Discussion

To understand and critically evaluate translational and clinical research studies, it is paramount to identify the level of evidence. Study design contributes to the quality of the data, the level of evidence, and the potential impact within a specified field. By understanding study design, an individual can appropriately interpret the results, potential generalizability, and whether the results should be applied to clinical practice. Within the level of evidence pyramid, there is a role and utility for each study design, but the applicability is dependent upon the quality of evidence and generalizability. We will discuss in greater detail within this chapter.

A well conducted randomised controlled trial (RCT) is extremely important in the field of cardiovascular medicine. At the same time, it is equally important to understand the strengths and limitations of any RCT, and internal variability is a concept in clinical trials that is poorly understood. Variability in a clinical trial may be introduced at an individual level or during measurement, sampling, or conduct of the trial. It is not the same as internal validity, which is a broader concept of accuracy; to be valid, a study should minimise variability and have sound methodology. There are various steps that may be followed to minimise the internal variability in a clinical trial. One aspect of great importance is the adjudication process, which should be done meticulously and is often a step that is overlooked. It is important to standardise each step as much as possible, to ensure consistency and reduce noise at all levels. The concepts discussed in this review may serve as a roadmap to limit the influence of internal variability and maximise internal validity of RCT results.

Un essai contrôlé randomisé (ECR) bien réalisé est extrêmement important dans le domaine de la médecine cardiovasculaire. En même temps, il est également important de comprendre les points forts et les limites de tout ECR. La variabilité interne est l'un de ces concepts des essais cliniques qui sont mal compris. La variabilité dans un essai clinique peut être introduite à l’échelle des personnes, dans les mesures, dans l’échantillonnage ou dans la conduite de l'essai. Il ne s'agit pas de la « validité interne », qui est un concept plus large concernant la précision. Pour être valide, une étude doit réduire au minimum la variabilité et avoir une bonne méthodologie. Il y a diverses étapes à suivre pour réduire au minimum la variabilité interne dans un essai clinique. Un aspect d'une grande importance est le pro-cessus d'attestation qui doit être effectué méticuleusement et est souvent une étape qui est négligée. Il est important de normaliser chaque étape autant que possible pour assurer l'uniformité et réduire le bruit à tous les niveaux. Les concepts traités dans cet article peuvent servir de feuille de route pour limiter l'influence de la variabilité interne et maximiser la validité interne des résultats des ECR.

Randomised controlled trials are the best way to study the evaluation of treatments. We have evaluated the quantity and quality of clinical trials in three of the main journals in the specialty of oral and maxillofacial surgery between January 2010 and December 2016, using a scientometric analysis, and evaluation by the Jadad scale. In this period, 303 randomised controlled trials (5% of the total) were identified; the largest number of studies were from Asia (45%) followed by Europe (32%). The subgroup that concerned most studies was oral surgery. The mean score on the Jadad scale was 3.06 points, which means that 32% of the total studies had a low risk of bias. Studies that declared funding and adherence to Consolidated Standards of Reporting Trials (CONSORT) were given significantly higher scores (p < 0.001) than studies that did not. We conclude that randomised controlled trials in oral maxillofacial surgery have evolved in both quality and quantity since previous surveys were published. The quality of trials was related to the presence of funding and adherence to CONSORT.

The fragility index (FI) is calculated by iteratively changing one outcome “event” to a “non-event” within a trial until the associated p-value exceeds 0.05.

To investigate the FI and fragility quotient (FQ) of trial endpoints referenced in the ACCF/AHA/SCAI guidelines in the management of ST-elevation myocardial infarctions. Secondarily, we assess the post-hoc power and risk of bias for these specific outcomes and whether differences exist between adequately and inadequately powered studies on fragility measures.

All citations referenced in the guideline were screened for inclusion criteria. The FI and FQ for all included trials were then calculated. The Cochrane ‘risk of bias’ Tool 2.0 was used to evaluate the likelihood and sources of bias in the included trials.

Forty-two randomized controlled trials were included for assessment. The median FI was 10 with a FQ of 0.0055. Seven trials were at a high risk of bias, all due to bias in the randomization process. Fifteen trials were found to be underpowered. Adequately powered studies had higher FIs and FQs compared to underpowered studies.

Our findings support the use of FI and FQ analyses with power analyses in future methodology of randomized control trials. With understanding and reporting of FI and FQ, evidence of studies can be readily available and quickly eliminate some readers' concern for possible study limitations.

The LRINEC score was introduced in 2004 but it has not yet been incorporated as a reliable diagnostic tool in the management of Necrotising Fasciitis. NF is uncommon and proportionately the evidence is limited, but the contradictory literature has resulted in varying individual concepts and clinical practices. This article is aimed to evaluate the reliability of the LRINEC score and determine its effectiveness based on the current available evidence.

Eighteen clinical studies published during 2004–2018 were identified after an extended literature search and critically appraised to determine the validity, reliability and applicability of the individual pieces of evidence. The recommendations for practice were formulated after summative analysis of the evidence.

The systemic review and observational studies report the LRINEC score has a variable range of sensitivity (43.2–80%), positive predictive value (57–64%) and negative predictive value (42–86%) which is lower than the initial results by Wong et al. (2004). The LRINEC score ≥6 correlates well with diagnosis of NF but 7.1 was found to be statistically significant. Higher LRINEC score correlates with higher SOFA score, prolonged ICU and hospital stay, and mortality. It is not sensitive in immunocompromised patients.

There is Level 3 evidence that LRINEC score is a reliable tool which can aid the clinical diagnosis of NF. It can stratify the high-risk patients and predict outcome; however, it should be correlated with the clinical assessment and radiological diagnostic modalities should be simultaneously used when doubt exists (grade C).