Nuanced Interpretation of Research Results

Editorial| December 2023 Nuanced Interpretation of Research Results Daniel I. Sessler, M.D. Daniel I. Sessler, M.D. 1Department of Outcomes Research, Cleveland Clinic, Cleveland, Ohio. https://orcid.org/0000-0001-9932-3077 Search for other works by this author on: This Site PubMed Google Scholar Author and Article Information Accepted for publication August 23, 2023. Address correspondence to Dr. Sessler: ds@ccf.org Anesthesiology December 2023, Vol. 139, 730–733. https://doi.org/10.1097/ALN.0000000000004753 Views Icon Views Article contents Figures & tables Video Audio Supplementary Data Peer Review Share Icon Share Facebook Twitter LinkedIn Email Cite Icon Cite Get Permissions Search Site Citation Daniel I. Sessler; Nuanced Interpretation of Research Results. Anesthesiology 2023; 139:730–733 doi: https://doi.org/10.1097/ALN.0000000000004753 Download citation file: Ris (Zotero) Reference Manager EasyBib Bookends Mendeley Papers EndNote RefWorks BibTex toolbar search Search Dropdown Menu toolbar search search input Search input auto suggest filter your search All ContentAll PublicationsAnesthesiology Search Advanced Search Topics: false-positive results, judgment, mortality, number needed to treat, observational studies, therapeutics, absolute risk reduction, antiemetic agents, cerebrovascular accident, false-negative results What fraction of published papers reporting statistically significant results is flat-out wrong remains unknown.1 But it is clearly higher than generally believed, and probably much higher.2 Poor methodology probably contributes most to unreliable research results. Observational research is especially prone to error—that is, conclusions that meaningfully diverge from ground truth—because it is impossible to completely control confounding. Some confounding factors are observed, but dichotomously—for example, history of hypertension with no gradations—although severity matters enormously. Other known confounding factors are unavailable from accessible registries. And of course, some important confounding factors are simply unknown. Perhaps consequently, observational research is often overturned by robust trials. Trials are also often wrong, although randomization and blinding provide substantial protection against selection bias, confounding, and measurement bias. The more serious issue with trials is that inadequate sample size makes many results fragile.3 Sample size becomes especially challenging when evaluating sparse dichotomous outcomes—including... You do not currently have access to this content.