This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. The ClinicalTrials.gov record reports one posted outcome measure, but do not provide a formal statistical analysis, effect estimate, confidence interval, or p-value for that outcome. Accordingly, this page distinguishes the registered endpoint and posted outcome measure from the statistical analysis that would ordinarily be used to evaluate it.
1. Trial at a Glance
PIVOT was a randomized, parallel-group phase 3 trial evaluating radical prostatectomy versus watchful waiting in patients with prostate cancer. The ClinicalTrials.gov record reports 731 participants, two arms, and all-cause mortality as the registered primary endpoint.
| Feature | PIVOT |
|---|---|
| Phase | Phase 3 |
| Condition | Prostate Cancer |
| Intervention | Radical prostatectomy |
| Comparator | Watchful Waiting |
| Design | Randomized, parallel |
| Masking | None |
| Primary purpose | Treatment |
| Enrollment | 731.0 |
| Lead sponsor | VA Office of Research and Development |
| Registry status | COMPLETED |
| ClinicalTrials.gov | NCT00007644 |
2. Clinical Question
The central clinical question was whether radical prostatectomy should be compared with watchful waiting with respect to all-cause mortality in patients enrolled in this phase 3 prostate cancer trial.
Population
Participants with prostate cancer enrolled in the PIVOT trial. The ClinicalTrials.gov record does not provide a more detailed baseline population description.
Intervention
Radical prostatectomy, recorded in the registry as a procedure.
Comparator
Watchful Waiting.
Primary question
What was the number of deaths from any cause from randomization through the end of the study, assessed for up to 16 years?
3. Trial Design
Radical Prostatectomy
- Intervention recorded as a procedure
- Randomized allocation
- Serious adverse events: 171/364 affected/at risk
Watchful Waiting
- Comparator arm
- Randomized allocation
- Serious adverse events: 183/367 affected/at risk
4. Trial Timeline
Study start
The registry records November 1994 as the study start.
Primary completion
The registry records January 2010 as the primary completion date.
Registry status
The ClinicalTrials.gov record lists the study status as completed.
5. Registered Primary Endpoint
The ClinicalTrials.gov record contains one registered primary endpoint: All Cause Mortality.
| Endpoint | Registry definition | Time frame | Results posted |
|---|---|---|---|
| All Cause Mortality | Number of deaths from any cause. | From date of randomization until date of death from any cause, assessed until end of study, up to 16 years | Yes |
Although the registered endpoint is labeled All Cause Mortality and its definition is the number of deaths from any cause, its time frame is explicitly defined from randomization until death and permits assessment through the end of the study for up to 16 years. That makes time-to-event methods particularly relevant when the individual event times and censoring information are available.
6. Results Status
The ClinicalTrials.gov record indicates that results have been posted and that one outcome measure is posted. However, no formal statistical analyses were posted, with 0 analyses omitted for length.
This distinction is important. A posted outcome measure establishes that the registry contains results information for the endpoint, but the absence of a registry-reported statistical analysis means that the evidence available in the registry record does not support reporting a numerical comparative treatment effect.
7. Planned Analysis
Because the ClinicalTrials.gov record does not provide a formal statistical analysis for the posted All Cause Mortality outcome, the appropriate statistical framework can be described without assigning an outcome to either treatment arm.
Time-to-event analysis
For an endpoint defined as time from randomization to death from any cause, a standard analysis would treat death as the event and patients who have not died by their last evaluable follow-up as right-censored observations. The analysis would preserve the time information rather than reducing follow-up to a simple proportion alive or dead.
Patients without a recorded death during available follow-up would ordinarily contribute information until their censoring time.
Kaplan-Meier estimation
A Kaplan-Meier estimator would ordinarily be used to estimate the survival function over time. This would provide a graphical and numerical description of the probability of remaining alive beyond a given time point while accounting for right censoring.
Between-arm comparison
A randomized two-arm trial with an all-cause mortality time-to-event endpoint would commonly compare the survival distributions using a log-rank test and estimate a relative treatment effect with a Cox proportional-hazards model. The precise analysis used for PIVOT cannot be asserted from the ClinicalTrials.gov record because the statistical-analysis field contains no entries.
What cannot be concluded from the ClinicalTrials.gov record
- No hazard ratio can be reported from the registry-reported statistical-analyses data.
- No confidence interval can be reported.
- No p-value can be reported.
- No median survival can be reported.
- No survival probability at a particular time point can be reported.
- No between-arm mortality difference can be calculated without the underlying event and follow-up information.
8. Statistical Methodology
Randomization
The study allocation is recorded as RANDOMIZED. Randomization is the central design feature that allows the two treatment groups to be compared while reducing systematic differences in measured and unmeasured baseline characteristics in expectation.
Randomization does not guarantee identical groups. It creates a probability mechanism for treatment assignment that supports causal comparison when the trial is conducted and analyzed appropriately.
Parallel-group design
The registry specifies a PARALLEL design model. In a parallel trial, participants remain associated with their randomized treatment strategy rather than crossing through a sequence of experimental conditions as in a crossover design.
Unmasked treatment assignment
The registry specifies NONE for masking. This means the trial was not described as masked. For a procedure-versus-management comparison, the lack of masking can be relevant to post-randomization care, outcome ascertainment, and behavior, although the ClinicalTrials.gov record does not document the magnitude of any such effect.
Primary purpose
The registry classifies the primary purpose as TREATMENT. The intervention is radical prostatectomy and the comparator is watchful waiting.
All-cause mortality as an endpoint
All-cause mortality is an especially direct time-to-event endpoint because the event is death from any cause rather than a composite of selected disease-related events. The registry definition is simply the number of deaths from any cause.
9. Serious Adverse Events
the ClinicalTrials.gov record includes serious adverse-event counts by arm. These are reported as affected participants divided by participants at risk.
| Arm | Affected | At risk |
|---|---|---|
| Radical Prostatectomy | 171 | 364 |
| Watchful Waiting | 183 | 367 |
The registry reports 171 affected participants among 364 at risk in the radical-prostatectomy arm and 183 among 367 in the watchful-waiting arm. These are the reported serious-adverse-event counts and denominators.
They should not be interpreted as a formal between-arm treatment effect because the ClinicalTrials.gov record does not provide a statistical analysis, confidence interval, or p-value for serious adverse events. The counts also do not establish whether an event occurred before or after a particular time point unless that timing information is separately available.
10. Statistical Methods Explained
Why is randomization important in PIVOT?
Randomization assigns treatment according to a predefined allocation mechanism rather than allowing treatment choice to determine group membership. In expectation, this balances both measured and unmeasured prognostic factors between treatment groups. That is the foundation for interpreting differences in subsequent outcomes as treatment-associated differences rather than simply differences in who received each strategy.
Why is all-cause mortality naturally suited to survival analysis?
The registered endpoint is assessed from the date of randomization until death, with follow-up potentially extending to 16 years. Participants therefore have different amounts of observed follow-up, and some may not experience death during their observable period. Survival methods incorporate both the event time and the censoring time rather than discarding incomplete follow-up.
What would a Kaplan-Meier curve show?
A Kaplan-Meier curve would estimate the probability of remaining alive over time separately for the randomized treatment groups. Each observed death would change the estimated survival function, while censored observations would contribute information up to their censoring time without being counted as deaths.
What would a hazard ratio mean here?
If a Cox model were used, the hazard ratio would describe the relative instantaneous rate of death between the randomized treatment groups under the fitted model. A hazard ratio below 1 would indicate a lower estimated hazard in the numerator treatment group; it would not mean that the same percentage of patients avoided death or that individual patients experienced a fixed percentage reduction in risk.
Why does censoring matter?
Right censoring occurs when a participant's exact event time is not observed within the available follow-up period. Survival methods rely on assumptions about the censoring mechanism. Informative censoring—where the likelihood of being censored is related to future event risk in a way not adequately handled by the analysis—can threaten interpretation.
Why can the number of deaths alone be insufficient?
The registered endpoint definition identifies the number of deaths, but the time frame begins at randomization and extends through the end of the study. Two groups could have the same number of deaths while having very different timing of those deaths. Time-to-event analysis preserves that timing information and can therefore answer a different and more informative question than a simple event count.
11. Understanding the Reported Safety Counts
The serious-adverse-event information posted on ClinicalTrials.gov for PIVOT uses the structure affected / at risk. This distinction is statistically important because the denominator defines the population to which the reported count applies.
| Measure | Radical Prostatectomy | Watchful Waiting |
|---|---|---|
| Affected participants | 171 | 183 |
| Participants at risk | 364 | 367 |
| Formal comparative analysis reported? | No | No |
A descriptive event count is not equivalent to a hypothesis test. A formal safety comparison could use a risk ratio, risk difference, odds ratio, time-to-first-event analysis, or another prespecified method depending on the endpoint definition and follow-up structure. The ClinicalTrials.gov record does not identify which method, if any, was used for this safety outcome.
12. Analysis Population and Denominators
The overall enrollment is reported as 731.0, while the serious-adverse-event data use arm-specific denominators of 364 and 367. The ClinicalTrials.gov recordset does not explain the relationship between the enrollment total and these safety denominators.
13. Missing Data, Imputation, and Follow-Up
The ClinicalTrials.gov record does not describe a missing-data strategy, an imputation method, or rules for handling loss to follow-up for the all-cause mortality endpoint.
For a mortality time-to-event analysis, investigators would ordinarily distinguish between a participant who has died and a participant whose follow-up ends without an observed death. The latter is generally handled through censoring rather than imputing a death time. The exact censoring rules, however, must come from the trial's statistical analysis plan or registry analysis information, neither of which is reported here.
14. Stratification and Covariate Adjustment
The ClinicalTrials.gov record identifies the allocation as randomized but do not list stratification factors or prespecified covariates for the primary analysis.
| Design or analysis feature | Supplied data |
|---|---|
| Randomization | Yes |
| Parallel design | Yes |
| Masking | None |
| Stratification factors | Not reported in the ClinicalTrials.gov record |
| Covariate-adjusted primary model | Not reported in the ClinicalTrials.gov record |
| Bayesian methods | Not reported in the ClinicalTrials.gov record |
| Interim analysis | Not reported in the ClinicalTrials.gov record |
| Multiplicity procedure | Not reported in the ClinicalTrials.gov record |
| Missing-data/imputation method | Not reported in the ClinicalTrials.gov record |
15. Non-Inferiority, Multiplicity, and Interim Analysis
The ClinicalTrials.gov record does not identify a non-inferiority design or margin, a multiplicity procedure, or an interim-analysis plan.
Non-inferiority
No non-inferiority margin is reported in the ClinicalTrials.gov record. The page therefore does not apply non-inferiority logic to the PIVOT endpoint.
Multiplicity
The ClinicalTrials.gov record identifies one registered primary endpoint. No formal multiplicity adjustment is documented in the ClinicalTrials.gov record.
Interim analysis
No interim-analysis procedure or alpha-spending method is reported in the ClinicalTrials.gov record.
Bayesian methods
No Bayesian analysis is identified in the ClinicalTrials.gov record.
16. Interpreting a Potential Hazard Ratio
If the full trial analysis reported a hazard ratio for all-cause mortality, that estimate would summarize a relative time-to-event comparison between the randomized groups. A value below 1 would correspond to a lower estimated hazard for the specified numerator group; a value above 1 would correspond to a higher estimated hazard.
The registry-reported PIVOT data do not provide such an estimate. Therefore, no numerical hazard ratio, confidence interval, or p-value is presented here.
This is an important statistical discipline: a familiar endpoint does not justify supplying a remembered result when the task requires every numerical result to come from the ClinicalTrials.gov record.
17. Why a Confidence Interval Would Matter
If a formal treatment-effect estimate were available, its confidence interval would be essential for interpreting precision. The point estimate alone would not show how much uncertainty surrounds the estimated treatment effect.
A confidence interval describes uncertainty around an estimated parameter under the specified statistical model and sampling framework. It does not describe the range of outcomes that individual patients will experience.
Because the registry-reported statistical-analyses field is empty, there is no PIVOT confidence interval to interpret for the registered all-cause mortality endpoint.
18. Why a P-Value Would Not Measure Effect Size
A p-value addresses compatibility of the observed data with a specified null hypothesis under the statistical model. It is not a measure of the magnitude of a treatment effect.
For a time-to-event endpoint, a small p-value could accompany a relatively modest effect in a large dataset, while a clinically meaningful effect could fail to reach a prespecified threshold in a small or information-limited dataset. Effect estimates and their confidence intervals therefore carry information that a p-value alone cannot provide.
No p-value for the PIVOT all-cause mortality comparison is present in the ClinicalTrials.gov record.
19. Limitations
- Incomplete statistical reporting in the ClinicalTrials.gov recordset: results are indicated as posted, but no formal statistical analyses were posted.
- No comparative effect estimate: no hazard ratio, confidence interval, p-value, or formal between-arm mortality analysis is available in the ClinicalTrials.gov record.
- Limited endpoint detail: the registered primary endpoint is defined as the number of deaths from any cause, with follow-up from randomization until death and assessment through the end of study for up to 16 years.
- No detailed analysis plan: stratification factors, covariate adjustment, interim monitoring, multiplicity, Bayesian methods, and missing-data procedures are not reported in the ClinicalTrials.gov record.
- Safety denominator differences: serious adverse events use denominators of 364 and 367, while overall enrollment is reported as 731.0; the ClinicalTrials.gov record does not explain the denominator derivation.
- No baseline table: demographic and disease characteristics are not provided in the ClinicalTrials.gov record and are not reported here.
- No subgroup estimates: subgroup analyses and forest-plot estimates are not included in the ClinicalTrials.gov record.
- No median survival: median survival times are not reported in the ClinicalTrials.gov record and are not inferred.
- No treatment recommendation: statistical evidence and clinical decision-making are separate questions; this page does not recommend either management strategy.
20. Why This Trial Matters Statistically
PIVOT is a useful statistical teaching case because the registry data combine a randomized two-arm design with a long-horizon mortality endpoint and an unmasked procedure-versus-management comparison.
| Concept | How it appears in the registry-reported PIVOT data |
|---|---|
| Randomization | The allocation is explicitly recorded as RANDOMIZED. |
| Parallel-group design | The design model is PARALLEL. |
| Unmasked trial | Masking is recorded as NONE. |
| Time-to-event endpoint | All Cause Mortality is assessed from randomization until death, through the end of study. |
| Long follow-up | The registered endpoint may be assessed for up to 16 years. |
| Survival analysis | Kaplan-Meier and Cox methods would ordinarily be relevant to this endpoint type, although the ClinicalTrials.gov record does not specify the actual method. |
| Censoring | Potentially relevant because the endpoint is followed over time; exact censoring rules are not reported. |
| Safety denominators | Serious adverse events are reported as affected/at risk separately by arm. |
| Analysis transparency | The registry indicates posted results, but no formal statistical analyses were posted. |
21. Related Tutorials
Learn more about the methods and statistical concepts relevant to this trial:
22. Related Statistical Calculators
23. Sources
- ClinicalTrials.gov: PIVOT — NCT00007644.
- Linked publication record: PubMed PMID 7577210.
- Linked publication record: PubMed PMID 7638470.
- Linked publication record: PubMed PMID 9268976.
- Linked publication record: PubMed PMID 9674725.
- Linked publication record: PubMed PMID 10866876.
Continue through the Clinical Biostats statistical pathway
Connect randomized trial design to survival analysis, time-to-event methods, and clinical-trial statistical tools.
24. Record Summary
PIVOT is a completed randomized phase 3 parallel-group trial in prostate cancer comparing radical prostatectomy with watchful waiting. The ClinicalTrials.gov record reports 731.0 enrolled participants, two arms, no masking, and one registered primary endpoint: All Cause Mortality, defined as the number of deaths from any cause and assessed from randomization until death through the end of study, up to 16 years.
The statistical interpretation of this record is necessarily constrained by the ClinicalTrials.gov record. ClinicalTrials.gov indicates that results have been posted, but the provided statistical-analyses field is empty. Instead, it explains how the registered mortality endpoint would ordinarily be analyzed using time-to-event methods and distinguishes those methodological principles from results that are actually documented in the ClinicalTrials.gov record.