← Clinical Trials
Cardiovascular Phase 3 SCORED NCT03315143

SCORED: Complete Statistical Analysis of Sotagliflozin in Type 2 Diabetes and Renal Impairment

An independent statistical analysis of the randomized phase 3 SCORED trial evaluating sotagliflozin versus placebo in participants with type 2 diabetes mellitus, moderate renal impairment, and cardiovascular risk, with a primary endpoint combining cardiovascular death, hospitalizations for heart failure, and urgent heart-failure visits.

Trial start: 2017-12-19  ·  Primary completion: 2020-07-08  ·  Enrollment: 10584
Scope of this record

This page separates reported trial results from statistical interpretation. Numerical results on this page are taken from the ClinicalTrials.gov record. ClinicalTrials.gov provides the official trial registry record.

Registry record: 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

SCORED was a randomized, parallel-group, quadruple-masked phase 3 trial evaluating sotagliflozin versus placebo in participants with type 2 diabetes mellitus and moderate renal impairment who were at cardiovascular risk. The registry identifies cardiovascular, heart-failure, and renal conditions within the ClinicalTrials.gov record and records a primary prevention purpose.

10584
Enrolled
Randomized trial
2
Treatment arms
Sotagliflozin vs placebo
0.74
Primary HR
95% CI 0.63–0.88
<0.001
Primary P-value
Two-sided analysis
FeatureSCORED
Trial nameSCORED
NCT identifierNCT03315143
PhasePhase 3
Therapeutic areaCardiovascular
ConditionsHeart Failure; Type 2 Diabetes Mellitus; Chronic Kidney Diseases
AllocationRandomized
Design modelParallel
MaskingQuadruple
Primary purposePrevention
Enrollment10584
InterventionsSotagliflozin; Placebo
Lead sponsorLexicon Pharmaceuticals
Sponsor typeIndustry
Study statusTerminated
Trial datesStart: 2017-12-19; Primary completion: 2020-07-08

2. Clinical Question

The central statistical question was whether participants randomized to sotagliflozin experienced a different rate of the registered composite cardiovascular and heart-failure endpoint than participants randomized to placebo over a time frame of up to 30 months.

Population

Participants with type 2 diabetes mellitus and moderate renal impairment who were at cardiovascular risk. the ClinicalTrials.gov record lists heart failure, type 2 diabetes mellitus, and chronic kidney diseases among the conditions.

Intervention

Sotagliflozin.

Comparator

Placebo.

Primary question

Does randomized assignment to sotagliflozin change the occurrence rate of the combined cardiovascular death, heart-failure hospitalization, and urgent heart-failure-visit endpoint?

3. Trial Design

01
Randomize10584 participants
02
AllocateSotagliflozin or placebo
03
FollowUp to 30 months
04
AnalyzeITT population
05
CompareCox hazard ratios
ARM 1 · SOTAGLIFLOZIN

Sotagliflozin

  • Randomized treatment assignment
  • Intervention classified as a drug
  • Primary comparison with placebo
ARM 2 · PLACEBO

Placebo

  • Randomized control assignment
  • Intervention classified as a drug
  • Comparator for the sotagliflozin group

The registry describes the allocation as randomized, the design model as parallel, and masking as quadruple. These features matter statistically because randomization establishes the basis for a causal treatment comparison, while masking can reduce the influence of treatment knowledge on participant care, outcome assessment, and other trial processes.

Termination: The registry limitation states that the study was terminated prematurely due to a business decision. This is important context when interpreting a trial whose planned follow-up and information accumulation may have been affected by early termination.

4. Endpoints

EndpointRegistry definitionTime frameAnalysis type
Primary Number of Total Occurrences of Cardiovascular (CV) Death, Hospitalizations for Heart Failure (HHF) and Urgent Visits for Heart Failure (HF) Up to 30 months Count / rate; reported as events per 100 person-years
Secondary Total Number of Occurrences of HHF and Urgent HF Visits Up to 30 months Time-to-event; hazard ratio
Secondary Number of Deaths From Cardiovascular Causes Up to 30 months Time-to-event; hazard ratio
Secondary Total Number of Occurrences of CV Death, HHF, Non-fatal Myocardial Infarction and Non-fatal Stroke Up to 30 months Time-to-event; hazard ratio
Secondary Total Number of Occurrences of HHF, Urgent HF Visit, CV Death, and HF While Hospitalized Up to 30 months Time-to-event; hazard ratio
Secondary Number of Occurrences After Randomization of the Composite of Sustained ≥50% Decrease in Estimated Glomerular Filtration Rate (eGFR) From Baseline (for ≥30 Days), Chronic Dialysis, Renal Transplant, or Sustained eGFR <15 mL/Min/1.73 m2 (for ≥30 Days Up to 30 months Time-to-event; hazard ratio
Secondary Number of Deaths From Any Cause Up to 30 months Time-to-event; hazard ratio
Secondary Total Number of Occurrences of CV Death, Non-fatal Myocardial Infarction and Non-fatal Stroke Up to 30 months Time-to-event; hazard ratio

The primary endpoint is unusual in one important respect: it is defined as the total number of occurrences, including first and potentially subsequent events, rather than simply the time to a participant's first event. The registry states that events occurring during the study were calculated as the total number of events per 100 person-years of follow-up.

5. Statistical Methodology

Intention-to-treat analysis

The statistical analyses used the ITT population, defined in the registry as all randomized participants analyzed according to the treatment group allocated by randomization. This preserves the treatment assignment established by randomization rather than redefining treatment groups according to treatment received.

Core analysis population
ITT = all randomized participants, analyzed by randomized treatment group

The ITT principle makes the randomized groups the basis of the efficacy comparison and helps preserve the comparability created by randomization.

Marginal Cox proportional-hazards model

The registry reports a Cox proportional-hazards model for the primary endpoint and the secondary time-to-event analyses. For the primary endpoint, the analysis notes specify a marginal Cox proportional hazard model stratified by region and ejection fraction, with non-cardiovascular death treated as a competing event.

The same model description is provided for the secondary heart-failure and cardiovascular endpoints, while the analysis of all-cause death is described as stratified by region and ejection fraction without the competing non-cardiovascular-death specification.

Conceptual Cox model
h(t | X) = h0(t) exp(βX)

The hazard ratio is obtained from the treatment coefficient: HR = exp(β). An HR below 1 represents a lower estimated instantaneous event rate for sotagliflozin relative to placebo under the fitted model.

Stratification

The reported primary analysis was stratified by region and ejection fraction. Stratification allows the baseline hazard to differ across these strata while estimating the treatment effect across the randomized comparison.

Competing events

For the primary endpoint and several secondary cardiovascular endpoints, the registry specifies that non-cardiovascular death was treated as a competing event. This matters because a participant who dies from a non-cardiovascular cause can no longer subsequently experience a cardiovascular event included in the endpoint.

Superiority framework

The registered hypothesis type is superiority. The reported primary analysis therefore estimates whether the event experience differs between randomized treatment groups rather than testing whether sotagliflozin stays within a prespecified non-inferiority margin.

6. Primary Endpoint Result

The primary endpoint was the number of total occurrences of cardiovascular death, hospitalizations for heart failure, and urgent visits for heart failure, assessed up to 30 months. The reported analysis compared sotagliflozin with placebo in the ITT population using a marginal Cox proportional-hazards model stratified by region and ejection fraction.

Primary hazard ratio

0.74

95% CI: 0.63–0.88   ·   P < 0.001

Effect measure: Hazard Ratio  ·  Two-sided 95% confidence interval  ·  Superiority analysis

Primary endpointSotagliflozin vs placebo
EndpointNumber of Total Occurrences of Cardiovascular (CV) Death, Hospitalizations for Heart Failure (HHF) and Urgent Visits for Heart Failure (HF)
Time frameUp to 30 months
Analysis populationITT population
ModelMarginal Cox proportional hazard model stratified by region and ejection fraction
Competing eventNon-cardiovascular death
Hazard ratio0.74
95% CI0.63–0.88
P-value<0.001
Outcome unitEvents per 100 person-years
Clinical Biostats interpretation

What the estimate means: An HR of 0.74 means that, under the fitted Cox model, the estimated instantaneous rate associated with the primary endpoint was about 74% of the corresponding rate in the placebo group. Equivalently, 1 − 0.74 = 0.26, so the model-based relative hazard is approximately 26% lower with sotagliflozin.

What it does not mean: The HR is not a statement that 26% of participants avoided an event, nor does it mean that each participant had exactly a 26% reduction in risk. It is a relative time-to-event measure derived from a statistical model.

What the confidence interval says: The two-sided 95% CI of 0.63–0.88 describes statistical uncertainty around the estimated HR under the model and sampling framework. It is not a range containing 95% of individual treatment effects or individual patient outcomes.

Why the P-value is different: The P < 0.001 result addresses evidence against the null hypothesis specified for the superiority comparison. It does not quantify the size or clinical importance of the treatment effect. The HR and its confidence interval are needed to describe magnitude and precision.

Important modeling caution: Because the estimate comes from a Cox proportional-hazards model, interpretation of a single HR depends on the model's proportional-hazards structure. The ClinicalTrials.gov record does not provide a time-varying hazard analysis, so the HR should be understood as the reported model-based summary rather than as a guarantee that the relative hazard was identical at every time point.

7. Secondary Endpoint Results

The registry contains seven posted secondary statistical analyses. Each uses the ITT population and compares sotagliflozin with placebo. The reported effect measure is a hazard ratio with a two-sided 95% confidence interval, generally using a marginal Cox proportional-hazards model stratified by region and ejection fraction.

Secondary endpointHR95% CIP-value
Total Number of Occurrences of HHF and Urgent HF Visits 0.67 0.55–0.82 <0.001
Number of Deaths From Cardiovascular Causes 0.90 0.73–1.12 =0.35
Total Number of Occurrences of CV Death, HHF, Non-fatal Myocardial Infarction and Non-fatal Stroke 0.72 0.63–0.83 Not reported in the ClinicalTrials.gov record
Total Number of Occurrences of HHF, Urgent HF Visit, CV Death, and HF While Hospitalized 0.76 0.65–0.89 Not reported in the ClinicalTrials.gov record
Composite renal endpoint 0.71 0.46–1.08 Not reported in the ClinicalTrials.gov record
Number of Deaths From Any Cause 0.99 0.83–1.18 Not reported in the ClinicalTrials.gov record
Total Number of Occurrences of CV Death, Non-fatal Myocardial Infarction and Non-fatal Stroke 0.77 0.65–0.91 Not reported in the ClinicalTrials.gov record

The ClinicalTrials.gov record provides formal P-values for the first three posted analyses listed in the source data, but do not provide P-values for the remaining four secondary analyses. The absence of a registry-reported P-value should not be converted into a significance claim or a nonsignificance claim.

Heart-failure events

HHF and urgent HF visits

0.67

95% CI: 0.55–0.82   ·   P < 0.001

The HR of 0.67 corresponds to an estimated instantaneous event rate approximately 67% of that in the placebo group under the reported Cox model, or approximately a 33% lower estimated hazard.

Cardiovascular death

Deaths from cardiovascular causes

0.90

95% CI: 0.73–1.12   ·   P = 0.35

The point estimate is below 1, but the 95% CI extends across 1.00. The registry-reported P-value is 0.35. This illustrates why the point estimate, confidence interval, and P-value should be considered together rather than interpreting an HR below 1 as sufficient evidence of a treatment difference.

8. Additional Cardiovascular and Renal Analyses

CV death + MI + stroke

The total number of occurrences of cardiovascular death, non-fatal myocardial infarction, and non-fatal stroke had an HR of 0.77 with a 95% CI of 0.65–0.91.

HF composite

The total number of occurrences of HHF, urgent HF visit, CV death, and HF while hospitalized had an HR of 0.76 with a 95% CI of 0.65–0.89.

Renal composite

The reported renal composite had an HR of 0.71 with a 95% CI of 0.46–1.08.

All-cause death

Deaths from any cause had an HR of 0.99 with a 95% CI of 0.83–1.18.

EndpointEstimatePrecisionStatistical reading
CV death + non-fatal MI + non-fatal stroke HR 0.77 95% CI 0.65–0.91 Estimated hazard below placebo; interval does not cross 1.00
HHF + urgent HF visit + CV death + HF while hospitalized HR 0.76 95% CI 0.65–0.89 Estimated hazard below placebo; interval does not cross 1.00
Renal composite HR 0.71 95% CI 0.46–1.08 Point estimate below placebo, but interval includes 1.00
All-cause death HR 0.99 95% CI 0.83–1.18 Point estimate close to 1.00; interval includes 1.00
CV death + HHF + urgent HF visit HR 0.74 95% CI 0.63–0.88 Primary endpoint; interval below 1.00

The pattern of estimates should not be treated as a set of interchangeable outcomes. Each composite has a different clinical definition and event structure. In particular, the primary endpoint permits first and potentially subsequent occurrences, while the statistical interpretation of the reported HR remains tied to the specified model and endpoint.

9. Safety Results

The ClinicalTrials.gov record provides serious adverse-event counts by randomized treatment arm. These figures are reported as affected participants divided by the participants at risk.

Safety measureSotagliflozinPlacebo
Serious adverse events1236 / 52911331 / 5286
Serious adverse events: affected participants
Sotagliflozin
1236
Placebo
1331

The affected/at-risk counts should not be interpreted as a comparative risk ratio or hazard ratio. The ClinicalTrials.gov record gives the number affected and the corresponding number at risk, but do not provide a formal statistical comparison for serious adverse events. The safety analysis therefore remains descriptive on this page.

Safety interpretation: Serious adverse events are a separate dimension of the trial evidence from the primary cardiovascular efficacy endpoint. A participant can contribute to an adverse-event summary and separately contribute to an efficacy endpoint, and the two analyses answer different questions.

10. Statistical Methods Explained

Why was a Cox proportional-hazards model used?

The registered and reported endpoints are time-to-event outcomes. A Cox model uses both whether an event occurred and when it occurred, while allowing participants who have not experienced the event during observation to contribute information through their follow-up time. The reported effect measure is a hazard ratio.

What does an HR of 0.74 mean?

An HR of 0.74 means the estimated instantaneous event rate under the fitted model is 74% of the comparator rate. The complementary calculation, 1 − 0.74, gives approximately 26%, so the estimate corresponds to an approximately 26% lower relative hazard. It does not mean a 26-percentage-point reduction in event probability.

Why was the analysis stratified by region and ejection fraction?

Stratification permits the baseline hazard to differ across region and ejection-fraction strata while estimating a common treatment effect across those strata. This can account for systematic differences in baseline event dynamics associated with the stratification variables.

Why is an ITT analysis important?

ITT analysis retains participants in the groups to which they were randomized. This protects the principal comparison created by randomization and avoids redefining the efficacy population based on treatment received or treatment adherence.

Why was non-cardiovascular death treated as a competing event?

For several cardiovascular endpoints, a participant who dies from a non-cardiovascular cause can no longer experience a later cardiovascular endpoint event. Treating that death as a competing event recognizes that this alternative event changes the participant's opportunity to experience the endpoint of interest.

Why should the confidence interval be reported with the HR?

The HR is only one estimate. The 95% CI provides information about its statistical precision. For example, the primary HR of 0.74 has a 95% CI of 0.63–0.88, whereas the renal composite HR of 0.71 has a substantially wider interval of 0.46–1.08. The point estimates alone do not communicate that difference in precision.

Why should P-values not be treated as effect sizes?

A P-value addresses evidence against a null hypothesis under the specified statistical model. It does not tell the reader how large the treatment effect is. Effect magnitude is communicated by measures such as the HR, while the confidence interval communicates uncertainty around that estimate.

11. Understanding the Primary Endpoint

The primary endpoint combines three clinically distinct event types: cardiovascular death, hospitalization for heart failure, and urgent visits for heart failure. Importantly, the registry defines it as the total number of occurrences, including first and potentially subsequent occurrences.

Endpoint structure
Primary endpoint = CV death + HHF + urgent HF visits

Events occurring during the study were calculated as total events per 100 person-years of follow-up.

This construction means that the endpoint is not simply asking, "How many participants had at least one event?" It incorporates the total number of occurrences. That distinction is statistically important because recurrent events can carry additional information, but they also make the event process more complex than a conventional first-event endpoint.

The registry-reported analysis nevertheless reports a Cox proportional-hazards model and a hazard ratio. The correct interpretation is therefore tied to the registry's specified model and endpoint definition rather than to a simple participant-level proportion.

12. Reading the Confidence Intervals

EndpointHR95% CIWidth of interval
Primary composite0.740.63–0.880.25
HHF + urgent HF visits0.670.55–0.820.27
CV death0.900.73–1.120.39
CV death + MI + stroke0.770.65–0.910.27
HF composite0.760.65–0.890.25
Renal composite0.710.46–1.080.63
All-cause death0.990.83–1.180.36
CV death + MI + stroke0.770.65–0.910.27

The confidence intervals illustrate a central principle of trial interpretation: a point estimate should never be read independently of its precision. The renal composite has a point estimate below 1, but its interval spans 1.00. The cardiovascular-death estimate is also below 1, with an interval that spans 1.00. These are different from the primary endpoint, whose reported interval is entirely below 1.00.

Statistical interpretation

A confidence interval that includes 1.00 for a hazard ratio means that the registry-reported interval is compatible with both a lower and a higher hazard relative to placebo, including the null value of 1.00. It does not prove that there is no treatment effect. Conversely, an interval below 1.00 does not tell us that every participant benefited or that the absolute effect was large.

13. Hazard Ratios and Absolute Risk

The SCORED ClinicalTrials.gov record emphasize hazard ratios and event rates per 100 person-years. Those measures answer questions about relative event dynamics and event occurrence over follow-up, but they do not by themselves provide an absolute probability that a particular participant will experience an event.

Relative measure

The hazard ratio compares estimated instantaneous event rates between the randomized groups under the Cox model.

Rate measure

The primary endpoint is reported in events per 100 person-years, reflecting the total number of events relative to accumulated follow-up.

Not an absolute risk

An HR cannot be converted directly into an absolute risk difference without additional information about baseline event experience and follow-up.

Not a patient-level guarantee

A population-level HR does not imply that each individual experiences the same proportional change in event risk.

This distinction is particularly important for composite endpoints. An overall HR summarizes the treatment comparison across the endpoint as defined, but it does not establish that each component contributed equally to the observed result.

14. Multiplicity and Multiple Secondary Endpoints

The ClinicalTrials.gov record identifies one primary endpoint and seven secondary statistical analyses. The hypothesis type is listed as superiority. The ClinicalTrials.gov record does not describe an alpha-allocation strategy, hierarchical testing procedure, multiplicity adjustment, or formal familywise error-control scheme for the secondary endpoints.

Analysis levelNumber in the ClinicalTrials.gov recordWhat can be concluded from the ClinicalTrials.gov record
Primary endpoint1Formal Cox analysis reported with HR, 95% CI, and P-value
Secondary endpoints7Formal HR and 95% CI reported for each; P-values posted on ClinicalTrials.gov for some analyses only
Multiplicity procedureNot reportedNo specific adjustment should be inferred
Hierarchical testingNot reportedNo hierarchy should be inferred

Consequently, the secondary results should be read as individual reported analyses rather than automatically treating every endpoint as an independently confirmatory hypothesis test. In particular, a P-value should not be interpreted as if it had a known familywise-error role when that role is not provided in the ClinicalTrials.gov record.

15. Missing Data, Censoring, and Follow-Up

The time-to-event framework naturally accommodates participants whose endpoint status is not observed for the entire study period through censoring. However, the ClinicalTrials.gov record does not specify a detailed missing-data or imputation strategy for the time-to-event analyses.

Censoring

Participants without an observed endpoint event by the end of available follow-up contribute information through their observed time under the survival-analysis framework.

Missing-data method

No specific imputation method is provided in the ClinicalTrials.gov record for these analyses.

Follow-up horizon

The registered primary endpoint and all registry-reported statistical analyses use a time frame of up to 30 months.

Interpretive caution

The study was terminated prematurely due to a business decision, which is a registry-documented limitation when considering the planned versus observed information available to the trial.

16. Stratified Analysis in SCORED

Stratification appears directly in the registry-reported statistical analysis descriptions. The primary endpoint uses a marginal Cox proportional hazard model stratified by region and ejection fraction, with non-cardiovascular death treated as a competing event. The same stratification is reported for the secondary analyses involving cardiovascular and heart-failure outcomes.

Stratification structure
Treatment effect | region, ejection fraction

The purpose is to estimate the treatment effect while allowing baseline hazard differences across the specified strata.

Stratification should not be confused with subgroup analysis. A stratification factor can be incorporated into the model to account for baseline hazard differences without asking whether the treatment effect itself differs between strata. Demonstrating treatment-effect heterogeneity requires a different analysis, typically involving an interaction term or an equivalent formal comparison.

17. Competing Risks: Why Non-CV Death Matters

The primary analysis notes explicitly state that non-cardiovascular death was treated as a competing event. This is a meaningful feature of the statistical analysis because the event process for cardiovascular outcomes is interrupted when a participant experiences an alternative fatal event.

Interpretive distinction: A competing event is not simply another form of censoring from a conceptual standpoint. Once a participant dies from a non-cardiovascular cause, that participant cannot subsequently experience a cardiovascular death, heart-failure hospitalization, or urgent heart-failure visit.

The ClinicalTrials.gov record does not describe a separate cumulative-incidence analysis or a Fine-Gray subdistribution-hazard model. Therefore, this page does not substitute another competing-risk method for the reported Cox approach.

18. Proportional-Hazards Assumption

The Cox proportional-hazards model assumes that the relative hazard represented by the treatment coefficient is appropriately summarized by a hazard ratio over time, subject to the model structure. This assumption does not mean that the absolute hazard is constant; rather, it concerns the relationship between the treatment-group hazards.

Why this matters

A hazard ratio of 0.74 should not automatically be read as meaning that the sotagliflozin group had exactly 26% lower hazard at every individual time point. If the relative treatment effect changes materially over time, a single HR can become a less complete description of the event histories.

The registry-reported SCORED data report the Cox model and its HR but do not provide a formal proportional-hazards diagnostic. Therefore, the reported HR should be interpreted as the prespecified model-based summary rather than as evidence that proportional hazards were empirically demonstrated.

19. Trial Timeline

2017-12-19

Study start

The SCORED trial began on December 19, 2017.

2020-07-08

Primary completion

The registry lists July 8, 2020 as the primary completion date.

Registry status

Terminated

The study is listed as terminated. The registry caveat states that it was terminated prematurely due to a business decision.

20. What the Primary Hazard Ratio Does — and Does Not — Mean

What it means

The primary HR of 0.74 indicates that the estimated instantaneous rate for the defined primary endpoint was approximately 26% lower under the sotagliflozin treatment assignment than under placebo, according to the reported Cox model.

What it does not mean

It does not mean that 26% of participants were protected from an event, that the absolute probability of the endpoint fell by 26 percentage points, or that each participant experienced an identical 26% reduction in risk.

Why the confidence interval matters

The 95% CI of 0.63–0.88 indicates the statistical precision of the primary HR estimate under the reported model. It is narrower than the registry-reported CI for the renal composite, which is 0.46–1.08, illustrating how point estimates and precision must be considered together.

Why the P-value does not measure effect size

The primary P < 0.001 indicates strong statistical evidence against the relevant null hypothesis under the specified analysis. It does not tell us whether the HR is 0.74 because of a large effect, a small effect with substantial information, or some combination of effect magnitude and precision. The HR and CI supply that complementary information.

21. Limitations

22. Why This Trial Matters Statistically

SCORED is a useful statistical teaching case because it combines randomized treatment assignment with a complex recurrent-event-oriented primary endpoint, time-to-event modeling, stratification, competing-event handling, multiple secondary endpoints, and an ITT analysis population.

ConceptHow it appears in SCORED
RandomizationParticipants were randomized to sotagliflozin or placebo.
Parallel designThe trial used a parallel-group design with two treatment arms.
Quadruple maskingThe registry classifies the trial as quadruple masked.
ITT analysisAll randomized participants were analyzed according to randomized treatment group.
Time-to-event analysisThe primary and secondary efficacy analyses use time-to-event methods.
Cox proportional-hazards modelThe reported statistical method for the primary and secondary analyses.
Hazard ratioThe treatment effect measure for the reported Cox analyses.
StratificationPrimary and several secondary models were stratified by region and ejection fraction.
Competing eventNon-cardiovascular death was treated as a competing event for the primary analysis.
Composite endpointThe primary endpoint combines CV death, HHF, and urgent HF visits.
Recurrent occurrencesThe primary endpoint counts first and potentially subsequent occurrences.
SuperiorityThe registry-reported hypothesis type is superiority.
Multiple endpointsSeven secondary statistical analyses accompany the primary analysis.
Premature terminationThe study was terminated prematurely due to a business decision.

23. Statistical Methods in Context

The statistical story of SCORED can be understood as a sequence of related decisions rather than as a single P-value.

1. Randomization

Random allocation establishes the primary basis for comparing treatment groups and supports the ITT analysis.

2. Endpoint construction

The primary endpoint combines several cardiovascular and heart-failure events and permits first and potentially subsequent occurrences.

3. Time-to-event modeling

The Cox framework incorporates event timing and follow-up rather than treating every observation as an ordinary binary outcome.

4. Stratification

Region and ejection fraction are incorporated as stratification variables in the primary model.

5. Competing event

Non-cardiovascular death is recognized as a competing event for the primary analysis.

6. Effect interpretation

The HR, confidence interval, and P-value each provide different pieces of information and should be interpreted together.

24. Related Tutorials

Learn more about the methods used in this trial:

25. Related Statistical Calculators

26. Sources

Continue through the Clinical Biostats knowledge graph

Use the trial's endpoints and statistical methods as a starting point for deeper study of survival analysis, hazard ratios, confidence intervals, randomization, and clinical-trial methodology.

27. Record Summary

SCORED provides a detailed example of randomized cardiovascular-trial analysis centered on a complex composite endpoint. The ClinicalTrials.gov record reports a primary hazard ratio of 0.74 with a two-sided 95% CI of 0.63–0.88 and P < 0.001, using an ITT population and a marginal Cox proportional hazard model stratified by region and ejection fraction, with non-cardiovascular death treated as a competing event. Secondary analyses extend the statistical framework to heart-failure, cardiovascular, renal, and all-cause outcomes.

The most informative way to read these results is to combine the effect estimate, confidence interval, P-value, endpoint definition, analysis population, and model assumptions. The registry's documentation of premature termination due to a business decision is also an important part of the statistical context.

Clinical Biostats methodology: A trial-results page should not merely repeat the headline estimate. The goal is to reconstruct the statistical story of the trial while clearly distinguishing the reported analysis from the interpretation required to understand what the estimate, confidence interval, P-value, endpoint construction, competing-event treatment, and modeling assumptions actually mean.