No abstract

The strategic formation of inter-individual collaboration networks. Evidence from co-invention patterns Nicolas CARAYOL and Pascale ROUX* ABSTRACT.-This paper contributes to an emerging literature aiming to understand the behavioral patterns that lead to the formation of social networks. We introduce a strategic model of inter-individual collaborations formation which is a variation of the Connections model. Heterogeneous agents benefit from knowledge spillovers flowing through the network and bear the costs of maintaining their direct links which increase with geographic distance. \Ne show that this simple model generates emergent networks that share the main structural properties observed in most real social networks. Moreover, we bring the model to co-invention networks data and find that the model fits quite well the data through the various structural measures used. |n particular, it provides a consistent explanation of the observed asymmetry in the distribution of neighborhood sizes and of the high concentration of connections in local areas while some distant connections are formed. R?SUM?.-Cet article contribue ? une litt?rature ?mergente d?di?e ? la mise en ?vidence des comportements qui pr?sident ? la formation des r?seaux sociaux. Nous introduisons un mod?le de formation strat?gique des collaborations inter-individuelles qui est une variante du mod?le des connections. Des agents h?t?rog?nes b?n?ficient des spillovers de connaissance qui circulent imparfaitement dans le r?seau et supportent les co?ts de maintien de leurs liens directs, lesquels s'accroissent avec la distance g?ographique. Nous montrons que ce mod?le simple g?n?re des r?seaux ?mergents qui disposent des principales propri?t?s structurelles observ?es dans la plupart des r?seaux sociaux. En outre, nous confrontons le mod?le aux donn?es de co-invention et trouvons que le mod?le th?orique est consistant avec les observations ? travers les diff?rentes mesures structurelles utilis?es. En particulier, le mod?le propose une explication de l'asym?trie de la distribution des tailles de voisinage et de la forte concentration des connections dans les espaces locaux alors que quelques connections distantes sont form?es.

No abstract

Blood-culture bottles (BCBs) are widely used to improve the diagnosis of orthopedic device-related infections. Data is scarce on the growth of Cutibacterium acnes and its genotypes in BCBs under real-life clinical conditions.We studied 39 cases of revision arthroplasty for which at least one intraoperative sample yielded a pure C. acnes culture from anaerobic BCBs (BD Bactec Lytic/10 Anaerobic/F [Lytic Ana]) and/or solid media. Genotyping of C. acnes isolates from the 39 cases allowed: i) the identification of 49 non-redundant isolates belonging to four clonal complexes (CCs): CC18, CC28, CC36, and CC53 and ii) the determination of infectant and contaminant strains. Under real-life clinical conditions, Lytic Ana alone was more often positive for contaminants than infectant strains (18/36 [50%] versus 2/13 [15.4%]; p = 0.047). The time to detection (TTD) values in Lytic Ana were shorter for CC53 than other CCs (mean [SD] TTD: 77 [15] versus 165 [71] hours; p = 0.02). CC53 was confirmed to grow faster than other CCs by studying an enlarged panel of 70 genotyped C. acnes strains inoculated in vitro into Lytic Ana vials (mean [SD] TTD: 73 [13] versus 122 [50] hours; p < 0.001).The use of Lytic Ana BCBs in orthopedics increases the recovery rate of C. acnes but leads to the isolation of proportionally more contaminants than true infectant strains. TTD values are much shorter for CC53 strains, irrespective of their being infectant or contaminant. TTD does not solely reflect the bacterial load of samples but also clonal complex-related traits.