Portal de Periódicos da CAPES

Empiric Antibiotic for Community Acquired Pneumonia in Adult Patients. A Network Meta-Analysis

2019; RELX Group (Netherlands); Linguagem: Inglês

10.2139/ssrn.3436307

1556-5068

Lara Montes-Andujar, Elena Tinoco, Orville Baez-Pravia, Pablo Llamazares Blanco, Cindy Paola Oviedo Segura, Stefania Prol, Victoria Reyes, Carmen Zurdo, Veronica V. Angeles, Olga Varona, José Antonio Gómez Valero, R. Suárez del Villar, Guillermo Ortíz, Justo Menéndez, Jesus Blanco, Antoní Torres, Pablo Cardinal‐Fernández,

Machine Learning in Healthcare

Objective: The main aim of this network meta-analysis was to identify, in patients with community acquired pneumonia (CAP) that required hospital admission, the empiric antibiotic (Em-ATB) with the highest probability of being the best (HPBB) in term of cure rate. The ancially aim was to identify the Em-ATB with the HPBB in term of mortality.Method: Included criteria: studies that analyze adult patients (>16 y.o.) diagnosed with a CAP that required hospitalization, randomized at least two different Em-ATB, reported cure rate and written in English or Spanish. Excluded criteria: ambiguous antibiotics protocol and studies published only in abstract or letter format. Data Sources: MEDLINE, EMBASE, COCHRANE, LILACS and citation review from 1/1/00 to 12/31/18. Risk of bias: Cochrane Collaboration's tool. Statistical analyses: Frequentist method performed with the "netmeta" library, R package.Results: 28 RCTs from the initial 41,307 citations that were initially screened were included. The quality of RCTs was high. For cure, two networks were constructed thus two Em-ATB have the HPBB: Cetaroline 600mg (BID) and Piperacillin 2000mg (BID). For mortality, three networks were constructed thus three Em-ATB have the HPBB: ceftriaxone 2000mg (QD) plus levofloxacine 500(BID), ertapenem 1000mg (BID) and amikacin 250mg (BID) plus clarithromycin 500mg (BID).Conclusion: This NMA reported that the knowledge regarding Em-ATB in CAP that require hospitalization is unconnected. Future RCTs that compare treatments with the HPBB in each event (cure or mortality) are necessary for unifying the knowledge (CRD42017060692).Clinical Trial Registration: This study was registered at PROSPERO database (CRD42017060692).Funding Statement: Supported by: SGR 2017/787, Ciber de Enfermedades Respiratorias (Ciberes CB06/06/0028), Pneumonia Corporate Research Program (CRP). The Ciberes is an iniciative of the ISCIII. SGR: Support to research groups of Catalunya.Declaration of Interests: The authors declare no conflict of interest.Ethical Approval Statement: The authors state ethics approval was not required.