Reviews: Rapid! Rapid! Rapid! …and systematic
Notice bibliographique
Résumé
“Although time is a reality, lack of time is not lack ofreality” - HOJESSystematic reviewsSystematic reviews are scientific investigations, with pre-planned methods and an assembly of original studies astheir “subjects” [1,2]. They synthesize the results of mul-tiple primary investigations by using strategies that limitbias and random error [1,2]. Systematic reviews aretransparent about how studies were identified and whichwere included or excluded, the risk of bias assessment,and the methods to summarize data and assess the cer-tainty in the evidence. Standards for the conduct of sys-tematic reviews have been made available by theCochrane Collaboration and other organizations. If sys-tematic reviews are done well, e.g., by adhering to con-duct (e.g., Cochrane Handbook for Systematic Reviewsof Interventions) and reporting (e.g., PRISMA State-ment) best practice standards, it is not sensible to ques-tion the value of systematic reviews as a source ofinformation for shaping decision making [3,4]. Thismethodology of systematic reviews—although laid outthree or more decades ago—is continuously and rapidlyupdated by scientists specializing in research synthesis.Now, Systematic Reviews is publishing a series of articlesincluding methods and examples of accelerating ap-proaches to conducting literature reviews. As a rule ofthumb, rapid systematic reviews should be conducted inless than 8 weeks, including protocol publication. Onthe whole, this is a saving of about 75% in terms of timecompared to what most researchers would propose asstandard timeline for systematic reviews. Examples willhighlight how health policy decisions can be influencedwhen a rapid review methodology is used.The challenge of traditional systematic reviewsFindings from a single randomized trial are often rapidlychallenged by succeeding studies, and rigorous systematicreviews help approximate “true evidence” and estimatesin effects [5]. High-quality systematic reviews are usedmore often and are considered more trustworthy byhealth professionals in terms of relevance to clinical prac-tice than other types of designs [6]. However, conductingand adhering to the standards of traditional systematic re-views can be time consuming. The reason for that lies inthe rigorous approach to methods ensuring that the bestavailable evidence is identified, assessed, and synthesized.But those demanding evidence syntheses for decisionmaking are increasingly living in faster paced times, influ-enced by innovative interventions and technology that ac-celerate communication and interaction. Decision makersoften do not appreciate the intricacies of researchmethods and the time needed to comply with the task.The argument that transparency is ensured by completinga traditional systematic reviews is often not convincingenough. But there are other reasons.The rationale for rapid —systematic reviewsThe concern regarding a timely decision on health care andpolicies is the driving force for rapid reviews. In fact, deci-sion making should not be delayed in most situations andcannot be delayed in some. In the face of a tragic Ebola epi-demic, we are reminded of how rapidly answers are re-quired. To base answers on the best available evidence, thisevidence must be synthesized without undue delays. Whiletypical systematic reviews can take years to complete (oneof the author was involved in a systematic review that took12 years to complete), rapid reviews are required when fa-cing such dramatic situations. Prior to the Ebola epidemic,the fear of avian influenza prompted the World HealthOrganization to offer rapid guidelines that were supportedby a rapid review methodology. From guideline panel for-mation to completion of the recommendations, only about12 weeks passed [7]. Another recent rapid systematic re-view was commissioned to inform decision making with re-gard to the safety of two drugs, bevacizumab (Avastin) andranibizumab (Lucentis), widely used to stabilize vision in
Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.
Comment cette classification a été obtenuedéplier
Prédiction machine sur la base complète
Imitation des enseignantsNi prévalence calibrée, ni vérité terrain. Validation humaine à venir. Le volet Gemma est une étiquette directe du modèle pour chaque travail de la base, lue sur la notice réduite au titre. Le volet Codex est un classifieur appris des 10 348 étiquettes directes de Codex et calibré sur les taux pondérés de l'échantillon; les champs sans appui suffisant ne portent aucun appel Codex. Le mode candidate est l'union des deux volets; le consensus est leur intersection. Ces sorties portent le statut machine_predicted_unvalidated et ne sont pas des étiquettes humaines.
Scores du classifieur distillé par catégorie (deux têtes)
| Catégorie | Codex | Gemma |
|---|---|---|
| Métarecherche | 0,112 | 0,298 |
| Méta-épidémiologie (sens strict) | 0,005 | 0,006 |
| Méta-épidémiologie (sens large) | 0,014 | 0,009 |
| Bibliométrie | 0,028 | 0,028 |
| Études des sciences et des technologies | 0,003 | 0,008 |
| Communication savante | 0,019 | 0,020 |
| Science ouverte | 0,008 | 0,011 |
| Intégrité de la recherche | 0,009 | 0,017 |
| Charge utile insuffisante (le modèle a refusé de juger) | 0,154 | 0,148 |
Scores machine (provisoires)
Les deux têtes enseignantes du modèle étudiant, lues sur ce travail. Un score ordonne la base pour la relecture; il n'affirme jamais une catégorie, et le statut de validation accompagne chaque rangée tel quel.
Scores de référence d'un modèle non mature (critères de maturité non atteints, 7 itérations). Un score ordonne; il n'affirme jamais une catégorie.
score_only:v0-immature-baseline · tel quel depuis la passe de notation : score_only signifie que le nombre peut ordonner les travaux, et qu'aucune étiquette de catégorie n'en découleClassification
machine, non validéePrédiction automatique; un appel candidat d’une seule source (Gemma direct ou Codex distillé), pas un consensus.
Le détail, modèle par modèle et score par score, se trouve en fin de page sous « Comment cette classification a été obtenue ».