Evidence-based medicine: A method for solving clinical problems in hepatology

Abstract

Evidence to solve clinical problems may be drawn from sources as diverse as individual clinical experience, the clinical experience of one' s colleagues, physiological research, observational studies, randomized clinical trials (RCTs), and meta-analyses.The strongest evidence comes when RCTs that examine patient-important outcomes are summarized in cumulative meta-analyses.An alternative to using evidence for solving clinical problems is to rely on expert recommendations from, for example, textbooks or review articles.Unfortunately, differences among experts are common, and there are delays and sometimes discrepancies between accumulating evidence and recommendations.Antman et al. provided a vivid illustration of this issue for a common medical problem, treatment of myocardial infarction.These investigators conducted cumulative meta-analyses and surveyed expert recommendations that were made as the evidence accumulated. 1 The first RCT of thrombolytic therapy for patients with myocardial infarction, conducted in 1965, enrolled 23 patients.The small sample size implies a very wide confidence interval around the estimate of treatment effect (odds ratio [OR], 0.5; 95% CI, 0.5-1.7;P Ͼ .05).As more trials were conducted, the confidence interval around the estimate from the cumulative meta-analysis narrows.By 1973, 10 RCTs had enrolled a total of 2,544 patients, showing a definite mortality reduction with thrombolytic agents (OR, 0.7; 95% CI, 0.6-0.9;P Ͻ .01),and by 1986, almost 22,000 patients had been randomized, and the relative-risk reduction was highly statistically significant (OR, 0.78; 95% CI, 0.69-0.90;P Ͻ .00001).Although the evidence supported thrombolytic therapy from the early 1970s onward, experts did not routinely endorse thrombolytic therapy in textbooks and peer-reviewed articles until the late 1980s. 1 Variability in recommendations persisted into the early 1990s; some suggested routine thrombolytic therapy, whereas others continued to recommend restricted use for specific indications.Similar discrepancies between accumulating evidence and expert recommendations can be found in hepatology.In the

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 12 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.