Risk Of Severe Hepatic Injury In Patients With Invasive Candidiasis Treated With Echinocandins: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Risk Of Severe Hepatic Injury In Patients With Invasive Candidiasis Treated With Echinocandins (NCT01213823), a registered phase not stated study. The primary outcome is Number of Any Severe Hepatic Injury Cases and Matched Controls.

2. Clinical Question

The study’s stated aim was to estimate the relative risk of severe hepatic injury among hospitalized patients with invasive candidiasis and candidemia who received anidulafungin, compared with patients who received caspofungin and/or micafungin. The record describes a case-control approach: patients with severe hepatic injury were identified as cases, and controls were selected from the same acute-care inpatient cohort.

The primary outcome title refers to the number of cases and matched controls, but outcome counts and a comparative risk estimate are not reported in the registry record. The record therefore states the question and the intended case-control comparison, but does not provide an answer to the relative-risk question.

3. Trial Design

This is a non-randomized study record, not a like-for-like randomized treatment comparison. The registry’s allocation information is not stated. Its brief summary describes an exposure comparison among hospitalized patients who received echinocandins, while the outcome description identifies cases and controls within an acute-care inpatient cohort.

Potential cases were patients with a severe hepatic injury diagnosis identified using ICD-9 codes associated with the case definition. Their case status was to be validated by a Consultant Gastroenterologist blinded to study drug exposure. The description is truncated after that point, so further details of validation are not reported in the registry record. Controls were described as patients without a diagnosis of severe hepatic injury, selected at random from the same acute-care inpatient cohort as cases. The record gives N=467 for this control description. Although the outcome title calls the controls matched, the matching procedure and variables are not reported in the registry record.

The enrollment is recorded as 536. The study start is listed as 2008-09, and primary completion as 2011-05. Its status is TERMINATED. A reason for termination and details about recruitment or follow-up are not reported in the registry record. The outcome timeframe is 01 June 2006 to 30 June 2008 (up to 25 Months), which is the period specified for the severe hepatic injury outcome.

The listed interventions are both Does not apply, and the record does not describe treatment assignment, dose, duration, or treatment switching. These omissions matter when interpreting a comparison based on observed drug exposure: patient characteristics and clinical circumstances may differ between treatment groups. The registry record does not describe adjustments intended to address such differences.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Overall Study

Analysis population for the primary outcome, as defined in the registry: Acute-care participants, with at least 1 dose of echinocandin antifungal therapy and a primary or secondary International Classification of Diseases 9 (ICD-9).

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

6. Primary Endpoint

The registry lists the primary outcome as Number of Any Severe Hepatic Injury Cases and Matched Controls, measured over 01 June 2006 to 30 June 2008 (up to 25 Months).

The primary endpoint is titled Number of Any Severe Hepatic Injury Cases and Matched Controls, with the unit given as participants and the parameter described as a number. It counts participants classified as severe hepatic injury cases and controls without a severe hepatic injury diagnosis. The endpoint’s timeframe is 01 June 2006 to 30 June 2008 (up to 25 Months).

The severe hepatic injury definition includes acute liver failure, Hy’s Law Criteria, elevated ALT thresholds, and cases classified by a clinician. Specifically, the description includes acute liver failure associated with encephalopathy and/or coagulopathy in the absence of underlying liver disease; Hy’s Law Criteria with ALT levels >3 times ULN and direct bilirubin >2 times ULN without alkaline phosphatase elevation; ALT levels >= 10 times ULN; ALT levels >3 times ULN and <10 times ULN; or clinician classification. Disease Related Group severity-of-illness coding was reported for cases and matched controls.

A case-control endpoint can help characterize how often an outcome was identified in a defined cohort and support a comparison of prior drug exposures. The broad case definition may capture different forms or degrees of hepatic injury; how those categories were distributed is not reported in the registry record. Neither case nor control counts are reported as outcome results, so the endpoint’s observed values cannot be assessed.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The record describes an observational case-control comparison rather than randomized allocation. In general, a case-control design identifies people with an outcome and compares their prior exposures with those of people without the outcome. This can be useful when the outcome is uncommon or when exposure information can be obtained from an existing clinical cohort. It does not, by itself, ensure that the groups are comparable in their underlying risk.

What does a relative risk comparison require?

A relative risk compares outcome risks across exposure groups. To interpret one, readers need the number of outcomes and the relevant denominators for each group, along with a clear account of how exposure groups and outcomes were defined. Here, the stated comparison is anidulafungin versus caspofungin and/or micafungin, but the registry record reports neither treatment-specific case counts nor denominators nor a relative-risk estimate. The endpoint is described as a number of cases and controls. A count alone is not a treatment-specific risk unless the corresponding group size and analytic method are also known.

In case-control research, groups may be selected according to outcome status, so the proportions sampled as cases and controls do not necessarily represent outcome risks in the source population. Depending on design and analysis, the exposure contrast may instead be expressed using another measure of association. The registry record does not report the analysis method, so it is not possible to determine what measure was calculated or how it was interpreted.

Why does non-randomized allocation matter?

Randomization is a general method for balancing measured and unmeasured characteristics between treatment groups, on average. Because allocation is not stated and the comparison is identified as non-randomized, differences between the drug groups may influence an observed association. Confounding occurs when a factor related to treatment choice is also related to the outcome. Statistical adjustment can address measured confounders when suitable variables and methods are available, but it cannot automatically remove unmeasured or poorly measured differences. The registry record does not report whether adjustment was performed or which covariates were considered.

Selection also matters. The cases and controls came from an acute-care inpatient cohort, and controls were described as selected at random from that cohort. The record does not explain the sampling fraction or matching procedure. If selection into the study depends on characteristics related to exposure or outcome, the observed comparison may not represent the intended source population in a straightforward way. These are general design considerations, not claims that a specific bias occurred.

How should the outcome definition be read?

The endpoint combines multiple forms of severe hepatic injury: acute liver failure, Hy’s Law Criteria, ALT thresholds, and clinician classification. A composite or broad definition can increase coverage of clinically recognized presentations, but its components may differ in severity and ascertainment. Separate counts by category would help readers understand what contributed to the outcome. Such counts are not reported in the registry record. The use of ICD-9 codes to identify potential cases, followed by described blinded specialist validation, may help with classification; the record does not give the validation findings or details of the adjudication process.

The outcome description includes ALT thresholds relative to ULN and specifies bilirubin and alkaline phosphatase criteria for Hy’s Law. These criteria help make the outcome operational, but the record does not state whether laboratory values were available consistently across the cohort or how missing measurements were handled. It also does not report the frequency of clinician-classified cases or the distribution of DRG severity coding.

What can uncertainty measures tell a reader?

In general, a confidence interval presents a range of effect estimates compatible with the data under the analysis assumptions. Its width reflects precision: a wider interval indicates less precise estimation, while a narrower interval indicates greater precision. A p-value, in general, measures how unusual data at least as incompatible with a specified null hypothesis would be under that hypothesis and the model assumptions. Neither measure alone establishes clinical importance or eliminates bias. No confidence interval or p-value is reported in the registry record, so statistical precision and the evidence against a null association cannot be assessed.

The record’s sample-size assumptions describe planning illustrations based on the observed effect, with a two-sided significance level of 0.05, equal allocation, and a normal approximation; the effect to detect is stated as a fraction of the observed effect. Such calculations are planning scenarios, not observed power and not evidence that the study had a particular probability of detecting an effect. The assumptions do not substitute for an outcome estimate, and the registry record does not report an analysis result or a sample-size justification for the original study.

What analysis details remain unknown?

No outcome analyses are listed. The statistical plan, model specification, treatment of repeated observations, handling of missing values, analysis populations, and any planned sensitivity analyses are not reported in the registry record. Those details would be needed to evaluate whether the analysis matched the design and endpoint. The description of a blinded validator is relevant to outcome assessment, but it does not establish that the overall analysis was blinded or that other sources of bias were controlled.

10. Multiplicity, Missing Data and Analysis Populations

The registry record describes one primary outcome measure, but does not report an analysis plan or any outcome analyses. It does not state whether additional comparisons, subgroup analyses, or secondary outcomes were planned. Therefore, multiplicity—the issue of interpreting results when several hypotheses or comparisons are tested—cannot be evaluated from the record. The number of analyses listed is 0; this indicates that no analyses are listed in the record, not that no analyses were conducted outside it.

The severe hepatic injury definition includes multiple qualifying forms, including acute liver failure, Hy’s Law Criteria, ALT thresholds, and clinician classification. These are components of a broad outcome definition, not necessarily separate statistical tests. Component-specific results and any plan for interpreting them are not reported in the registry record. It is not possible to determine whether the components were analyzed separately or how any such analyses would have been handled.

Missing-data procedures are not reported. The record does not describe the extent of missing clinical or laboratory information, how incomplete records were treated, or whether imputation or other methods were used. It also does not describe exclusions, censoring, or how the analysis population was defined beyond the acute-care cohort and the stated echinocandin and ICD-9 eligibility information.

The record does not identify an intention-to-treat, per-protocol, or other named analysis population. In this non-randomized exposure comparison, the exact rules for assigning patients to drug groups would be important, but are not reported. Without the analysis population and missing-data details, the relationship between the enrolled group, the cases and controls, and any estimate cannot be assessed.

11. What the Result Does, and Does Not, Mean

The study asks a clinically focused comparative safety question about severe hepatic injury in hospitalized patients with invasive candidiasis and candidemia who received echinocandins. It defines a case-control approach, names anidulafungin and caspofungin and/or micafungin as the exposure groups of interest, and describes a broad severe hepatic injury definition. Those details clarify the intended research question.

However, the registry record does not report case counts, treatment-specific denominators, an effect estimate, or uncertainty measures. It therefore cannot establish whether the observed data favored either exposure group, how large any association might have been, or how precisely it was estimated. The recorded enrollment and the control count do not fill those gaps.

The absence of a posted result must not be interpreted as evidence of equal risk or as evidence that severe hepatic injury did not occur. Nor does it show that an association was present. A result requires reported outcome data and an analysis that connects those data to the intended comparison.

Even with a reported association, this non-randomized design would require careful consideration of confounding, outcome ascertainment, selection, and the definition of the analysis population. The record describes blinded specialist validation of potential cases, but does not give its findings. It also does not report adjustment methods or matching details. These limitations would affect the strength of any causal interpretation.

The status TERMINATED gives the administrative status recorded for the study, not an explanation for its termination and not a clinical conclusion. The reason is not reported in the registry record. The available information therefore supports understanding what the study set out to examine and how cases and controls were described, but does not support a conclusion about which antifungal exposure was associated with greater or lesser risk.

12. Why This Trial Matters Statistically

This study matters statistically because comparative safety questions often concern outcomes that are identified in routine clinical records rather than assigned as study treatments. The record combines a clinically defined injury outcome with an observational comparison of drug exposures. That structure makes clear distinctions between outcome ascertainment, exposure classification, group selection, and statistical adjustment especially important.

Its case-control description illustrates why counts and denominators must be reported together with an effect measure. A case count and a control count do not alone reveal treatment-specific risk, and a sampled case-control dataset may need an analysis suited to its selection process. The registry record does not give the details needed to evaluate that analysis or its result.

The outcome definition also raises a useful statistical issue: a broad composite may combine components with different clinical features and measurement requirements. Reporting component-level results and the rules used to classify cases can help readers judge what the overall endpoint represents. Those component results are not reported in the registry record.

Finally, non-randomized exposure comparisons can be informative while remaining vulnerable to confounding and selection effects. A transparent account of covariates, matching, missing data, analysis populations, and uncertainty would help readers distinguish what the data estimate from what they cannot establish. Here, those analysis details and the outcome findings are not reported. The record is therefore informative about the question and design, but not about the comparative result.

13. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The most consequential limitation for interpreting this specific record is that outcome results are absent. No case count, treatment-specific count, comparative estimate, confidence interval, or p-value is reported. As a result, the planned comparison cannot be evaluated for direction, magnitude, or precision.

The design is non-randomized, and allocation is not stated. The record does not report potential confounders, adjustment methods, matching factors, or how exposure groups were assigned. Although the outcome title refers to matched controls, the description says controls were selected at random from the same cohort; the relationship between these descriptions is not explained.

Potential cases were identified through ICD-9 codes and described as subject to blinded validation by a Consultant Gastroenterologist. The validation description is truncated, and no validation results are reported. The record also does not provide component-specific counts for its broad severe hepatic injury definition, or explain the completeness of laboratory measurements used for threshold-based criteria.

The eligibility description refers to acute-care participants who received at least 1 dose of echinocandin antifungal therapy and had a primary or secondary ICD-9 diagnosis. Further eligibility rules, treatment exposure details, and analytic population definitions are not reported in the registry record. Missing-data handling and the statistical analysis plan are also not reported.

The record lists enrollment as 536 and controls as N=467, but does not report the case count or treatment-specific denominators. The study is marked TERMINATED, while the reason for termination is not reported. These omissions prevent assessment of the completeness of the study and the representativeness of the analyzed cohort.

14. Related Tutorials

15. Related Statistical Calculators

16. Sources

• ClinicalTrials.gov: Risk Of Severe Hepatic Injury In Patients With Invasive Candidiasis Treated With Echinocandins — NCT01213823.
• Registry record last updated 2016-03-30; status last verified 2016-03.
• Results first posted 2016-02-04.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

17. Record Summary

The record describes a terminated, non-randomized study sponsored by Pfizer of severe hepatic injury in hospitalized patients with invasive candidiasis and candidemia receiving echinocandin therapy. Its stated comparison is anidulafungin versus caspofungin and/or micafungin. Potential cases were identified using ICD-9 codes and described as undergoing blinded specialist validation; controls without a qualifying diagnosis were selected at random from the same acute-care inpatient cohort. The control description gives N=467.

Enrollment is recorded as 536. The study start is 2008-09, primary completion is 2011-05, and the outcome timeframe is 01 June 2006 to 30 June 2008 (up to 25 Months). The primary endpoint counts severe hepatic injury cases and matched controls under a definition covering several clinical and laboratory criteria. Outcome counts, comparative estimates, uncertainty measures, analysis methods, missing-data procedures, and the reason for termination are not reported in the registry record.