====== Recommendations ====== {{rss>https://pubmed.ncbi.nlm.nih.gov/rss/search/1lIZdG7__acOiOnbLyjnbtjg5c5eh691Ei7_riG6xLOiVTVs0O/?limit=15&utm_campaign=pubmed-2&fc=20240828050009}} Representation in favor of a [[management]]. In [[guideline]]s and other [[publication]]s, recommendation for a clinical service is classified by the balance of risk versus benefit of the service and the [[level of evidence]] on which this information is based. ---- [[Prophylactic antibiotics]] is an [[option]] and not a [[recommendation]] ((JAUME, Alejandra et al. Fístula de líquido cefalorraquídeo postraumática: propuesta de algoritmo diagnóstico y terapéutico. Arch. Med Int [online]. 2015, vol.37, n. 1 [citado 2022-08-24], pp.47-52. Disponível em: . ISSN 0250-3816)). ---- Clinicians' trust [[level of evidence 1]] [[recommendation]]s, issued on preponderantly solid [[randomized clinical trial]]s (RCTs), to guide [[best practice]] [[decision-making]]. However, sometimes physicians following one CPG find themselves in a situation in which they do not follow another, issued on the same strong evidence base. The aim of Volovici et al. is to reflect on the [[consistency]] of [[recommendation]]s in different [[guideline]]s (between-guideline consistency). They also consider within-guideline consistency (or durability), defined as the number of recommendations carried over from one edition to another in consecutive editions of the same CPG. For illustration purposes, they use two examples: hypertension guidelines and traumatic brain injury (TBI) guidelines. They conclude that just like research, CPGs also need to have between-guideline and within-guideline consistency (akin to the [[reproducibility]] of studies). Clinicians and researchers should take into account the lower [[consistency]] of guidelines that are not based on at least one strong RCT ((Volovici V, Steyerberg EW. Lost in translation between [[evidence]] and [[recommendation]]s: Expert opinion is needed to define "level I". World Neurosurg. 2021 Mar 25:S1878-8750(21)00465-4. doi: 10.1016/j.wneu.2021.03.095. Epub ahead of print. PMID: 33775869.)). ===== Strength ===== [[Level of evidence 1]] [[Level of evidence 2]] [[Level of evidence 3]] ---- ----