Mortality predictors and definition proposal for complicated coagulase-negative Staphylococcus bacteraemia: a multicentre prospective cohort study

Fo Pee Re iew Mo ali y p edic o s and de ini ion p oposal o complica ed coagulase-nega i e S aphylococcus bac e emia. A mul icen e p ospec i e coho s udy Jou nal: Clinical Mic obiology and In ec ion Manusc ip ID CLM-24-28675.R1 A icle Type: O iginal A icle Da e Submi ed by he Au ho : n/a Comple e Lis o Au ho s: Va isco, Benede a; ASST San i Paolo e Ca lo, Clinic o In ec ious and T opical Diseases, Depa men o Heal h Sciences Ma ínez Pé ez-C espo, Ped o Ma ía; Hospi al Uni e si a io de Valme, Unidad de En e medades In ecciosas y Mic obiología Re ama , Pila ; Hospi al Uni e si a io Vi gen Maca ena, Unidad Clínica de En e medades In ecciosas y Mic obiología López-He nández, Inmaculada; Hospi al Uni e si a io Vi gen Maca ena, Unidad Clínica de En e medades In ecciosas y Mic obiología Fa iñas, Ma ia Ca men; Uni e si y Hospi al Ma qués de Valdecilla, In ec ious Diseases Fe nández-Na al, Isabel; Complejo Asis encial Uni e si a io de León, Clinical Mic obiology Ma ía Te esa, Pé ez-Rod íguez; Complexo Hospi ala io Uni e si a io de Vigo, Se icio de Medicina In e na Goikoe xea, Josune; Hospi al Uni e si a io de C uces, En e medades In ecciosas Sánchez Cal o, Juan Manuel; Hospi al Uni e si a io de Je ez de la F on e a, UGC En e medades In ecciosas y Mic obiología Clínica Buzon-Ma in, Luis; Hospi al Uni e si a io de Bu gos, Medicina In e na León , E a; Hospi al Uni e si a io Nues a Seño a de Valme, Unidad de En e medades In ecciosas y Mic obiología Vinuesa Ga cía, Da id; Hospi al Uni e si a io Clínico San Cecilio., Unidad de Ges ión Clínica de En e medades In ecciosas. Regue a, José; Regional Uni e si y Hospi al o Malaga, Depa men o In ec ious Diseases Bahamonde, Albe o; Hospi al El Bie zo, Depa men o In e nal Medicine. Fe nandez, Jona han; Hospi al Uni e si a io Cen al de As u ias, Mic obiology Uni Rod íguez-Baño, Jesús; Hospi al Uni e si a io Vi gen Maca ena, En e medades In ecciosas, Mic obiologa y Medicina P e en i a; Uni e sidad de Se illa, Medicina López-Co és, Luis Edua do ; Hospi al Uni e si a io Vi gen Maca ena, Unidad Clínica de En e medades In ecciosas y Mic obiología; Uni e sidad de Se illa, Facul ad de Medicina Key Wo ds: Coagulase-nega i e s aphylococci, Bloods eam in ec ions, Complica ed bac e emia, Mo ali y p edic o s, Ca he e - ela ed bloods eam in ec ions Clinical Mic obiology and In ec ion Fo Pee Re iew Abs ac : Objec i e. To explo e a de ini ion o complica ed coagulase-nega i e s aphylococci bloods eam in ec ions (CoNS BSI), and o iden i y p edic o s o mo ali y. Me hods. P ospec i e coho s udy conduc ed om Oc obe 2016 o Ma ch 2017 in 26 Spanish hospi als. Complica ed CoNS BSI c i e ia included lack o ea ly ca he e emo al in ca he e - ela ed cases, o eign indwelling implan , pe sis en bac e emia, e e ≥72 hou s on ac i e he apy, me as a ic in ec ion o deep-sea ed ocus and in ec i e endoca di is. Independen p edic o s o 30-day mo ali y we e e alua ed by Cox eg ession, and he impac o he de ini ion o complica ed bac e emia assessed. Resul s. O e all, 445 CoNS BSI cases we e included; ca he e - ela ed in ec ions we e p edominan (336/445, 75.5%). Complica ed bac e emia was iden i ied in 240/445 pa ien s (53.9%); 30-day mo ali y in complica ed and uncomplica ed cases we e 53/240 (22.1%) and 24/205 (11.7%), espec i ely (p=0.004). P edic o s o 30-day mo ali y iden i ied in he mul i a ia e analysis included age (HR 1.03, 95%CI 1.01-1.05), ce eb o ascula disease (HR 2.58, 95%CI 1.45-4.58), immunosupp essi e he apy (HR 2.16, 95%CI 1.22-3.84), SOFA sco e (HR 1.09, 95%CI 1.03-1.16), and complica ed bac e emia (HR 2.14, 95%CI 1.29-3.53). A ca he e - ela ed sou ce o bac e emia was ound o be p o ec i e (HR 0.49, 95%CI 0.30-0.80). When speci ic c i e ia o de ine complica ed bac e emia we e included, e e ≥72h was associa ed wi h inc eased isk o dea h (HR 2.52, 95%CI 1.52-4.17) and ea ly ca he e emo al was p o ec i e (HR 0.47, 95%CI 0.26-0.83). Conclusions. A high p opo ion o pa ien s p esen ed complica ed bac e emia acco ding o he p oposed c i e ia; hese pa ien s had highe haza ds o mo ali y. O he mo ali y p edic o s we e iden i ied. Fu he s udies would be needed o alida e he p oposed c i e ia. Page 1 o 108 Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew Se ille, Decembe 2024 Dea P o . Paul, Edi o , Clinical Mic obiology and In ec ion Thank you o he cons uc i e eedback and de ailed sugges ions ega ding ou manusc ip , "Mo ali y p edic o s and de ini ion p oposal o complica ed coagulase-nega i e S aphylococcus bac e emia. A mul icen e p ospec i e coho s udy" (manusc ip ID: CLM-24-28675). We app ecia e he ime and e o ha bo h you and he e iewe s ha e dedica ed o his p ocess. Based on you sugges ions, we ha e made ca e ul e isions o imp o e he cla i y and s eng h o ou wo k. The acked and clean copies o he e ised manusc ip a e submi ed he e o you conside a ion. Kind ega ds, Luis Edua do López-Co és On behal o all au ho s Page 2 o 108Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew EDITOR COMMENTS TO THE AUTHOR Commen (1): Among inclusion c i e ia “ he e we e sys emic symp oms like e e o chills” – was his he ac ual c i e ion used o inclusion? As de ined, his c i e ion is non-speci ic, canno be eplica ed and p obably leads o he e ogenei y in pa ien inclusion. I a mo e p ecise c i e ion was used please de ine i in he manusc ip . I his was he c i e ion, we ha e li le con idence ha he pa ien s in he coho has a clinically-signi ican CONS bac e emia. Response: Thank you o he impo an eedback. To be b ie , we acknowledge ha he c i e ia we e no desc ibed wi h su icien speci ici y. We ha e now cla i ied he speci ic inclusion c i e ia in he manusc ip and Supplemen a y ma e ial, aligning hem wi h hose ou lined in he s udy p o ocol. Commen (2): “Da a we e collec ed ia an elec onic case epo ” – please de ine da a sou ces and collec ion me hods. We e da a manually collec ed om elec onic pa ien s iles o pape cha s? Was any au oma ed da a que y ool used? Response: We ha e e ised he manusc ip o cla i y he da a collec ion me hods, which now eads: "Da a we e collec ed p ospec i ely om ei he elec onic heal h eco ds o pape cha s a each cen e acco ding o local a ailabili y. Da a o each en olled pa ien we e en e ed in o a s anda dized online case epo o m.” Commen (3): Unde “De ini ions and a iables” please de ine “empi ic he apy” – i s 24 hou s? 48 hou s” wha is he s a ing ime poin – blood cul u e collec ion ime? In me hods you de ined App op ia e an imic obial he apy and in he esul s you e e o “ac i e” he apy – please use consis en e minology. Response: Thank you o you eedback. We ha e added he de ini ion o "empi ic he apy," which is now s a ed as " he an imic obial egimen ini ia ed p io o he iden i ica ion o he in ec ing mic oo ganism and i s suscep ibili y p o ile; since G am/MALDI epo s we e ypically a ailable wi hin 24 hou s, he apy was usually conside ed empi ical du ing his ime ame". Rega ding he e minology inconsis ency, we app ecia e you obse a ion. We ind he e m "ac i e he apy" mo e accu a e han "app op ia e he apy." As such, we ha e e ised he manusc ip o ensu e he consis en use o "ac i e he apy" h oughou he sec ions. Commen (4): Follow-up blood cul u es we e aken only o <30% o pa ien s by physicians’ disc e ion. CONS a e low i ulence pa hogens who do no necessa y cause ma ked sepsis signs, Page 3 o 108 Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew i is likely ha some o he pa ien s in whom ollow-up blood cul u es we e no aken (p esumably because hey had no e e o o he sepsis signs) also had p olonged bac e emia. I p opose lis ing he indica ions o aking ollow-up blood cul u es (p ospec i e c i e ia o e oac i e easons om he da a) and e-naming he a iable “Posi i e ollow-up blood cul u es” o “Posi i e ollow-up blood cul u es when indica ed” o simila . Response: Thank you o highligh ing his impo an poin . Gi en he lack o guidelines o ollow-up blood cul u es (FUBC) in CoNS bac e emia, ou obse a ional s udy e lec s clinical p ac ice, wi h FUBC pe o med a he disc e ion o he ea ing physician. We ecognize ha posi i e FUBC esul s may ep esen only a subse o ac ual pe sis en bac e emia cases, as some pa ien s no es ed may ha e had unde ec ed posi i e esul s. To add ess his, we ha e included his poin in he Limi a ions sec ion o he s udy. In esponse o you sugges ion, we conduc ed a e ospec i e analysis o ac o s ha may ha e in luenced he decision o pe o m FUBC. As shown in Supplemen a y Table S1, pa ien s wi h endoca di is o ca diac/endo ascula de ices we e signi ican ly mo e likely o unde go FUBC. We ha e summa ized hese indings in he Resul s sec ion. Al hough we hypo hesize ha hese lis ed ac o s likely in luenced FUBC decisions, ou da ase does no include in o ma ion on he speci ic clinical indica ions o each FUBC, as no ed in he Limi a ions. Rega ding he a iable name, please no e ha in Table 1 we include he a iable “Blood cul u es” (now called “Follow-up blood cul u es pe o med”), and we s a e a Table oo ha FUBC we e pe o med a he disc e ion o a ending physician, and e e o ou analysis. The o he a iable is now called "Posi i e ollow-up blood cul u es among hose pe o med." Commen (5): The e iewe s we e c i ical o he ca he e ex ac ion measu e among he measu es de ining complica ed bac e emia, since he e is a consensus ha ca he e e aining is possible in he ini ial managemen o CONS CRABSI bac e emia. I ag ee ha since he e is no good e idence o his ecommenda ion his emains a ques ion. Howe e , his a iable is ele an only o he subg oup o pa ien s wi h a CVC. The a iable “ea ly (<48h) ca he e emo al” in model 2 mixes wo a iables: Yes – emo ed in CRBSI, No – no emo ed in CRABSI o no CRABSI. This is no a use ul p edic o clinically, because o he indi idual pa ien he sou ce o he bac e emia is known and i unknown, he a iable is no ele an . To app op ia ely analyse he associa ion be ween ca he e ex ac ion and mo ali y he analysis needs o be limi ed o pa ien s wi h a CVC. I ecommend e ising model 2 o include a 3- ca ego y a iable o CRBSI ex ac ed, CRBSI e ained and no CVC, o allow a mo e app op ia e e alua ion o hese ac o s in a model including all indi idual componen s o complica ed bac e emia. Al e na i ely, o concomi an ly, i would be o in e es o e alua e ca he e ex ac ion in he subg oup o pa ien s wi h a CVC, among whom isk ac o s o mo ali y need Page 4 o 108Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew o be e-e alua ed o selec he app op ia e con ounde s o he eg ession analysis including ca he e ex ac ion. Response: We ully ag ee; in ac , Model 2 was pe o med using pa ien s wi h CRBSI wi hou ea ly ca he e emo al as he e e ence ca ego y o ea ly ca he e emo al (es ima ion p o ided in he model) and o non-CRBSI (we did no include he es ima e o non-CRBSI due o he lack o associa ion, bu we ag ee i should be included). We now include his as Model 2a, and explained he s a i ica ion in he i le o Table 4. We also add essed you second sugges ion abou adding an analysis o pa ien s wi h a CVC and he e o e we added a Model 2b including only pa ien s wi h CRBSI. Commen (6): Please de ine he a iables en e ed in o he eg ession analyses, hose emo ed in he s epwise selec ion p ocess and all hose e ained wi h hei signi icance. Desc ibe he a iables ha we e no en e ed due o co ela ions. Please documen he numbe o pa ien s included in each eg ession model and he s a is ics o he model – i ness and calib a ion. See ou guidance o epo ing o mul i a iable logis ic eg ession models. h ps://www.clinicalmic obiologyandin ec ion.com/a icle/S1198-743X(19)30592-0/ ull ex . As a e iewe no ed he model including all componen s o complica ed bac e emia migh ha e highe dependen o independen a io han ecommended esul ing is o e i ing. Response: Thank you o you aluable eedback. We added in he ex ha a de ailed explana ion o models’ de elopmen is explained in he Supplemen a y ma e ial (please see sec ion "Mul i a iable eg ession model me hodology"). Commen (7): Please include dispe sion measu es (SD, min-max, IQR, CI) o all con inuous and ou comes and a io maeasu es epo ed in abs ac , ex and able and c ude numbe s (nume a o / denomina o ) o all pe cen ages. Response: We ha e included he dispe sion measu es (IQR and CI) whe e e hey we e missing in he abs ac , ex , and ables. Addi ionally, we speci ied all absolu e numbe s (nume a o /denomina o ) o each pe cen age. Commen (8): App op ia e empi ic he apy was no associa ed wi h mo ali y in you coho . You ollowing s a emen s in he discussion a e unsubs an ia ed: “Howe e , empi ic he apy should be ini ia ed i isk ac o s o ad e se ou comes a e iden i ied” and lines 244-247 “Addi ionally, nea ly hal o he empi ic he apies we e ine ec i e agains he isola ed s ain, unde sco ing he impo ance o s a ing ea men ha co e s me hicillin- esis an s ains when CoNS in ec ions a e suspec ed. This app oach is suppo ed by ou inding ha o e 70% o isola es in ou coho we e me hicillin- esis an , consis en wi h o he epo s.” Please e ise – I Page 5 o 108 Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew sugges dele ing hese pa s o he discussion. In addi ion, you no e o he i s ime in he discussion ha only 90% p pa ien s ecei ed a ge ed he apy once cul u e esul s we e a ailable – please include his a iable in he ables. Response: We ag ee wi h he Edi o . We ha e made he eques ed changes by dele ing he speci ied pa s o he discussion. Addi ionally, we ha e included he in o ma ion on a ge ed he apy in he ables ha was missing. Commen (9): Please check and e e ence lines 234-235 in he discussion on adhe ence o guidelines ela ing o ca he e ex ac ion. Mos people ollow he IDSA guidelines ha as no ed abo e allow e aining he ca he e in CONS CRBSI. Response: Following he Edi o ’s commen , we ha e comple ely e ised ha pa o he Discussion, ecognizing ha IDSA guidelines conside ca he e e en ion in CoNS CRBSI. Page 6 o 108Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew REVIEWER #1 COMMENTS Commen (1): In his s udy, Spanish in es iga o s we e looking a pa ien s wi h coagulase- nega i e s aphylococcus (CoNS) bac e emia in a p ospec i e s udy o bac e emia ha en olled a mul iple hospi als in 2017. They sough o look a p edic o s o mo ali y and p oposed an ope a ional de ini ion o complica ed bac e emia. Table 1 shows he cha ac e is ics o he 445 pa ien s included in he s udy. The au ho s de ined complica ed in ec ions acco ding o clinical c i e ia p e iously applied o S. au eus bac e emia plus hey included lack o ca he e emo al wi hin 48 hou s i CRBSI was p esen . 30 day mo ali y was highe in he complica ed bac e emia g oup. Table 4 shows mul i- a ia e analysis whe e ea ly ca he e emo al was associa ed wi h lowe mo ali y. O e all: The au ho s a e add essing an impo an opic, namely bloods eam in ec ions due o coagulase-nega i e s aphylococci. The main s eng hs o hei s udy a e he la ge sample size and mul i-cen e design. The main weakness is ha i is obse a ional wi h all he d awbacks such a s udy design en ails. Thus, al hough associa ions can be made, he clinical impac o his s udy is limi ed. Mo eo e , he au ho s ha e combined in ec ion/pa ien cha ac e is ics, such as pe sis en e e /p esence o in-dwelling de ice wi h p ac i ione decisions (i.e. ea ly ca he e emo al) which ul ima ely obscu es hei de ini ion o “complica ed bac e emia”. Response: Thank you o you commen . We acknowledge he signi ican limi a ions inhe en in obse a ional s udies. Howe e , we belie e ha ou s udy design, ea u ing p ospec i e da a collec ion, p o ides a aluable e lec ion o eal-wo ld p ac ices in managing coagulase-nega i e s aphylococcal bloods eam in ec ions. Rega ding he de ini ion o "complica ed bac e emia," we unde s and you conce n abou he in eg a ion o pa ien /in ec ion cha ac e is ics and he physician's decision ega ding ea ly ca he e emo al. We would like o emphasize ha ou de ini ion is no in ended o be pa hophysiologic bu a he p ac ical o clinical use. I lack o ca he e emo al impac s ou comes, we belie e ha conside ing i is use ul om a clinical pe spec i e. Commen (2): The au ho s could ha e ac ually de i ed isk ac o s o poo ou come a he han deciding on hem be o ehand and hen seeing i hey we e associa ed wi h mo ali y. This would ha e s eng hened hei de ini ion o “complica ed in ec ion”. Response: We app ecia e you insigh ul commen , which allows us o cla i y ou s udy's pu pose. Ac ually, ou objec i e was hypo hesis-based, speci ically aiming o assess whe he a p e-es ablished de ini ion o “complica ed bac e emia” was associa ed wi h an inc eased isk o 30-day mo ali y ins ead o jus explo ing which a iables a e associa ed wi h ou come. While iden i ying b oade isk ac o s o poo ou comes would indeed be a ele an app oach, i aligns wi h a di e en objec i e han ou s. By applying p ede ined c i e ia, we sough o e alua e hei associa ions wi h mo ali y, bo h collec i ely and independen ly, wi hin ou ope a ional de ini ion. Page 7 o 108 Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew To e lec his mo e clea ly, we ha e e ised he s udy objec i es in he In oduc ion as ollows: “This s udy aims o desc ibe a la ge coho o CoNS BSI cases, es whe he p ede ined c i e ia o complica ed bac e emia a e associa ed wi h 30-day mo ali y and iden i y o he ou come p edic o s.” Commen (3): The e a e s a is ical e o s, such as in Table 1 when he au ho s compa e uncomplica ed and complica ed in ec ion in e ms o whe he pa ien s ha e ea ly ca he e emo al. Gi en ha ea ly ca he e emo al is pa o he de ini ion, i is no su p ising ha he e is a s a is ically signi ican di e ence be ween he wo g oups. Gi en ha his manusc ip has complex s a is ics a i s hea , his kind o compa ison aises conce n o he alidi y o he o he indings. Response: The Re iewe is igh ; we dele ed he s a is ical compa isons in Table 1 dealing wi h ac o s which a e c i e ia o complica ed bac e emia, as pa ien s wi h uncomplica ed bac e emia canno mee hese c i e ia. We also ook his oppo uni y o echeck all he da a be o e esubmission. We also cla i y ha in he Table oo . Commen (4): On line 233 he au ho s s a e ha lack o ea ly emo al o a ca he e in CoNS CRBSI indica es poo adhe ence o CRBSI managemen s anda ds wi hou gi ing a e e ence – howe e , he 2009 IDSA guidelines (which a e mos cu en ) s a e ha is app op ia e o lea e a ca he e in place ini ially o CoNS CRBSI and see how pa ien s do. The as majo i y o pa ien s a e cu ed wi h his app oach (see PMID 19780661 which is no e e enced he e) which in u n sa es pa ien s om unde going unnecessa y ca he e change. Thus, I would sugges he au ho s empe his c i icism unless hey ha e o he da a o suppo i . Response: Thank you o you insigh ul commen ; we ully ag ee wi h you obse a ions. This conce n was also aised by he Edi o . We app ecia e you sugges ion ega ding he e e ence, which has helped us cla i y ou pe spec i e. Following you and he Edi o ’s commen , we ha e e ised in deep he Discussion sec ion dealing wi h his aspec . Commen (5): The au ho s should be ca e ul abou including ca he e e en ion in any model wi hou speci ically unde s anding why ca he e s we e e ained. Fo example, i a pa ien is dying o ano he p ocess, ca he e e en ion may simply e lec he unwillingness o he ea ing physicians o pu he pa ien h ough mo e p ocedu es. Response: Thank you o you commen . We ag ee ha unde s anding he speci ic easons o ca he e e en ion is impo an ; howe e , ou da ase does no include his in o ma ion. Despi e his limi a ion, he associa ion o ea ly ca he e emo al wi h mo ali y in ou coho emains Page 8 o 108Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew Commen (6): Majo poin s. Me hods. Lines 142-144: How many a iables wen in o he mul i a ia e Cox eg ession analysis? Was he numbe in p opo ion o 77 e en s/dea hs a day30? Please cla i y and commen . Response: Thank you o aising his poin , which was also highligh ed in Commen (6) om he Edi o . We hope ha ou modi ica ions ha e cla i ied his aspec . Commen (7): Majo poin . Line 144: In e ac ions we e e alua ed – could he au ho s please speci y his impo an s ep in he espec i e s eps. Response: In e ac ions be ween complica ed bac e emia and he o he a iables in he inal model we e assessed, as well as in e ac ions be ween he a iables and each indi idual c i e ion de ining complica ed bac e emia. No clinically o s a is ically signi ican in e ac ions we e obse ed. The explana ion can now be ead in he Sec ion “Me hods [3]. Mul i a iable eg ession model me hodology” in Supplemen a y ma e ial. Commen (8): Majo poin . Resul s. I ‘me as a ic complica ion’ occu ed in <3% o cases, could his c i e ion be omi ed? Response: Thank you o he commen . Since he de ini ion o complica ed bac e emia was p e-de ined, we decided o e ain all he c i e ia conside ed. To no e, 5 o he 7 pa ien s wi h me as a ic complica ions me o he c i e ia o "complica ed bac e emia": 1 was a ascula p os he ic de ice ca ie , 1 had an a icula de ice, 2 had pe sis en e e , and 1 had a CRBSI wi hou ea ly ca he e emo al. Only wo pa ien s did no mee any o he "complica ed c i e ia," and he de ini ion was solely a ibu ed o he p esence o me as a ic oci. We ha e added his explana ion in he Supplemen a y Ma e ial unde he sec ion "Me hods [2]: Jus i ica ion o he inclusion o “me as a ic complica ions” as a c i e ion o “complica ed bac e emia”." Commen (9): Mino poin . Gene al aspec s. The manusc ip could bene i om some language edi ing. Response: Thank you o you eedback. We ha e e ised he manusc ip o cla i y and language quali y o enhance i s o e all eadabili y. Commen (10): Mino poin . Abs ac and h oughou he manusc ip . Sugges o use ‘in ec i e’ endoca di is ins ead o ‘in ec ious’. Page 15 o 108 Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew Response: Thank you o he sugges ion. We ag ee and ha e made he modi ica ion o use "in ec i e" endoca di is ins ead o "in ec ious" h oughou he manusc ip and addi ional ma e ial. Commen (11): Mino poin . Line 103: which (in a ascula ) ca he e s we e included – please speci y. Response: Cul u es d awn om any ype o in a ascula ca he e could be sen o mic obiology o iden i ica ion o CRBSI, i submi ed alongside pe iphe al blood cul u es. This in o ma ion was added o he Supplemen a y Ma e ial unde he sec ion “S udy mic obiological and clinical inclusion c i e ia o CoNS bloods eam in ec ion episodes,” speci ically wi hin he mic obiological c i e ia. Commen (12): Mino poin . Me hods and Resul s. Please speci y which implan ed de ices we e included (so a ou lined in able, bu no in ex ). Response: Since his in o ma ion is in he Table, we hink is i su icien , also gi en he wo ds limi ; howe e , we a e open o ollowing any speci ic guidance om he Edi o on his ma e . Commen (13): Mino poin . Discussion. Line 214: please elabo a e o gi e e e ences o he ‘p e ious s udies’ men ioned. Response: The e e ence was indeed p o ided a he end o he sen ence, bu we will mo e i wi hin he sen ence o imp o ed cla i y. We would inally like o decla e he ollowing: •In Table 2, he c i e ion o “p ima y endoca di is” was omi ed in he p e ious e sion. While i was conside ed in he coun , i was inad e en ly excluded om he able. This has been co ec ed, and he c i e ion is now included in he e ised able. •In esponse o he au ho s' and e iewe s' commen s, he cu en e sion o he manusc ip is 2,593 wo ds, which sligh ly exceeds he wo d limi . Page 16 o 108Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 1 1Mo ali y p edic o s and de ini ion p oposal o complica ed coagulase-nega i e S aphylococcus 2bac e emia. A mul icen e p ospec i e coho s udy 3 4Au ho s 5Benede a Va isco1, Ped o Ma ía Ma ínez Pé ez-C espo2, Pila Re ama -Gen il3,4, Inmaculada López 6He nandez3,4, Mª Ca men Fa iñas-Ál a ez4,5, Isabel Fe nández-Na al6, Ma ía Te esa Pé ez- 7Rod íguez7, Ane Josune Goikoe xea Agui e8, Juan Manuel Sánchez-Cal o9, Luis Buzón Ma ín10, 8E a León-Jiménez2, Da id Vinuesa Ga cía11, José Ma ía Regue a-Iglesias12, Albe o Bahamonde- 9Ca asco13, Jona han Fe nández Suá ez14, Jesús Rod íguez-Baño3,4*, Luis Edua do López-Co és3,4*, 10 on behal o he PROBAC REIPI/GEIH-SEIMC/SAEI G oup 11 12 A ilia ions 13 1Depa men o Heal h Sciences, Clinic o In ec ious Diseases, Uni e si y o Milan, ASST San i Paolo 14 e Ca lo, Via A. Di Rudinì, 8, 20142, Milan, I aly. 15 2Unidad de En e medades In ecciosas y Mic obiología, Hospi al Uni e si a io de Valme, 41014, 16 Se illa, Spain. 17 3Unidad Clínica de En e medades In ecciosas y Mic obiología, Hospi al Uni e si a io Vi gen 18 Maca ena; Depa amen os de Medicina y Mic obiología, Facul ad de Medicina, Uni e sidad de 19 Se illa; Ins i u o de Biomedicina de Se illa (IBiS)/CSIC, 41009, Se ille, Spain. 20 4CIBERINFEC, Ins i u o de Salud Ca los III, 28029, Mad id, Spain. 21 5Unidad de En e medades In ecciosas, Hospi al Uni e si a io Ma qués de Valdecilla, Uni e sidad de 22 Can ab ia, IDIVAL, 39008, San ande , Spain. Page 17 o 108 Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 2 23 6Depa men o Clinical Mic obiology, Complejo Asis encial Uni e si a io de León (CAULE), 24071, 24 León, Spain. 25 7 In ec ious Diseases Uni , Depa men o In e nal Medicine, Complexo Hospi ala io Uni e si a io de 26 Vigo, 36312, Vigo, Spain; Ins i u o de In es igación Biomédica Galicia Su , Spain. 27 8 Unidad de En e medades In ecciosas, Hospi al Uni e si a io de C uces, 48903, Bizkaia, Spain. 28 9 Uni o In ec ious Diseases and Clinical Mic obiology. Je ez De La F on e a Uni e si y Hospi al, 29 11407, Je ez De La F on e a, Cádiz, Spain. 30 10 Unidad de Ges ión Clínica de En e medades In ecciosas, Hospi al Uni e si a io de Bu gos, 09006, 31 Bu gos, Spain. 32 11 Unidad Ges ión Clínica En e medades In ecciosas, Hospi al Uni e si a io Clínico San Cecilio, 33 18016, G anada, Spain. 34 12Se icio de En e medades In ecciosas, Hospi al Regional Uni e si a io de Málaga, IBIMA Málaga, 35 29010, Málaga, Spain. 36 13Depa amen o de Medicina In e na, Hospi al de El Bie zo, 24404, Pon e ada, Spain. 37 14Unidad de Mic obiología, Ins i u o de In es igación Sani a ia del P incipado de As u ias (ISPA), 38 Hospi al Uni e si a io Cen al de As u ias, 33011, O iedo, Spain. 39 *These au ho s con ibu ed equally as senio au ho s. 40 41 Co esponding au ho 42 Luis Edua do López-Co és 43 In ec ious Diseases and Mic obiology Uni 44 Hospi al Uni e si a io Vi gen Maca ena 45 Depa men o Medicine, Uni e si y o Se illa Page 18 o 108Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 3 46 Biomedicines Ins i u e o Se illa 47 41009 Se illa, Spain 48 Phone: +34 670 946 430 49 Email: [email p o ec ed] Page 19 o 108 Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 4 50 Keywo ds. Coagulase-nega i e s aphylococci; bloods eam in ec ions; complica ed bac e emia 51 A icle ype. O iginal a icle 52 Subjec sec ion. Bac e ial In ec ions; An imic obial S ewa dship 53 Abs ac wo d coun . 245 54 Tex wo d coun . 2593 55 Page 20 o 108Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 5 56 ABSTRACT 57 Objec i e. To explo e a de ini ion o complica ed coagulase-nega i e s aphylococci bloods eam 58 in ec ions (CoNS BSI), and o iden i y p edic o s o mo ali y. 59 60 Me hods. P ospec i e coho s udy conduc ed om Oc obe 2016 o Ma ch 2017 in 26 Spanish 61 hospi als. Complica ed CoNS BSI c i e ia included lack o ea ly ca he e emo al in ca he e - ela ed 62 cases, o eign indwelling implan , pe sis en bac e emia, e e ≥72 hou s on ac i e he apy, me as a ic 63 in ec ion o deep-sea ed ocus and in ec i e endoca di is. Independen p edic o s o 30-day mo ali y 64 we e e alua ed by Cox eg ession, and he impac o he de ini ion o complica ed bac e emia assessed. 65 66 Resul s. O e all, 445 CoNS BSI cases we e included; ca he e - ela ed in ec ions we e p edominan 67 (336/445, 75.5%). Complica ed bac e emia was iden i ied in 240/445 pa ien s (53.9%); 30-day 68 mo ali y in complica ed and uncomplica ed cases we e 53/240 (22.1%) and 24/205 (11.7%), 69 espec i ely (p=0.004). P edic o s o 30-day mo ali y iden i ied in he mul i a ia e analysis included 70 age (HR 1.03, 95%CI 1.01-1.05), ce eb o ascula disease (HR 2.58, 95%CI 1.45-4.58), 71 immunosupp essi e he apy (HR 2.16, 95%CI 1.22-3.84), SOFA sco e (HR 1.09, 95%CI 1.03-1.16), 72 and complica ed bac e emia (HR 2.14, 95%CI 1.29-3.53). A ca he e - ela ed sou ce o bac e emia was 73 ound o be p o ec i e (HR 0.49, 95%CI 0.30-0.80). When speci ic c i e ia o de ine complica ed 74 bac e emia we e included, e e ≥72h was associa ed wi h inc eased isk o dea h (HR 2.52, 95%CI 75 1.52-4.17) and ea ly ca he e emo al was p o ec i e (HR 0.47, 95%CI 0.26-0.83). 76 Page 21 o 108 Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 6 77 Conclusions. A high p opo ion o pa ien s p esen ed complica ed bac e emia acco ding o he 78 p oposed c i e ia; hese pa ien s had highe haza ds o mo ali y. O he mo ali y p edic o s we e 79 iden i ied. Fu he s udies would be needed o alida e he p oposed c i e ia. Page 22 o 108Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 7 80 INTRODUCTION 81 82 Coagulase-nega i e s aphylococci (CoNS) ha e adi ionally been ega ded as low- i ulence 83 o ganisms due o hei ole as skin commensals and equen con aminan s in mic obiological 84 samples. They a e now ecognized as a leading cause o nosocomial bloods eam in ec ions (BSI), 85 pa icula ly ca he e - ela ed (CRBSI), due o hei abili y o o m bio ilms.1 Despi e hei high 86 p e alence, esea ch in o he mo bidi y and mo ali y o CoNS bac e emia emains limi ed, wi h 87 epo ed mo ali y a es anging om 4% o 24%.2–7 Fu he mo e, ea men ecommenda ions ely 88 on limi ed e idence, e lec ing he challenge o dis inguishing ue in ec ions om con amina ion.8 89 Ne e heless, a combina ion o accu a e clinical e alua ion and in e p e a ion o mic obiological 90 indings (pa icula ly he numbe o posi i e blood cul u es)9 can help educe misdiagnoses. 91 In e ms o ea men , he In ec ious Diseases Socie y o Ame ica (IDSA) guidelines10 92 ecommend ea ing uncomplica ed CoNS CRBSI o 5-7 days i he ca he e is emo ed, and o 10- 93 14 days wi h an ibio ic lock he apy i he ca he e is e ained (le el BIII e idence). Howe e , o he 94 bes o ou knowledge, he impac o delayed ca he e emo al in high- isk pa ien s has no been 95 sys ema ically assessed. Addi ionally, c i e ia o complica ed bac e emia used in he guidelines 96 include suppu a i e h ombophlebi is, endoca di is, os eomyeli is, and me as a ic seeding, hough 97 hese c i e ia a e no speci ic o CoNS in ec ions.10 Whe he addi ional c i e ia, such as p olonged 98 e e , pe sis en bac e emia, and lack o sou ce con ol (conside ed o S aphylococcus au eus )11,12 99 a e associa ed wi h wo se ou comes in CoNS BSI emains uns udied. 100 This s udy aims o desc ibe a la ge coho o CoNS BSI cases, e alua e he associa ion 101 be ween p ede ined c i e ia o complica ed bac e emia and 30-day mo ali y, and iden i y o he 102 ou come p edic o s. Page 23 o 108 Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 8 103 104 METHODS 105 106 Design and se ing 107 This s udy is pa o he PROBAC p ojec , a p ospec i e coho s udy in ol ing pa ien s >14 108 yea s admi ed o 26 hospi als in Spain om Oc obe 1, 2016, o Ma ch 31, 2017, wi h clinically 109 signi ican BSI.13,14 The coho included 18 e ia y ca e and 8 communi y hospi als. 110 111 Inclusion and exclusion c i e ia 112 To ensu e he clinical signi icance o CoNS bloods eam in ec ions (BSI), episodes we e 113 included in he s udy i hey me bo h mic obiological and clinical p especi ied c i e ia (Me hods [1], 114 Supplemen a y Ma e ial). In summa y, mic obiological c i e ia equi ed he isola ion o he same 115 CoNS species wi h an iden ical suscep ibili y p o ile om a leas wo dis inc blood cul u e se s, wi h 116 addi ional c i e ia o ca he e - ela ed in ec ions.10 Clinical c i e ia included he p esence o a leas 117 wo sys emic in ec ion signs ( e e o hypo he mia, achyca dia, achypnoea, leukocy osis o 118 leukopenia). Addi ionally, o he causes o in ec ion we e easonably excluded; o pa ien s in whom 119 he sou ce o bac e emia was excep ional o CoNS (i.e., pneumonia, in a-abdominal in ec ion, 120 meningi is, o unknown sou ce in pa ien s wi hou ca he e s o p os he ic ma e ial), he opinion o a 121 second local in es iga o was sough . 122 The ollowing exclusion c i e ia we e applied: polymic obial BSI; in ec ions by 123 S aphylococcus lugdunensis ; 15 ecu en episodes o bac e emia, de ined by he isola ion o he same 124 species wi h clinical signs and symp oms consis en wi h in ec ion, wi hin h ee mon hs a e esolu ion 125 o he ini ial episode. Page 24 o 108Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 15 261 disease, and immunosupp essi e he apy we e linked o highe mo ali y, likely e lec ing pa ien 262 ail y, while he SOFA sco e indica ed acu e se e i y. Addi ionally, ca he e - ela ed in ec ions we e 263 associa ed wi h lowe mo ali y compa ed o in ec ions om o he sou ces. 264 In he absence o speci ic c i e ia o CoNS BSI, ou s udy used modi ied c i e ia o iginally 265 designed o S. au eus complica ed bac e emia. The inding ha complica ed bac e emia is 266 independen ly associa ed wi h highe mo ali y sugges s ha hese c i e ia migh be use ul o CoNS 267 BSI as well. Howe e , he sepa a e analysis in he second model, whe e no all c i e ia we e linked o 268 mo ali y, indica es he need o a ailo ed app oach o CoNS BSI in u u e esea ch. 269 FUBC we e pe o med in ewe han 30% o cases in ou coho . Al hough speci ic guidelines 270 a e lacking, ob aining hem in high- isk pa ien s wi h complica ed bac e emia may be bene icial, 271 especially gi en he obse ed associa ion be ween mo ali y and pe sis en e e , which could indica e 272 ongoing bac e emia. In con as , FUBC may be unnecessa y o low- isk cases, as sugges ed by Badia- 273 Cebada e al.23 Thei ial compa ing an ibio ics e sus no an ibio ics in low- isk CoNS CRBSI ound 274 only one case o pe sis en bac e emia in he no-an ibio ics g oup. Howe e , he small sample size due 275 o s udy in e up ion limi s he s eng h o hese indings. 276 Al hough ou s udy ound ea ly ca he e emo al o be p o ec i e agains mo ali y, i was 277 ca ied ou in ewe han 60% o CRBSI cases. IDSA guidelines conside he possibili y o ca he e 278 e en ion in uncomplica ed CoNS CRBSI cases.10 Howe e , despi e he ac ha some s udies sugges 279 ha ca he e e en ion in CoNS bloods eam in ec ions may no signi ican ly impac mo ali y, i 280 emains an independen isk ac o o ecu ence,24 while ecu ence is a e a e ca he e emo al.25 281 Ou inding ha ea ly ca he e emo al is independen ly associa ed wi h a p o ec i e e ec on 30-day 282 mo ali y sugges s he need o u he s udies o be e iden i y he subg oups o pa ien s in whom 283 e aining he ca he e is sa e. Page 31 o 108 Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 16 284 In ou s udy, abou one- hi d o pa ien s did no ecei e empi ic he apy a he onse o in ec ion; 285 howe e , app oxima ely 90% o hese pa ien s we e ini ia ed on a ge ed he apy once cul u e esul s 286 became a ailable. Omi ing empi ic he apy may be jus i ied i no isk ac o s a e p esen and sou ce 287 con ol (e.g. ca he e emo al) is achie ed, po en ially making subsequen a ge ed he apy 288 unnecessa y. 23,26 289 Ou s udy has limi a ions. Fi s , he absence o a compa a i e g oup es ic s ou abili y o 290 es ablish causal ela ionships and assess he p opo ion o mo ali y di ec ly a ibu able o CoNS BSI 291 e sus unde lying condi ions. Al hough a mino i y o pa ien s did no comple e he 30-day ollow-up, 292 sensi i i y analyses sugges he esul s a e obus . We acknowledge ha posi i e FUBC esul s likely 293 ep esen only a subse o pe sis en bac e emia cases, as some pa ien s wi hou FUBC may ha e had 294 unde ec ed posi i e esul s; al hough we e ospec i ely analyzed ac o s in luencing he decision o 295 pe o m FUBC, he da a lacks clea indica ions o each case. None heless, we belie e including his 296 a iable is impo an o highligh he need o u he s udies o cla i y FUBC indica ions in CoNS 297 bloods eam in ec ions. Managemen o pa ien s was pe o med acco ding o a ending physicians and 298 he e o e was no s anda dized; none heless, we belie e he da a e lec eal-wo ld clinical p ac ice. 299 Despi e e o s like in es iga o aining and da a moni o ing, some da a collec ion e o s may s ill 300 exis . Howe e , he s udy's s eng hs include a la ge sample size, and a mul icen e , p ospec i e design 301 conduc ed by in ec ious disease expe s in hospi als wi h bac e emia p og ams, which enhances he 302 c edibili y o ou indings. 303 In summa y, ou s udy iden i ies key mo ali y ac o s in CoNS BSI and emphasizes he need 304 o imp o ed isk s a i ica ion. The p oposed de ini ion o complica ed CoNS BSI needs alida ion 305 in u he s udies. 306 Page 32 o 108Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 17 307 AUTHOR CONTRIBUTIONS 308 309 Concep ualiza ion: LELC, JRB. Me hodology and s a is ical analysis: BV, LELC, JRB. 310 O iginal d a : BV. Re iew & edi ing: BV, LELC, JRB (all au ho s). Supe ision: JRB. P ojec 311 adminis a ion: LELC. Funding acquisi ion: LELC, JRB. All au ho s con ibu ed o he inal 312 manusc ip . 313 314 FUNDING 315 316 This esea ch was suppo ed by g an s om Plan Nacional de I+D+i 2013-2016, Ins i u o de 317 Salud Ca los III, Subdi ección Gene al de Redes y Cen os de In es igación Coope a i a, Minis e io 318 de Ciencia, Inno ación y Uni e sidades (PI16/01432); he Spanish Ne wo k o Resea ch in In ec ious 319 Diseases (REIPI) (RD16/0016/0001; RD16/ 0016/0008); and co- inanced by he Eu opean Regional 320 De elopmen Fund unde he 'A way o achie e Eu ope' Ope a ional P og am o In elligen G ow h 321 2014-2020. 322 323 TRANSPARENCY STATEMENT 324 325 PRG has se ed as a scien i ic ad iso o Shionogi and Ad anz, speake o Angelini, and 326 Mena ini. LELC has se ed as a scien i ic ad iso o No a is, speake o MSD, P ize , Angelini, 327 and ViiV, and aine o MSD and ViiV. The o he au ho s decla e no con lic s o in e es . Page 33 o 108 Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 18 328 REFERENCES 329 1. Heilmann C, Ziebuh W, Becke K. A e coagulase-nega i e s aphylococci i ulen ? Clinical 330 Mic obiology and In ec ion. 2019;25(9):1071-1080. doi:10.1016/j.cmi.2018.11.012 331 332 2. Molina J, Peñuela I, Lepe JA, e al. Mo ali y and hospi al s ay ela ed o coagulase-nega i e 333 S aphylococci bac e emia in non-c i ical pa ien s. Jou nal o In ec ion. 2013;66(2):155-162. 334 doi:10.1016/j.jin .2012.10.021 335 336 337 3. Pa k SY, Kwon KH, Chung JW, Huh HJ, Chae SL. Coagulase-nega i e s aphylococcal 338 bac e emia: isk ac o s o mo ali y and impac o ini ial app op ia e an imic obial he apy 339 on ou come. Eu opean Jou nal o Clinical Mic obiology and In ec ious Diseases. 340 2015;34(7):1395-1401. doi:10.1007/s10096-015-2364-3 341 342 4. Hen zien M, S ady C, Ve ne -Ga nie V, e al. P ognos ic ac o s associa ed wi h 30-day in- 343 hospi al mo ali y in coagulase-nega i e S aphylococcus bac e aemia: no impac o 344 ancomycin minimum inhibi o y concen a ion. In ec Dis. 2017;49(9):664-673. 345 doi:10.1080/23744235.2017.1323346 346 347 5. Cui J, Liang Z, Mo Z, Zhang J. The species dis ibu ion, an imic obial esis ance and isk 348 ac o s o poo ou come o coagulase-nega i e s aphylococci bac e aemia in China. 349 An imic ob Resis In ec Con ol. 2019;8(1):1-10. doi:10.1186/s13756-019-0523-5 350 351 6. Yamada K, Namikawa H, Fujimo o H, e al. Clinical cha ac e is ics o me hicillin- esis an 352 coagulase-nega i e s aphylococcal bac e emia in a e ia y hospi al. In e nal Medicine. 353 2017;56(7):781-785. doi:10.2169/in e nalmedicine.56.7715 354 355 7. Rosa RG, Dos San os RP, Goldani LZ. Mo ali y ela ed o coagulase-nega i e s aphylococcal 356 bac e emia in eb ile neu openia: A coho s udy. Canadian Jou nal o In ec ious Diseases 357 and Medical Mic obiology. 2014;25(1):14-18. doi:10.1155/2014/702621 358 359 8. Rahkonen M, Lu inen S, Koskela M, Hau ala T. T ue bac e emias caused by coagulase 360 nega i e S aphylococcus a e di icul o dis inguish om blood cul u e con aminan s. 361 Eu opean Jou nal o Clinical Mic obiology and In ec ious Diseases. 2012;31(10):2639-2644. 362 doi:10.1007/s10096-012-1607-9 363 364 9. Mi e S, Weins ein MP, Reime LG, Wilson ML, Relle LB. Rele ance o he numbe o 365 posi i e bo les in de e mining clinical signi icance o coagulase-nega i e s aphylococci in 366 blood cul u es. J Clin Mic obiol. 2001;39(9):3279-3281. doi:10.1128/JCM.39.9.3279- 367 3281.2001 368 369 10. Me mel LA, Allon M, Bouza E, e al. Clinical p ac ice guidelines o he diagnosis and 370 managemen o in a ascula ca he e - ela ed in ec ion: 2009 upda e by he In ec ious 371 Diseases Socie y o Ame ica. Clinical In ec ious Diseases. 2009;49(1):1-45. 372 doi:10.1086/599376 373 374 11. Jung N, Rieg S. Essen ials in he managemen o S. au eus bloods eam in ec ion. In ec ion. 375 2018;46(4):441-442. doi:10.1007/s15010-018-1130-8 376 Page 34 o 108Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 19 377 12. López-Co és LE, Del To o MD, Gál ez-Acebal J, e al. Impac o an e idence-based bundle 378 in e en ion in he quali y-o -ca e managemen and ou come o S aphylococcus au eus 379 bac e emia. Clinical In ec ious Diseases. 2013;57(9):1225-1233. doi:10.1093/cid/ci 499 380 381 13. Calo F, Re ama P, Pe ez-C espo PMM, e al. Ca he e - ela ed bloods eam in ec ions: 382 P edic i e ac o s o G am-nega i e bac e ia ae iology and 30 day mo ali y in a mul icen e 383 p ospec i e coho . Jou nal o An imic obial Chemo he apy. 2020;75(10):3056-3061. 384 doi:10.1093/jac/dkaa262 385 386 14. Mussa M, Ma ínez Pé ez-C espo PM, Lopez-Co es LE, e al. Risk Fac o s and P edic i e 387 Sco e o Bac e emic Bilia y T ac In ec ions Due o En e ococcus aecalis and En e ococcus 388 aecium: a Mul icen e Coho S udy om he PROBAC P ojec . Mic obiol Spec . 389 2022;10(4). doi:10.1128/spec um.00051-22 390 391 15. F ank KL, Del Pozo JL, Pa el R. F om clinical mic obiology o in ec ion pa hogenesis: How 392 da ing o be di e en wo ks o S aphylococcus lugdunensis. Clin Mic obiol Re . 393 2008;21(1):111-133. doi:10.1128/CMR.00036-07 394 395 16. Rod íguez Díaz JC, Guna Se ano M del R, La osa Esca ín N, Ma ín A iaza M. Diagnós ico 396 mic obiológico de la bac e iemia y la ungemia: hemocul i os y mé odos molecula es. 397 Sociedad Española de En e medades In ecciosas y Mic obiología Clínica. 2017. Accessed 398 No embe 6, 2024. 399 h ps://seimc.o g/con enidos/documen oscien i icos/p ocedimien osmic obiologia/seimc- 400 p ocedimien omic obiologia62.pd 401 402 17. Cha lson ME, Pompei P, Ales KL, MacKenzie CR. A new me hod o classi ying p ognos ic 403 como bidi y in longi udinal s udies: De elopmen and alida ion. J Ch onic Dis. 404 1987;40(5):373-383. doi:h ps://doi.o g/10.1016/0021-9681(87)90171-8. 405 406 18. F iedman ND, Kaye KS, S ou JE, e al. Heal h ca e-associa ed bloods eam in ec ions in 407 adul s: A eason o change he accep ed de ini ion o communi y-acqui ed in ec ions. Ann 408 In e n Med. 2002;137(10):791-797. doi:10.7326/0003-4819-137-10-200211190-00007 409 410 19. Hil M, Yu VL, Sha p J, Zu a le JJ, Ko ick JA, Mude RR. An ibio ic he apy o 411 Pseudomonas ae uginosa bac e emia: Ou come co ela ions in a p ospec i e s udy o 200 412 pa ien s. Am J Med. 1989;87(5):540-546. doi:10.1016/S0002-9343(89)80611-4 413 414 20. Fe nando SM, Rochwe g B, Seely AJE. Clinical implica ions o he hi d in e na ional 415 consensus de ini ions o sepsis and sep ic shock (Sepsis-3). JAMA. 2018;190(36):E1058- 416 E1059. doi:10.1503/cmaj.170149 417 418 21. Cha es F, Ga nacho-Mon e o J, del Pozo JL, e al. Diagnosis and ea men o ca he e - ela ed 419 bloods eam in ec ion: Clinical guidelines o he Spanish Socie y o In ec ious Diseases and 420 Clinical Mic obiology and (SEIMC) and he Spanish Socie y o Spanish Socie y o In ensi e 421 and C i ical Ca e Medicine a. Med In ensi a. 2018;42(1):5-36. 422 doi:10.1016/j.medin.2017.09.012 423 424 22. O ihuela-Ma ín J, Rod íguez-Núñez O, Mo a a L, e al. Pe o mance o di e en ial ime o 425 posi i i y as a ou ine diagnos ic es o ca he e - ela ed bloods eam in ec ions: a single- 426 cen e expe ience. Clinical Mic obiology and In ec ion. 2020;26(3):383.e1-383.e7. Page 35 o 108 Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 20 427 doi:10.1016/j.cmi.2019.07.001 428 429 23. Badia-Cebada L, Ca mezim J, Pé ez-Rod íguez MT, e al. Randomized Clinical T ial o he 430 Need o An ibio ic T ea men o Low-Risk Ca he e -Rela ed Bloods eam In ec ion Caused 431 by Coagulase-Nega i e S aphylococci. An ibio ics. 2023;12(5). 432 doi:10.3390/an ibio ics12050839 433 434 24. Raad I, Kassa R, Ghannam D, Cha a i AM, Hachem R, Jiang Y. Managemen o he ca he e 435 in documen ed ca he e - ela ed coagulase-nega i e s aphylococcal bac e emia: Remo e o 436 e ain? Clinical In ec ious Diseases. 2009;49(8):1187-1194. doi:10.1086/605694 437 438 25. Hebeisen U, Babouee Flu y B, A kinson A, Ma schall J, Bue i N. Ca he e - ela ed 439 bloods eam in ec ions due o coagulase-nega i e s aphylococci managed wi h ca he e 440 emo al: Recu ences a e a e. Am J In ec Con ol. 2020;48(7):837-839. 441 doi:10.1016/j.ajic.2019.10.013 442 443 26. Hebeisen UP, A kinson A, Ma schall J, Bue i N. Ca he e - ela ed bloods eam in ec ions wi h 444 coagulase-nega i e s aphylococci: A e an ibio ics necessa y i he ca he e is emo ed? 445 An imic ob Resis In ec Con ol. 2019;8(1):1-8. doi:10.1186/s13756-019-0474-x 446 Page 36 o 108Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 1 1Mo ali y p edic o s and de ini ion p oposal o complica ed coagulase-nega i e S aphylococcus 2bac e emia. A mul icen e p ospec i e coho s udy 3 4Au ho s 5Benede a Va isco1, Ped o Ma ía Ma ínez Pé ez-C espo2, Pila Re ama -Gen il3,4, Inmaculada López 6He nandez3,4, Mª Ca men Fa iñas-Ál a ez4,5, Isabel Fe nández-Na al6, Ma ía Te esa Pé ez- 7Rod íguez7, Ane Josune Goikoe xea Agui e8, Juan Manuel Sánchez-Cal o9, Luis Buzón Ma ín10, 8E a León-Jiménez2, Da id Vinuesa Ga cía11, José Ma ía Regue a-Iglesias12, Albe o Bahamonde- 9Ca asco13, Jona han Fe nández Suá ez14, Jesús Rod íguez-Baño3,4*, Luis Edua do López-Co és3,4*, 10 on behal o he PROBAC REIPI/GEIH-SEIMC/SAEI G oup 11 12 A ilia ions 13 1Depa men o Heal h Sciences, Clinic o In ec ious Diseases, Uni e si y o Milan, ASST San i Paolo 14 e Ca lo, Via A. Di Rudinì, 8, 20142, Milan, I aly. 15 2Unidad de En e medades In ecciosas y Mic obiología, Hospi al Uni e si a io de Valme, 41014, 16 Se illa, Spain. 17 3Unidad Clínica de En e medades In ecciosas y Mic obiología, Hospi al Uni e si a io Vi gen 18 Maca ena; Depa amen os de Medicina y Mic obiología, Facul ad de Medicina, Uni e sidad de 19 Se illa; Ins i u o de Biomedicina de Se illa (IBiS)/CSIC, 41009, Se ille, Spain. 20 4CIBERINFEC, Ins i u o de Salud Ca los III, 28029, Mad id, Spain. 21 5Unidad de En e medades In ecciosas, Hospi al Uni e si a io Ma qués de Valdecilla, Uni e sidad de 22 Can ab ia, IDIVAL, 39008, San ande , Spain. Page 37 o 108 Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 2 23 6Depa men o Clinical Mic obiology, Complejo Asis encial Uni e si a io de León (CAULE), 24071, 24 León, Spain. 25 7 In ec ious Diseases Uni , Depa men o In e nal Medicine, Complexo Hospi ala io Uni e si a io de 26 Vigo, 36312, Vigo, Spain; Ins i u o de In es igación Biomédica Galicia Su , Spain. 27 8 Unidad de En e medades In ecciosas, Hospi al Uni e si a io de C uces, 48903, Bizkaia, Spain. 28 9 Uni o In ec ious Diseases and Clinical Mic obiology. Je ez De La F on e a Uni e si y Hospi al, 29 11407, Je ez De La F on e a, Cádiz, Spain. 30 10 Unidad de Ges ión Clínica de En e medades In ecciosas, Hospi al Uni e si a io de Bu gos, 09006, 31 Bu gos, Spain. 32 11 Unidad Ges ión Clínica En e medades In ecciosas, Hospi al Uni e si a io Clínico San Cecilio, 33 18016, G anada, Spain. 34 12Se icio de En e medades In ecciosas, Hospi al Regional Uni e si a io de Málaga, IBIMA Málaga, 35 29010, Málaga, Spain. 36 13Depa amen o de Medicina In e na, Hospi al de El Bie zo, 24404, Pon e ada, Spain. 37 14Unidad de Mic obiología, Ins i u o de In es igación Sani a ia del P incipado de As u ias (ISPA), 38 Hospi al Uni e si a io Cen al de As u ias, 33011, O iedo, Spain. 39 *These au ho s con ibu ed equally as senio au ho s. 40 41 Co esponding au ho 42 Luis Edua do López-Co és 43 In ec ious Diseases and Mic obiology Uni 44 Hospi al Uni e si a io Vi gen Maca ena 45 Depa men o Medicine, Uni e si y o Se illa Page 38 o 108Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 3 46 Biomedicines Ins i u e o Se illa 47 41009 Se illa, Spain 48 Phone: +34 670 946 430 49 Email: [email p o ec ed] Page 39 o 108 Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 4 50 Keywo ds. Coagulase-nega i e s aphylococci; bloods eam in ec ions; complica ed bac e emia 51 A icle ype. O iginal a icle 52 Subjec sec ion. Bac e ial In ec ions; An imic obial S ewa dship 53 Abs ac wo d coun . 245 54 Tex wo d coun . 2593 55 Page 40 o 108Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 11 171 implan , pe sis en bac e emia, e e ≥72 hou s on ea men , me as a ic in ec ion o deep-sea ed ocus, 172 in ec ious in ec i e endoca di is ( hese we e p e iously applied o S. au eus bac e emia)11, o , in 173 CRBSI cases, absence o ca he e emo al wi hin 48 hou s. Da a quali y was moni o ed, wi h que ies 174 sen o local in es iga o s o inconsis encies o missing da a. 175 176 Da a analysis 177 Ca ego ical a iables we e p esen ed as coun s/pe cen ages, and con inuous a iables as 178 median wi h in e qua ile ange (IQR). Ca ego ical a iables we e compa ed using Chi-squa e o 179 Fishe exac es s, and con inuous a iables wi h S uden ’s - es s o Mann-Whi ney U es s. S a i ied 180 analyses we e pe o med o accoun o po en ial con ounde s and e ec modi ie s. 181 Su i al cu es o pa ien s wi h complica ed and uncomplica ed bac e emia we e plo ed and 182 compa ed using he log- ank es . Pa ien s wi hou a 30-day assessmen we e censo ed a hei las 183 ollow-up. Independen p edic o s o 30-day mo ali y we e iden i ied using Cox eg ession, wi h 184 a iables selec ed h ough a s epwise backwa d p ocedu e, including hose wi h a p- alue <0.1 in 185 uni a ia e analysis and hose o clinical o li e a u e-based ele ance. (Me hods [3], Supplemen a y 186 Ma e ial). In e ac ions we e e alua ed. 187 To assess he impac o complica ed bac e emia on mo ali y, we de eloped wo models: one 188 including he composi e a iable "complica ed bac e emia" and ano he including ins ead each 189 c i e ion o complica ed bac e emia indi idually. To assess he impac o complica ed bac e emia on 190 mo ali y, wo models we e de eloped: one wi h a composi e a iable o complica ed bac e emia and 191 ano he analyzing each c i e ion indi idually. 192 Sensi i i y analyses we e pe o med using logis ic eg ession o es he obus ness o indings, 193 accoun ing o 21 pa ien s censo ed be o e day 30. Assump ions included all 21 su i ing and andom Page 47 o 108 Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 12 194 impu a ion o wo dea hs in 10 i e a ions (Me hods [4], Supplemen a y ma e ial). Addi ionally, we 195 epea ed he mul i a iable models excluding pa ien s classi ied as ha ing a apidly a al unde lying 196 disease acco ding o he McCabe classi ica ion (i.e., conside ed o ha e a high p obabili y o dying 197 wi hin he nex h ee mon hs) o accoun o cases whe e a e minal condi ion migh p eclude ce ain 198 aspec s o managemen and in whom dea h would likely be p ima ily a ibu able o he unde lying 199 condi ion (Table S5). S a is ical analyses we e conduc ed using SPSS e sion 29.0.1.0 (SPSS Inc., 200 Chicago, IL, USA). 201 202 RESULTS 203 204 Cha ac e is ics o he s udy coho 205 The PROBAC coho included 6,313 episodes o bac e emia, wi h CoNS isola ed om blood 206 cul u es and no conside ed con aminan s in 481 cases. A e excluding 36 cases (28 polymic obial 207 and 8 due o S. lugdunensis ), 445 pa ien s (7.0% o he coho ) we e included in he analysis (Table 208 1). 209 Among hese pa ien s, he majo i y (268/445, 60.2%) we e male, wi h a median age o 67 yea s 210 (IQR 54-76). The mos common sou ce o in ec ion was ca he e - ela ed (336/445, 75.5%). A onse , 211 109/445 pa ien s (24.5%) had a SOFA sco e ≥2, and 327/445 isola es (73.5%) we e me hicillin- 212 esis an . FUBC Follow-up blood cul u es we e pe o med in 122/445 pa ien s (27.4%), e ealing 213 pe sis en bac e emia in 44 o hemse cases (36.1% o pa ien s wi h FUBCs), which ep esen s 9.8% 214 o he o al coho . Compa a i e analysis be ween pa ien s wi h and wi hou FUBC (Table S1) showed 215 ha pa ien s wi h endoca di is (9.8% s. 2.2%, espec i ely; p<0.001), ca diac p os he ic al es (7.4% 216 s. 2.2%, p=0.008), and pacemake s (6.6% s. 1.5%, p=0.005) mo e equen ly had FUBC. Ac i e Page 48 o 108Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 13 217 empi ic he apy was adminis e ed o 155 pa ien s (34.8%). Empi ic he apy was adminis e ed in 67.2% 218 o cases (299/445), and i was ac i e in i o in 151 (33.9%). 219 In-hospi al mo ali y was 14.8% (66/445), while 30-day mo ali y was 17.3% (77/445). Da a 220 o 21/445 pa ien s (4.7%) who we e discha ged ali e and we e no assessed a day 30 be o e day 30 221 we e censo ed a hei las ollow-up da e.. 222 223 Complica ed and uncomplica ed CoNS BSI 224 Among he 445 pa ien s, 240 (53.9%) me he c i e ia o complica ed BSI (Table 2). The mos 225 common issue was delayed ca he e emo al in CRBSI (141/240 pa ien s, 58.8%), ollowed by 226 pe sis en e e o ≥72 hou s (77/240 pa ien s, 32.1%) and he p esence o a p os he ic de ice (64/240 227 pa ien s, 26.7%). 228 Pa ien s wi h complica ed BSI had a highe p e alence o mode a e- o-se e e CKD (13.3% s. 229 7.3%, p=0.045) and we e mo e likely o be on immunosupp essi e he apy (18.3% s. 10.2%, 230 p=0.016). Sep ic shock was also mo e equen (10% s. 4.4%, p=0.024). Ca he e - ela ed in ec ions 231 we e mo e p e alen in complica ed cases (80.8% s. 69.3%, p=0.005), while an unknown bac e emia 232 sou ce was mo e common in uncomplica ed cases (18.1% s. 5%, p<0.001). Addi ionally, oxacillin 233 esis ance was highe among hose wi h complica ed BSI (80.4% s. 65.4%, p<0.001). 234 Mo ali y was signi ican ly highe in he complica ed g oup (22.1% s. 11.7%, p=0.004). The 235 p opo ion o censo ed pa ien s was simila be ween g oups (4.5% s. 4.8%, p=0.88). Pa ien s wi h 236 complica ed BSI had a longe hospi al s ay (median 17 days [IQR 7-34] s. 11 days [IQR 7-23], 237 p=0.07), bu he e we e no signi ican di e ences in ICU admission o ecu ence. Su i al cu es a e 238 shown in Figu e 1. 239 Page 49 o 108 Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 14 240 P edic o s o mo ali y in CoNS BSI cases 241 Bi a ia e analysis esul s a e in Table 3, and mul i a ia e esul s a e in Table 4. Iin mul i a ia e 242 model 1 (Table 4), which conside ed “complica ed BSI” as a single en i y, inc eased mo ali y was 243 associa ed wi h age (HR 1.03 pe yea , 95% CI 1.01-1.05), ce eb o ascula disease (HR 2.58, 95% CI 244 1.45-4.58), immunosupp essi e he apy (HR 2.16, 95% CI 1.22-3.84), SOFA sco e (HR 1.09 pe sco e 245 poin , 95% CI 1.03-1.16), and complica ed bac e emia (HR 2.14, 95% CI 1.29-3.53). Con e sely, a 246 ca he e sou ce o bac e emia was p o ec i e (HR 0.49, 95% CI 0.30-0.80) (Table 4). The AUROC o 247 he model was 0.75 (95%CI 0.70 – 0.81). 248 In models 2, he speci ic c i e ia o complica ed bac e emia we e included sepa a ely ins ead 249 o he composi e a iable. In model 2a, which included all pa ien s e alua ed each c i e ion o 250 “complica ed BSI” sepa a ely, ca he e emo al was ca ego ized as CRBSI wi hou ea ly ca he e 251 emo al ( e e ence), CRBSI wi h ea ly ca he e emo al, and non-CRBSI; , pe sis en e e (HR 2.52, 252 95% CI 1.52-4.17) was a isk ac o o mo ali y, while ea ly ca he e emo al in CRBSI was 253 p o ec i e (HR 0.47, 95% CI 0.26-0.83). The ca he e - ela ed o igin was no linked o mo ali y in his 254 model.In Model 2b, which included only he CRBSI cases, he same componen s we e ound o be 255 associa ed wi h mo ali y, wi h simila haza d a ios. The AUROC o models 2a and 2b we e 0.75 256 (95%CI 0.69–0.81) and 0.75 (95%CI 0.67–0.82), espec i ely. 257 Tables S21 and S32 show he bi a ia e analysis o in-hospi al mo ali y and he mul i a ia e 258 logis ic eg ession models. P edic o s o in-hospi al mo ali y we e consis en wi h hose o 30-day 259 mo ali y. Sensi i i y analysis, which impu ed wo dea hs o pa ien s discha ged ali e be o e day 30, 260 e ealed adjus ed odds a ios (aORs) o complica ed bac e emia anging om 1.90 o 2.06, wi h all 261 95% CIs abo e 1 (Table S43). Addi ionally, in he sensi i i y analysis excluding he 28 pa ien s wi h 262 a apidly a al unde lying disease, complica ed bac e emia emained associa ed wi h inc eased haza ds Page 50 o 108Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 15 263 o mo ali y (HR 2.34, 95%CI 1.37–3.99), while ca he e emo al wi h lowe mo ali y (HR 0.46, 264 95%CI 0.25–0.85) (Table S5). 265 266 DISCUSSION 267 268 This p ospec i e coho s udy analyzed pa ien demog aphics, clinical ea u es, complica ions, 269 and ou comes o CoNS BSI. A de ini ion o complica ed bac e emia was explo ed, consis en ly linked 270 oassocia ed wi h highe 30-day mo ali y, wi h wo o he de ining c i e ia showing indi idual 271 associa ion: pe sis en e e and delayed ca he e emo al. 272 To ou knowledge, his is he la ges coho ocused solely on CoNS BSI, unde sco ing he 273 ela i e lack o esea ch compa ed o he ex ensi e li e a u e on S. au eus BSI. The pa ien 274 cha ac e is ics align wi h hose ypical o ca he e - ela ed BSI, such as ad anced age and high 275 como bidi y a es. No ably, abou 25% o pa ien s had sepsis a onse . The 30-day all-cause mo ali y 276 a e was 17.3%, wi h p e ious s udies epo ing a wide ange o mo ali y a es om 4% o 24% 2–7, 277 likely due o a ia ions in pa ien popula ions and c i e ia o de ining CoNS bac e emia. 278 P e ious s udies ha e iden i ied isk ac o s o mo ali y including age, lack o ac i e an ibio ic 279 ea men , and como bidi ies like enal o li e disease. 2–5 In ou s udy, ad anced age, ce eb o ascula 280 disease, and immunosupp essi e he apy we e linked o highe mo ali y, likely e lec ing pa ien 281 ail y, while he SOFA sco e indica ed acu e se e i y. Addi ionally, ca he e - ela ed in ec ions we e 282 associa ed wi h lowe mo ali y compa ed o in ec ions om o he sou ces.2–5 283 In he absence o speci ic c i e ia o CoNS BSI, ou s udy used modi ied c i e ia o iginally 284 designed o S. au eus complica ed bac e emia. The inding ha complica ed bac e emia is Page 51 o 108 Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 16 285 independen ly associa ed wi h highe mo ali y sugges s ha hese c i e ia migh be use ul o CoNS 286 BSI as well. Howe e , he sepa a e analysis in he second model, whe e no all c i e ia we e linked o 287 mo ali y, indica es he need o a ailo ed app oach o CoNS BSI in u u e esea ch. 288 Follow-up blood cul u esFUBC we e pe o med in ewe han 30% o cases in ou coho . 289 Al hough speci ic guidelines a e lacking, ob aining hem in conside ing ollow-up blood cul u es o 290 high- isk pa ien s wi h complica ed bac e emia may be bene icial, especially gi en he obse ed 291 associa ion be ween mo ali y and pe sis en e e , which could indica e ongoing bac e emia. In 292 con as , FUBC ollow-up blood cul u es may be unnecessa y o low- isk cases, as sugges ed by 293 Badia-Cebada e al.23 Thei ial compa ing an ibio ics e sus no an ibio ics in low- isk CoNS CRBSI 294 ound only one case o pe sis en bac e emia in he no-an ibio ics g oup. Howe e , he small sample 295 size due o s udy in e up ion limi s he s eng h o hese indings. 296 Al hough ou s udy ound ea ly ca he e emo al o be p o ec i e agains mo ali y, i was 297 ca ied ou in ewe han 60% o CRBSI cases. IDSA guidelines conside he possibili y o ca he e 298 e en ion in uncomplica ed CoNS CRBSI cases.10 Howe e , despi e he ac ha some s udies sugges 299 ha ca he e e en ion in CoNS bloods eam in ec ions may no signi ican ly impac mo ali y, i 300 emains an independen isk ac o o ecu ence,24 while ecu ence is a e a e ca he e emo al.25 301 Ou inding ha ea ly ca he e emo al is independen ly associa ed wi h a p o ec i e e ec on 30-day 302 mo ali y sugges s he need o u he s udies o be e iden i y he subg oups o pa ien s in whom Page 52 o 108Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 17 303 e aining he ca he e is sa e. This indica es poo adhe ence o CRBSI managemen s anda ds, e en in 304 hospi als wi h dedica ed bac e emia p og ams, whe e ea ly ca he e emo al is c ucial. In cases wi hou 305 isk ac o s o complica ions, emo ing he in ec ed ca he e alone migh be su icien and could 306 po en ially educe he need o an ibio ics. Howe e , suppo ing da a o his app oach emains limi ed 307 In ou s udy, abou one- hi d o pa ien s did no ecei e empi ic he apy a he onse o in ec ion 308 onse ; howe e , hough app oxima ely 90% o hese pa ien s we e s a ed ini ia ed on on a ge ed 309 he apy once cul u e esul s we e became a ailable. Omi ing empi ic he apy may be jus i ied i no 310 isk ac o s a e p esen and sou ce con ol (e.g. ca he e emo al) is achie ed, po en ially making 311 subsequen a ge ed he apy unnecessa y , as p e iously men ioned. 23,2625–27 Howe e , empi ic he apy 312 should be ini ia ed i isk ac o s o ad e se ou comes a e iden i ied. These indings highligh he need 313 o ca e ul ini ial assessmen and isk s a i ica ion. 314 Addi ionally, nea ly hal o he empi ic he apies we e ine ec i e agains he isola ed s ain, 315 unde sco ing he impo ance o s a ing ea men ha co e s me hicillin- esis an s ains when CoNS 316 in ec ions a e suspec ed. This app oach is suppo ed by ou inding ha o e 70% o isola es in ou 317 coho we e me hicillin- esis an , consis en wi h o he epo s. 318 Ou s udy has limi a ions. Fi s , he absence o a compa a i e g oup es ic s ou abili y o 319 es ablish causal ela ionships and assess he p opo ion o mo ali y di ec ly a ibu able o CoNS BSI 320 e sus unde lying condi ions. Al hough a mino i y o pa ien s did no comple e he 30-day ollow-up, 321 sensi i i y analyses sugges he esul s a e obus . We acknowledge ha posi i e FUBC esul s likely Page 53 o 108 Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 18 322 ep esen only a subse o pe sis en bac e emia cases, as some pa ien s wi hou FUBC may ha e had 323 unde ec ed posi i e esul s; al hough we e ospec i ely analyzed ac o s in luencing he decision o 324 pe o m FUBC, he da a lacks clea indica ions o each case. None heless, we belie e including his 325 a iable is impo an o highligh he need o u he s udies o cla i y FUBC indica ions in CoNS 326 bloods eam in ec ions. Managemen o pa ien s was pe o med acco ding o a ending physicians and 327 he e o e wasa no s anda dized; none heless, we belie e he da a e lec eal-wo ld clinical p ac ice. 328 Despi e e o s like in es iga o aining and da a moni o ing, some da a collec ion e o s may s ill 329 exis . Howe e , he s udy's s eng hs include a la ge sample size, and a mul icen e , p ospec i e design 330 conduc ed by in ec ious disease expe s in hospi als wi h bac e emia p og ams, which enhances he 331 c edibili y o ou indings. 332 In summa y, ou s udy iden i ies key mo ali y ac o s in CoNS BSI and emphasizes he need 333 o imp o ed isk s a i ica ion. The p oposed de ini ion o complica ed CoNS BSI needs alida ion 334 in u he s udies. 335 336 AUTHOR CONTRIBUTIONS 337 338 Concep ualiza ion: LELC, JRB. Me hodology and s a is ical analysis: BV, LELC, JRB. O iginal d a : 339 BV. Re iew & edi ing: BV, LELC, JRB (all au ho s). Supe ision: JRB. P ojec adminis a ion: 340 LELC. Funding acquisi ion: LELC, JRB. All au ho s con ibu ed o he inal manusc ip . 341 342 FUNDING 343 Page 54 o 108Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 19 344 This esea ch was suppo ed by g an s om Plan Nacional de I+D+i 2013-2016, Ins i u o de Salud 345 Ca los III, Subdi ección Gene al de Redes y Cen os de In es igación Coope a i a, Minis e io de 346 Ciencia, Inno ación y Uni e sidades (PI16/01432); he Spanish Ne wo k o Resea ch in In ec ious 347 Diseases (REIPI) (RD16/0016/0001; RD16/ 0016/0008); and co- inanced by he Eu opean Regional 348 De elopmen Fund unde he 'A way o achie e Eu ope' Ope a ional P og am o In elligen G ow h 349 2014-2020. 350 351 TRANSPARENCY STATEMENT 352 353 PRG has se ed as a scien i ic ad iso o Shionogi and Ad anz, speake o Angelini, and Mena ini. 354 LELC has se ed as a scien i ic ad iso o No a is, speake o MSD, P ize , Angelini, and ViiV, 355 and aine o MSD and ViiV. The o he au ho s decla e no con lic s o in e es . Page 55 o 108 Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 20 356 REFERENCES 357 1. Heilmann C, Ziebuh W, Becke K. A e coagulase-nega i e s aphylococci i ulen ? Clinical 358 Mic obiology and In ec ion. 2019;25(9):1071-1080. doi:10.1016/j.cmi.2018.11.012 359 360 2. Molina J, Peñuela I, Lepe JA, e al. Mo ali y and hospi al s ay ela ed o coagulase-nega i e 361 S aphylococci bac e emia in non-c i ical pa ien s. Jou nal o In ec ion. 2013;66(2):155-162. 362 doi:10.1016/j.jin .2012.10.021 363 364 3. Pa k SY, Kwon KH, Chung JW, Huh HJ, Chae SL. Coagulase-nega i e s aphylococcal 365 bac e emia: isk ac o s o mo ali y and impac o ini ial app op ia e an imic obial he apy 366 on ou come. Eu opean Jou nal o Clinical Mic obiology and In ec ious Diseases. 367 2015;34(7):1395-1401. doi:10.1007/s10096-015-2364-3 368 369 4. Hen zien M, S ady C, Ve ne -Ga nie V, e al. P ognos ic ac o s associa ed wi h 30-day in- 370 hospi al mo ali y in coagulase-nega i e S aphylococcus bac e aemia: no impac o 371 ancomycin minimum inhibi o y concen a ion. In ec Dis. 2017;49(9):664-673. 372 doi:10.1080/23744235.2017.1323346 373 374 5. Cui J, Liang Z, Mo Z, Zhang J. The species dis ibu ion, an imic obial esis ance and isk 375 ac o s o poo ou come o coagulase-nega i e s aphylococci bac e aemia in China. 376 An imic ob Resis In ec Con ol. 2019;8(1):1-10. doi:10.1186/s13756-019-0523-5 377 378 6. Yamada K, Namikawa H, Fujimo o H, e al. Clinical cha ac e is ics o me hicillin- esis an 379 coagulase-nega i e s aphylococcal bac e emia in a e ia y hospi al. In e nal Medicine. 380 2017;56(7):781-785. doi:10.2169/in e nalmedicine.56.7715 381 382 7. Rosa RG, Dos San os RP, Goldani LZ. Mo ali y ela ed o coagulase-nega i e s aphylococcal 383 bac e emia in eb ile neu openia: A coho s udy. Canadian Jou nal o In ec ious Diseases 384 and Medical Mic obiology. 2014;25(1):14-18. doi:10.1155/2014/702621 385 386 8. Rahkonen M, Lu inen S, Koskela M, Hau ala T. T ue bac e emias caused by coagulase 387 nega i e S aphylococcus a e di icul o dis inguish om blood cul u e con aminan s. 388 Eu opean Jou nal o Clinical Mic obiology and In ec ious Diseases. 2012;31(10):2639-2644. 389 doi:10.1007/s10096-012-1607-9 390 391 9. Mi e S, Weins ein MP, Reime LG, Wilson ML, Relle LB. Rele ance o he numbe o 392 posi i e bo les in de e mining clinical signi icance o coagulase-nega i e s aphylococci in 393 blood cul u es. J Clin Mic obiol. 2001;39(9):3279-3281. doi:10.1128/JCM.39.9.3279- 394 3281.2001 395 396 10. Me mel LA, Allon M, Bouza E, e al. Clinical p ac ice guidelines o he diagnosis and 397 managemen o in a ascula ca he e - ela ed in ec ion: 2009 upda e by he In ec ious 398 Diseases Socie y o Ame ica. Clinical In ec ious Diseases. 2009;49(1):1-45. 399 doi:10.1086/599376 400 401 11. Jung N, Rieg S. Essen ials in he managemen o S. au eus bloods eam in ec ion. In ec ion. 402 2018;46(4):441-442. doi:10.1007/s15010-018-1130-8 403 Page 56 o 108Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew Table 1. Baseline pa ien and in ec ion cha ac e is ics wi h bi a ia e analysis o complica ed e sus uncomplica ed bac e emia. Va iables N (%) excep whe e speci ied All pa ien s N = 445 Uncomplica ed N = 205 (46.1%) Complica ed N = 240 (53.9%) p - alue1 Age, median (IQR) 67 (54 - 76) 67 (53 – 76) 67 (54 – 76) 0.950 Male gende 268 (60.2) 124 (60.5) 144 (60) 0.766 Age-adjus ed Cha lson, median (IQR) 4 (2 - 6) 4 (2 - 6) 4 (2 - 7) 0.281 Age-adjus ed Cha lson >3 261 (58.7) 118 (57.6) 143 (59.6) 0.666 Como bidi ies and condi ions Diabe es melli us 65 (14.6) 31 (15.1) 34 (14.2) 0.776 Ch onic li e disease 39 (8.8) 23 (11.2) 16 (6.7) 0.177 Ch onic kidney disease, mode a e-se e e 47 (10.6) 15 (7.3) 32 (13.3) 0.045 Hemodialysis 9 (2.0) 1 (0.5) 8 (3.3) 0.077 Ce eb o ascula disease 50 (11.2) 25 (12.2) 25 (10.4) 0.554 COPD 40 (9.0) 18 (8.8) 22 (9.2) 0.887 Ch onic hea ailu e 37 (8.3) 16 (7.8) 21 (8.8) 0.719 Solid cance 115 (25.8) 46 (22.4) 69 (28.8) 0.130 Immunosupp essi e he apy 65 (14.6) 21 (10.2) 44 (18.3) 0.016 Recen exposu es (p e ious mon h) An ibio ic he apy 239 (53.7) 106 (51.7) 133 (55.4) 0.434 Su ge y 124 (27.9) 61 (29.8) 63 (26.3) 0.411 Pa en e al nu i ion 97 (21.8) 48 (23.4) 49 (20.4) 0.445 In ec ion acquisi ion Communi y 12 (2.7) 5 (2.4) 7 (2.9) 0.757 Heal hca e-associa ed 69 (15.5) 28 (13.7) 41 (17) 0.320 Nosocomial 364 (81.8) 171 (83.4) 193 (80.4) 0.354 P e ious leng h o hospi aliza ion in days, median (IQR) 15 (9 - 27) 12 (5 – 21) 12 (3 – 24) 0.891 Hospi al depa men o admission a onse Eme gency 3 (0.7) 2 (1) 1 (0.4) 0.473 In ensi e ca e 58 (13) 24 (11.7) 34 (14.2) 0.442 Page 63 o 108 Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew Su ge y 123 (27.6) 69 (33.7) 54 (22.5) 0.009 Medicine 241 (54.2) 100 (48.8) 141 (58.8) 0.035 O he s 20 (4.5) 10 (4.9) 10 (4.2) 0.718 Vascula access de ices PVC 229 (51.5) 113 (55.1) 116 (48.3) 0.153 Sho - e m CVC 229 (51.5) 102 (49.8) 127 (52.9) 0.506 PICC 74 (16.6) 27 (13.2) 47 (19.6) 0.070 Tunneled CVC 12 (2.7) 0 12 (5) - Po -a-Ca h 6 (1.4) 0 6 (2.5) - O he de ices Pacemake 13 (2.9) 0 13 (5.4) - Vascula p os he ic de ice 9 (2) 0 9 (3.8) - P os he ic join 15 (3.4) 0 15 (6.3) - Hea al ula p os hesis 16 (3.6) 0 16 (6.7) - Se e i y a onse SOFA sco e, median (IQR) 3 (1 - 6) 3 (1 - 5) 3 (1 - 6) 0.533 SOFA sco e ≥2 109 (24.5) 43 (21) 66 (27.5) 0.337 Pi sco e, median (IQR) 0 (0-3) 1 (0 - 2) 1 (0 - 3) 0.828 Pi sco e >3 87 (19.6) 39 (19) 48 (20) 0.815 Vasop esso s suppo 27 (6.1) 8 (3.9) 19 (7.9) 0.087 Sep ic shock 33 (7.4) 9 (4.4) 24 (10) 0.024 Sou ce o in ec ion Ca he e - ela ed 336 (75.5) 142 (69.3) 194 (80.8) 0.005 Unknown 49 (11.0) 37 (18.1) 12 (5) <0.001 Endoca di is 19 (4.3) 0 19 (7.9) - Skin and so issues 14 (3.2) 9 (4.4) 5 (2.1) 0.165 Abdominal 7 (1.6) 4 (2) 3 (1.3) 0.533 Bone and join s 6 (1.4) 4 (2) 2 (0.8) 0.308 O he s 14 (3.2) 9 (4.4) 5 (2.1) 0.264 CoNS species S aphylococcus epide midis 332 (74.6) 145 (70.7) 187 (77.9) 0.083 S aphylococcus hominis 60 (13.5) 32 (15.6) 28 (11.7) 0.225 Page 64 o 108Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew S aphylococcus haemoly icus 30 (6.7) 16 (7.8) 14 (5.8) 0.408 S aphylococcus capi is 16 (3.6) 8 (3.9) 8 (3.3) 0.748 S aphylococcus wa ne i 3 (0.7) 1 (0.5) 2 (0.8) 0.657 S aphylococcus schlei e i 2 (0.5) 1 (0.5) 1 (0.4) 0.279 S aphylococcus simulans 1 (0.2) 1 (0.5) 0 0.911 S aphylococcus au icula is 1 (0.2) 1 (0.5) 0 0.279 Oxacillin esis ance 327 (73.5) 134 (65.4) 193 (80.4) <0.001 Follow-up blood cul u es pe o med2 122 (27.4) 56 (27.3) 66 (27.5) 0.873 Posi i e ollow-up blood cul u es among hose pe o med 44/122 (36.1) 0/56 (0) 44/66 (66.7) - Sou ce con ol (only CRBSI) Ea ly (<48 h) ca he e emo al 195/336 (58) 142/142 (100) 53/194 (27.3) - Days un il ca he e emo al, median (IQR) 3 (2 – 8) 1 (0 – 3) 2 (2 – 7) 0.132 An imic obial ea men Empi ic an ibio ic he apy, yes 299 (67.2) 134 (65.4) 165 (68.8) 0.449 Ac i e empi ic he apy, yes 151/299 (50.5) 76/134 (56.7) 75/165 (45.4) 0.060 Empi ic an ibio ic he apy, no 146 (32.8) 71 (34.6) 75 (31.3) 0.714 Empi ic no, a ge ed yes 129/146 (88.4) 59/71 (83.1) 70/75 (93.3) 0.054 Du a ion in days, median (IQR) 10 (7-17) 10 (6 – 16) 12 (7 – 18) 0.002 Ou comes 30-day mo ali y 77 (17.3) 24 (11.7) 53 (22.1) 0.004 ICU admission 90 (20.2) 40 (19.5) 50 (20.8) 0.999 Leng h o hospi al s ay a e BSI in days, median (IQR) 13 (7 – 28) 11 (7 – 23) 17 (7 – 34) 0.079 Recu ence 8 (1.8) 3 (1.5) 5 (2.1) 0.459 COPD: ch onic obs uc i e pulmona y disease; PVC: pe iphe al enous ca he e ; CVC: cen al enous ca he e ; PICC: pe iphe ally inse ed cen al ca he e ; CRBSI: ca he e - ela ed bloods eam in ec ion; CoNS: coagulase-nega i e s aphylococci; ICU: in ensi e ca e uni ; BSI: bloods eam in ec ion Page 65 o 108 Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 1 S a is ical analyses we e no pe o med o a iables ha a e c i e ia o complica ed bac e emia; hey a e included solely o p o ide da a o uncomplica ed bac e emia cases. 2 Follow-up blood cul u es we e pe o med a he disc e ion o he a ending physician, wi h a speci ic analysis o pa ien cha ac e is ics associa ed wi h ob aining hem p o ided in he Supplemen a y Ma e ial. Page 66 o 108Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew Table 2. Complica ed bac e emia. C i e ia o complica ed bac e emia among he 240 pa ien s wi h episodes classi ied as such. Pa ien s may ha e had mo e han one c i e ion. C i e ia o complica ed bac e emia N (%) Absence o ea ly (<48h) ca he e emo al in CRSBI 141 (58.8) Pe sis en e e ≥72h 77 (32.1) Implan ed de ice 64 (26.7) Pe sis en bac e emia 44 (18.3) Me as a ic complica ions 7 (2.9) P ima y endoca di is 19 (7.9) Page 67 o 108 Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew Table 2. Complica ed bac e emia. C i e ia o complica ed bac e emia among he 240 pa ien s wi h episodes classi ied as such. Pa ien s may ha e had mo e han one c i e ion. C i e ia o complica ed bac e emia N (%) Absence o ea ly (<48h) ca he e emo al in CRSBI 141 (58.8) Pe sis en e e ≥72h 77 (32.1) Implan ed de ice 64 (26.7) Pe sis en bac e emia 44 (18.3) Me as a ic complica ions 7 (2.9) P ima y endoca di is 19 (7.9) Page 68 o 108Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew Table 3. Bi a ia e analysis o he associa ion o di e en a iables wi h 30-day mo ali y. Va iables N (%) excep whe e speci ied Ali e N = 368 (82.7%) Dead N = 77 (17.3%) HR 95%CI p - alue Age, median (IQR) 66 (53 – 75) 75 (62 – 81) 1.03 1.01 – 1.04 0.001 Male gende 222 (60.3) 46 (59.7) 1.02 0.64 – 1.61 0.943 Age-adjus ed Cha lson, median (IQR) 4 (2 - 6) 2 (1 - 5) 1.09 1.01 – 1.18 0.039 Age-adjus ed Cha lson >3 208 (56.5) 53 (68.8) 1.62 0.99 – 2.65 0.053 Como bidi ies and condi ions Diabe es melli us 55 (15) 10 (13) 0.83 0.43 – 1.61 0.575 Ch onic li e disease 33 (9) 6 (7.8) 0.86 0.38 – 1.99 0.732 Ch onic kidney disease, mode a e-se e e 34 (9.2) 13 (16.9) 1.94 1.07 – 3.53 0.029 Hemodialysis 7 (1.9) 2 (2.6) 1.13 0.26 – 4.79 0.874 Ce eb o ascula disease 34 (9.2) 16 (20.8) 2.22 1.28 – 3.85 0.005 COPD 31 (8.4) 9 (11.7) 1.42 0.71 – 2.84 0.326 Ch onic hea ailu e 25 (6.8) 12 (15.6) 2.25 1.21 – 4.16 0.010 Solid cance 91 (24.7) 24 (31.2) 1.31 0.81 – 2.13 0.269 Immunosupp essi e he apy 48 (13) 17 (22.1) 1.70 0.99 – 2.92 0.054 Recen exposu es (p e ious mon h) An ibio ic he apy 191 (51.9) 48 (62.3) 1.37 0.86 – 2.18 0.188 Su ge y 107 (29.1) 17 (22.1) 0.68 0.39 – 1.17 0.167 Pa en e al nu i ion 82 (22.3) 15 (19.5) 0.82 0.47 – 1.45 0.497 In ec ion acquisi ion Communi y 11 (3) 1 (1.3) 0.45 0.06 – 3.21 0.422 Heal hca e-associa ed 51 (13.9) 18 (23.4) 1.80 1.06 – 3.06 0.029 Nosocomial 305 (82.9) 58 (75.3) 0.65 0.39 – 1.09 0.107 P e ious leng h o hospi aliza ion in days, median (IQR) 12 (5 – 23) 11 (3 – 24) 1.01 0.99 – 1.01 0.492 Hospi al depa men o admission a onse Eme gency 3 (0.8) 0 - - 0.606 In ensi e ca e 41 (11.1) 17 (22.1) 2.03 1.18 – 3.48 0.010 Page 69 o 108 Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew Su ge y 111 (30.2) 12 (15.6) 0.45 0.25 – 0.84 0.012 Medicine 195 (53) 46 (59.7) 1.31 0.83 – 2.08 0.246 O he s 18 (4.9) 2 (2.6) 0.35 0.05 – 2.51 0.295 Vascula access de ices PVC 184 (50) 45 (58.4) 1.33 0.84 – 2.10 0.223 Sho - e m CVC 191 (51.9) 38 (49.4) 0.87 0.55 – 1.36 0.530 PICC 63 (17.1) 11 (14.3) 0.79 0.42 – 1.51 0.490 Tunneled CVC 8 (2.2) 4 (5.2) 2.22 0.81 – 6.09 0.119 Po -a-Ca h 4 (1.1) 3 (3.9) 3.15 0.99 – 10.01 0.051 O he de ices Pacemake 8 (2.2) 5 (6.5) 3.18 1.29 – 7.89 0.012 Vascula p os he ic de ice 6 (1.6) 3 (3.9) 2.26 0.71 – 7.16 0.168 P os he ic join 11 (3) 4 (5.2) 1.60 0.59 – 4.38 0.359 Hea al ula p os hesis 11 (3) 5 (6.5) 2.05 0.83 – 5.07 0.122 Se e i y a onse SOFA sco e, median (IQR) 3 (1 - 5) 4 (1 - 8) 1.08 1.01 – 1.14 0.015 SOFA sco e ≥2 80 (21.7) 29 (37.7) 1.96 1.24 – 3.12 0.004 Pi sco e, median (IQR) 1 (0 - 3) 0 (0 – 2) 0.99 0.91 – 1.09 0.893 Pi sco e >3 74 (20.1) 13 (16.9) 0.86 0.47 – 1.56 0.619 Vasop esso s suppo 16 (4.4) 11 (14.3) 2.29 0.97 – 5.39 0.059 Sep ic shock 16 (4.4) 17 (22.1) 5.09 2.96 – 8.75 <0.001 Sou ce o in ec ion Ca he e - ela ed 287 (78) 49 (63.6) 0.472 0.29 – 0.75 0.002 Unknown 40 (10.9) 9 (11.7) 1.38 0.69 – 2.76 0.369 P ima y eEndoca di is 13 (3.5) 6 (7.8) 2.03 0.88 – 4.66 0.097 Skin and so issues 10 (2.7) 4 (5.2) 1.86 0.68 – 5.08 0.228 Abdominal 6 (1.6) 1 (1.3) 0.78 0.11 – 5.61 0.805 Bone and join s 5 (1.4) 1 (1.3) 0.99 0.14 – 7.18 0.999 O he s 1 (0.3) 0 - - 0.766 CoNS species S aphylococcus epide midis 278 (75.5) 54 (70.1) 0.75 0.46 – 1.22 0.246 S aphylococcus hominis 48 (13) 12 (15.6) 1.21 0.65 – 2.24 0.545 Page 70 o 108Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew S aphylococcus haemoly icus 22 (6) 8 (10.4) 1.93 0.93 – 4.01 0.080 S aphylococcus capi is 14 (3.8) 2 (2.6) 0.72 0.18 – 2.95 0.652 S aphylococcus wa ne i 2 (0.5) 1 (1.3) 1.89 0.26 – 13.65 0.525 S aphylococcus schlei e i 2 (0.5) 0 - - 0.674 S aphylococcus simulans 1 (0.3) 0 - - 0.766 S aphylococcus au icula is 1 (0.3) 0 - - 0.766 Oxacillin esis ance 266 (72.3) 61 (79.2) 1.42 0.81 – 2.49 0.228 Follow-up blood cul u es pe o med 101 (27.5) 21 (27.3) 0.99 0.59 – 1.67 0.998 Posi i e ollow-up blood cul u es among hose pe o med 29 (7.9) 15 (19.5) 2.34 1.36 – 4.20 0.003 Sou ce con ol (only CRBSI) Ea ly ca he e emo al 173/287 (60.3) 22/49 (44.9) 0.49 0.28 – 0.88 0.016 Days un il ca he e emo al, median (IQR) 2 (1 – 6) 4 (1 – 8) 0.85 0.73 – 0.99 0.033 Complica ed bac e emia 187 (50.8) 53 (68.8) 2.09 1.28 – 3.40 0.003 Absence o ea ly (<48h) ca he e emo al in CRSBI 114/287 (39.7) 27/49 (55.1) 2.04 1.14 - 3.57 0.016 Pe sis en e e ≥72h 52 (14.1) 25 (32.5) 2.48 1.54 – 4.01 <0.001 Implan ed de ice 44 (12) 20 (26) 2.44 1.47 – 4.07 <0.001 Pe sis en bac e emia 29 (7.9) 15 (19.5) 2.34 1.36 – 4.20 0.003 Me as a ic complica ions 4 (1.1) 3 (3.9) 2.39 0.76 – 7.60 0.138 An imic obial ea men Empi ic an ibio ic he apy, yes 249 (67.7) 50 (64.9) 0.39 0.22 – 0.69 0.441 Ac i e empi ic he apy, yes 128/249 (51.4) 23/50 (46) 0.88 0.67 - 1.17 0.383 Empi ic an ibio ic he apy, no 119 (32.3) 27 (35.1) 1.21 0.75 - 194 0.549 Empi ic no, a ge ed yes 104/119 (87.4) 25/27 (92.6) 0.98 0.23 – 4.14 0.976 Du a ion in days, median (IQR) 11 (7 – 18) 8 (5 – 13) 0.96 0.92 – 0.99 0.009 Ou comes ICU admission 66 (17.9) 24 (31.2) 1.89 1.17 – 3.07 0.010 Leng h o hospi al s ay a e BSI in days, median (IQR) 14 (7 – 31) 12 (5 – 24) 0.99 0.99 – 1.00 0.219 Recu ence 5 (1.4) 3 (3.9) 2.39 0.75 – 7.58 0.140 Page 71 o 108 Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew CKD: ch onic kidney disease; COPD: ch onic obs uc i e pulmona y disease; PICC: pe iphe ally inse ed cen al ca he e ; PVC: pe iphe al enous ca he e ; CVC: cen al enous ca he e ; CRBSI: ca he e - ela ed bloods eam in ec ion. Page 72 o 108Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew Table 4. Mul i a ia e analysis o ac o s associa ed wi h 30-day mo ali y. Model 1 included complica ed bloods eam in ec ion, and Models 2 included he indi idual componen s o complica ed bloods eam in ec ion. In Model 2a, he “ea ly ca he e emo al” a iable was s a i ied in o h ee ca ego ies o include all pa ien s. Model 2b included only pa ien s wi h CRBSI. Model 1 Va iables HR 95% CI p - alue Age 1.03 1.01 – 1.05 0.002 Ce eb o ascula disease 2.58 1.45 – 4.58 0.001 Immunosupp essi e he apy 2.16 1.22 – 3.84 0.008 SOFA sco e 1.09 1.03 – 1.16 0.005 Ca he e - ela ed bac e emia 0.49 0.30 – 0.80 0.004 Complica ed bac e emia 2.14 1.29 – 3.53 0.003 Model 2a Va iables HR 95% CI p - alue Age 1.03 1.01 – 1.05 <0.001 Ce eb o ascula disease 2.49 1.41 – 4.39 0.002 Immunosupp essi e he apy 1.99 1.11 – 3.57 0.o020 SOFA sco e 1.09 1.03 – 1.16 0.005 Fe e ≥72 hou s 2.52 1.52 – 4.17 <0.001 Ca he managemen CRBSI wi hou ea ly ca he e emo al Re e ence - - CRBSI wi h ea ly ca he e emo al 0.47 0.26 – 0.83 0.010 CoNS BSI o non-ca he e o igin 1.14 0.66 – 1.98 0.638 Model 2b Page 79 o 108 Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew Va iables HR 95% CI p - alue Age 1.03 1.01 – 1.06 0.003 Ce eb o ascula disease 2.92 1.49 – 5.70 0.002 Immunosupp essi e he apy 2.22 1.09 – 4.52 0.028 SOFA sco e 1.10 1.02 – 1.18 0.019 Fe e ≥72 hou s 2.20 1.14 – 4.23 0.018 Ea ly ca he e emo al 0.47 0.26 – 0.84 0.010 Page 80 o 108Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew Figu e 1. Cumula i e su i al o pa ien s wi h complica ed ( ed line) and non-complica ed (blue line) bac e emia due o coagulase-nega i e S aphylococcus. P alue (log ank es ) = 0.004 Day 0 5 10 15 20 25 30 NC 205 193 198 180 177 174 170 C 240 220 206 198 192 186 175 Numbe a isk (NC: non complica ed; C: complica ed) Page 81 o 108 Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 1 1SUPPLEMENTARY MATERIAL 2 3 4Me hods [1]. S udy mic obiological and clinical inclusion c i e ia o CoNS bloods eam in ec ion 5episodes 6 7Episodes o CoNS BSI we e included in he s udy i hey me bo h mic obiological and clinical c i e ia as ollows: 8 91. Mic obiological c i e ia. The same CoNS species wi h iden ical an ibio ic suscep ibili y p o iles was 10 isola ed om a leas wo dis inc se s o blood cul u es d awn om sepa a e si es. In he case o 11 ca he e - ela ed CoNS BSI: 12 -I he ca he e was emo ed and cul u ed, a diagnosis o CoNS CRBSI was con i med i he same 13 CoNS species wi h an iden ical suscep ibili y p o ile was isola ed om a leas one pe iphe al blood 14 cul u e and om a cul u e o he ca he e ip, in ei he a semiquan i a i e ( oll pla e, wi h p esence 15 o 15 CFU pe pla e o mo e) o quan i a i e ( o ex wi h a coun o a leas 103 CFU/segmen , o 16 sonica ion me hods, wi h coun s abo e 102 CFU/segmen ) echniques 17 -I he ca he e was no emo ed, he same CoNS species had o be isola ed om a leas one 18 pe iphe al blood cul u e and om a sample d awn om he ca he e hub. Diagnosis equi ed ei he 19 a di e en ial ime o posi i i y ( ime o de ec ed mic obial g ow h) o a leas 120 minu es lowe in 20 he ca he e sample compa ed o he sample om pe iphe al enipunc u e; o quan i a i e blood 21 cul u es, de ined as a h ee- old highe colony coun in he blood sample ob ained om he ca he e 22 compa ed o he pe iphe al blood sample. 23 24 2. Clinical signs o in ec ion. A leas wo o he ollowing clinical signs had o be p esen : 25 -Fe e (T >38ºC) o hypo he mia (T <36ºC) 26 -Tachyca dia (>90 bea s pe minu e) 27 -Tachypnea (>20 b ea hs pe minu e) o PaCO₂ < 32 mmHg 28 -Leukocy osis (>12,000 cells/mm³) o leukopenia (<4,000 cells/mm³) 29 30 3. O he in ec ion causes we e excluded. Fo pa ien s wi h an unusual sou ce o bac e emia o CoNS 31 (i.e., pneumonia, in aabdominal in ec ion, meningi is, unknown sou ce in pa ien s wi hou ca he e s o 32 p os he ic ma e ial), he opinion o a second local in es iga o was sough . 33 Page 82 o 108Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 2 34 Me hods [2]. Jus i ica ion o he inclusion o “me as a ic complica ions” as a c i e ion o “complica ed 35 bac e emia” 36 37 Al hough me as a ic complica ions occu ed in ewe han 3% o cases, his was e ained as a c i e ion o 38 complica ed bac e emia due o he p e-de ined na u e o he de ini ion. No ably, 5 o he 7 pa ien s wi h 39 me as a ic complica ions also me o he c i e ia o "complica ed bac e emia" as ollows: 40 -1 pa ien was a ascula p os he ic de ice ca ie 41 -1 had an a icula de ice 42 -2 had pe sis en e e 43 -1 had a CRBSI wi hou ea ly ca he e emo al. 44 Only 2 pa ien s did no mee any o he "complica ed c i e ia," and he de ini ion o complica ed bac e emia in 45 hese cases was solely a ibu ed o he p esence o me as a ic oci. 46 47 Page 83 o 108 Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 3 49 Me hods [3]. Mul i a iable eg ession model me hodology 50 51 In he eg ession analysis, we employed a backwa d s epwise p ocedu e. Va iables we e ini ially included based 52 on clinical ele ance (e.g., hei po en ial con ounding o modi ying e ec on he associa ion be ween 53 complica ed bac e emia and mo ali y, o hei di ec associa ion wi h mo ali y) and s a is ical signi icance (i.e., 54 p- alue < 0.1 in bi a ia e analysis). The a iables included we e age, Cha lson como bidi y index, ch onic kidney 55 disease, ce eb o ascula disease, ch onic hea ailu e, immunosupp essi e he apy, heal hca e-associa ed 56 acquisi ion, wa d o admission, p esence o a po -a-ca h o pacemake , SOFA sco e, asop esso use, sep ic 57 shock, ca he e - ela ed sou ce, endoca di is, and complica ed bac e emia. Collinea i y was es ed, and collinea 58 a iables we e emo ed based on clinical easoning and a iance in la ion ac o (VIF) alues. Speci ically, 59 asop esso use, ICU admission a onse , and sep ic shock we e excluded due o collinea i y wi h he SOFA 60 sco e. The Cha lson como bidi y index was eplaced by indi idual como bidi ies. All a iables we e ea ed as 61 ca ego ical, excep o age and SOFA sco e, which we e ea ed as con inuous a iables. 62 63 In Model 1, he a iables ha cons i u e he c i e ia o complica ed bac e emia we e excluded, while 64 complica ed bac e emia was e ained in all s eps. Va iables we e disca ded sequen ially based on he backwa d 65 s epwise p ocedu e, elimina ing he a iable wi h he lowes in luence on model pe o mance, as de e mined by 66 he Akaike In o ma ion C i e ion (AIC). The inal model included 424 pa ien s, wi h 21 excluded due o missing 67 da a. 68 69 In Model 2, we included indi idual c i e ia o complica ed bac e emia and excluded he composi e a iable. The 70 selec ion p ocedu e o a iables was iden ical o Model 1. Model 2a included 424 pa ien s (21 missing da a), 71 conside ing "ea ly ca he e emo al" as a 3-ca ego y a iable: ca he e - ela ed bac e emia wi hou ea ly ca he e 72 emo al ( e e ence ca ego y), ca he e - ela ed bac e emia wi h ea ly ca he e emo al, and non-ca he e - ela ed 73 bac e emia. Model 2b included 324 pa ien s (12 missing alues), ocusing solely on pa ien s wi h ca he e - 74 ela ed bac e emia. 75 76 In e ac ions be ween complica ed bac e emia and he o he a iables in he inal model we e assessed, as well 77 as in e ac ions be ween he a iables and each indi idual c i e ion de ining complica ed bac e emia. 78 79 The p edic i e abili y o he inal models was assessed by calcula ing he a ea unde he ecei e ope a ing 80 cha ac e is ic (AUROC) cu e wi h 95% con idence in e als. 81 82 Page 84 o 108Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 4 83 Me hods [4]:. Sensi i i y analysis 84 85 Fo 21 pa ien s, con i ma ion o su i al a day 30 was lacking. Consequen ly, hese pa ien s we e excluded 86 om he p ima y mul i a iable analysis o mo ali y p edic o s. All hese pa ien s we e discha ged ali e. Fo 87 pa ien s wi h comple e da a, 8 ou o 379 pa ien s who we e discha ged ali e had died by day 30 (2%; 99% CI 88 0-5%). Based on his, an an icipa ion o 0-1 dea hs among he 21 pa ien s was es ablished. Subsequen ly, wo 89 scena ios we e examined: 90 91 (a) assuming all 21 pa ien s su i ed a day 30; a mul i a ia e analysis o mo ali y p edic o s was pe o med 92 using logis ic eg ession conside ing ha all 21 pa ien s we e ali e a day 30. 93 94 (b) conside ing he possibili y o up o 2 dea hs among hem ( wo- old he uppe limi o 99% CI). Dea h was 95 andomly assigned o 2 pa ien s om he 21 wi h 10 i e a ions, and mul i a ia e logis ic eg ession was 96 conduc ed o each i e a ion. 97 Page 85 o 108 Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 5 98 Table S1. Fea u es o pa ien s acco ding o whe he ollow-up blood cul u es we e pe o med 99 100 FUBC pe o med N=122 FUBC no pe o med N=323 OR (95% CI) p- alue Fe e >72h 27 (22.1) 50 (15.5) 1.55 (0.92 - 2.62) 0.098 Lack o ea ly ca he e emo al (CRBSI only) 21/89 (23.6) 120/247 (48.6) 0.50 (0.33 - 0.77) <0.001 Endoca di is 12 (9.8) 7 (2.2) 4.93 (1.89 - 12.82) <0.001 Ca diac p os he ic al e 9 (7.4) 7 (2.2) 3.60 (1.31 – 9.88) 0.008 Pacemake 8 (6.6) 5 (1.5) 4.46 (1.43 – 13.92) 0.005 A icula p os he ic de ice 6 (4.9) 9 (2.8) 1.81 (0.63 – 5.18) 0.266 Endo ascula p os hesis 5 (4.1) 4 (1.2) 3.41 (0.90 – 12.91) 0.056 Me as a ic oci 3 (2.5) 4 (1.2) 2.01 (0.44 – 9.12) 0.356 SOFA sco e – median (IQR) 3 (1 – 5) 3 (1 – 6) 0.97 (0.91 - 1.04) 0.407 101 Page 86 o 108Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 6 102 103 Table S21. Bi a ia e analysis o he associa ion o di e en exposu es wi h in-hospi al mo ali y. 104 Va iable OR 95% CI p- alue Age, median (IQR) 1.03 1.01 – 1.05 <0.001 Age > 65 2.48 1.39 - 4.42 0.002 Male gende 1.26 0.74 – 2.15 0.402 Age-adjus ed Cha lson 1.07 0.90 – 1.26 0.442 Age-adjus ed Cha lson > 3 2.40 1.34 – 4.27 0.002 Diabe es melli us 1.03 0.50 – 2.14 0.936 Ch onic li e disease 0.82 0.31 – 2.17 0.683 Mode a e- o-se e e CKD 2.14 1.05 – 4.37 0.034 Hemodialysis 1.44 0.28 – 7.53 0.648 Ce eb o ascula disease 2.79 1.43 – 5.45 0.002 COPD 1.47 0.64 – 3.34 0.359 Ch onic hea ailu e 3.13 1.49 – 6.59 0.002 Rheuma ologic disease 0.51 0.07 – 3.98 1.000 Solid malignancy 1.27 0.73 – 2.29 0.416 Immunosupp essi e he apy 1.89 0.99 – 3.61 0.050 Recen an ibio ic he apy (30 days) 1.48 0.87 – 2.52 0.145 Recen su ge y (30 days) 0.53 0.27 – 1.03 0.049 Communi y-acqui ed 0.97 0.95 – 0.99 0.228 Hospi al-acqui ed 0.61 0.33 – 1.13 0.112 Heal hca e-associa ed 1.92 1.02 – 3.62 0.040 PICC 1.00 0.49 – 2.02 0.998 PVC 1.41 0.83 – 2.38 0.201 Sho - e m CVC 0.87 0.52 – 1.46 0.511 Tunneled ca he e 3.42 0.97 – 12.02 0.068 Po -a-Ca h® 0.96 0.11 – 8.06 1.000 Pacemake 2.65 0.79 – 8.85 0.112 Vascula p os he ic de ice 1.66 0.34 – 8.15 0.628 A icula p os hesis 1.45 0.39 – 5.29 0.476 Ca diac al ula p os hesis 1.97 0.62 – 6.29 0.275 Pa en e al nu i ion 0.77 0.39 – 1.50 0.437 Sep ic shock 5.07 2.40 – 10.72 <0.001 SOFA sco e 1.20 1.07 – 1.34 0.002 SOFA sco e ≥ 2 1.92 1.11 – 3.35 0.019 Pi sco e 1.26 1.12 – 1.43 <0.001 Pi sco e > 3 3.57 1.57 – 8.12 0.004 ICU admission 1.75 0.97 – 3.15 0.072 Amine suppo 2.55 0.88 – 7.40 0.080 Oxacillin esis ance 1.54 0.73 - 3.27 0.260 Ca he e - ela ed ocus 0.55 0.31 – 0.96 0.034 Unknown ocus 1.14 0.51 – 2.55 0.755 U ina y ocus 0.99 0.98 – 0.99 0.348 Skin and so issues ocus 1.59 0.43 – 5.87 0.480 Endoca di is ocus 2.82 1.03 – 7.69 0.036 Respi a o y ocus 6.05 1.47 – 24.82 0.005 Abdominal ocus 0.96 0.11 – 8.08 0.967 Bone and join s ocus 1.15 0.13 – 10.01 0.899 No empi ic he apy ecei ed 1.03 0.59 - 1.79 0.922 Adequacy o empi ic he apy 0.79 0.42 – 1.50 0.468 Du a ion empi ic, days 0.99 0.97 – 1.01 0.481 Du a ion a ge ed, days 0.99 0.99 – 1.01 0.353 Page 87 o 108 Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 7 Du a ion o he apy o e all, days 0.95 0.92 – 0.99 0.013 Ca he e emo al in CRBSI 0.59 0.31 - 1.11 0.101 Pe sis en e e a 72h 3.19 1.787 – 5.71 <0.001 Pe sis en bac e emia 2.74 1.350 – 5.57 0.005 Sep ic complica ions 4.45 0.974 – 20.37 0.036 Complica ed BSI 2.19 1.25 – 3.85 0.005 Days o ca he e emo al 0.99 0.84 – 1.17 0.888 ICU s ay, days 0.99 0.98 – 1.01 0.758 BSI onse o discha ge, days 0.99 0.99 – 1.01 0.327 Admission o BSI, days 1.00 0.99 – 1.01 0.447 105 CKD: ch onic kidney disease; COPD: ch onic obs uc i e pulmona y disease; AIDS: acqui ed 106 immunode iciency synd ome; PICC: pe iphe ally inse ed cen al ca he e ; PVC: pe iphe al enous ca he e ; 107 CVC: cen al enous ca he e ; ICU: in ensi e ca e uni ; CRBSI: ca he e - ela ed bloods eam in ec ion; BSI: 108 bloods eam in ec ion 109 Page 88 o 108Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 14 (a) Gi e unadjus ed es ima es and, i applicable, con ounde -adjus ed es ima es and hei p ecision (eg, 95% con idence in e al). Make clea which con ounde s we e adjus ed o and why hey we e included 10- 1112- 13 (b) Repo ca ego y bounda ies when con inuous a iables we e ca ego ized 12- 1310- 11 Main esul s 16 (c) I ele an , conside ansla ing es ima es o ela i e isk in o absolu e isk o a meaning ul ime pe iod - O he analyses 17 Repo o he analyses done—eg analyses o subg oups and in e ac ions, and sensi i i y analyses 1113, S3 Discussion Key esul s 18 Summa ize key esul s wi h e e ence o s udy objec i es 11- 1213- 14 Limi a ions 19 Discuss limi a ions o he s udy, aking in o accoun sou ces o po en ial bias o imp ecision. Discuss bo h di ec ion and magni ude o any po en ial bias 13- 1415 In e p e a ion 20 Gi e a cau ious o e all in e p e a ion o esul s conside ing objec i es, limi a ions, mul iplici y o analyses, esul s om simila s udies, and o he ele an e idence 11- 1414- 16 Gene alizabili y 21 Discuss he gene alizabili y (ex e nal alidi y) o he s udy esul s 1415 O he in o ma ion Funding 22 Gi e he sou ce o unding and he ole o he unde s o he p esen s udy and, i applicable, o he o iginal s udy on which he p esen a icle is based 1417 198 199 *Gi e in o ma ion sepa a ely o exposed and unexposed g oups. 200 201 No e: An Explana ion and Elabo a ion a icle discusses each checklis i em and gi es me hodological 202 backg ound and published examples o anspa en epo ing. The STROBE checklis is bes used in 203 conjunc ion wi h his a icle ( eely a ailable on he Web si es o PLoS Medicine a 204 h p://www.plosmedicine.o g/, Annals o In e nal Medicine a h p://www.annals.o g/, and Epidemiology a 205 h p://www.epidem.com/). In o ma ion on he STROBE Ini ia i e is a ailable a h p://www.s obe- 206 s a emen .o g. Page 95 o 108 Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 1 1SUPPLEMENTARY MATERIAL 2 3 4Me hods [1]. S udy mic obiological and clinical inclusion c i e ia o CoNS bloods eam in ec ion 5episodes 6 7Episodes o CoNS BSI we e included in he s udy i hey me bo h mic obiological and clinical c i e ia as ollows: 8 91. Mic obiological c i e ia. The same CoNS species wi h iden ical an ibio ic suscep ibili y p o iles was 10 isola ed om a leas wo dis inc se s o blood cul u es d awn om sepa a e si es. In he case o 11 ca he e - ela ed CoNS BSI: 12 -I he ca he e was emo ed and cul u ed, a diagnosis o CoNS CRBSI was con i med i he same 13 CoNS species wi h an iden ical suscep ibili y p o ile was isola ed om a leas one pe iphe al blood 14 cul u e and om a cul u e o he ca he e ip, in ei he a semiquan i a i e ( oll pla e, wi h p esence 15 o 15 CFU pe pla e o mo e) o quan i a i e ( o ex wi h a coun o a leas 103 CFU/segmen , o 16 sonica ion me hods, wi h coun s abo e 102 CFU/segmen ) echniques 17 -I he ca he e was no emo ed, he same CoNS species had o be isola ed om a leas one 18 pe iphe al blood cul u e and om a sample d awn om he ca he e hub. Diagnosis equi ed ei he 19 a di e en ial ime o posi i i y ( ime o de ec ed mic obial g ow h) o a leas 120 minu es lowe in 20 he ca he e sample compa ed o he sample om pe iphe al enipunc u e; o quan i a i e blood 21 cul u es, de ined as a h ee- old highe colony coun in he blood sample ob ained om he ca he e 22 compa ed o he pe iphe al blood sample. 23 24 2. Clinical signs o in ec ion. A leas wo o he ollowing clinical signs had o be p esen : 25 -Fe e (T >38ºC) o hypo he mia (T <36ºC) 26 -Tachyca dia (>90 bea s pe minu e) 27 -Tachypnea (>20 b ea hs pe minu e) o PaCO₂ < 32 mmHg 28 -Leukocy osis (>12,000 cells/mm³) o leukopenia (<4,000 cells/mm³) 29 30 3. O he in ec ion causes we e excluded. Fo pa ien s wi h an unusual sou ce o bac e emia o CoNS 31 (i.e., pneumonia, in aabdominal in ec ion, meningi is, unknown sou ce in pa ien s wi hou ca he e s o 32 p os he ic ma e ial), he opinion o a second local in es iga o was sough . 33 Page 96 o 108Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 2 34 Me hods [2]. Jus i ica ion o he inclusion o “me as a ic complica ions” as a c i e ion o “complica ed 35 bac e emia” 36 37 Al hough me as a ic complica ions occu ed in ewe han 3% o cases, his was e ained as a c i e ion o 38 complica ed bac e emia due o he p e-de ined na u e o he de ini ion. No ably, 5 o he 7 pa ien s wi h 39 me as a ic complica ions also me o he c i e ia o "complica ed bac e emia" as ollows: 40 -1 pa ien was a ascula p os he ic de ice ca ie 41 -1 had an a icula de ice 42 -2 had pe sis en e e 43 -1 had a CRBSI wi hou ea ly ca he e emo al. 44 Only 2 pa ien s did no mee any o he "complica ed c i e ia," and he de ini ion o complica ed bac e emia in 45 hese cases was solely a ibu ed o he p esence o me as a ic oci. 46 47 Page 97 o 108 Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 3 49 Me hods [3]. Mul i a iable eg ession model me hodology 50 51 In he eg ession analysis, we employed a backwa d s epwise p ocedu e. Va iables we e ini ially included based 52 on clinical ele ance (e.g., hei po en ial con ounding o modi ying e ec on he associa ion be ween 53 complica ed bac e emia and mo ali y, o hei di ec associa ion wi h mo ali y) and s a is ical signi icance (i.e., 54 p- alue < 0.1 in bi a ia e analysis). The a iables included we e age, Cha lson como bidi y index, ch onic kidney 55 disease, ce eb o ascula disease, ch onic hea ailu e, immunosupp essi e he apy, heal hca e-associa ed 56 acquisi ion, wa d o admission, p esence o a po -a-ca h o pacemake , SOFA sco e, asop esso use, sep ic 57 shock, ca he e - ela ed sou ce, endoca di is, and complica ed bac e emia. Collinea i y was es ed, and collinea 58 a iables we e emo ed based on clinical easoning and a iance in la ion ac o (VIF) alues. Speci ically, 59 asop esso use, ICU admission a onse , and sep ic shock we e excluded due o collinea i y wi h he SOFA 60 sco e. The Cha lson como bidi y index was eplaced by indi idual como bidi ies. All a iables we e ea ed as 61 ca ego ical, excep o age and SOFA sco e, which we e ea ed as con inuous a iables. 62 63 In Model 1, he a iables ha cons i u e he c i e ia o complica ed bac e emia we e excluded, while 64 complica ed bac e emia was e ained in all s eps. Va iables we e disca ded sequen ially based on he backwa d 65 s epwise p ocedu e, elimina ing he a iable wi h he lowes in luence on model pe o mance, as de e mined by 66 he Akaike In o ma ion C i e ion (AIC). The inal model included 424 pa ien s, wi h 21 excluded due o missing 67 da a. 68 69 In Model 2, we included indi idual c i e ia o complica ed bac e emia and excluded he composi e a iable. The 70 selec ion p ocedu e o a iables was iden ical o Model 1. Model 2a included 424 pa ien s (21 missing da a), 71 conside ing "ea ly ca he e emo al" as a 3-ca ego y a iable: ca he e - ela ed bac e emia wi hou ea ly ca he e 72 emo al ( e e ence ca ego y), ca he e - ela ed bac e emia wi h ea ly ca he e emo al, and non-ca he e - ela ed 73 bac e emia. Model 2b included 324 pa ien s (12 missing alues), ocusing solely on pa ien s wi h ca he e - 74 ela ed bac e emia. 75 76 In e ac ions be ween complica ed bac e emia and he o he a iables in he inal model we e assessed, as well 77 as in e ac ions be ween he a iables and each indi idual c i e ion de ining complica ed bac e emia. 78 79 The p edic i e abili y o he inal models was assessed by calcula ing he a ea unde he ecei e ope a ing 80 cha ac e is ic (AUROC) cu e wi h 95% con idence in e als. 81 82 Page 98 o 108Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 4 83 Me hods [4]. Sensi i i y analysis 84 85 Fo 21 pa ien s, con i ma ion o su i al a day 30 was lacking. Consequen ly, hese pa ien s we e excluded 86 om he p ima y mul i a iable analysis o mo ali y p edic o s. All hese pa ien s we e discha ged ali e. Fo 87 pa ien s wi h comple e da a, 8 ou o 379 pa ien s who we e discha ged ali e had died by day 30 (2%; 99% CI 88 0-5%). Based on his, an an icipa ion o 0-1 dea hs among he 21 pa ien s was es ablished. Subsequen ly, wo 89 scena ios we e examined: 90 91 (a) assuming all 21 pa ien s su i ed a day 30; a mul i a ia e analysis o mo ali y p edic o s was pe o med 92 using logis ic eg ession conside ing ha all 21 pa ien s we e ali e a day 30. 93 94 (b) conside ing he possibili y o up o 2 dea hs among hem ( wo- old he uppe limi o 99% CI). Dea h was 95 andomly assigned o 2 pa ien s om he 21 wi h 10 i e a ions, and mul i a ia e logis ic eg ession was 96 conduc ed o each i e a ion. 97 Page 99 o 108 Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 5 98 Table S1. Fea u es o pa ien s acco ding o whe he ollow-up blood cul u es we e pe o med 99 100 FUBC pe o med N=122 FUBC no pe o med N=323 OR (95% CI) p- alue Fe e >72h 27 (22.1) 50 (15.5) 1.55 (0.92 - 2.62) 0.098 Lack o ea ly ca he e emo al (CRBSI only) 21/89 (23.6) 120/247 (48.6) 0.50 (0.33 - 0.77) <0.001 Endoca di is 12 (9.8) 7 (2.2) 4.93 (1.89 - 12.82) <0.001 Ca diac p os he ic al e 9 (7.4) 7 (2.2) 3.60 (1.31 – 9.88) 0.008 Pacemake 8 (6.6) 5 (1.5) 4.46 (1.43 – 13.92) 0.005 A icula p os he ic de ice 6 (4.9) 9 (2.8) 1.81 (0.63 – 5.18) 0.266 Endo ascula p os hesis 5 (4.1) 4 (1.2) 3.41 (0.90 – 12.91) 0.056 Me as a ic oci 3 (2.5) 4 (1.2) 2.01 (0.44 – 9.12) 0.356 SOFA sco e – median (IQR) 3 (1 – 5) 3 (1 – 6) 0.97 (0.91 - 1.04) 0.407 101 Page 100 o 108Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 6 102 103 Table S2. Bi a ia e analysis o he associa ion o di e en exposu es wi h in-hospi al mo ali y. 104 Va iable OR 95% CI p- alue Age, median (IQR) 1.03 1.01 – 1.05 <0.001 Age > 65 2.48 1.39 - 4.42 0.002 Male gende 1.26 0.74 – 2.15 0.402 Age-adjus ed Cha lson 1.07 0.90 – 1.26 0.442 Age-adjus ed Cha lson > 3 2.40 1.34 – 4.27 0.002 Diabe es melli us 1.03 0.50 – 2.14 0.936 Ch onic li e disease 0.82 0.31 – 2.17 0.683 Mode a e- o-se e e CKD 2.14 1.05 – 4.37 0.034 Hemodialysis 1.44 0.28 – 7.53 0.648 Ce eb o ascula disease 2.79 1.43 – 5.45 0.002 COPD 1.47 0.64 – 3.34 0.359 Ch onic hea ailu e 3.13 1.49 – 6.59 0.002 Rheuma ologic disease 0.51 0.07 – 3.98 1.000 Solid malignancy 1.27 0.73 – 2.29 0.416 Immunosupp essi e he apy 1.89 0.99 – 3.61 0.050 Recen an ibio ic he apy (30 days) 1.48 0.87 – 2.52 0.145 Recen su ge y (30 days) 0.53 0.27 – 1.03 0.049 Communi y-acqui ed 0.97 0.95 – 0.99 0.228 Hospi al-acqui ed 0.61 0.33 – 1.13 0.112 Heal hca e-associa ed 1.92 1.02 – 3.62 0.040 PICC 1.00 0.49 – 2.02 0.998 PVC 1.41 0.83 – 2.38 0.201 Sho - e m CVC 0.87 0.52 – 1.46 0.511 Tunneled ca he e 3.42 0.97 – 12.02 0.068 Po -a-Ca h 0.96 0.11 – 8.06 1.000 Pacemake 2.65 0.79 – 8.85 0.112 Vascula p os he ic de ice 1.66 0.34 – 8.15 0.628 A icula p os hesis 1.45 0.39 – 5.29 0.476 Ca diac al ula p os hesis 1.97 0.62 – 6.29 0.275 Pa en e al nu i ion 0.77 0.39 – 1.50 0.437 Sep ic shock 5.07 2.40 – 10.72 <0.001 SOFA sco e 1.20 1.07 – 1.34 0.002 SOFA sco e ≥ 2 1.92 1.11 – 3.35 0.019 Pi sco e 1.26 1.12 – 1.43 <0.001 Pi sco e > 3 3.57 1.57 – 8.12 0.004 ICU admission 1.75 0.97 – 3.15 0.072 Amine suppo 2.55 0.88 – 7.40 0.080 Oxacillin esis ance 1.54 0.73 - 3.27 0.260 Ca he e - ela ed ocus 0.55 0.31 – 0.96 0.034 Unknown ocus 1.14 0.51 – 2.55 0.755 U ina y ocus 0.99 0.98 – 0.99 0.348 Skin and so issues ocus 1.59 0.43 – 5.87 0.480 Endoca di is ocus 2.82 1.03 – 7.69 0.036 Respi a o y ocus 6.05 1.47 – 24.82 0.005 Abdominal ocus 0.96 0.11 – 8.08 0.967 Bone and join s ocus 1.15 0.13 – 10.01 0.899 No empi ic he apy ecei ed 1.03 0.59 - 1.79 0.922 Adequacy o empi ic he apy 0.79 0.42 – 1.50 0.468 Du a ion empi ic, days 0.99 0.97 – 1.01 0.481 Du a ion a ge ed, days 0.99 0.99 – 1.01 0.353 Page 101 o 108 Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 7 Du a ion o he apy o e all, days 0.95 0.92 – 0.99 0.013 Ca he e emo al in CRBSI 0.59 0.31 - 1.11 0.101 Pe sis en e e a 72h 3.19 1.787 – 5.71 <0.001 Pe sis en bac e emia 2.74 1.350 – 5.57 0.005 Sep ic complica ions 4.45 0.974 – 20.37 0.036 Complica ed BSI 2.19 1.25 – 3.85 0.005 Days o ca he e emo al 0.99 0.84 – 1.17 0.888 ICU s ay, days 0.99 0.98 – 1.01 0.758 BSI onse o discha ge, days 0.99 0.99 – 1.01 0.327 Admission o BSI, days 1.00 0.99 – 1.01 0.447 105 CKD: ch onic kidney disease; COPD: ch onic obs uc i e pulmona y disease; AIDS: acqui ed 106 immunode iciency synd ome; PICC: pe iphe ally inse ed cen al ca he e ; PVC: pe iphe al enous ca he e ; 107 CVC: cen al enous ca he e ; ICU: in ensi e ca e uni ; CRBSI: ca he e - ela ed bloods eam in ec ion; BSI: 108 bloods eam in ec ion 109 Page 102 o 108Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 8 110 Table S3. Mul i a ia e analysis o ac o s associa ed wi h in-hospi al mo ali y. 111 112 Model 1 113 Va iable OR 95% CI p- alue Age > 65 yea s old 2.87 1.53 – 5.38 < 0.001 Ce eb o ascula disease 3.55 1.72 – 7.35 < 0.001 Immunosupp essi e he apy 3.24 1.55 – 6.79 0.002 Pi sco e > 3 3.54 1.45 – 8.64 0.005 Complica ed BSI 1.99 1.10 – 3.59 0.023 114 BSI: bloods eam in ec ion 115 AUROC o 0.74 (95%CI 0.67 – 0.80) 116 117 Model 2 118 Va iable OR 95% CI p- alue Age > 65 yea s old 2.99 1.57 – 5.71 < 0.001 Ce eb o ascula disease 3.59 1.71 – 7.53 < 0.001 Immunosupp essi e he apy 2.98 1.34 - 6.41 0.005 Pi sco e > 3 3.52 1.38 – 8.97 0.009 Fe e > 72 hou s 3.29 1.72 – 6.28 < 0.001 Ea ly ca he e emo al in CRBSI 0.50 0.28 - 0.90 0.020 119 CRBSI: ca he e - ela ed bloods eam in ec ion 120 AUROC o 0.76 (95%CI 0.69 – 0.82) 121 Page 103 o 108 Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Fo Pee Re iew 9 122 Table S4. Sensi i i y analysis o 30-day mo ali y p edic ion. 123 124 125 (a) Conside ing ha all 21 pa ien s su i ed a day 30 126 Va iable OR 95% CI p- alue Age > 65 yea s 2.47 1.39 – 4.38 0.002 Ce eb o ascula disease 3.07 1.53 – 6.19 0.002 Immunosupp essi e he apy 2.92 1.46 – 5.85 0.003 Pi sco e > 3 3.39 1.43 – 8.00 0.005 Complica ed BSI 1.95 1.12 – 3.37 0.017 127 BSI: bloods eam in ec ion 128 AUROC 0.72 (95%CI 0.66 – 0.78) 129 130 (b) Conside ing ha up o 2 pa ien s o hem died (chosen andomly). We p o ide a summa y o 131 he es ima es in he 10 models. 132 1. Va iable OR 95% CI p- alue Age > 65 yea s 2.47 1.40 – 4.38 0.002 Ce eb o ascula disease 3.07 1.53 – 6.19 0.002 Immunosupp essi e he apy 2.92 1.46 – 5.85 0.003 Pi sco e > 3 3.39 1.43 – 8.00 0.005 Complica ed BSI 1.95 1.12 – 3.37 0.017 133 BSI: bloods eam in ec ion 134 AUROC 0.73 (95%CI 0.66 – 0.78) 135 136 2. Va iable OR 95% CI p- alue Age > 65 yea s 2.58 1.46 – 4.57 0.001 Ce eb o ascula disease 2.93 1.46 – 5.91 0.003 Immunosupp essi e he apy 2.87 1.43 – 5.75 0.003 Pi sco e > 3 3.90 1.67 – 9.12 0.002 Complica ed BSI 1.90 1.10 – 3.25 0.021 137 BSI: bloods eam in ec ion 138 AUROC 0.73 (95%CI 0.66 – 0.78) 139 3. 140 Va iable OR 95% CI p- alue Age > 65 yea s 2.59 1.47 – 4.59 0.001 Ce eb o ascula disease 3.33 1.67 – 6.67 < 0.001 Immunosupp essi e he apy 2.84 1.42 – 5.69 0.003 Pi sco e > 3 3.22 1.36 – 7.63 0.009 Complica ed BSI 2.01 1.20 – 3.59 0.009 141 BSI: bloods eam in ec ion 142 AUROC 0.73 (95%CI 0.67 – 0.79) 143 144 4. 145 Va iable OR 95% CI p- alue Page 104 o 108Clinical Mic obiology and In ec ion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60