This page separates reported trial information from statistical interpretation. The ClinicalTrials.gov record contains the registered primary endpoint and serious adverse-event counts, but do not contain formal statistical-analysis results for the primary endpoint. Accordingly, no treatment-effect estimate, confidence interval, or p-value is presented here.
1. Trial at a Glance
OUTBACK was a completed, randomized phase 3 trial evaluating cisplatin and radiation therapy with or without carboplatin and paclitaxel in patients with locally advanced cervical cancer. The registry reports 926 participants, two parallel treatment arms, an open-label design, and one registered primary time-to-event endpoint: overall survival rate at 5 years.
| Feature | OUTBACK |
|---|---|
| Trial name | OUTBACK |
| Phase | Phase 3 |
| Status | Completed |
| Population | Patients with locally advanced cervical cancer represented by the registered cervical cancer histologies and stages |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | None |
| Primary purpose | Treatment |
| Enrollment | 926 |
| Arms | 2 |
| Primary endpoint | Overall Survival Rate at 5 Years |
| Primary endpoint type | Time-to-event |
| Results posted | Yes |
| Outcome measures posted | 6 |
| Lead sponsor | GOG Foundation |
| Sponsor type | Network |
| ClinicalTrials.gov | NCT01414608 |
2. Clinical Question
The central clinical question is whether the treatment strategy evaluated in the OUTBACK trial should be assessed against standard cisplatin and radiation therapy using a long-term overall-survival endpoint. The registry describes the trial as comparing cisplatin and radiation therapy with or without carboplatin and paclitaxel in patients with locally advanced cervical cancer.
Population
The registered conditions include cervical adenocarcinoma, cervical adenosquamous carcinoma, cervical squamous cell carcinoma not otherwise specified, and the specified stage categories of cervical cancer in the registry record.
Intervention framework
The registered interventions include carboplatin, cisplatin, paclitaxel, external beam radiation therapy, and brachytherapy.
Comparator framework
The trial title identifies cisplatin and radiation therapy as the common treatment framework, with the comparison involving the addition or absence of carboplatin and paclitaxel.
Primary question
Does the randomized treatment strategy produce a different probability of overall survival at 5 years from study randomization?
3. Trial Design
Standard Chemoradiation
- The registry's serious-adverse-event data identify this as one of the two treatment groups.
- The trial title identifies cisplatin and radiation therapy as the standard treatment framework.
- Serious adverse events affected 98 of 453 participants in this arm.
Standard Chemoradiation With Adjuvant Ch
- The registry's serious-adverse-event data identify this as the second treatment group.
- The trial title identifies carboplatin and paclitaxel as the additional treatment components being evaluated.
- Serious adverse events affected 107 of 361 participants in this arm.
4. Trial Timeline
| Milestone | Date |
|---|---|
| Study start | 2012-01-09 |
| Primary completion | 2021-07-09 |
| Status | Completed |
The interval between study start and primary completion is part of the registry record, but the ClinicalTrials.gov record does not provide enrollment dates, data-cutoff dates, follow-up distributions, or dates associated with the reported outcome measures.
5. Endpoints
| Endpoint | Registry definition | Time frame | Type |
|---|---|---|---|
| Overall Survival Rate at 5 Years | Estimate for probability of overall survival at 5 years by Kaplan-Meier method, where overall survival is defined as the time from randomization to time of death due to any cause or the date of last contact, whichever occurs first. | 5 years from study randomization | Time-to-event |
This endpoint has an important statistical feature: the registry does not simply ask whether a participant was alive at a fixed visit. It defines survival as time from randomization until death or the last contact date, with the 5-year survival probability estimated using Kaplan-Meier methodology.
6. Results
Primary Endpoint: Overall Survival Rate at 5 Years
The ClinicalTrials.gov record identifies the primary endpoint and state that results are posted, but the ClinicalTrials.gov record contains no formal treatment-effect estimate, confidence interval, or p-value.
The registry defines the primary endpoint as the probability of overall survival at 5 years, estimated by the Kaplan-Meier method. The ClinicalTrials.gov record contains no numerical 5-year survival estimates for either randomized group and no formal comparison between the groups.
The appropriate interpretation is therefore limited to the endpoint definition and the analysis framework. A 5-year Kaplan-Meier survival estimate is an absolute probability estimate at a specified time point. It is not itself a hazard ratio, risk ratio, odds ratio, or p-value.
A formal between-group comparison would normally report an effect estimate together with an uncertainty interval and a hypothesis test.
The confidence interval, if reported in a complete analysis, would quantify statistical uncertainty around the estimated treatment effect or survival estimate, depending on the analysis being described. It would not describe the range of outcomes expected for an individual patient.
A p-value, if reported, would address evidence against a specified null hypothesis under the statistical model and testing framework. It would not measure the size or clinical importance of the treatment effect.
Because the endpoint is time-to-event, censoring is central to the analysis. Patients who have not died by their last available contact can contribute survival information up to that point. The ClinicalTrials.gov record does not describe the censoring rules beyond the endpoint definition, so no additional assumptions are made here.
7. Secondary and Other Posted Outcome Measures
The registry record states that 6 outcome measures are posted. The ClinicalTrials.gov record identifies the number of posted outcome measures but provide no definitions, estimates, confidence intervals, p-values, or formal analyses for the additional outcomes.
Accordingly, this page does not assign numerical results to secondary endpoints. A complete statistical interpretation of a secondary outcome would require its registered definition, time frame, analysis population, and reported estimates.
8. Statistical Methodology
Kaplan-Meier estimation
The registered primary endpoint explicitly specifies estimation of the probability of overall survival at 5 years by the Kaplan-Meier method. Kaplan-Meier estimation is designed for time-to-event data in which follow-up can be incomplete for some participants.
Here, di represents the number of observed events at an event time and ni represents the number of participants at risk immediately before that time.
The estimator builds the survival curve sequentially at observed event times. Participants who are censored contribute information while they remain under observation, but they are not counted as deaths after their censoring time.
Why the 5-year estimate is a time-specific quantity
A 5-year survival rate answers a specific question: what is the estimated probability of remaining alive through 5 years after randomization under the observed survival experience? It is therefore an absolute measure at a defined time point.
It should not be confused with the median survival time. The median is the time at which the estimated survival probability reaches 0.50, whereas a 5-year survival estimate fixes the time point first and estimates survival at that time. The ClinicalTrials.gov record does not report a median survival estimate.
Time origin and event definition
The registered endpoint uses study randomization as time zero. The event is death due to any cause. If a participant is not known to have died, the definition uses the date of last contact, whichever occurs first.
This time origin is statistically important because treatment assignment occurs at randomization. Beginning follow-up there aligns the primary time-to-event analysis with the randomized comparison.
Typical treatment-group comparison
For two randomized groups with a time-to-event endpoint, a conventional analysis would commonly compare Kaplan-Meier curves using a log-rank test and estimate a relative treatment effect with a Cox proportional-hazards model. However, the registry-reported OUTBACK data do not specify that these particular comparative procedures were used for the between-arm analysis.
The only method explicitly attached to the registered primary endpoint in the ClinicalTrials.gov record is the Kaplan-Meier method. The additional procedures described here are therefore methodological context, not claims about an unreported OUTBACK analysis.
9. Statistical Methods Explained
Why was Kaplan-Meier used for the primary endpoint?
Overall survival is a time-to-event endpoint. Participants can have different lengths of observed follow-up, and some may still be alive at their last contact. Kaplan-Meier estimation is designed to use the available follow-up while accounting for right censoring rather than requiring every participant to have complete 5-year follow-up.
What does a 5-year survival rate mean?
A 5-year survival estimate is the estimated probability of surviving to 5 years after randomization. It is an absolute time-specific quantity. It does not tell us the instantaneous event rate at a particular moment and does not by itself quantify the relative effect of one randomized treatment versus another.
Why is censoring important?
If a participant is alive at the last available contact but is not observed indefinitely, the analysis does not treat that participant as having died. Instead, the participant contributes information until the censoring time. The validity of the Kaplan-Meier estimate depends on the assumptions underlying that censoring process.
What would a hazard ratio add?
A hazard ratio would provide a relative comparison of the estimated instantaneous event rates between treatment groups under a specified survival model. It would answer a different question from the 5-year survival probability. The registry-reported OUTBACK statistical-analysis data do not contain a hazard ratio, so no hazard ratio is reported on this page.
Why is a p-value not enough?
A p-value is evidence relative to a null hypothesis and testing framework. It does not describe how large the treatment difference is. For a clinically interpretable time-to-event analysis, an absolute survival estimate, an appropriate relative effect measure, and a measure of uncertainty are more informative when available together.
Why does randomization matter statistically?
Randomization creates the basis for comparing outcomes between treatment assignments without relying on observational adjustment to establish the initial treatment groups. The OUTBACK registry identifies the allocation as randomized and the design as parallel. The ClinicalTrials.gov record does not provide the randomization ratio or stratification variables.
10. Safety Results
The ClinicalTrials.gov record provides serious adverse-event counts by arm. These are reported as affected participants divided by participants at risk.
| Treatment group | Participants with serious AEs | Participants at risk | Reported affected / at risk |
|---|---|---|---|
| Standard Chemoradiation | 98 | 453 | 98/453 |
| Standard Chemoradiation With Adjuvant Ch | 107 | 361 | 107/361 |
The registry data therefore show the number of participants affected by serious adverse events and the corresponding number at risk in each reported arm. The ClinicalTrials.gov record does not provide a confidence interval, p-value, rate ratio, risk ratio, or formal statistical comparison for these safety outcomes.
The reported quantities are counts over denominators. They should not be interpreted as evidence of a statistically significant or nonsignificant difference between the arms because no formal comparison is reported.
For safety endpoints, the analysis population and exposure definition matter. The denominator can differ from the total randomized enrollment, as it does here. The ClinicalTrials.gov record identifies the denominators but does not explain the precise eligibility or exposure rules used to construct them.
Serious adverse events are also distinct from all adverse events, treatment-emergent adverse events, and grade-specific adverse events. The ClinicalTrials.gov record should therefore not be generalized to those other safety categories.
11. Missing Data, Censoring, and Analysis Populations
The primary endpoint is inherently subject to incomplete follow-up because overall survival is observed over time. The registry definition explicitly uses the date of last contact when death has not occurred by that point.
Censoring
The endpoint definition permits participants to contribute survival information through their last contact when death has not occurred.
Imputation
The ClinicalTrials.gov record does not report an imputation method for missing data, and no imputation procedure is inferred.
Analysis population
The ClinicalTrials.gov record does not identify the exact population used for the posted primary endpoint analysis.
Follow-up
The ClinicalTrials.gov record provides the study start and primary completion dates but do not provide participant-level follow-up distributions.
This distinction is important because randomized enrollment, safety denominators, and the population contributing to a time-to-event estimate are not necessarily interchangeable quantities. The ClinicalTrials.gov record should therefore not be used to infer an analysis population that has not been explicitly reported.
12. Randomization and Allocation
OUTBACK is registered as a randomized, parallel trial with none masking. These three descriptors establish the broad structure of the comparison.
| Design element | Registry information | Statistical implication |
|---|---|---|
| Allocation | Randomized | Provides the basis for a prospective comparison of treatment assignments. |
| Design model | Parallel | Participants are evaluated within one of the trial's treatment groups rather than through a crossover design described in the ClinicalTrials.gov record. |
| Masking | None | The ClinicalTrials.gov record describes an unmasked study. |
| Number of arms | 2 | The primary comparison is structured between two randomized treatment groups. |
| Enrollment | 926 | The registered study enrollment provides the overall scale of the trial. |
The ClinicalTrials.gov record does not specify the allocation ratio, randomization blocks, stratification factors, concealment procedure, or sequence-generation method. These design details can materially affect implementation but should not be from assumptions.
13. Multiplicity, Interim Analysis, and Bayesian Methods
The ClinicalTrials.gov record does not report an interim-analysis strategy, alpha-spending method, multiplicity procedure, or Bayesian analysis for OUTBACK.
Multiplicity
No multiplicity adjustment or endpoint hierarchy is provided in the ClinicalTrials.gov record. The single registered primary endpoint should therefore be distinguished from the six total posted outcome measures.
Interim analysis
No interim-analysis schedule, stopping boundary, or alpha-spending approach is reported.
Bayesian methods
No Bayesian model, prior distribution, posterior estimate, or Bayesian decision rule is identified in the ClinicalTrials.gov record.
Non-inferiority
No non-inferiority margin or non-inferiority testing framework is posted on ClinicalTrials.gov for the registered primary endpoint.
These are not merely technical omissions. A formal clinical-trial interpretation depends on knowing whether multiple hypotheses were tested, whether interim looks were planned, and whether the inferential framework was superiority, non-inferiority, equivalence, or another design. The registry-reported OUTBACK data do not establish those features beyond the registered primary endpoint and randomized design.
14. Crossover and Treatment Switching
The ClinicalTrials.gov record does not report crossover or treatment switching. Consequently, no adjustment for crossover is described and no claim is made about its effect on the interpretation of overall survival.
This is particularly relevant for time-to-event trials because treatment received after randomization can affect subsequent survival. If crossover occurred, a simple intention-to-treat comparison and a treatment-received comparison could answer different questions. The present record does not contain enough information to determine whether such an issue applies to OUTBACK.
15. Limitations
- No formal primary-endpoint analysis reported: the primary endpoint is defined and results are marked as posted, but the ClinicalTrials.gov recordset contains no estimate, confidence interval, or p-value.
- No numerical 5-year survival estimate reported: the registry definition identifies the requested Kaplan-Meier quantity but does not provide its numerical value in the ClinicalTrials.gov record.
- No comparative effect measure: a hazard ratio, risk ratio, survival difference, or other treatment-effect estimate is not present in the ClinicalTrials.gov record.
- Additional outcome measures are incompletely described: the registry states that 6 outcome measures are posted, but only the primary endpoint is reported in endpoint-level detail here.
- Safety analysis is limited: serious adverse-event affected/at-risk counts are available, but no formal statistical comparison is reported.
- Analysis population is not specified: the ClinicalTrials.gov record does not establish whether the primary result used all randomized participants or another defined analysis population.
- Missing-data procedures are not specified: no imputation strategy is provided.
- Design details are incomplete: the ClinicalTrials.gov record does not specify stratification, randomization ratio, interim monitoring, multiplicity control, crossover, or Bayesian methodology.
- Follow-up information is limited: study start and primary completion dates are available, but participant-level follow-up and censoring distributions are not.
16. Why This Trial Matters Statistically
OUTBACK is a useful teaching example because its registered primary endpoint illustrates a central problem in clinical-trial statistics: a clinically intuitive endpoint such as 5-year overall survival still requires careful treatment of time-to-event structure, censoring, and uncertainty.
| Concept | How it appears in OUTBACK |
|---|---|
| Randomization | The study is registered as randomized. |
| Parallel-group design | The registry identifies a parallel design model with 2 arms. |
| Time-to-event analysis | The registered primary endpoint is a time-to-event endpoint. |
| Kaplan-Meier estimation | The 5-year overall survival probability is explicitly specified to be estimated by Kaplan-Meier. |
| Censoring | The primary endpoint definition incorporates date of last contact when death has not occurred. |
| Absolute survival probability | The primary endpoint focuses on the probability of overall survival at 5 years. |
| Comparative inference | A formal between-arm analysis would require an explicitly reported effect estimate and uncertainty measure. |
| Safety denominators | Serious adverse-event counts are reported relative to arm-specific numbers at risk. |
| Evidence boundaries | The ClinicalTrials.gov record demonstrate that results are posted but do not supply the formal primary-endpoint analysis. |
17. Absolute Survival and Relative Treatment Effects
A central statistical lesson from OUTBACK is the distinction between an absolute survival probability and a relative treatment effect.
Absolute measure
The registered endpoint asks for the probability of overall survival at 5 years. This directly describes survival at a clinically specified time point.
Relative measure
A hazard ratio would summarize a relative difference in instantaneous event rates under a Cox-type model. No such estimate is reported here.
These measures should not be substituted for one another. Two trials could have similar hazard ratios but different absolute 5-year survival probabilities if their baseline risks and follow-up patterns differ. Conversely, a difference in 5-year survival does not by itself establish a proportional-hazards relationship.
18. Understanding the Registered Endpoint
The registered endpoint asks for an estimate of the probability of overall survival at 5 years from randomization, using the Kaplan-Meier method.
The endpoint definition does not itself provide the magnitude of the treatment effect, the uncertainty around that magnitude, or the statistical significance of the comparison. Those require the actual results and analysis output.
Separating the endpoint definition from its numerical result prevents a common analytical error: treating the existence of a registered outcome as evidence for a particular observed treatment effect.
19. Interpreting Serious Adverse-Event Counts
The safety data posted on ClinicalTrials.gov for OUTBACK give two affected/at-risk quantities:
| Arm | Affected | At risk | Interpretive role |
|---|---|---|---|
| Standard Chemoradiation | 98 | 453 | Descriptive serious-adverse-event count |
| Standard Chemoradiation With Adjuvant Ch | 107 | 361 | Descriptive serious-adverse-event count |
The denominators differ from one another and from the overall enrollment of 926. That difference is itself a reminder that enrollment is not automatically the same as the denominator for every safety analysis.
A formal safety comparison would normally require additional information, such as the precise safety population, event definition, exposure period, handling of multiple events per participant, and statistical method. None of those details is reported here beyond the affected/at-risk counts.
20. What a Complete Time-to-Event Result Would Contain
For an endpoint such as OUTBACK's registered 5-year overall survival rate, a complete statistical presentation would ideally make several pieces of information visible together.
| Component | Purpose | Available in the ClinicalTrials.gov record? |
|---|---|---|
| 5-year Kaplan-Meier survival estimate | Describes absolute survival probability at 5 years. | No numerical estimate reported. |
| Confidence interval | Describes statistical uncertainty around the estimate. | No confidence interval reported. |
| Between-arm effect estimate | Quantifies the relative or absolute treatment difference. | No formal effect estimate reported. |
| Hypothesis test | Addresses the specified null hypothesis under the testing framework. | No p-value reported. |
| Number at risk | Shows how much follow-up information remains at successive times. | Not posted on ClinicalTrials.gov for the survival endpoint. |
| Censoring information | Clarifies how incomplete follow-up contributes to the analysis. | Only the endpoint definition is reported. |
This framework is useful because it shows exactly what is missing from the ClinicalTrials.gov record without filling those gaps with estimates from outside the ClinicalTrials.gov record.
21. Sources
- ClinicalTrials.gov: OUTBACK, NCT01414608.
- PubMed: PMID 37080223.
- PubMed: PMID 34201380.
22. Related Tutorials
Learn more about the statistical concepts that provide the foundation for interpreting this trial:
23. Related Statistical Calculators
Continue through the Clinical Biostats statistical tutorials
Explore the underlying trial-design concepts and statistical methods used to understand randomized clinical trials and time-to-event endpoints.
24. Record Summary
OUTBACK is a completed randomized phase 3 trial with 926.0 enrolled participants, two parallel treatment arms, no masking, and a registered primary time-to-event endpoint of overall survival rate at 5 years. The registry defines that endpoint as the Kaplan-Meier estimate of the probability of overall survival at 5 years from study randomization, with overall survival defined from randomization to death from any cause or the date of last contact, whichever occurs first.
The ClinicalTrials.gov record also report serious adverse-event counts of 98/453 in the Standard Chemoradiation group and 107/361 in the Standard Chemoradiation With Adjuvant Ch group. However, the ClinicalTrials.gov record contains no formal analysis results for the primary endpoint and no comparative effect estimates, confidence intervals, or p-values.