; Nephropathy; Type 2 Diabetes |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Double |
| Primary purpose | Treatment |
| Enrollment | 1,448 |
| Lead sponsor | US Department of Veterans Affairs |
| Trial status | Terminated |
| Start | July 2008 |
| Primary completion | September 2014 |
2. Clinical Question
The central statistical question was whether combination therapy with an ARB and ACEI changed the time to a clinically important renal or mortality event compared with ARB monotherapy in the study population.
Population
Participants enrolled in a phase 3 treatment trial for kidney disease, nephropathy, and type 2 diabetes.
Intervention
Combination therapy with an angiotensin receptor blocker (ARB) and an angiotensin-converting enzyme inhibitor (ACEI).
Comparator
Monotherapy with an ARB.
Primary question
Does combination ARB and ACEI therapy alter the time to the first primary renal or mortality event compared with ARB monotherapy?
3. Trial Design
Combination ARB + ACEI
- Combination of an angiotensin receptor blocker and an angiotensin-converting enzyme inhibitor
- Compared with ARB monotherapy
- Evaluated for time to first primary event
ARB monotherapy
- Monotherapy with an angiotensin receptor blocker
- Comparator treatment group
- Evaluated using the same time-to-event framework
The registry identifies the interventions as losartan and lisinopril. The posted statistical analysis describes the randomized comparison as combination ARB and ACEI versus monotherapy ARB.
4. Registered Primary Endpoint
| Endpoint | Registry definition | Time frame |
|---|---|---|
| Composite time-to-event endpoint | A composite endpoint of reduction in estimated GFR of 30 ml/min/1.73m*m in individuals with a baseline estimated GFR >= 60 ml/min/1.73m*m, reduction in estimated GFR >50% in individuals with baseline estimated GFR <60 ml/min/1.73m*m; ESRD or death | From enrollment to time of first primary event, up to 4.5 years |
The endpoint is therefore not a single laboratory measurement. It is a time-to-first-event composite in which the first qualifying renal deterioration event, ESRD, or death determines the event time for that participant.
Why the endpoint is composite
Composite endpoints can increase the number of observed events by combining clinically relevant event types. Statistically, however, interpretation depends on understanding exactly what qualifies as the first event and whether the treatment effect is reasonably consistent across the components.
For this trial, the registry definition uses different estimated-GFR thresholds depending on baseline estimated GFR. That makes the endpoint more specific than simply defining an event as any decline in estimated GFR.
5. Secondary Endpoint
| Endpoint | Definition | Time frame | Analysis |
|---|---|---|---|
| Renal composite endpoint | Reduction in estimated GFR of >50% for individuals with baseline GFR <60, or reduction in GFR of >30 for individuals with baseline GFR >=60, or ESRD | From enrollment to time of first event, up to 4.5 years | Log-rank test; hazard ratio |
The secondary endpoint is similar to the primary composite but excludes death from its definition. That distinction is statistically important: an analysis that includes death asks about a broader composite of renal deterioration, ESRD, or mortality, whereas the renal composite isolates renal progression and ESRD.
6. Statistical Methodology
Time-to-event analysis
The primary and secondary endpoints are time-to-event outcomes. Each participant contributes follow-up from enrollment until the first qualifying event or censoring. The statistical comparison therefore needs to account not only for whether an event occurred, but also for when it occurred.
The analysis compares the distributions of event times between the randomized treatment groups rather than reducing every participant to a simple yes/no event indicator.
Log-rank test
The registry reports the log-rank test as the statistical method for both posted analyses. The log-rank test compares the event-time experience between groups across the observed follow-up while accounting for censoring.
Conceptually, the test compares observed and expected numbers of events between treatment groups at the event times. Participants who remain event-free and are censored still contribute information before their censoring time.
Cox proportional-hazards effect measure
The posted primary analysis reports a Cox proportional-hazards effect measure, summarized as a hazard ratio. The hazard ratio expresses the relative instantaneous event rate between the two randomized groups under the Cox model.
An HR below 1 corresponds to a lower estimated instantaneous event rate in the combination group under the fitted model. An HR above 1 corresponds to a higher estimated instantaneous event rate.
Kaplan-Meier estimation
Kaplan-Meier estimation is the standard descriptive framework for time-to-event data of this type. It estimates the probability of remaining event-free over time while allowing for right censoring. The registry's normalized primary method is the log-rank test, while the reported effect measure is the hazard ratio.
Here, di is the number of events at event time ti, and ni is the number at risk immediately before that time.
Superiority framework
The registry identifies the hypothesis type as superiority. Thus, the statistical question is whether the randomized treatment groups differ in the direction and magnitude supported by the prespecified superiority framework, rather than whether the combination is merely sufficiently similar to monotherapy.
7. Primary Result
The registry reports a formal statistical analysis for the primary composite time-to-event endpoint. The comparison was combination ARB and ACEI versus ARB monotherapy.
Primary composite endpoint
Hazard ratio for the first primary event
95% CI: 0.70–1.12 · P = 0.30
Analysis: log-rank test; effect measure: Cox proportional-hazards estimate
| Primary endpoint | Combination ARB + ACEI | ARB monotherapy | Effect estimate |
|---|---|---|---|
| Composite time to first primary event | Randomized comparison | Randomized comparison | HR 0.88; 95% CI 0.70–1.12; P = 0.30 |
The estimated hazard ratio of 0.88 means that, under the Cox proportional-hazards model, the combination group had an estimated instantaneous rate of the first primary event that was approximately 12% lower than the ARB-monotherapy group. This is a relative hazard measure, not a 12-percentage-point reduction in the probability of an event.
The 95% confidence interval of 0.70–1.12 describes the statistical uncertainty around the estimated hazard ratio. Because the interval includes 1, the data are compatible with a range of treatment effects extending from a lower estimated hazard to a higher estimated hazard for the combination group.
The P = 0.30 value addresses evidence against the null hypothesis in the specified statistical test; it does not measure the size, clinical importance, or probability that the treatment has no effect. The hazard ratio and its confidence interval are needed to describe the estimated effect and its precision.
Because this is a time-to-event analysis, the hazard-ratio interpretation also depends on the proportional-hazards framework of the Cox model. A hazard ratio is not equivalent to a risk ratio at 4.5 years and does not imply that every participant experienced the same relative change in event risk.
8. Secondary Renal Composite Result
Hazard ratio for the renal composite endpoint
95% CI: 0.58–1.05 · P = 0.10
Analysis: log-rank test; effect measure: hazard ratio
| Secondary endpoint | Comparison | Hazard ratio | 95% CI | P-value |
|---|---|---|---|---|
| Renal composite endpoint | Combination ARB + ACEI vs ARB monotherapy | 0.78 | 0.58–1.05 | 0.10 |
The hazard ratio of 0.78 corresponds to an estimated 22% lower instantaneous rate of the renal composite event in the combination group under the reported hazard-ratio framework.
The 95% CI of 0.58–1.05 is wider than the primary endpoint's confidence interval and crosses 1. This means the estimated treatment effect is uncertain: the interval includes both a potentially lower event hazard and a value slightly above the null value.
The P = 0.10 value is a measure of statistical evidence under the reported test, not a measure of effect magnitude. It should not be interpreted as a 10% probability that the null hypothesis is true or as a 90% probability that the treatment works.
This endpoint also differs conceptually from the primary composite because death is not included in the registered secondary definition. The two estimates therefore answer related but distinct questions.
9. Comparing the Two Time-to-Event Results
The point estimate for the secondary renal composite is lower than the point estimate for the primary composite, but these estimates should not be treated as directly interchangeable. The endpoints have different definitions, particularly because the primary endpoint includes death while the secondary renal composite does not.
Both confidence intervals include 1.0. The primary estimate has a 95% CI of 0.70–1.12, while the secondary estimate has a 95% CI of 0.58–1.05. The secondary estimate is therefore more favorable numerically, but its confidence interval still encompasses the null value.
10. Serious Adverse Events
The registry reports serious adverse events by randomized treatment arm. The counts are presented as affected participants over participants at risk.
| Treatment group | Participants with serious adverse events | At risk |
|---|---|---|
| Combination of ARB and ACEI | 416 | 724 |
| Monotherapy ARB | 380 | 724 |
These safety counts are descriptive. They should not be interpreted as an efficacy endpoint or combined with the hazard-ratio results into a single treatment-benefit statistic. Safety and efficacy address different outcomes and can have different analysis populations, follow-up structures, and statistical considerations.
11. Why the Hazard Ratio Requires Careful Interpretation
An HR of 0.88 means the fitted model estimates a lower instantaneous rate of the primary event in the combination group relative to ARB monotherapy. Expressed as a relative difference, 0.88 corresponds to an estimated 12% lower hazard.
It does not mean that 12% fewer participants experienced the endpoint, that each participant had exactly a 12% reduction in risk, or that the probability of remaining event-free at every time point was exactly 12% higher.
The 95% CI of 0.70–1.12 communicates uncertainty around the estimated hazard ratio. A point estimate alone can give a false impression of precision. The interval shows that the data are compatible with materially different underlying hazard ratios.
The P-value of 0.30 quantifies the statistical evidence under the specified test against the null hypothesis. It does not tell us whether the observed effect is clinically important, nor does it provide the probability that the null hypothesis is true.
12. Early Study Termination
The registry records an important design and interpretation event: in October 2012, the Data Monitoring Committee recommended that study treatment be stopped primarily because of safety concerns along with low conditional power to detect a treatment effect on the primary outcome.
Trial start
The phase 3 randomized treatment study began.
Data Monitoring Committee recommendation
The committee recommended stopping study treatment primarily because of safety concerns together with low conditional power to detect a treatment effect on the primary outcome.
Primary completion
The registry records September 2014 as the primary completion date.
Conditional power is a forward-looking probability calculation based on the information available at an interim analysis and assumptions about what might occur with additional follow-up. Low conditional power indicates that, given the interim information and the assumptions used for the calculation, the trial had limited projected ability to demonstrate the targeted treatment effect if it continued.
13. Statistical Methods Explained
Why was a time-to-event endpoint used?
The primary endpoint records the time to the first qualifying event, rather than simply whether an event ever occurred. This preserves information about when renal deterioration, ESRD, or death occurred and allows participants with incomplete follow-up to contribute information before censoring.
Why use a log-rank test?
The log-rank test is designed for comparing time-to-event distributions between groups. Rather than comparing only the total number of events, it evaluates how observed events are distributed over follow-up relative to what would be expected under the null hypothesis of no difference between treatment groups.
What does a hazard ratio of 0.88 mean?
A hazard ratio of 0.88 indicates an estimated 12% lower instantaneous event rate for the combination group relative to ARB monotherapy under the Cox proportional-hazards model. It is a relative rate measure over time, not an absolute risk reduction.
Why does the confidence interval include 1?
The null value for a hazard ratio is 1.00 because a ratio of 1 represents equal hazards between groups. The primary 95% CI extends from 0.70 to 1.12, so it includes the null value and reflects uncertainty that spans both lower and higher hazards for the combination group.
Why does the secondary HR of 0.78 not automatically establish a treatment effect?
The point estimate is below 1, but the 95% CI extends from 0.58 to 1.05 and therefore includes the null. The P-value is 0.10. Statistical interpretation requires considering the entire interval and prespecified testing framework rather than the point estimate alone.
Why is early termination important for interpretation?
The registry states that treatment was stopped following a Data Monitoring Committee recommendation because of safety concerns together with low conditional power for detecting a treatment effect on the primary outcome. Stopping early changes the amount of information available for estimating treatment effects and therefore forms an important part of the statistical context.
14. Composite Endpoints: What Is Actually Being Compared?
The primary endpoint combines several possible first events:
| Component | Registry-defined role |
|---|---|
| Reduction in estimated GFR of 30 ml/min/1.73m*m | Applies to individuals with baseline estimated GFR >= 60 ml/min/1.73m*m |
| Reduction in estimated GFR >50% | Applies to individuals with baseline estimated GFR <60 ml/min/1.73m*m |
| ESRD | Primary composite component |
| Death | Primary composite component |
For a participant who experiences more than one component, the registered time-to-event framework uses the first primary event. Consequently, the primary hazard ratio summarizes the treatment comparison for the composite endpoint as a whole rather than separately estimating the effect on each component.
This distinction matters because a composite can be driven by its more frequent components. The registry does not provide component-specific statistical analyses in the information summarized here, so the reported primary hazard ratio should be understood as the effect estimate for the complete registered composite.
15. Primary Versus Secondary Endpoint Interpretation
| Feature | Primary composite | Secondary renal composite |
|---|---|---|
| Death included? | Yes | No |
| Time frame | From enrollment to first primary event, up to 4.5 years | From enrollment to first event, up to 4.5 years |
| Method | Log-rank test | Log-rank test |
| Effect measure | Hazard ratio | Hazard ratio |
| Estimate | 0.88 | 0.78 |
| 95% CI | 0.70–1.12 | 0.58–1.05 |
| P-value | 0.30 | 0.10 |
The two results are directionally similar in that both hazard-ratio point estimates are below 1. However, the endpoints are not identical, and neither confidence interval excludes 1. The appropriate interpretation is therefore to report the estimates and their uncertainty rather than reducing the results to the direction of the point estimates alone.
16. Safety and Efficacy Are Separate Statistical Questions
The trial illustrates why an efficacy result should not be interpreted in isolation from safety. The primary statistical endpoint concerns time to a composite renal or mortality event, while serious adverse events are summarized separately by treatment group.
Efficacy question
How does randomized treatment assignment affect time to the first qualifying primary or secondary event?
Safety question
How frequently were serious adverse events observed in each treatment group?
Relative effect
The primary efficacy comparison is summarized with a hazard ratio and confidence interval.
Descriptive safety data
Serious adverse events are reported as affected participants over participants at risk for each arm.
A hazard ratio for efficacy and a serious-adverse-event count do not live on the same statistical scale. They should therefore remain separate when describing the evidence.
17. What the P-values Tell Us
The primary endpoint has a reported P = 0.30, while the secondary renal composite has P = 0.10. Both are statistical measures associated with their respective hypothesis tests.
| Endpoint | P-value | What it addresses | What it does not provide |
|---|---|---|---|
| Primary composite | 0.30 | Evidence against the null hypothesis under the reported log-rank test | The probability that the null hypothesis is true; effect size; clinical importance |
| Secondary renal composite | 0.10 | Evidence against the null hypothesis under the reported log-rank test | The probability that the treatment has no effect; magnitude of benefit; individual patient risk |
For time-to-event analyses, the most informative statistical description combines the hazard ratio, its confidence interval, and the hypothesis-test result. Reporting the P-value alone would omit the magnitude and precision of the estimated treatment effect.
18. Confidence Intervals as Measures of Precision
Primary endpoint
HR 0.88 with a 95% CI of 0.70–1.12. The interval includes both a lower estimated hazard and a value above the null.
Secondary endpoint
HR 0.78 with a 95% CI of 0.58–1.05. The interval also crosses the null value of 1.
A confidence interval should not be read as the range of treatment effects that occurred among individual participants. It quantifies uncertainty around the estimated population-level parameter under the statistical model and confidence procedure used.
The widths of the intervals also matter. The primary estimate is surrounded by a 0.70–1.12 interval, while the secondary estimate is surrounded by a 0.58–1.05 interval. The estimates should therefore be interpreted as uncertain quantities rather than exact measures of treatment effect.
19. Proportional-Hazards Considerations
The posted effect measure is a Cox proportional-hazards estimate. The proportional-hazards framework assumes that the relative hazard between treatment groups is adequately represented by a common hazard ratio over the relevant follow-up.
This issue is especially relevant when interpreting long follow-up from time-to-event trials because the hazard ratio compresses a potentially complex set of event and censoring patterns into one relative measure.
20. Censoring and Follow-Up
Time-to-event analysis permits participants to contribute information even when they do not experience an event during observed follow-up. A participant who remains event-free when follow-up ends can be censored rather than treated as if the event will never occur.
This is one reason survival analysis differs from a simple comparison of event proportions. The analysis uses both event occurrence and the timing of available follow-up.
For VA NEPHRON-D, the registered endpoints are explicitly defined as time from enrollment to the first qualifying event, with a time frame of up to 4.5 years.
21. Trial Status and Statistical Context
The registry identifies the study status as TERMINATED. It records a start date of July 2008 and a primary completion date of September 2014.
The most important statistical context surrounding termination is the October 2012 Data Monitoring Committee recommendation. The registry states that treatment was stopped primarily because of safety concerns along with low conditional power to detect a treatment effect on the primary outcome.
These facts affect how the reported estimates should be read. A trial stopped for safety and low conditional power is not statistically interchangeable with a trial that continues until a planned final information target is reached.
22. Why This Trial Matters Statistically
VA NEPHRON-D is a useful teaching case because it brings together several core principles of clinical-trial biostatistics: randomized comparison, double masking, composite time-to-event endpoints, log-rank testing, Cox hazard ratios, confidence intervals, censoring, and the interpretation of an early stopping decision.
| Statistical concept | How it appears in VA NEPHRON-D |
|---|---|
| Randomization | Participants were allocated randomly to two treatment groups. |
| Double masking | The registry identifies the study as double masked. |
| Parallel design | The registry identifies a parallel trial design with two treatment arms. |
| Time-to-event endpoint | The primary endpoint measures time to the first qualifying renal, ESRD, or death event. |
| Composite endpoint | Multiple clinically defined event types contribute to the primary endpoint. |
| Log-rank test | The registry reports log-rank testing for the primary and secondary posted analyses. |
| Hazard ratio | The treatment effect is summarized using a Cox proportional-hazards estimate. |
| Confidence interval | Both posted analyses include two-sided 95% confidence intervals. |
| Superiority hypothesis | The registry identifies superiority as the hypothesis type. |
| Early stopping | The Data Monitoring Committee recommended stopping treatment in October 2012 because of safety concerns and low conditional power. |
| Safety analysis | Serious adverse events are reported separately for the two treatment groups. |
23. Limitations and Interpretation Issues
- Early termination: the registry states that study treatment was stopped following a Data Monitoring Committee recommendation because of safety concerns along with low conditional power to detect a treatment effect on the primary outcome.
- Uncertainty of the primary estimate: the primary HR of 0.88 has a 95% CI of 0.70–1.12, which includes the null value of 1.
- Composite endpoint: the primary outcome combines different renal and mortality events. The overall hazard ratio summarizes the composite rather than isolating the effect on each component.
- Different secondary endpoint: the secondary renal composite excludes death, so its HR should not be interpreted as interchangeable with the primary HR.
- Hazard-ratio assumptions: Cox proportional-hazards interpretation depends on the model's proportional-hazards framework.
- Censoring: time-to-event estimates depend on follow-up and censoring information, not simply the number of participants experiencing an event.
- Safety context: serious adverse events are a separate outcome domain and should not be combined mathematically with the efficacy hazard ratio.
- Reported statistical scope: the registry provides formal analyses for the primary and secondary time-to-event outcomes, but the available record summarized here does not provide additional subgroup, component-specific, or long-term statistical estimates.
24. A Practical Reading of the Primary Result
The primary estimand is a relative time-to-event comparison for the first occurrence of the registered composite endpoint.
The HR of 0.88 is below 1, corresponding to an estimated 12% lower instantaneous event rate in the combination group under the Cox model.
The 95% CI of 0.70–1.12 spans the null value of 1 and therefore indicates substantial uncertainty around the point estimate.
The P-value is 0.30 for the reported log-rank analysis. It summarizes statistical evidence under the test; it is not an estimate of treatment benefit or harm.
The result should be interpreted alongside the randomized, double-masked, parallel design and the registry's documentation that treatment was stopped in October 2012 because of safety concerns together with low conditional power.
25. Related Tutorials
Learn more about the statistical methods used in this trial:
26. Related Calculators
27. Sources
- ClinicalTrials.gov: VA NEPHRON-D, NCT00555217.
- PubMed: PMID 24206457.
- PubMed: PMID 24552328.
- PubMed: PMID 25014549.
- PubMed: PMID 25157242.
- PubMed: PMID 40548378.
Continue through the Clinical Biostats statistical library
Connect this trial's time-to-event methods with deeper tutorials and practical statistical tools.
28. Record Summary
VA NEPHRON-D provides a useful example of how a randomized clinical trial can combine a composite time-to-event endpoint with log-rank testing and Cox hazard-ratio estimation. The primary analysis reported an HR of 0.88 with a 95% CI of 0.70–1.12 and P = 0.30. The secondary renal composite reported an HR of 0.78 with a 95% CI of 0.58–1.05 and P = 0.10. Both estimates require interpretation through their confidence intervals, endpoint definitions, and time-to-event framework.
The trial's statistical context also includes an important early-stopping decision. The registry states that in October 2012 the Data Monitoring Committee recommended stopping study treatment primarily because of safety concerns along with low conditional power to detect a treatment effect on the primary outcome. That design history is an essential part of understanding the evidence alongside the reported efficacy and safety results.