Epidemiology and clinical ea u es o communi y-acqui ed,
heal hca e-associa ed and nosocomial bloods eam in ec ions
in e ia y-ca e and communi y hospi als
J. Rod ı
´guez-Ban
˜o
1
,M.D.Lo
´pez-P ie o
2
, M. M. Po illo
1
, P. Re ama
1
, C. Na e a
3
, E. Nun
˜o
4
, M. He e o
5
, A. del A co
6
,
A
´. Mun
˜oz
7
,F.Te
´llez
8
, M. To es-To osa
9
, A. Ma ı
´n-Aspas
10
, A. A oyo
11
, A. Ruiz
12
, R. Moya
13
, J. E. Co zo
14
, L. Leo
´n
15
and J. A. Pe
´ ez-Lo
´pez
16
, on behal o he SAEI/SAMPAC Bac e aemia G oup
1) Seccio
´n de En e medades In ecciosas, Hospi al Uni e si a io Vi gen Maca ena, Se ille, 2) Se icio de Mic obiologı
´a, Hospi al del SAS, Je ez de la F on e a
(Ca
´diz), 3) Seccio
´n de En e medades In ecciosas, Hospi al Uni e si a io Reina So ı
´a, Co
´ doba, 4) Seccio
´n de En e medades In ecciosas, Hospi al Uni e si a io
Vi gen de la Vic o ia, Ma
´laga, 5) Se icio de En e medades In ecciosas, Hospi ales Uni e si a ios Vi gen del Rocı
´o, Se ille, 6) Se icio de Medicina In e na,
Hospi al Cos a del Sol, Ma bella (Ma
´laga), 7) Se icio de Medicina In e na, Hospi al de la Se anı
´a, Ronda (Ma
´laga), 8) Unidad de En e medades In eccio-
sas, Hospi al de La Lı
´nea, La Lı
´nea de la Concepcio
´n (Ca
´diz), 9) Seccio
´n de En e medades In ecciosas, Hospi al Pun a de Eu opa, Algeci as (Ca
´diz),
10) Se icio de Medicina In e na, Hospi al Pue a del Ma , Ca
´diz, 11) Seccio
´n de En e medades In ecciosas, Complejo Hospi ala io de Jae
´n, Jae
´n,
12) Unidad de En e medades In ecciosas, Hospi al del SAS, Je ez de la F on e a (Ca
´diz), 13) Se icio de Medicina In e na, Hospi al de An eque a, Ma
´laga,
14) Seccio
´n de En e medades In ecciosas, Hospi al Uni e si a io de Valme, Se ille, 15) Unidad de En e medades In ecciosas, Hospi al To eca
´ denas, Alme ı
´a
and 16) Se icio de Mic obiologı
´a, Hospi al Uni e si a io San Cecilio, G anada, Spain
Abs ac
Classi ica ion o bloods eam in ec ions (BSIs) as communi y-acqui ed (CA), heal hca e-associa ed (HCA) and hospi al-acqui ed (HA) has
been p oposed. The epidemiology and clinical ea u es o BSI acco ding o ha classi ica ion in e ia y-ca e (TH) and communi y (CH)
hospi als we e in es iga ed in a p ospec i e coho o 821 BSI episodes om 15 hospi als ( en TH and i e CH hospi als) in Andalucı
´a,
Spain. Eigh een pe cen we e CA, 24% we e HCA and 58% we e HA. The incidence o CA and HCA BSI was highe in CH han in TH
(CA: 3.9 episodes pe 1000 admissions s. 2.2, p <0.01; HCA: 5.0 s. 2.9, p <0.01), whe eas he incidence o HA BSI was lowe (7.7 s.
8.7, p <0.01). In CA and HCA BSI, he espi a o y ac was mo e equen ly he sou ce in CH han in TH (CA: 30% s. 15%; HCA: 20%
s. 9%, p £0.03). In HCA BSI, ch onic enal insu iciency and unnelled ca he e s we e less equen in CH han in TH (11% s. 26% and
7% s. 19%, p £0.03), al hough ch onic ulce s we e mo e equen (22% s. 8%, p 0.008). BSIs as a esul o me hicillin- esis an S aphylo-
coccus au eus o Pseudomonas ae uginosa we e e y a e in CA episodes, al hough ex ended-spec um b-lac amase-p oducing Esche ichia
coli (ESBLEC) caused a simila p opo ion o all BSIs in CA, HCA and HA episodes. Mul i a ia e analysis e ealed no signi ican di e ence
in mo ali y a es in CH and TH. HCA in ec ions should be conside ed as a sepa a e class o BSI in bo h TH and CH, al hough di e ences
be ween hospi als mus be conside ed. CA BSIs we e no caused by mul id ug- esis an pa hogens, excep o ESBLEC.
Keywo ds: Bloods eam in ec ions, communi y-acqui ed in ec ions, ex ended-spec um b-lac amases, heal hca e-associa ed in ec ions,
me hicillin- esis an S aphylococcus au eus, mul icen e s udy, nosocomial in ec ions
O iginal Submission: 17 July 2009; Re ised Submission: 9 Oc obe 2009; Accep ed: 9 Oc obe 2009
Edi o : M. Paul
Clin Mic obiol In ec
Co esponding au ho and ep in eques s: J. Rod ı
´guez Ban
˜o,
Seccio
´n de En e medades In ecciosas, Hospi al Uni e si a io Vi gen
Maca ena, A da D Fed iani 3, 41009 Se ille, Spain
E-mail: [email p o ec ed]m
In oduc ion
Bloods eam in ec ions (BSI) a e an impo an cause o mo -
bidi y and mo ali y in hospi alized pa ien s [1–3]. T adi ionally,
BSI ha e been di ided in o communi y and nosocomial
episodes [4]. Howe e , subsequen o he implemen a ion o
ambula o y al e na i es o inpa ien ca e, signi ican changes
in he epidemiology o BSI ha e been no ed and a p oposal
was made o u he subdi ide communi y-onse BSI in o
heal hca e-associa ed (HCA) episodes o pa ien s wi h signi -
ican ecen heal hca e con ac and p ocedu es, and s ic ly
communi y-acqui ed (CA) episodes o pa ien s wi hou [5].
Howe e , his new classi ica ion has been e alua ed in
only a ew s udies [5–9], in which HCA episodes we e
CLM 3089
B
Dispa ch: 2.2.10 Jou nal: CLM
CE: Shiyamala
Jou nal Name Manusc ip No.
Au ho Recei ed: No. o pages: 6 PE: Gaya h i
ª2010 The Au ho s
Jou nal Compila ion ª2010 Eu opean Socie y o Clinical Mic obiology and In ec ious Diseases
ORIGINAL ARTICLE 10.1111/j.1469-0691.2010.03089.x
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
shown o be somewha mo e simila o hospi al-acqui ed
(HA) han o CA BSI in ae iology and p edisposing condi-
ions; hese s udies included e ospec i e analysis o
included cases om a limi ed numbe o hospi als, and used
di e en c i e ia. I s applicabili y o bo h e ia y-ca e hospi-
als and communi y hospi als has no been s udied speci i-
cally, no is i known whe he his classi ica ion is help ul in
iden i ying pa ien s wi h a low isk o acqui ing some eme -
gen communi y-onse in ec ion-causing, an ibio ic- esis an
pa hogens, such as ex ended-spec um b-lac amase
(ESBL)-p oducing Esche ichia coli [10]. The p esen s udy
aimed o in es iga e he popula ion-based incidence, epidemi-
ology, ae iology and clinical ea u es o BSI acco ding o he
epidemiological ype o in ec ion, and o e alua e he newly-
p oposed scheme wi hin a wide sample o hospi als ha
included bo h e ia y-ca e and communi y se ings.
Ma e ials and Me hods
Si e and design
A mul icen e p ospec i e coho s udy was conduc ed in 15
public hospi als ( en e ia y-ca e and i e communi y hospi-
als) in Andalucı
´a, Spain. The pa icipa ing hospi als p o ide
heal h ca e o >90% o hei ca chmen a eas (>4 million)
[11].
All episodes o clinically signi ican BSI in adul pa ien s
(>14 yea s) be ween 15 Oc obe 2006 and 15 Decembe
2006 we e included. Because he aim o he s udy was o
include a signi ican numbe o cases om communi y cen-
es, he s udy pe iod was ex ended un il 15 Ma ch 2007 in
hese hospi als.
Blood cul u es we e pe o med, p ocessed and in e -
p e ed in acco dance wi h he ecommenda ions o he
Spanish Socie y o In ec ious Diseases and Clinical Mic obiol-
ogy [12]. Blood cul u es (a leas wo blood d aws) we e
ob ained om pe iphe al eins, excep in a ew cases whe e
he cen al enous ca he e was he suspec ed cause, when
a leas one blood d aw was ob ained om a pe iphe al ein
and ano he om he ca he e . Suscep ibili y esul s we e
in e p e ed acco ding o he CLSI ecommenda ions [13]; in
his analysis, we ocussed on me hicillin esis ance in S aphylo-
coccus au eus, ancomycin- esis ance in en e ococci, ESBL
p oduc ion in E. coli and Klebsiella spp., and ca bapenem
esis ance in En e obac e iaceae and Pseudomonas ae uginosa.
Cases we e de ec ed by daily e iew o blood cul u e
esul s. All pa ien s wi h posi i e blood cul u es we e consid-
e ed eligible. Di e en episodes occu ing in he same
pa ien s we e included only when caused by di e en spe-
cies. Fo po en ial con aminan s (coagulase-nega i e s aphylo-
cocci, diph he oids), only episodes in which he o ganism
had been isola ed om wo o mo e blood d aws we e
included [4,12]. Pa ien s we e ollowed un il discha ge o
dea h. The s udy was app o ed by he E hics Commi ees o
he pa icipa ing cen es, which wai ed he need o ob ain-
ing in o med consen .
Va iables and de ini ions
The da a collec ed included: ype o hospi al; epidemiologic
ype o in ec ion (CA, HCA o HA); demog aphics; ype and
se e i y o ch onic unde lying disease [14]; sou ce o BSI
(see below); ascula o u ina y ca he e and mechanical en-
ila ion a onse , endoscopic p ocedu es pe o med wi hin
he p eceding week, and majo su ge y in he p eceding
3 mon hs; an imic obial use; ae iology; empi ical he apy and
ou come.
Episodes we e classi ied as HA i he episode occu ed
mo e han 48 h a e admission [4]; all o he episodes we e
conside ed communi y-onse , which we e hen classi ied as
HCA o CA, acco ding o he c i e ia used by F iedman
e al. [5]. In b ie , episodes we e classi ied as HCA when any
o he ollowing was p esen : in a enous he apy o special-
is nu sing ca e a home, haemodialysis in he 30 days be o e
he BSI; hospi aliza ion o >2 days in an acu e ca e hospi al
in he 90 days be o e he BSI; o esiden s a us in a nu sing
home o long- e m ca e acili y.
The BSI sou ce was de e mined om clinical and labo a-
o y da a, using CDC c i e ia o seconda y BSI [4]. Empi i-
cal he apy was conside ed app op ia e when an ac i e
an imic obial agen (acco ding o suscep ibili y da a) was
adminis e ed a ecommended doses wi hin he i s 24 h
a e he blood cul u es we e pe o med. C ude mo ali y
was eco ded a days 14 and 30.
S a is ical analysis
Popula ion-based incidence a es we e calcula ed acco ding
o he assigned e e ence popula ion o each cen e [15]
and ex apola ed o 1 yea . The e e ence popula ions a e
de ined by speci ic geog aphic limi s; pa ien s needing hospi al
admission a e ini ially hospi alized in hei co esponding
cen e and a e only ans e ed o e e ence cen es when
needed. A compa ison o a es was pe o med using a Pois-
son eg ession model. Compa isons be ween CA and HCA,
and be ween HCA and HA episodes we e pe o med simi-
la ly o p e iously epo ed me hods [5]; a iables included
in he de ini ions we e no compa ed. Simila me hods we e
used o compa isons be ween e ia y-ca e and communi y
hospi als. Ca ego ical and con inuous a iables we e com-
pa ed using Fishe ’s exac es and he he Mann–Whi ney
U- es , espec i ely. Whene e a signi ican c ude associa ion
2Clinical Mic obiology and In ec ion CMI
ª2010 The Au ho s
Jou nal Compila ion ª2010 Eu opean Socie y o Clinical Mic obiology and In ec ious Diseases, CMI
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
was ound be ween he ype o hospi al and mo ali y, a mul-
i a ia e analysis was pe o med by logis ic eg ession o
con ol o con oude s using a s epwise o wa d me hod.
Resul s
Incidence o BSI and epidemiological ype o in ec ion
Du ing he s udy pe iod, 821 BSI episodes in adul s we e
included; 660 (80%) occu ed in e ia y-ca e hospi als and
161 (20%) occu ed in communi y hospi als. O e all, 476
(58%) we e classi ied as HA, 195 (24%) as HCA, and 150
(18%) as CA. The dis ibu ion o BSI acco ding o p e-
de ined epidemiological ypes o in ec ion was somewha di -
e en in he wo ypes o hospi als; compa ed o e ia y
hospi als, he e was a highe equency o CA episodes in
communi y cen es (25% s. 17%, p 0.01), a lowe equency
o HA episodes (47% s. 60%, p 0.002) and a simila e-
quency o HCA episodes (28% s. 23%, p 0.1). The incidence
o BSIs is shown in Table 1. The o e all popula ion-based
incidence was highe in e ia y-ca e cen es ela ed o he
highe incidence o nosocomial episodes; he admissions-
based incidence was highe in communi y hospi als. Because
only i e pa ien s had been admi ed om o he cen es ou -
side hei ca chmen a ea, he exclusion o hese pa ien s
om he analysis did no change he esul s.
O pa ien s wi h HCA episodes, 131 (67%) had been p e-
iously admi ed, 28 (14%) we e esiden s in long- e m ca e
acili ies, 28 (14%) we e ecei ing haemodialysis and wo
(1%) we e ecei ing pe i oneal dialysis, 41 (21%) a ended a
day hospi al, ou (2%) ecei ed home in a enous he apy,
and i e (3%) we e ecei ing o he ypes o specialized home
ca e (se e al pa ien s had mo e han one o hese p edispos-
ing ac o s).
P edisposing ea u es, sou ces and ou come acco ding o
epidemiological ype o in ec ion
Table 2 shows pa ien cha ac e is ics, clinical ea u es and
he ou come o BSI. Ch onic enal insu iciency, haema ologi-
cal cance and neu openia we e mo e equen in pa ien s
wi h HCA han in CA BSI. Compa ed o HA episodes, HCA
occu ed mo e equen ly in pa ien s wi h ch onic enal
insu iciency, li e disease and cu aneous ulce , whe eas p e-
ious an ibio ic use and in asi e p ocedu es (wi h he excep-
ion o unnelled enous ca he e s) we e mo e equen in
pa ien s wi h HA BSI. Wi h ega d o he sou ce o in ec ion,
u ina y ac in ec ion was mo e equen in CA episodes
han in HCA BSI. Endoca di is was mo e equen in HCA
han in HA cases, whe eas ascula ca he e - ela ed in ec-
ions we e mo e equen in HA BSI. Inapp op ia e empi ical
he apy was mo e equen in HA BSI. Al hough he e was a
end owa d highe mo ali y in HCA and HA episodes,
hese di e ences we e no s a is ically signi ican .
TABLE 1. Popula ion-based and admissions-based incidence
a es o bloods eam in ec ions
All
hospi als
Te ia y
hospi als
Communi y
hospi als
Episodes pe 100 000 popula ion/yea
All episodes 109.2 116.4
a
75.3
CA 19.2 19.2 18.0
HCA 25.2 25.2 27.3
HA 64.8 72.0
a
30.0
Episodes pe 1000 admissions
All episodes 14.7 13.9
a
16.8
CA 2.7 2.2
a
3.9
HCA 3.5 2.9
a
5.0
HA 8.4 8.7
a
7.7
CA, communi y-acqui ed; HCA, heal hca e associa ed; HA, hospi al-acqui ed.
a
Fo compa ison wi h communi y hospi als, p <0.01
TABLE 2. Pa ien cha ac e is ics, clinical ea u es and
mo ali y o bloods eam in ec ions by epidemiological ype
o in ec ion
CA
(n= 150)
HCA
(n= 195)
HA
(n= 476)
Age in yea s, median (IQ ange) 67 (46–77) 69 (56–70) 66 (53–75)
Male gende 80 (53) 116 (60) 286 (60)
Non a al unde lying disease 106 (71)
a
91 (47) 326 (68)
Ch onic enal insu iciency 5 (3)
a
33 (17)
b
42 (9)
Ch onic li e disease 17 (11) 26 (13)
b
31 (7)
Solid cance 21 (14) 34 (17) 112 (24)
Haema ological cance 4 (3)
a
24 (12) 43 (9)
HIV in ec ion 6 (4) 5 (3) 5 (1)
Ch onic cu aneous ulce –
c
12 (6)
d
9 (2)
Neu openia 2 (1)
a
13 (7) 32 (7)
Non unnelled cen al enous ca he e – 4 (2)
b
182 (38)
Tunnelled enous ca he e – 32 (16)
b
35 (7)
U ina y ca he e 19 (13) 25 (13)
b
238 (50)
Mechanical en ila ion – 1 (1)
b
93 (20)
Endoscopic p ocedu e 1 (1) 1 (1)
b
28 (6)
Su ge y – 9 (5)
b
106 (22)
P e ious an imic obials 26 (17)
a
62 (32)
b
253 (53)
Sou ce
Unknown 22 (15) 45 (23) 127 (27)
U ina y ac 46 (31)
a
41 (21) 73 (15)
Respi a o y ac 28 (19) 22 (11) 55 (12)
In a-abdominal in ec ion 15 (10) 22 (11) 44 (9)
Bilia y ac 18 (12) 18 (9) 26 (5)
Endoca di is 8 (5) 9 (5)
b
6 (1)
Skin and skin s uc u es 8 (5) 10 (5) 21 (4)
Vascula ca he e – 24 (12)
b
116 (24)
O he s 5 (3) 4 (2) 8 (2)
App op ia e empi ical he apy 115 (77)
b
141 (72)
b
296 (62)
Mo ali y
14-day 22 (15) 36 (18) 97 (20)
30-day 29 (19) 43 (22) 116 (24)
Da a a e exp essed as he numbe o cases (%), excep whe e speci ied. p alue
o all s a is ical compa isons ‡0.05, excep whe e speci ied.
CA, communi y-acqui ed. HCA, heal hca e-associa ed. HA, hospi al-acqui ed.
a
Fo compa ison wi h heal hca e-associa ed episodes, p <0.01 (Fishe ’s exac
es ).
b
Fo compa ison wi h hospi al-acqui ed episodes, p £0.01 (Fishe ’s exac es ).
c
Fo compa ison wi h heal hca e-associa ed episodes, p 0.02 (Fishe ’s exac
es ).
d
Fo compa ison wi h hospi al-acqui ed episodes, p 0.04 (Fishe ’s exac es ). 2
CMI Rod ı
´guez-Ban
˜oe al. Bac e aemia in e ia y-ca e and communi y hospi als 3
ª2010 The Au ho s
Jou nal Compila ion ª2010 Eu opean Socie y o Clinical Mic obiology and In ec ious Diseases, CMI
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
Ae iology and an imic obial suscep ibili y
The mic oo ganisms causing BSI a e shown in Table 3. O e -
all, G am-nega i es we e mo e equen in HCA han in CA
and HA in ec ions. S. au eus, coagulase-nega i e s aphylo-
cocci, En e ococcus spp., En e obac e spp. and P. ae uginosa
we e less equen in CA in ec ions, whe eas S aphylococcus
pneumoniae and E. coli we e mo e equen ly associa ed wi h
his g oup. None o he CA cases we e caused by me hicillin-
esis an S. au eus, S eno ophomonas mal ophilia o Acine obac-
e baumannii, and only one by P. ae uginosa. By con as , he
pe cen ages o isola es p oducing ESBLs among E. coli o Kle-
bsiella spp. we e simila among CA, HCA and HA BSI. The e
we e no episodes as a esul o ancomycin- esis an en e o-
cocci o ca bapenem- esis an En e obac e iceae.
Analysis o BSI by hospi al ype
Among pa ien s wi h CA episodes (40 om communi y hos-
pi als and 110 om e ia y hospi als), he e we e no signi i-
can di e ences in demog aphics, unde lying condi ions,
p edisposing ac o s, sou ce, mic oo ganisms o mo ali y
by ype o hospi al (da a no shown), wi h he ollowing
excep ions: he espi a o y ac was a mo e equen sou ce
in communi y cen es (30% s. 15%, p 0.01), whe eas endo-
ca di is was a he less equen (0 s. 7%, p 0.07); E. coli
was less equen in communi y cen es (23% s. 44%,
p 0.01), whe eas S. pneumoniae was mo e equen (28% s.
15%, p 0.06).
Mo e di e ences we e ound in pa ien s wi h HCA epi-
sodes (45 om communi y and 150 om e ia y-ca e hospi-
als). Ch onic enal insu iciency was less equen in
communi y cen es (11% s. 26%, p 0.03) and ch onic cu a-
neous ulce s we e mo e equen (22% s. 8%, p 0.008).
Tunnelled enous ca he e s we e mo e equen in e ia y-
ca e cen es (19% s. 7%, p 0.04), as was haemodialysis (18%
s. 2%, p 0.008). The espi a o y ac was he mos equen
sou ce o BSI in communi y hospi als (20% s. 9%, p 0.03).
Mo ali y a day 14 was highe in communi y cen es (16%
s. 29%, p 0.03). Howe e , he di e ence was no longe s a-
is ically signi ican when adjus ed o p esen a ion wi h
se e e sepsis o sep ic shock, sou ce o in ec ion and se e -
i y o unde lying disease (adjus ed OR 2.5, 95% CI 0.8–6.9,
p 0.1). The e was no signi ican di e ence in mo ali y a day
30 (20% s. 13%, p 0.2).
Fo pa ien s wi h HA episodes (76 om communi y cen-
es and 400 om e ia y hospi als), he ollowing di e -
ences we e ound: haema ological cance and neu openia
we e mo e equen in e ia y-ca e cen es (11% s. 1%,
p 0.01, and 8% s. 1%, p 0.04); enous ca he e s we e also
mo e equen (90% s. 73%, p <0.001) bu p e ious endo-
scopic p ocedu es we e less equen (5% s. 12%, p 0.01).
Skin and so issue in ec ions we e he mos equen
sou ce o BSI in e ia y-ca e hospi als (21% s. 0, p 0.04),
whe eas u ina y ac and in a-abdominal in ec ions we e
less equen (14% s. 24%, p 0.02, and 7% s. 20%,
p <0.001). The e we e no signi ican di e ences in ae iology
o mo ali y.
Discussion
The conside a ion o HCA as a speci ic ca ego y o BSI has
been ound o be clinically ele an in he ew s udies
add essing he issue ha we a e awa e o [5–9] because i
was disco e ed o be mo e simila o HA han o CA epi-
sodes wi h espec o unde lying condi ions and ae iology.
We we e in e es ed in in es iga ing whe he he HCA ca e-
go y was jus as ele an in a b oade sample o hospi als
ha included e ia y-ca e and communi y hospi als. We
used he c i e ia empi ically de eloped by F iedman e al. [5]
because hey ha e been shown o be o p ognos ic impo -
ance [5,8] and p edic i e o ine ec i e ini ial he apy [16].
TABLE 3. Mic oo ganisms isola ed om blood cul u es, by
epidemiological ype o in ec ion
CA
(n= 150)
HCA
(n= 195)
HA
(n= 476)
G am-posi i e 67 (45) 74 (38)
a
240 (50)
S aphylococcus au eus 10 (7) 22 (11) 67 (14)
Me hicillin- esis an S. au eus
b
0
c
6 (3%) 24 (5%)
Coagulase-nega i e s aphylococci 8 (5) 20 (10)
a
118 (25)
En e ococcus 5 (3) 8 (4)
a
47 (10)
S aphylococcus pneumoniae 27 (18)
d
11 (6)
a
3 (<1)
G am-nega i e 85 (57) 124 (64)
a
237 (50)
Esche ichia coli 57 (38) 72 (37)
a
96 (20)
ESBL-p oducing E. coli
e
5 (3) 6 (3) 18 (4)
Klebsiella spp.12 (8) 13 (7) 39 (8)
ESBL-p oducing Klebsiella spp.
1 (<1) 1 (<1) 3 (<1)
En e obac e spp.1 (<1) 8 (4) 17 (4)
Pseudomonas ae uginosa
g
1 (<1)
d
13 (7) 26 (5)
Acine obac e baumannii 0 2 (1) 23 (5)
S eno ophomonas mal ophilia 0 1 (<1) 5 (1)
Anae obes 4 (3) 3 (2) 6 (1)
Fungi 0 1 (1)
h
15 (3)
Polymic obial bac e aemia 12 (8) 10 (5) 38 (8)
Da a a e exp essed as he numbe o cases (%), excep whe e speci ied. The
p alue o all s a is ical compa isons ‡0.05, excep whe e speci ied.
CA, communi y-acqui ed; HCA, heal hca e associa ed; HA, hospi al-acqui ed.
a
Fo compa ison wi h hospi al-acqui ed episodes, p £0.01 (Fishe ’s exac es ).
b
Me hicillin- esis an isola es we e 0, 27% and 36% o all S. au eus in CA, HCA
and HA bloods eam in ec ion (BSI), espec i ely.The p alue o CA s. HA
was 0.02 (Fishe ’s exac es ).
c
Fo compa ison wi h heal hca e-associa ed episodes, p 0.03 (Fishe ’s exac
es ).
d
Fo compa ison wi h heal hca e-associa ed episodes, p £0.01 (Fishe ’s exac
es ).
e
Ex ended-spec um b-lac amase (ESBL)-p oducing isola es we e 9%, 8% and
19% o all E. coli in CA, HCA, and HA BSI, espec i ely. The p alue o HCA
s. HA was 0.05 (Fishe ’s exac es ).
ESBL-p oducing isola es we e 8% o all Klebsiella spp. in CA, HCA, and HA BSI.
g
Ca bapenem- esis an isola es we e 0, 15% and 12% all P. ae uginosa in CA,
HCA, and HA BSI, espec i ely. The p alue o HCA s. HA was 0.05 (Fishe ’s
exac es ).
h
Fo compa ison wi h hospi al-acqui ed episodes, p 0.04 (Fishe ’s exac es ). 3
4Clinical Mic obiology and In ec ion CMI
ª2010 The Au ho s
Jou nal Compila ion ª2010 Eu opean Socie y o Clinical Mic obiology and In ec ious Diseases, CMI
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
Twen y- ou pe cen o BSI in ec ions epo ed in his
mul icen e s udy including public e ia y-ca e and commu-
ni y hospi als in Andalucı
´a, Spain, we e o he HCA ype.
This is a simila igu e o ha epo ed by Valle
´se al. [8] in
hei s udy pe o med in wo eaching hospi als in Ba celona
(24.5%), and lowe han ha epo ed by F iedman e al. [5]
in ou hospi als in No h Ca olina (37%), using simila de i-
ni ions. Mos o ou HCA BSI pa ien s had ecen ly been
admi ed o an acu e ca e hospi al, and some we e esiden s
in a long- e m ca e acili y o ecei ing haemodialysis. The
equency o BSI in pa ien s wi h ambula o y in a enous
ea men o specialized home ca e was low, whe eas i was
much highe in No h Ca olina [5]. This p obably e lec s
signi ican di e ences be ween he Spanish and US heal hca e
sys ems in luencing he epidemiology o BSI. We ound ha
he pe cen age o HCA episodes was simila in communi y
and e ia y-ca e cen es, indica ing ha his is a ca ego y
o BSI ha should be aken in o accoun in bo h ypes o
hospi als.
The da a ob ained in he p esen s udy con i m ha HCA
and CA BSIs show signi ican di e ences ha physicians
a ending pa ien s in he eme gency oom should bea in
mind when a ending pa ien s wi h communi y-onse sepsis
[5–9,17]. Beyond he ea u es included in he de ini ion o
HCA BSI, pa ien s in his g oup mo e equen ly showed
a al unde lying condi ions, ch onic enal insu iciency, haema-
ological cance , p e ious an imic obial ea men and an
unknown sou ce o in ec ion. Mo e impo an ly, some ypical
heal hca e-associa ed an ibio ic- esis an o ganisms, such as
P. ae uginosa o me hicillin- esis an S. au eus (MRSA), we e
no p esen o a e in CA episodes bu should be consid-
e ed in pa ien s wi h a suspec ed HCA BSI (CA MRSA is s ill
anecdo al in Spain) [18]. Cu iously, o he an ibio ic- esis an
o ganisms (such as ESBL p oduce s among E. coli o K. pneu-
moniae) we e ound a simila equencies in CA, HCA and
HA episodes, e lec ing he ac ha ESBL-p oducing en e o-
bac e ia a e now signi ican causes o CA BSI [10] and ha
he c i e ia used o de ine an HCA episode a e no use ul
o uling ou hese o ganisms in pa ien s wi h communi y-
onse BSI.
As a as we a e awa e, he cha ac e is ics o pa ien s wi h
HCA BSI episodes by ype o hospi al ha e no p e iously
been s udied. Signi ican di e ences we e ound be ween
HCA episodes occu ing in e ia y-ca e and communi y cen-
es, e lec ing he ype o ou pa ien s ca ed o in bo h
ypes o cen es. Thus, haemodialysis and unnelled ca he e s
we e mo e equen in e ia y-ca e hospi als, whe eas
pa ien s wi h HCA BSI in communi y hospi als we e mainly
ecen ly admi ed pa ien s wi h ch onic condi ions o
pa ien s wi h solid cance ecei ing in a enous chemo-
he apy in a day hospi al. The e was a highe c ude mo ali y
a e o HCA BSI pa ien s ea ed in communi y cen es
compa ed o e ia y-ca e hospi als, al hough he esul s o
he mul i a ia e analysis sugges ha he c ude associa ion
was con ounded by o he a iables.
Al hough benchma k a es exis s o HA BSI and o al BSI
[1–3], we ha e p o ided incidence a es o HCA BSI.
Because he mos app op ia e denomina o o HCA BSI (i.e.
he numbe o pa ien s o pa ien -days a isk) is e y di i-
cul o ob ain, we op ed o p o ide bo h popula ion-based
and admission-based a es. We conside popula ion-based
a es o be mo e app op ia e because no e e y pa ien wi h
HCA BSI is admi ed o hospi al. Because hospi als pa icipa -
ing in he s udy a end >90% o pa ien s in hei ca chmen
a ea needing hospi aliza ion, he popula ion-based a es es i-
ma ed in he p esen s udy should be conside ed as mini-
mum a es.
The p esen s udy has se e al limi a ions. The numbe o
cases om communi y hospi als was lowe compa ed o ha
om e ia y-ca e cen es, which limi s he compa isons
be ween he wo ypes o hospi al. Also, he s udy pe iod
was longe in he communi y hospi als and, because commu-
ni y-onse BSI may show seasonal di e ences, his may ha e
in luenced he esul s. Howe e , he s udy pe iods we e
chosen aiming o minimize he impac o he di e en
pa ame e s. In addi ion, e en hough blood cul u es we e
p ocessed in all cen es using s anda dized p o ocols [12,13],
he e may ha e been some di e ences among he hospi als.
Finally, he esul s ob ained a e p obably no applicable o
a eas wi h di e en a epidemiology o esis an bac e ia (such
as endemic CA MRSA) and di e en heal hca e sys ems.
In summa y, HCA BSI should be conside ed as a dis inc
class o BSI, in bo h e ia y-ca e and communi y hospi als,
al hough di e ences in he p edisposing condi ions o he
pa ien s and clinical ea u es should be aken in o accoun .
Acknowledgemen s
The pa icipan s om he SAEI/SAMPAC Bac e aemia G oup
we e: F. Rod ı
´guez (Hospi al Uni e si a io Reina So ı
´a, Co
´ -
doba), Ma ina de Cue o (Hospi al Uni e si a io Vi gen Maca-
ena, Se illa), Ma ı
´a V. Ga cı
´a (Hospi al Uni e si a io Vi gen
de la Vic o ia, Ma
´laga), Ve o
´nica Gonza
´lez-Gala
´n (Hospi ales
Uni e si a ios Vi gen del Rocı
´o, Se illa), Fe nando Fe na
´n-
dez-Sa
´nchez (Hospi al Cos a del Sol, Ma bella, Ma
´laga), Ma ı
´a
J. Gu ie
´ ez (Hospi al de la Se anı
´a, Ronda, Ma
´laga), An onio
Sa
´nchez-Po o (Hospi al de La Lı
´nea, La Lı
´nea de la Concep-
cio
´n, Ca
´diz), Be a Bece il (Hospi al Pun a de Eu opa,
Algeci as, Ca
´diz), Ana Ga cı
´a-Tapia (Hospi al Pue a del Ma ,
CMI Rod ı
´guez-Ban
˜oe al. Bac e aemia in e ia y-ca e and communi y hospi als 5
ª2010 The Au ho s
Jou nal Compila ion ª2010 Eu opean Socie y o Clinical Mic obiology and In ec ious Diseases, CMI
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

Ca
´diz), Juan C. Alados (Hospi al de Je ez, Je ez de la
F on e a, Ca
´diz), Fede ico Acos a (Hospi al de An eque a,
Ma
´laga), Ca men Flo ez (Hospi al Uni e si a io de Valme,
Se illa), Pe a Na as (Hospi al To eca
´ denas, Alme ı
´a),
Ma ı
´a A. Ma ı
´nez-Pe
´ ez (Hospi al Uni e si a io San Cecilio,
G anada), Inmaculada Ca azo (Hospi al de Jae
´n, Jae
´n).
Au ho con ibu ions
JRB and MDLP we e he coo dina o s o he s udy. JRB,
MDLP, MMP and PR a e esponsible o he s udy design,
which was ac i ely discussed wi h all au ho s and modi ied
acco dingly. MMP, PR, CN, EN, MH, AA, AM, FT, MTT,
AMA, AA, AR, RM, JEC, LL and JAPL coo dina ed he me h-
odology and da a collec ion in each hospi al and e i ied he
da abases. JRB, MMP and PR pe o med he p elimina y anal-
ysis and w o e he d a o he manusc ip , which was dis-
cussed wi h all au ho s and modi ied acco dingly.
T anspa ency Decla a ion
This s udy was unded by Conseje ı
´a de Salud, Jun a de And-
alucı
´a (0063/2006 and PI0048/2008), Minis e io de Sanidad y
Consumo, Ins i u o de Salud Ca los III-FEDER, Spanish Ne -
wo k o he Resea ch in In ec ious Diseases (REIPI RD06/
0008) and FIS (PI070190). All au ho s decla e ha he e a e
no con lic s o in e es .
Re e ences
1. Weins ein MP, Towns ML, Qua e y SM e al. The clinical signi icance
o posi i e blood cul u es in he 1990s: a p ospec i e comp ehensi e
e alua ion o he mic obiology, epidemiology, and ou come o bac e -
emia and ungemia in adul s. Clin In ec Dis 1997; 24: 584–602.
2. Wisplingho H, Bischo T, Tallen SM e al. Nosocomial bloods eam
in ec ions in US hospi als: analysis o 24,179 cases om a p ospec i e
na ionwide su eillance s udy. Clin In ec Dis 2004; 39: 309–317.
3. Rod ı
´guez-C eixems M, Alcala
´L, Mun
˜oz P e al. Bloods eam in ec-
ions. E olu ion and ends in he mic obiology wo kload, incidence
and e iology, 1985–2006. Medicine 2008; 87: 234–249.
4. Ga ne JS, Ja is WR, Emo i TG e al. CDC de ini ions o nosoco-
mial in ec ions. Am J In ec Con ol 1988; 16: 128–140.
5. F iedman ND, Kaye KS, S ou JE e al. Heal h ca e-associa ed blood-
s eam in ec ions in adul s; a eason o change he accep ed de ini-
ion o communi y-acqui ed in ec ions. Ann In e n Med 2002; 137:
791–797.
6. Siegman-Ig a Y, Fou e B, O ni-Wassenlau R e al. Reapp aisal o
communi y-acqui ed bac e emia: a p oposal o a new classi ica ion
o he spec um o acquisi ion o bac e emia. Clin In ec Dis 2002; 34:
1431–1439.
7. Sho AF, Tabak YP, Killiam AD, Gup a V, Liu LZ, Kolle MH. Heal h-
ca e-associa ed bloods eam in ec ion: a dis inc en i y? Insigh s om
a la ge U.S. da abase C i Ca e Med 2006; 34: 1588–1595.
8. Valles J, Calbo E, Ano o A e al. Bloods eam in ec ions in adul s:
impo ance o heal hca e-associa ed in ec ions. J In ec 2008; 56: 27–
34.
9. Raymond NJ, Blackmo e TK, Humble MW e al. Bloods eam in ec-
ions in a seconda y and e ia y ca e hospi al se ing. In e n Med J
2006; 36: 765–772.
10. Rod ı
´guez-Ban
˜o J, Na a o MD, Rome o L e al. Bac e emia due o
ex ended-spec um be a-lac amase-p oducing Esche ichia coli in he
CTX-M e a: a new clinical challenge. Clin In ec Dis 2006; 43: 1407–
1414.
11. Es adı
´s ica de es ablecimien os sani a ios con e
´gimen de in e nado.
Agencia de Calidad del Sis ema Nacional de Salud, Ins i u o de
In o macio
´n Sani a ia, Minis e io de Sanidad y Consumo 2005. A ail-
able a : h p://www.msc.es/es adEs udios/es adis icas/docs/es Hosp05/
publicacionESCRI2005.pd 1
.
12. Loza Fe na
´ndez de Bobadilla E, Planes Reig A, Rod ı
´guez-C eixems
M. Hemocul i os. In: P ocedimien os en Mic obiologı
´a Clı
´nica. Socie-
dad Espan
˜ola de En e medades In ecciosas y Mic obiologı
´a Clı
´nica,
2003. A ailable a : h p://www.seimc.o g/documen os/p o ocolos/mi-
c obiologia. Accessed 15 June 2009.
13. Clinical and Labo a o y S anda ds Ins i u e 2005. Pe o mance s an-
da ds o an imic obial suscep ibili y es ing. 15 h in o ma ional supple-
men . App o ed s anda d M100-S15. Wayne, PA: CLSI, 2005.
14. McCabe WR, Jackson GG. G am-nega i e bac e emia I. E iology and
ecology. A ch In e n Med 1962; 110: 847–855.
15. Se icio Andaluz de Salud. Memo ia 2006. Se ille, Spain: Conseje ı
´a
de Salud, Jun a Andalucı
´a, 2007. A ailable a : h p://www.sas.jun a-
andalucia.es/publicaciones/lis ado.asp?ma e =7.
16. McDonald JR, F iedman ND, S ou JE e al. Risk ac o s o ine ec i e
he apy in pa ien s wi h bloods eam in ec ion. A ch In e n Med 2005;
165: 308–313.
17. Cisne os-He e os JM, Cobo-Reinoso J, Pujol-Rojo M e al. Guı
´a
pa a el diagno
´s ico y a amien o del pacien e con bac e iemia.
Guı
´as de la Sociedad Espan
˜ola de En e medades In ecciosas y
Mic obiologı
´a Clı
´nica (SEIMC). En e m In ecc Mic obiol Clin 2007; 25:
111–130.
18. Rod ı
´guez-Ban
˜o J, Domı
´nguez MA, Milla
´nAe al. Clinical and molecu-
la epidemiology o communi y, heal h ca e-associa ed and nosoco-
mial me hicillin- esis an S aphylococus au eus in Spain. Clin Mic obiol
In ec 2009; 15: 1111–1118.
6Clinical Mic obiology and In ec ion CMI
ª2010 The Au ho s
Jou nal Compila ion ª2010 Eu opean Socie y o Clinical Mic obiology and In ec ious Diseases, CMI
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
Au ho Que y Fo m
Jou nal: CLM
A icle: 3089
Dea Au ho ,
Du ing he copy-edi ing o you pape , he ollowing que ies a ose. Please espond o hese by ma king up you
p oo s wi h he necessa y changes/addi ions. Please w i e you answe s on he que y shee i he e is insu icien
space on he page p oo s. Please w i e clea ly and ollow he con en ions shown on he a ached co ec ions
shee . I e u ning he p oo by ax do no w i e oo close o he pape ’s edge. Please emembe ha illegible
ma k-ups may delay publica ion.
Many hanks o you assis ance.
Que y
e e ence
Que y Rema ks
Q1 AUTHOR: Please check his websi e add ess and con i m ha i
is co ec . (Please no e ha i is he esponsibili y o he
au ho (s) o ensu e ha all URLs gi en in his a icle a e co ec
and useable.)
Q2 AUTHOR: some % alues ha e been co ec ed: HCA, 17, 17 and
2, HA, 68 – please e i y.
Q3 AUTHOR: some % alues ha e been co ec ed: CA, 57; HCA, 64
and HA, 5 – please e i y.
MARKED PROOF
Please co ec and e u n his se
Ins uc ion o p in e
Lea e unchanged unde ma e o emain
h ough single cha ac e , ule o unde line
New ma e ollowed by
o
o
o
o
o
o
o
o
o
and/o
and/o
e.g.
e.g.
unde cha ac e
o e cha ac e
new cha ac e
new cha ac e s
h ough all cha ac e s o be dele ed
h ough le e o
h ough cha ac e s
unde ma e o be changed
unde ma e o be changed
unde ma e o be changed
unde ma e o be changed
unde ma e o be changed
Enci cle ma e o be changed
(As abo e)
(As abo e)
(As abo e)
(As abo e)
(As abo e)
(As abo e)
(As abo e)
(As abo e)
linking cha ac e s
h ough cha ac e o
whe e equi ed
be ween cha ac e s o
wo ds a ec ed
h ough cha ac e o
whe e equi ed
o
indica ed in he ma gin
Dele e
Subs i u e cha ac e o
subs i u e pa o one o
mo e wo d(s)
Change o i alics
Change o capi als
Change o small capi als
Change o bold ype
Change o bold i alic
Change o lowe case
Change i alic o up igh ype
Change bold o non-bold ype
Inse ‘supe io ’ cha ac e
Inse ‘in e io ’ cha ac e
Inse ull s op
Inse comma
Inse single quo a ion ma ks
Inse double quo a ion ma ks
Inse hyphen
S a new pa ag aph
No new pa ag aph
T anspose
Close up
Inse o subs i u e space
be ween cha ac e s o wo ds
Reduce space be ween
cha ac e s o wo ds
Inse in ex he ma e
Tex ual ma k Ma ginal ma k
Please use he p oo co ec ion ma ks shown below o all al e a ions and co ec ions. I you
in da k ink and a e made well wi hin he page ma gins.
wish o e u n you p oo by ax you should ensu e ha all amendmen s a e w i en clea ly