Type-II generalized family-wise error rate formulas with application to sample size determination

Interleukin-receptor antagonist and tumour necrosis factor inhibitors for the primary and secondary prevention of atherosclerotic cardiovascular diseases

BACKGROUND

Atherosclerotic cardiovascular disease (ACVD) is worsened by chronic inflammatory diseases. Interleukin receptor antagonists (IL-RAs) and tumour necrosis factor-alpha (TNF) inhibitors have been studied to see if they can prevent cardiovascular events.

OBJECTIVES

The purpose of this study was to assess the clinical benefits and harms of IL-RAs and TNF inhibitors in the primary and secondary prevention of ACVD.

SEARCH METHODS

The Cochrane Heart Specialised Register, the Cochrane Central Register of Controlled Trials (CENTRAL), Ovid MEDLINE (including In-Process & Other Non-Indexed Citations), Ovid Embase, EBSCO CINAHL plus, and clinical trial registries for ongoing and unpublished studies were searched in February 2024. The reference lists of relevant studies, reviews, meta-analyses and health technology reports were searched to identify additional studies. No limitations on language, date of publication or study type were set.

SELECTION CRITERIA

RCTs that recruited people with and without pre-existing ACVD, comparing IL-RAs or TNF inhibitors versus placebo or usual care, were selected. The primary outcomes considered were all-cause mortality, myocardial infarction, unstable angina, and adverse events.

DATA COLLECTION AND ANALYSIS

Two or more review authors, working independently at each step, selected studies, extracted data, assessed the risk of bias and used GRADE to judge the certainty of evidence.

MAIN RESULTS

We included 58 RCTs (22,053 participants; 21,308 analysed), comparing medication efficacy with placebo or usual care. Thirty-four trials focused on primary prevention and 24 on secondary prevention. The interventions included IL-1 RAs (anakinra, canakinumab), IL-6 RA (tocilizumab), TNF-inhibitors (etanercept, infliximab) compared with placebo or usual care. The certainty of evidence was low to very low due to biases and imprecision; all trials had a high risk of bias. Primary prevention: IL-1 RAs The evidence is very uncertain about the effects of the intervention on all-cause mortality(RR 0.33, 95% CI 0.01 to 7.58, 1 trial), myocardial infarction (RR 0.71, 95% CI 0.04 to 12.48, I² = 39%, 2 trials), unstable angina (RR 0.24, 95% CI 0.03 to 2.11, I² = 0%, 2 trials), stroke (RR 2.42, 95% CI 0.12 to 50.15; 1 trial), adverse events (RR 0.85, 95% CI 0.59 to 1.22, I² = 54%, 3 trials), or infection (rate ratio 0.84, 95% 0.55 to 1.29, I² = 0%, 4 trials). Evidence is very uncertain about whether anakinra and cankinumab may reduce heart failure (RR 0.21, 95% CI 0.05 to 0.94, I² = 0%, 3 trials). Peripheral vascular disease (PVD) was not reported as an outcome. IL-6 RAs The evidence is very uncertain about the effects of the intervention on all-cause mortality (RR 0.68, 95% CI 0.12 to 3.74, I² = 30%, 3 trials), myocardial infarction (RR 0.27, 95% CI 0.04 to1.68, I² = 0%, 3 trials), heart failure (RR 1.02, 95% CI 0.11 to 9.63, I² = 0%, 2 trials), PVD (RR 2.94, 95% CI 0.12 to 71.47, 1 trial), stroke (RR 0.34, 95% CI 0.01 to 8.14, 1 trial), or any infection (rate ratio 1.10, 95% CI: 0.88 to 1.37, I2 = 18%, 5 trials). Adverse events may increase (RR 1.13, 95% CI 1.04 to 1.23, I² = 33%, 5 trials). No trial assessed unstable angina. TNF inhibitors The evidence is very uncertain about the effects of the intervention on all-cause mortality (RR 1.78, 95% CI 0.63 to 4.99, I² = 10%, 3 trials), myocardial infarction (RR 2.61, 95% CI 0.11 to 62.26, 1 trial), stroke (RR 0.46, 95% CI 0.08 to 2.80, I² = 0%; 3 trials), heart failure (RR 0.85, 95% CI 0.06 to 12.76, 1 trial). Adverse events may increase (RR 1.13, 95% CI 1.01 to 1.25, I² = 51%, 13 trials). No trial assessed unstable angina or PVD. Secondary prevention: IL-1 RAs The evidence is very uncertain about the effects of the intervention on all-cause mortality (RR 0.94, 95% CI 0.84 to 1.06, I² = 0%, 8 trials), unstable angina (RR 0.88, 95% CI 0.65 to 1.19, I² = 0%, 3 trials), PVD (RR 0.85, 95% CI 0.19 to 3.73, I² = 38%, 3 trials), stroke (RR 0.94, 95% CI 0.74 to 1.2, I² = 0%; 7 trials), heart failure (RR 0.91, 95% 0.5 to 1.65, I² = 0%; 7 trials), or adverse events (RR 0.92, 95% CI 0.78 to 1.09, I² = 3%, 4 trials). There may be little to no difference between the groups in myocardial infarction (RR 0.88, 95% CI 0.0.75 to 1.04, I² = 0%, 6 trials). IL6-RAs The evidence is very uncertain about the effects of the intervention on all-cause mortality (RR 1.09, 95% CI 0.61 to 1.96, I² = 0%, 2 trials), myocardial infarction (RR 0.46, 95% CI 0.07 to 3.04, I² = 45%, 3 trials), unstable angina (RR 0.33, 95% CI 0.01 to 8.02, 1 trial), stroke (RR 1.03, 95% CI 0.07 to 16.25, 1 trial), adverse events (RR 0.89, 95% CI 0.76 to 1.05, I² = 0%, 2 trials), or any infection (rate ratio 0.66, 95% CI 0.32 to 1.36, I² = 0%, 4 trials). No trial assessed PVD or heart failure. TNF inhibitors The evidence is very uncertain about the effect of the intervention on all-cause mortality (RR 1.16, 95% CI 0.69 to 1.95, I² = 47%, 5 trials), heart failure (RR 0.92, 95% 0.75 to 1.14, I² = 0%, 4 trials), or adverse events (RR 1.15, 95% CI 0.84 to 1.56, I² = 32%, 2 trials). No trial assessed myocardial infarction, unstable angina, PVD or stroke. Adverse events may be underestimated and benefits inflated due to inadequate reporting.

AUTHORS' CONCLUSIONS

This Cochrane review assessed the benefits and harms of using interleukin-receptor antagonists and tumour necrosis factor inhibitors for primary and secondary prevention of atherosclerotic diseases compared with placebo or usual care. However, the evidence for the predetermined outcomes was deemed low or very low certainty, so there is still a need to determine whether these interventions provide clinical benefits or cause harm from this perspective. In summary, the different biases and imprecision in the included studies limit their external validity and represent a limitation to determining the effectiveness of the intervention for both primary and secondary prevention of ACVD.

Graphical approaches for the control of generalized error rates

When simultaneously testing multiple hypotheses, the usual approach in the context of confirmatory clinical trials is to control the familywise error rate (FWER), which bounds the probability of making at least one false rejection. In many trial settings, these hypotheses will additionally have a hierarchical structure that reflects the relative importance and links between different clinical objectives. The graphical approach of Bretz et al (2009) is a flexible and easily communicable way of controlling the FWER while respecting complex trial objectives and multiple structured hypotheses. However, the FWER can be a very stringent criterion that leads to procedures with low power, and may not be appropriate in exploratory trial settings. This motivates controlling generalized error rates, particularly when the number of hypotheses tested is no longer small. We consider the generalized familywise error rate (k-FWER), which is the probability of making k or more false rejections, as well as the tail probability of the false discovery proportion (FDP), which is the probability that the proportion of false rejections is greater than some threshold. We also consider asymptotic control of the false discovery rate, which is the expectation of the FDP. In this article, we show how to control these generalized error rates when using the graphical approach and its extensions. We demonstrate the utility of the resulting graphical procedures on three clinical trial case studies.

Citicoline for treating people with acute ischemic stroke

BACKGROUND

Stroke is one of the leading causes of long-lasting disability and mortality and its global burden has increased in the past two decades. Several therapies have been proposed for the recovery from, and treatment of, ischemic stroke. One of them is citicoline. This review assessed the benefits and harms of citicoline for treating patients with acute ischemic stroke.

OBJECTIVES

To assess the clinical benefits and harms of citicoline compared with placebo or any other control for treating people with acute ischemic stroke.

SEARCH METHODS

We searched in the Cochrane Stroke Group Trials Register, CENTRAL, MEDLINE Ovid, Embase Ovid, LILACS until 29 January 2020. We searched the World Health Organization Clinical Trials Search Portal and ClinicalTrials.gov. Additionally, we also reviewed reference lists of the retrieved publications and review articles, and searched the websites of the US Food and Drug Administration (FDA) and European Medicines Agency (EMA).

SELECTION CRITERIA

We included randomized controlled trials (RCTs) in any setting including participants with acute ischemic stroke. Trials were eligible for inclusion if they compared citicoline versus placebo or no intervention.

DATA COLLECTION AND ANALYSIS

We selected RCTs, assessed the risk of bias in seven domains, and extracted data by duplicate. Our primary outcomes of interest were all-cause mortality and the degree of disability or dependence in daily activities at 90 days. We estimated risk ratios (RRs) for dichotomous outcomes. We measured statistical heterogeneity using the I² statistic. We conducted our analyses using the fixed-effect and random-effects model meta-analyses. We assessed the overall quality of evidence for six pre-specified outcomes using the GRADE approach.

MAIN RESULTS

We identified 10 RCTs including 4281 participants. In all these trials, citicoline was given either orally, intravenously, or a combination of both compared with placebo or standard care therapy. Citicoline doses ranged between 500 mg and 2000 mg per day. We assessed all the included trials as having high risk of bias. Drug companies sponsored six trials. A pooled analysis of eight trials indicates there may be little or no difference in all-cause mortality comparing citicoline with placebo (17.3% versus 18.5%; RR 0.94, 95% CI 0.83 to 1.07; I² = 0%; low-quality evidence due to risk of bias). Four trials found no difference in the proportion of patients with disability or dependence in daily activities according to the Rankin scale comparing citicoline with placebo (21.72% versus 19.23%; RR 1.11, 95% CI 0.97 to 1.26; I² = 1%; low-quality evidence due to risk of bias). Meta-analysis of three trials indicates there may be little or no difference in serious cardiovascular adverse events comparing citicoline with placebo (8.83% versus 7.77%; RR 1.04, 95% CI 0.84 to 1.29; I² = 0%; low-quality evidence due to risk of bias). Overall, either serious or non-serious adverse events - central nervous system, gastrointestinal, musculoskeletal, etc. - were poorly reported and harms may have been underestimated. Four trials assessing functional recovery with the Barthel Index at a cut-off point of 95 points or more did not find differences comparing citicoline with placebo (32.78% versus 30.70%; RR 1.03, 95% CI 0.94 to 1.13; I² = 24%; low-quality evidence due to risk of bias). There were no differences in neurological function (National Institutes of Health Stroke Scale at a cut-off point of ≤ 1 points) comparing citicoline with placebo according to five trials (24.31% versus 22.44%; RR 1.08, 95% CI 0.96 to 1.21; I² = 27%, low-quality evidence due to risk of bias). A pre-planned Trial Sequential Analysis suggested that no more trials may be needed for the primary outcomes but no trial provided information on quality of life.

AUTHORS' CONCLUSIONS

This review assessed the clinical benefits and harms of citicoline compared with placebo or any other standard treatment for people with acute ischemic stroke. The findings of the review suggest there may be little to no difference between citicoline and its controls regarding all-cause mortality, disability or dependence in daily activities, severe adverse events, functional recovery and the assessment of the neurological function, based on low-certainty evidence. None of the included trials assessed quality of life and the safety profile of citicoline remains unknown. The available evidence is of low quality due to either limitations in the design or execution of the trials.

A comparison of different antibiotic regimens for the treatment of infective endocarditis

BACKGROUND

Infective endocarditis is a microbial infection of the endocardial surface of the heart. Antibiotics are the cornerstone of treatment, but due to the differences in presentation, populations affected, and the wide variety of micro-organisms that can be responsible, their use is not standardised. This is an update of a review previously published in 2016.

OBJECTIVES

To assess the existing evidence about the clinical benefits and harms of different antibiotics regimens used to treat people with infective endocarditis.

SEARCH METHODS

We searched the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, Embase Classic and Embase, LILACS, CINAHL, and the Conference Proceedings Citation Index - Science on 6 January 2020. We also searched three trials registers and handsearched the reference lists of included papers. We applied no language restrictions.

SELECTION CRITERIA

We included randomised controlled trials (RCTs) assessing the effects of antibiotic regimens for treating definitive infective endocarditis diagnosed according to modified Duke's criteria. We considered all-cause mortality, cure rates, and adverse events as the primary outcomes. We excluded people with possible infective endocarditis and pregnant women.

DATA COLLECTION AND ANALYSIS

Two review authors independently performed study selection, 'Risk of bias' assessment, and data extraction in duplicate. We constructed 'Summary of findings' tables and used GRADE methodology to assess the quality of the evidence. We described the included studies narratively.

MAIN RESULTS

Six small RCTs involving 1143 allocated/632 analysed participants met the inclusion criteria of this first update. The included trials had a high risk of bias. Three trials were sponsored by drug companies. Due to heterogeneity in outcome definitions and different antibiotics used data could not be pooled. The included trials compared miscellaneous antibiotic schedules having uncertain effects for all of the prespecified outcomes in this review. Evidence was either low or very low quality due to high risk of bias and very low number of events and small sample size. The results for all-cause mortality were as follows: one trial compared quinolone (levofloxacin) plus standard treatment (antistaphylococcal penicillin (cloxacillin or dicloxacillin), aminoglycoside (tobramycin or netilmicin), and rifampicin) versus standard treatment alone and reported 8/31 (26%) with levofloxacin plus standard treatment versus 9/39 (23%) with standard treatment alone; risk ratio (RR) 1.12, 95% confidence interval (CI) 0.49 to 2.56. One trial compared fosfomycin plus imipenem 3/4 (75%) versus vancomycin 0/4 (0%) (RR 7.00, 95% CI 0.47 to 103.27), and one trial compared partial oral treatment 7/201 (3.5%) versus conventional intravenous treatment 13/199 (6.53%) (RR 0.53, 95% CI 0.22 to 1.31). The results for rates of cure with or without surgery were as follows: one trial compared daptomycin versus low-dose gentamicin plus an antistaphylococcal penicillin (nafcillin, oxacillin, or flucloxacillin) or vancomycin and reported 9/28 (32.1%) with daptomycin versus 9/25 (36%) with low-dose gentamicin plus antistaphylococcal penicillin or vancomycin; RR 0.89, 95% CI 0.42 to 1.89. One trial compared glycopeptide (vancomycin or teicoplanin) plus gentamicin with cloxacillin plus gentamicin (13/23 (56%) versus 11/11 (100%); RR 0.59, 95% CI 0.40 to 0.85). One trial compared ceftriaxone plus gentamicin versus ceftriaxone alone (15/34 (44%) versus 21/33 (64%); RR 0.69, 95% CI 0.44 to 1.10), and one trial compared fosfomycin plus imipenem versus vancomycin (1/4 (25%) versus 2/4 (50%); RR 0.50, 95% CI 0.07 to 3.55). The included trials reported adverse events, the need for cardiac surgical interventions, and rates of uncontrolled infection, congestive heart failure, relapse of endocarditis, and septic emboli, and found no conclusive differences between groups (very low-quality evidence). No trials assessed quality of life.

AUTHORS' CONCLUSIONS

This first update confirms the findings of the original version of the review. Limited and low to very low-quality evidence suggests that the comparative effects of different antibiotic regimens in terms of cure rates or other relevant clinical outcomes are uncertain. The conclusions of this updated Cochrane Review were based on few RCTs with a high risk of bias. Accordingly, current evidence does not support or reject any regimen of antibiotic therapy for the treatment of infective endocarditis.

Sequential trials in the context of competing risks: Concepts and case study, with R and SAS code

Sequential designs and competing risks methodology are both well established. Their combined use has recently received some attention from a theoretical perspective, but their joint application in practice has been discussed less. The aim of this paper is to provide the applied statistician with a basic understanding of both sequential design theory and competing risks methodology and how to combine them in practice. Relevant references to more detailed theoretical discussions are provided, and all discussions are illustrated using a real case study. Extensive R and SAS code is provided in the online Supplementary Material.

CPMCGLM: an R package for p-value adjustment when looking for an optimal transformation of a single explanatory variable in generalized linear models

BACKGROUND

In medical research, explanatory continuous variables are frequently transformed or converted into categorical variables. If the coding is unknown, many tests can be used to identify the "optimal" transformation. This common process, involving the problems of multiple testing, requires a correction of the significance level. Liquet and Commenges proposed an asymptotic correction of significance level in the context of generalized linear models (GLM) (Liquet and Commenges, Stat Probab Lett 71:33-38, 2005). This procedure has been developed for dichotomous and Box-Cox transformations. Furthermore, Liquet and Riou suggested the use of resampling methods to estimate the significance level for transformations into categorical variables with more than two levels (Liquet and Riou, BMC Med Res Methodol 13:75, 2013).

RESULTS

CPMCGLM provides to users both methods of p-value adjustment. Futhermore, they are available for a large set of transformations. This paper aims to provide insight the user an overview of the methodological context, and explain in detail the use of the CPMCGLM R package through its application to a real epidemiological dataset.

CONCLUSION

We present here the CPMCGLMR package providing efficient methods for the correction of type-I error rate in the context of generalized linear models. This is the first and the only available package in R providing such methods applied to this context. This package is designed to help researchers, who work principally in the field of biostatistics and epidemiology, to analyze their data in the context of optimal cutoff point determination.

Acetyl-L-carnitine for patients with hepatic encephalopathy

BACKGROUND

Hepatic encephalopathy is a common and devastating neuropsychiatric complication of acute liver failure or chronic liver disease. Ammonia content in the blood seems to play a role in the development of hepatic encephalopathy. Treatment for hepatic encephalopathy is complex. Acetyl-L-carnitine is a substance that may reduce ammonia toxicity. This review assessed the benefits and harms of acetyl-L-carnitine for patients with hepatic encephalopathy.

OBJECTIVES

To assess the benefits and harms of acetyl-L-carnitine for patients with hepatic encephalopathy.

SEARCH METHODS

We searched the Cochrane Hepato-Biliary Group Controlled Trials Register, CENTRAL, MEDLINE Ovid, Embase Ovid, LILACS, and Science Citation Index Expanded for randomised clinical trials. We sought additional randomised clinical trials from the World Health Organization Clinical Trials Search Portal and ClinicalTrials.gov. We performed all electronic searches until 10 September 2018. We looked through the reference lists of retrieved publications and review articles, and we searched the FDA and EMA websites.

SELECTION CRITERIA

We searched for randomised clinical trials in any setting, recruiting people with hepatic encephalopathy. Trials were eligible for inclusion if they compared acetyl-L-carnitine plus standard care (e.g. antibiotics, lactulose) versus placebo or no acetyl-L-carnitine plus standard care. We are well aware that by selecting randomised clinical trials, we placed greater focus on potential benefits than on potential harms.

DATA COLLECTION AND ANALYSIS

We selected randomised clinical trials, assessed risk of bias in eight domains, and extracted data in a duplicate and independent fashion. We estimated risk ratios (RRs) for dichotomous outcomes and mean differences (MDs) for continuous outcomes. We measured statistical heterogeneity using I² and D² statistics. We subjected our analyses to fixed-effect and random-effects model meta-analyses. We assessed bias risk domains to control systematic errors. We assessed overall quality of the data for each individual outcome by using the GRADE approach.

MAIN RESULTS

We identified five randomised clinical trials involving 398 participants. All trials included only participants with cirrhosis as the underlying cause of hepatic encephalopathy. Trials included participants with covert or overt hepatic encephalopathy. All trials were conducted in Italy by a single team and assessed acetyl-L-carnitine compared with placebo. Oral intervention was the most frequent route of administration. All trials were at high risk of bias and were underpowered. None of the trials were sponsored by the pharmaceutical industry.None of the identified trials reported information on all-cause mortality, serious adverse events, or days of hospitalisation. Only one trial assessed quality of life using the Short Form (SF)-36 scale (67 participants; very low-quality evidence). The effects of acetyl-L-carnitine compared with placebo on general health at 90 days are uncertain (MD -6.20 points, 95% confidence interval (CI) -9.51 to -2.89). Results for additional domains of the SF-36 are also uncertain. One trial assessed fatigue using the Wessely and Powell test (121 participants; very low-quality evidence). The effects are uncertain in people with moderate-grade hepatic encephalopathy (mental fatigue: MD 0.40 points, 95% CI -0.21 to 1.01; physical fatigue: MD -0.20 points, 95% CI -0.92 to 0.52) and mild-grade hepatic encephalopathy (mental fatigue: -0.80 points, 95% CI -1.48 to -0.12; physical fatigue: 0.20 points, 95% CI -0.72 to 1.12). Meta-analysis showed a reduction in blood ammonium levels favouring acetyl-L-carnitine versus placebo (MD -13.06 mg/dL, 95% CI -17.24 to -8.99; 387 participants; 5 trials; very low-quality evidence). It is unclear whether acetyl-L-carnitine versus placebo increases the risk of non-serious adverse events (8/126 (6.34%) vs 3/120 (2.50%); RR 2.51, 95% CI 0.68 to 9.22; 2 trials; very low-quality evidence). Overall, adverse events data were poorly reported and harms may have been underestimated.

AUTHORS' CONCLUSIONS

This Cochrane systematic review analysed a heterogeneous group of five trials at high risk of bias and with high risk of random errors conducted by only one research team. We assessed acetyl-L-carnitine versus placebo in participants with cirrhosis with covert or overt hepatic encephalopathy. Hence, we have no data on the drug for hepatic encephalopathy in acute liver failure. We found no information about all-cause mortality, serious adverse events, or days of hospitalisation. We found no clear differences in effect between acetyl-L-carnitine and placebo regarding quality of life, fatigue, and non-serious adverse events. Acetyl-L-carnitine reduces blood ammonium levels compared with placebo. We rated all evidence as of very low quality due to pitfalls in design and execution, inconsistency, small sample sizes, and very few events. The harms profile for acetyl-L-carnitine is presently unclear. Accordingly, we need further randomised clinical trials to assess acetyl-L-carnitine versus placebo conducted according to the SPIRIT statements and reported according to the CONSORT statements.

Shift work, night work, and the risk of prostate cancer: A meta-analysis based on 9 cohort studies

BACKGROUND

Epidemiology studies suggested that shift work or night work may be linked to prostate cancer (PCa); the relationship, however, remains controversy.

METHODS

PubMed, ScienceDirect, and Embase (Ovid) databases were searched before (started from the building of the databases) February 4, 2017 for eligible cohort studies. We pooled the evidence included by a random- or fixed-effect model, according to the heterogeneity. A predefined subgroup analysis was conducted to see the potential discrepancy between groups. Sensitivity analysis was used to test whether our results were stale.

RESULTS

Nine cohort studies were eligible for meta-analysis with 2,570,790 male subjects. Our meta-analysis showed that, under the fixed-effect model, the pooled relevant risk (RR) of PCa was 1.05 (95% confidence interval [CI]: 1.00, 1.11; P = .06; I = 24.00%) for men who had ever engaged in night shift work; and under the random-effect model, the pooled RR was 1.08 (0.99, 1.17; P = .08; I = 24.00%). Subgroup analysis showed the RR of PCa among males in western countries was 1.05 (95% CI: 0.99, 1.11; P = .09; I = 0.00%), while among Asian countries it was 2.45 (95% CI: 1.19, 5.04; P = .02; I = 0.00%); and the RR was 1.04 (95% CI: 0.95, 1.14; P = .40; I = 29.20%) for the high-quality group compared with 1.21 (95% CI: 1.03, 1.41; P = .02; I = 0.00%) for the moderate/low-quality group. Sensitivity analysis showed robust results.

CONCLUSIONS

Based on the current evidence of cohort studies, we found no obvious association between night shift work and PCa. However, our subgroup analysis suggests that night shift work may increase the risk of PCa in Asian men. Some evidence of a small study effect was observed in this meta-analysis.

Design, data monitoring, and analysis of clinical trials with co-primary endpoints: A review

We review the design, data monitoring, and analyses of clinical trials with co-primary endpoints. Recently developed methods for fixed-sample and group-sequential settings are described. Practical considerations are discussed, and guidance for the application of these methods is provided.