ORIGINAL RESEARCH
Ra es and P edic o s o T ea men Failu e
in S aphylococcus au eus P os he ic Join In ec ions
Acco ding o Di e en Managemen S a egies:
A Mul ina ional Coho S udy—The ARTHR-IS S udy
G oup
Reinaldo Espı
´ndola .Venanzio Vella .Na i idad Beni o .Isabel Mu .Sa a Tedeschi .
Eleono a Zampa ini .Johannes G. E. Hend iks .Luisa So lı
´.Osca Mu illo .Lau a Solde ila .
Ma hew Sca bo ough .Clai e Sca bo ough .Jan Kluy mans .Ma eo Ca lo Fe a i .
Ma hias W. Ple z .Iain Mcnama a .Rosa Escude o-Sanchez .Ced ic A ieux .
Cecile Ba aille .F e
´de
´ ic-An oine Dauchy .Wai-Yan Liu .Jaime Lo a-Tamayo .Julia P aena .
And ew Us ianowski .Elisa Cinconze .Michele Pelleg ini .Fabio Bagnoli .Jesu
´s Rod ı
´guez-Ban
˜o.
Ma ia Dolo es del To o
Recei ed: June 23, 2022 / Accep ed: Sep embe 20, 2022 / Published online: Oc obe 15, 2022
ÓThe Au ho (s) 2022
ABSTRACT
In oduc ion: Guidelines ha e imp o ed he
managemen o p os he ic join in ec ions (PJI).
Howe e , i is necessa y o eassess he inci-
dence and isk ac o s o ea men ailu e (TF)
o S aphylococcus au eus PJI (SA-PJI) including
unc ional loss, which has so a been neglec ed
as an ou come.
Me hods: A e ospec i e coho s udy o SA-PJI
was pe o med in 19 Eu opean hospi als
be ween 2014 and 2016. The ou come a iable
was TF, including ela ed mo ali y, clinical
ailu e and unc ional loss bo h a e he ini ial
su gical p ocedu e and a e all p ocedu es a
18 mon hs. P edic o s o TF we e iden i ied by
logis ic eg ession. Landma k analysis was used
Jesu
´s Rod ı
´guez-Ban
˜o and Ma ia Dolo es del To o
con ibu ed equally as senio au ho s o he s udy.
Membe s o The ARTHR-IS G oup a e lis ed in
Acknowledgemen s sec ion.
Supplemen a y In o ma ion The online e sion
con ains supplemen a y ma e ial a ailable a h ps://
doi.o g/10.1007/s40121-022-00701-0.
R. Espı
´ndola
In ec ious Diseases and Mic obiology Clinical Uni ,
Hospi al Uni e si a io Vi gen Maca ena, Se ille,
Spain
V. Vella E. Cinconze M. Pelleg ini F. Bagnoli
GlaxoSmi hKline (GSK), Siena, I aly
N. Beni o I. Mu
In ec ious Diseases Uni , Hospi al de la San a C eu i
San Pau, Ba celona, Spain
N. Beni o I. Mu
San Pau Ins i u e o Biomedical Resea ch,
Ba celona, Spain
N. Beni o I. Mu
Depa men o Medicine, Uni e si a Au o
`noma de
Ba celona, Ba celona, Spain
S. Tedeschi
Depa men o Medical and Su gical Sciences,
Uni e si y o Bologna, Bologna, I aly
S. Tedeschi E. Zampa ini
In ec ious Diseases Uni , IRCCS Azienda
Ospedalie o Uni e si a ia di Bologna, Bologna, I aly
J. G. E. Hend iks
Depa men o O hopaedic Su ge y and T auma,
Ma
´xima MC, Eindho en, The Ne he lands
.The ARTHR-IS G oup
In ec Dis The (2022) 11:2177–2203
h ps://doi.o g/10.1007/s40121-022-00701-0
o a oid immo al ime bias wi h i ampicin
when deb idemen , an ibio ics and implan
e en ion (DAIR) was pe o med.
Resul s: One hund ed wen y cases o SA-PJI
we e included. TF a es a e he i s and all
su gical p ocedu es pe o med we e 32.8% and
24.2%, espec i ely. A e all p ocedu es, unc-
ional loss was 6.0% o DAIR and 17.2% o
p os hesis emo al. Va iables independen ly
associa ed wi h TF o he i s p ocedu e we e
Cha lson C2, haemoglobin 10 g/dL, bac e -
aemia, polymic obial in ec ion and addi ional
deb idemen (s). Fo DAIR, TF was also associ-
a ed wi h a body mass index (BMI) [30 kg/m
2
and delay o DAIR, while i ampicin use was
p o ec i e. Fo all p ocedu es, he a iables
associa ed wi h TF we e haemoglobin 10 g/
dL, hip ac u e and addi ional join su ge y no
ela ed o pe sis en in ec ion.
Conclusions: TF emains common in SA-PJI.
Func ional loss accoun ed o a subs an ial
p opo ion o ea men ailu es, pa icula ly
a e p os hesis emo al. Use o i ampicin a e
DAIR was associa ed wi h a p o ec i e e ec .
Among he isk ac o s iden i ied, anaemia and
obesi y ha e no equen ly been epo ed in
p e ious s udies.
T ial egis a ion: This s udy is egis e ed a
clinical ials.go , egis a ion no.
NCT03826108.
L. So lı
´
Depa men o In ec ious Diseases, Hospi al del Ma ,
Ins i u Hospi al del Ma d’In es igacions Me
`diques
(IMIM), Ba celona, Spain
L. So lı
´
Uni e si a Pompeu Fab a, Ba celona, Spain
O. Mu illo L. Solde ila
Depa men o In ec ious Diseases, Hospi al
Uni e si a i Bell i ge, IDIBELL, Ba celona, Spain
M. Sca bo ough C. Sca bo ough
Bone In ec ion Uni , Nu ield O hopaedic Cen e,
Ox o d Uni e si y Hospi als NHS Founda ion T us ,
Ox o d, UK
J. Kluy mans
Depa men o In ec ion Con ol, Amphia Hospi al,
B eda, The Ne he lands
J. Kluy mans
Depa men o Medical Mic obiology, Uni e si y
Medical Cen e U ech , U ech Uni e si y,
U ech , The Ne he lands
M. C. Fe a i
P os he ic-Join Replacemen Uni , Humani as
Resea ch Hospi al, Milan, I aly
M. W. Ple z
Cen e o In ec ious Diseases and In ec ion
Con ol, Jena Uni e si y Hospi al, Jena, Ge many
I. Mcnama a
Depa men o O hopaedics, No olk and No wich
Uni e si y Hospi al, No wich, UK
R. Escude o-Sanchez
Depa men o In ec ious Diseases, Hospi al
Uni e si a io Ramo
´n y Cajal, Mad id, Spain
R. Escude o-Sanchez J. Rod ı
´guez-Ban
˜o
M. D. del To o (&)
Cen o de In es igacio
´n en Red de En e medades
In ecciosas (CIBERINFEC), Ins i u o de Salud Ca los
III, Mad id, Spain
e-mail: [email p o ec ed]
C. A ieux
Depa men o In ec ious Diseases, Cen e
Hospi alie Uni e si ai e de Rennes, Rennes, F ance
C. Ba aille
O hopedic Su ge y Depa men , C oix Rousse
Hospi al, Lyon, F ance
F.-A. Dauchy
Depa men o In ec ious and T opical Diseases,
Cen e Hospi alie Uni e si ai e de Bo deaux,
Bo deaux, F ance
W.-Y. Liu
Depa men o O hopaedic Su ge y and T auma,
Ca ha ina Hospi al, Eindho en, The Ne he lands
2178 In ec Dis The (2022) 11:2177–2203
G aphical abs ac :
W.-Y. Liu
Depa men o O hopaedic Su ge y and T auma,
Ma
´xima MC, Eindho en, The Ne he lands
J. Lo a-Tamayo
Depa men o In e nal Medicine, Hospi al
Uni e si a io, 12 de Oc ub e, Mad id, Spain
J. P aena
Clinical Uni o In ec ious Diseases and
Mic obiology, Hospi al Uni e si a io Vi gen del
Rocı
´o, Se ille, Spain
A. Us ianowski
In ec ious Diseases Uni , No h Manches e Gene al
Hospi al, Manches e , UK
J. Rod ı
´guez-Ban
˜oM. D. del To o
In ec ious Diseases and Mic obiology Clinical Uni ,
Hospi al Uni e si a io Vi gen Maca ena, A da D
Fed iani 3, 41009 Se ille, Spain
In ec Dis The (2022) 11:2177–2203 2179
PLAIN LANGUAGE SUMMARY
S aphylococcus au eus is one o he mos i ulen
bac e ia and equen ly causes p os he ic join
in ec ions.
Knowledge o he ea men o his ype o
in ec ion has ad anced in ecen yea s, and
ea men guidelines ha e led o imp o ed
managemen . Typically, he success ul ea -
men o hese in ec ions has been de e mined by
clinical cu e, ha is, he symp oms o in ec ion
ha e disappea ed, bu has no aken in o accoun
loss o unc ion (such as signi ican di icul ies
walking), which is c i ical o he pa ien ’s qual-
i y o li e. Ou aim in his s udy was o e alua e
he success o cu en managemen s a egies o
S. au eus p os he ic join in ec ion, including
eco e y o unc ionali y, and he ac o s ha
p edic why some o hese in ec ions a e no
cu ed, o iden i y a eas o imp o emen .
In a mul ina ional coho o 128 pa ien s wi h
S. au eus p os he ic join in ec ion, a es o
ea men ailu e we e ound o be high, wi h
signi ican a es o loss o unc ion, especially
when he p os hesis needed o be emo ed. Loss
o unc ion was less equen when he in ec ion
was ini ially ea ed wi h su gical cleaning wi h-
ou emo al o he p os hesis, e en when his
p ocedu e ailed a i s . We ound ha anaemia
and obesi y we e associa ed wi h lowe ea men
success, and ha he p obabili y o ea men
success inc eased when su gical cleaning wi h-
ou p os hesis emo al was pe o med ea ly, and
when he an ibio ic i ampicin was used in
combina ion wi h ano he an ibio ic.
Keywo ds: P os he ic join in ec ion;
S aphylococcus au eus; Ou come; Clinical
ailu e; Func ional ailu e
Key Summa y Poin s
Why ca y ou his s udy?
S aphylococcus au eus has been associa ed
wi h a highe a e o p os he ic join
in ec ion (PJI) and ea men ailu e
compa ed wi h o he pa hogens.
Changes in he managemen o S. au eus
PJI (SA-PJI) in ecen decades make i
necessa y o eassess he incidence and
isk ac o s o ea men ailu e, including
unc ional loss which has p e iously been
neglec ed as an ou come.
Wha was lea ned om he s udy?
A conside able p opo ion o SA-PJIs ailed
a e ini ial su gical ea men s, al hough
a subs an ial pa o hem can be escued
by addi ional p ocedu es.
Signi ican unc ional loss mus be
conside ed in addi ion o clinical ailu e,
mainly in pa ien s in whom he p os hesis
was emo ed, wi h no possibili y o
eco e y.
In pa ien s managed wi h deb idemen ,
an ibio ic and p os hesis e en ion
(DAIR), he loss o unc ion was lowe
e en i his p ocedu e ails.
Anaemia and obesi y we e isk ac o s o
ea men ailu e ha a e a ely epo ed.
The impo ance o some isk ac o s o
ea men ailu e was ein o ced,
including he p o ec i e ole o
i ampicin-based ea men in DAIR.
DIGITAL FEATURES
This a icle is published wi h digi al ea u es,
including [lis digi al ea u es a ailable e.g. a
ideo abs ac and slide deck], o acili a e
unde s anding o he a icle. To iew digi al
J. Rod ı
´guez-Ban
˜oM. D. del To o
Depa men o Medicine, School o Medicine,
Uni e si y o Se illa, Se ille, Spain
J. Rod ı
´guez-Ban
˜oM. D. del To o
Biomedicine Ins i u e o Se illa (IBiS), Se ille, Spain
P esen Add ess:
R. Espı
´ndola
Hospi al Uni e si a io de Valme, Se ille, Spain
2180 In ec Dis The (2022) 11:2177–2203
ea u es o his a icle go o h ps://doi.o g/10.
6084/m9. igsha e.21163246.
INTRODUCTION
S aphylococcus au eus is widely ecognized as a
key mic oo ganism causing p os he ic join
in ec ion (PJI), and has been associa ed wi h
highe a es o ea men ailu e (TF) compa ed
wi h o he ae iologies [1]. Howe e , s udies
e alua ing TF ha e adi ionally ocused on
ou comes ela ed p ima ily wi h con ol o he
in ec ion [2–8], while unc ional ou comes ha e
been gi en li le conside a ion. A de ini ion o
TF ha in eg a es clinical and unc ional
aspec s could p o ide a mo e ealis ic measu e-
men o he consequences o S. au eus-PJI (SA-
PJI).
TF a es may a y acco ding o pa ien
cha ac e is ics, in ec ion ype (ea ly, delayed o
la e) and su gical managemen . Implemen a-
ion o he ecommenda ions included in he
guidelines o he managemen o PJI [9,10]
may ha e posi i ely in luenced changes in he
a es and p edic o s o TF in gene al, and pa -
icula ly in pa ien s managed wi h deb ide-
men , an ibio ics and implan e en ion (DAIR).
Fo SA-PJI, howe e , he e is li le published
da a on he ou comes associa ed wi h di e en
he apeu ic s a egies and hei p edic o s, and
he iming and ole o i ampicin in combina-
ion wi h o he an ibio ics, pa icula ly in
pa ien s unde going DAIR, emain
con o e sial.
Assessmen s o TF end o be analysed o he
i s su gical p ocedu e pe o med o ea he
in ec ion, which is impo an o e alua e he
ou comes associa ed o he decision abou
which p ocedu e mus be pe o med. Howe e ,
addi ional p ocedu es a e usually pe o med i
ha i s p ocedu e ails o cu e he in ec ion,
which may escue some pa ien s bu could also
a ec he unc ional ou come. E alua ions o TF
conside ing all he p ocedu es pe o med a e
no usually made. Such an e alua ion would
p o ide a mo e global iew o he inal conse-
quences o SA-PJI.
The objec i es o his s udy we e o p o ide
upda ed a es o TF in SA-PJI, aking in o
accoun clinical and unc ional aspec s, and o
in es iga e he p edic o s o TF acco ding o
di e en managemen s a egies. The assess-
men o bo h objec i es will be ca ied ou o
bo h he i s su gical p ocedu e and all addi-
ional p ocedu es pe o med on pa ien s in
whom he i s one ailed.
METHODS
S udy Design, Si es and Pe iod
This s udy is pa o he ARTHR-IS p ojec ( eg-
is e ed a clinical ials.go : NCT03826108) and
was designed o e alua e he incidence, isk
ac o s and p edic o s o SA-PJI TF a e p ima y
hip and knee a h oplas y. A e ospec i e
coho s udy was conduc ed in 19 Eu opean
hospi als o iden i y pa ien s olde han 18 yea s
who ecei ed a p ima y a h oplas y be ween 1
Janua y 2014 and 31 Decembe 2016 and
de eloped pos -su gical knee o hip PJI due o
S. au eus wi hin he i s yea a e he p oce-
du e. The pa icipa ing si es, loca ed in Spain,
I aly, F ance, Ge many, UK and he Ne he -
lands, we e selec ed using he CLIN-NET ne -
wo k (h ps://www.combac e.com/abou /clin-
ne /), based on hei esea ch expe ience and
da a collec ion capabili y.
The STROBE ecommenda ions we e ol-
lowed o he epo ing o he s udy (Supple-
men a y Table S1).
Pa icipan s
Pa ien s wi h pos -su gical hip o knee PJI due
o S. au eus diagnosed wi hin he i s yea a e
p ima y a h oplas y we e included. The c i e ia
used o de ine SA-PJI we e as ollows: p esence
o a leas one sign o symp om o PJI, including
join pain and/o swelling, o a sinus ac
communica ing wi h he p os hesis; and he
isola ion o S. au eus om (a) Cone join aspi-
a e cul u e, (b) C wo pe ip os he ic issue
samples and (c) blood cul u es wi h no o he
ob ious sou ce o in ec ion. The pa ien s we e
iden i ied by e iewing medical eco ds om
In ec Dis The (2022) 11:2177–2203 2181

mic obiological labo a o y da abases, local PJI
su eillance da abases and discha ge epo s.
Va iables and De ini ions
The p ima y ou come a iable was TF un il
mon h 18 a e he i s su gical p ocedu e pe -
o med, and was analysed sepa a ely bo h o
he i s su gical p ocedu e pe o med (mim-
icking an in en ion- o- ea analysis o ha
p ocedu e) and o all p ocedu es pe o med
(including hose pe o med a e ailu e o he
i s one). TF was de ined as a composi e a i-
able including SA-PJI- ela ed mo ali y, clinical
ailu e and unc ional ailu e. Clinical ailu e
was de ined as pe sis ence o ecu ence o signs
o symp oms o in ec ion. Fo he analysis o
he i s p ocedu e, his also included he need
o an addi ional cou se o an ibio ics beyond
he ini ial one, he need o use long- e m sup-
p essi e an ibio ic he apy and need o p os-
hesis emo al i no pe o med as he ini ial
su gical p ocedu e. Func ional ailu e was
de ined as impeded o signi ican ly impai ed
walking due o p os he ic loosening o he need
o pe o m a Gi dles one p ocedu e o
a h odesis. Finally, TF was also analysed in he
subg oup o pa ien s who unde wen DAIR as
he i s p ocedu e.
Po en ial p edic o s o TF we e selec ed
acco ding o p e ious s udies [2–8] and addi-
ional hypo heses de eloped by he p ojec
eam, and a e included in Tables 1and 2.
DAIR as p ima y ea men p ocedu e was
conside ed app op ia e i i was pe -
o med 21 days om onse o PJI symp oms,
he e was no sinus ac communica ing wi h
he join p os hesis and eplacemen o poly-
e hylene o mobile componen s was pe o med
acco ding o IDSA guidelines [9]. The de ini-
ions o o he a iables a e included in he
Supplemen a y Table S2.
Da a Collec ion and E hical Aspec s
Da a collec ion was supe ised locally by s a
wi h ele an expe ise in he ield. Da a we e
en e ed in o an anonymized elec onic case
epo o m and checked o missing alues and
inconsis encies. The s udy was app o ed by he
E hics Commi ees a each si e (Supplemen a y
Table S5). The need o ob ain w i en in o med
consen was wai ed owing o he e ospec i e
na u e o he s udy and anonymized da a,
excep in he case o he F ench hospi als whe e
a le e o non-opposi ion was sen o eligible
pa ien s. All pa ien s included in hese cen es
he e o e ga e hei au ho iza ion o
pa icipa e.
S a is ical Analysis
Fo bi a ia e analysis o he associa ion o
exposu e a iables wi h TF, ela i e isks wi h
95% con idence in e al (CI) we e calcula ed;
p alues we e calcula ed by Chi-squa e o Fish-
e ’s exac es , as app op ia e. Con inuous a i-
ables we e ca ego ized a e analysing o
s a i ied associa ions wi h TF. Mul i a iable
analyses we e pe o med by logis ic eg ession:
he e ec o s udy si e was con olled o using a
gene alized linea mixed model in which s udy
si es we e conside ed andom e ec s. Va iables
wi h p alue 0.15 in bi a ia e analysis, and
hose conside ed as po en ially ele an om
clinical judgemen , we e en e ed in o he
models and selec ed using a manual s epwise
backwa d p ocedu e. Va iables wi h
p alue 0.1 we e kep in he models.
Collinea i y and modi ica ion e ec s be ween
a iables we e s udied when clinically sound.
The p edic i e abili y o each model was
examined by calcula ing hei a eas unde he
ecei e ope a ing cha ac e is ic (AUROC)
cu es wi h 95% CIs. Fo he e ec o i ampicin
in he subg oup o pa ien s who unde wen
DAIR as i s su gical ea men , a p opensi y
sco e (PS) was calcula ed using a non-pa simo-
nious mul i a ia e logis ic eg ession model, in
which he ou come a iable was combina ion
he apy wi h i ampicin. As u he sensi i i y
analyses o he impac o i ampicin he apy,
pa simonious mul i a ia e logis ic eg ession
models we e pe o med, in which one o wo
o he a iables we e emo ed. In addi ion, o
a oid immo al ime bias, landma k analysis
was used, excluding pa ien s who died o ailed
ea men in he i s 21 days a e deb idemen .
2182 In ec Dis The (2022) 11:2177–2203
Table 1 Cha ac e is ics o 128 pa ien s wi h S aphylococ-
cus au eus p os he ic join in ec ions (SA-PJI)
Va iables No. o pa ien s
(pe cen age), excep
whe e specified
Spain 52 (40.6)
F ance 15 (11.7)
Ge many 7 (5.5)
Uni ed Kingdom 16 (12.5)
I aly 17 (13.3)
The Ne he lands 21 (16.4)
Male sex 65 (50.8)
Bo n ab oad 2 (1.5)
Age in yea s; median (IQR) 73 (59.3–80.8)
Body mass index uni s; median
(IQR)
31.2 (25.6–35.1)
ASA 3-4 assessmen o p ima y
a h oplas y
64 (50%)
Cha lson como bidi y index;
median (IQR)
1 (1–2)
Como bidi ies
Ch onic hea ailu e 17 (13.3)
Ch onic pulmona y disease 30 (23.4)
Diabe es melli us 29 (22.7)
Ch onic enal insu ficiency 5 (3.9)
Reason o a h oplas y
Os eoa h i is 83 (64.8)
Hip ac u e 34 (26.6)
Os eonec osis 5 (3.9)
O he s 6 (4.6)
Type o a h oplas y
To al hip a h oplas y 50 (39.1)
Pa ial hip a h oplas y 27 (21.1)
To al knee a h oplas y 47 (36.7)
Pa ial knee a h oplas y 4 (3.1)
Table 1 con inued
Va iables No. o pa ien s
(pe cen age), excep
whe e specified
Me hicillin- esis an
S aphylococcus au eus
28 (21.9)
Polymic obial in ec ion 36 (28.1)
Bac e aemia 25 (19.5)
Days om a h oplas y o onse o
SA-PJI symp oms; median
(IQR)
24 (15–36)
Symp oms and signs o SA-PJI
Fe e 32 (25.0)
Join pain 70 (54.7)
Suppu a ion 89 (69.5)
Celluli is 38 (29.7)
Wound dehiscence 56 (43.8)
A icula swelling 18 (14.1)
Sinus ac 8 (6.3)
Labo a o y da a a diagnosis o
SA-PJI; median (IQR)
Haemoglobin (g/dL) 10.5 (9.5–11.4)
Blood leucocy es (cells/lL) 9600 (7550–12,950)
C- eac i e p o ein (mg/L) 84.5 (23.2–224.2)
E y h ocy e sedimen a ion a e
(mm/h)
68.5 (39.8–92.3)
Type o ini ial su gical p ocedu e
pe o med o ea SA-PJI
Deb idemen and p os hesis
e en ion
99 (77.3)
Pa ial emo al and
eimplan a ion
6 (4.7)
One-s age eplacemen and
eimplan a ion
4 (3.1)
Two-s age eplacemen and
eimplan a ion
13 (10.2)
In ec Dis The (2022) 11:2177–2203 2183
RESULTS
Pa ien Cha ac e is ics and T ea men
Failu e Ra es
A o al o 130 cases o SA-PJI we e de ec ed, and
128 we e included ( ele an ollow-up da a we e
missing o he o he wo). The median numbe
o cases pe hospi al was 7 (in e qua ile ange
[IQR] 5–9). The median age o pa ien s was
73 yea s (IQR 59–81 yea s); 65 (50.8%) we e
males; 77 (60.2%) had hip a h oplas y (50 o al
and 27 pa ial) and 51 (39.8%) knee a h o-
plas y (47 o al and 4 pa ial). Pa ien
cha ac e is ics a e p esen ed in Table 1. In ec-
ion- ela ed symp oms s a ed a median o 24
(IQR 15–36) days a e he p ima y a h oplas y
(Supplemen a y Fig. S1), while he i s su gical
p ocedu e o ea men o in ec ion was pe -
o med a median o 4 days (IQR 1–11) a e
symp om onse . Bac e aemia occu ed in 25
cases (19.5%). O e all, ou o 128 PJI cases, 28
(21.9%) we e due o me hicillin- esis an S. au-
eus (MRSA) s ains.
Figu e 1shows pa ien ou comes acco ding
o he i s and addi ional su gical p ocedu es
pe o med. The a e o TF a e he i s p oce-
du e was 32.8% (42 pa ien s; 95% CI
25.2–41.3%). TF was due o clinical ailu e in 27
cases (21.1%), ela ed dea hs in 9 (7%) and loss
o unc ion in 6 (4.7%). Median days un il ail-
u e was 126 (IQR 34–335). Dea hs occu ed a a
median o 21 (IQR 13–48) days a e he i s
su gical p ocedu e pe o med.
A e u he su gical in e en ions, 11
pa ien s who ailed he i s p ocedu e (9 wi h
pe sis en in ec ion, 2 wi h unc ional ailu e
due o p os hesis loosening a e DAIR) we e
escued. A e 18 mon hs o ollow-up, TF was
24.2% (95% CI 17.5–32.3%). The easons o
ailu e we e 9 ela ed dea hs (7.0%), 11 cases o
clinical ailu e and 11 cases o unc ional loss
(8.5%, espec i ely). Excluding unc ional loss,
he ailu e a e was 15.5%.
O he 99 pa ien s who ecei ed DAIR as a
i s in e en ion o ea he SA-PJI (Fig. 1), 31
(31.3%) ailed ea men due o dea h (n=6),
clinical ailu e (n= 23) o loss o unc ion
(n= 2). Despi e u he in e en ions, 15 we e
s ill ailing a he end o he 18-mon h ollow-
up.
O he 29 pa ien s who ecei ed o he ypes
o i s in e en ions o ea SA-PJI, 11 (37%)
ailed ea men , b oken down as ollows: ela-
ed dea h (n= 3), clinical ailu e (n= 4) and loss
o unc ion (n= 4). No wi hs anding u he
in e en ions, 7 we e s ill ailing a he end o
he 18-mon hs ollow-up.
A summa y o he a es and easons o
ea men ailu e is p o ided in Supplemen a y
Table S3, including he a e o TF o p os hesis
emo al as i s p ocedu e.
Table 1 con inued
Va iables No. o pa ien s
(pe cen age), excep
whe e specified
Gi dles one p ocedu e (hip
esec ion a h oplas y)
6 (4.7)
Days om onse o SA-PJI
symp oms o fi s su gical
p ocedu e pe o med; median
(IQR)
4 (1–11)
Days o an ibio ic ea men o
ea PJI; median (IQR)
Empi ical in a enous 2 (1–4)
Ta ge in a enous 16 (11–34)
Ta ge o al 50 (33–80)
To al 73 (56–96)
Ta ge o al an ibio ics
Ri ampicin 103 (80.5)
Fluo oquinolones 73 (63.5)
Clindamycin 17 (14.8)
T ime hop im–sul ame hoxazole 15 (13.0)
Te acyclines 3 (2.6)
Linezolid 4 (3.5)
O he s 3 (2.6)
2184 In ec Dis The (2022) 11:2177–2203
Table 2 Bi a ia e analysis o po en ial p edic o s o ea men ailu e among pa ien s wi h SA-PJI: a e he fi s su gical p ocedu e, a e DAIR, and a e all su gical
p ocedu es pe o med
Failu e a e he ini ial su gical
p ocedu e (n= 128; ailu es = 42)
Failu e a e DAIR as fi s su gical
p ocedu e (n= 99; ailu es = 31)
Failu e a 18 mon hs a e all
p ocedu es pe o med (n= 128;
ailu es = 31)
Va iables No.
ailu e (%)
Rela i e isk
(95% CI)
p-
Value
No.
ailu e (%)
Rela i e isk
(95% CI)
p-
Value
No.
ailu e (%)
Rela i e isk
(95% CI)
p-
Value
Age
C80 yea s 13 (40.6) 1.6 (0.7–3.6) 0.277 8 (38.1) 1.5 (0.5–4.1) 0.450 12 (37.5) 2.4 (1.1–5.8) 0.043
80 yea s 29 (30.2) 23 (29.5) 19 (19.8)
Sex
Male 22 (33.8) 1.1 (0.5–2.3) 0.800 15 (30.6) 1.07 (0.4–2.5) 0.880 17 (26.2) 1.2 (0.5–2.8) 0.378
Female 20 (31.7) 16 (32.0) 14 (22.2)
Cha lson index C2
Yes 21 (47.7) 2.7 (1.3–5.9) 0.009 16 (44.4) 2.6 (1.1–6.1) 0.033 15 (34.1) 2.2 (0.9–5.0) 0.059
No 21 (25.0) 15 (23.8) 16 (19.0)
Haemoglobin 10 mg/dl
Yes 23 (48.9) 3.1 (1.4–6.7) 0.003 18 (50.0) 3.8 (1.6–9.4) 0.002 18 (38.3) 3.2 (1.4–7.5) 0.005
No 19 (23.5) 13 (20.6) 13 (16.0)
Leukocy es C7500/lL
Yes 29 (31.9) 0.9 (0.4–2.2) 0.882 19 (27.9) 1.9 (0.5–7.6) 0.383 24 (26.4) 2.3 (0.7–7.4) 0.211
No 10 (33.3) 9 (37.5) 4 (13.3)
C- eac i e p o ein C150 mg/L
a
Yes 18 (36.7) 1.3 (0.6–2.8) 0.478 10 (31.3) 1.0 (0.4–2.5) 0.992 15 (30.6) 1.7 (0.7–3.8) 0.221
No 22 (30.6) 19 (31.1) 15 (20.8)
In ec Dis The (2022) 11:2177–2203 2185
pa ien s who died o ailed in he i s 21 days
a e deb idemen , use o i ampicin emained
p o ec i e o TF [adjus ed ORs (95%CI), 0.22
(0.06–0.80); AUROC cu e o he model, 0.82
(95% CI 0.72–0.91)].
Table 2also p esen s he bi a ia e analysis o
isk ac o s o pa ien s who ailed a e all p o-
cedu es pe o med. In addi ion o he a iables
iden i ied o he i s p ocedu e, TF was also
associa ed wi h age [80 yea s, p os hesis
emo al as i s p ocedu e, need o pe o m
addi ional join su ge y no due o pe sis en
in ec ion and non-use o i ampicin and luo-
oquinolones in combina ion. On mul i a ia e
analysis, hip ac u e [aOR 4.6 (95% CI
1.6–12.9)], haemoglobin le el 10 g/dL [aOR
2.5 (95% CI 1.0–6.6)] and need o pe o m
addi ional join su ge y no due o pe sis en
in ec ion [aOR 3.2 (95% IC 1.1–8.9)] we e
independen ly associa ed wi h TF a 18 mon hs
(Table 3). The AUROC o he model o he
obse ed da a was 0.80 (95% CI 0.71–0.90).
DISCUSSION
In his s udy, we ound ha nea ly a hi d o
ini ial su gical p ocedu es esul ed in TF. The TF
a e dec eased when u he su gical p ocedu es
we e pe o med. Impo an ly, we es ima ed he
impac o signi ican unc ional loss. When
DAIR was used as he i s p ocedu e, e en
hough i is a less agg essi e s a egy, addi ional
p ocedu es escued a signi ican p opo ion o
ini ial ailu es wi hou inc easing loss o unc-
ion. In addi ion, he p edic o s o TF o SA-PJIs
we e iden i ied, and he ole o i ampicin in
pa ien s unde going DAIR was con i med.
A e iew o he li e a u e on SA-PJI s udies
ocusing on ea men ou comes highligh ed
he di icul ies o compa ing di e en s udy
esul s owing o he e ogenei y in s udy design,
case de ini ions adop ed, leng h o ollow-up
and ypes o analyses used (Table 4). O e all,
p e iously epo ed TF a es anged om 0% o
16.6% o p os hesis emo al [4,5,8,11] and
13.6% o 63% o DAIR [2,4–8,12–15]. I should
also be no ed ha , o he bes o he au ho s’
knowledge, unc ional ou come was no con-
side ed a all in p e ious s udies, despi e i being
c i ical o he quali y o li e o pa ien s. Func-
ional loss is signi ican ly in luenced by he
su gical p ocedu es pe o med and his in o -
ma ion is he e o e ele an o he decision-
making p ocess.
When DAIR was analysed as he ini ial p o-
cedu e, he TF a e was 31.3% (29.2% wi hou
conside ing unc ional loss), which is sligh ly
highe han epo ed in mo e ecen obse a-
ions [7,12] bu lowe han in olde publica-
ions [2–4,6,13] (Table 4). Howe e , he TF a e
dec eased o 21.2% a e addi ional p ocedu es,
and o 15% i unc ional loss was no aken in o
accoun (which is mo e consis en wi h de ini-
ions in p e ious epo s). The lowe TF a es o
DAIR epo ed by he la es s udies (and by his
one) could be a ibu ed o be e pa ien
selec ion o his ea men s a egy and o he
in ol emen o mul idisciplina y eams in he
managemen o PJI. Indeed, Bouaziz e al. [13]
ound an o e all TF a e o 42%, which
dec eased o 30% when DAIR was pe o med
acco ding o he la es guidelines [9]. Ou da a
u he sugges ha app op ia e pa ien selec-
ion a ou s mo e posi i e ou comes, and ha
an ini ial TF can be escued wi hou signi ican
unc ional loss in a conside able numbe o
pa ien s.
The o e all TF a es, including clinical and
unc ional upda es, o p os hesis emo al in
ou coho (34.4%) may seem ela i ely high
when compa ed wi h o he s udies, bu when
only clinical cu e was conside ed: he a e o TF
o SA-PJI was 17.2%, which is simila o ha
ound in o he se ies [4,5,11]. The high p o-
po ion o TF a e p os hesis emo al ela ed o
unc ional loss is no ewo hy and ein o ces he
impo ance o ea ly diagnosis o SA-PJI o
inc ease he likelihood o being ea ed wi h
DAIR.
Since he decision o pe o m DAIR o
emo e he p os hesis as i s p ocedu e is
s ongly in luenced by pa ien and in ec ion
cha ac e is ics, we did no y o compa e he
ou comes o he wo p ocedu es as hey a e no
compa able. Ins ead, we ocused ou analysis on
iden i ying po en ial p edic o s o TF. Rega d-
ing he a iables iden i ied, he Cha lson index
is a p edic o o su i al and also o p ognosis o
many in ec ions; simila ly, o he s udies ha e
2192 In ec Dis The (2022) 11:2177–2203

Table 4 Summa y o published s udies on he ou come and managemen o PJI including C20 cases ocusing on S. au eus
Re e ences Yea s o
diagnosis and
design
Types o in ec ion and
numbe o pa ien s
Numbe o pa ien s and
managemen
Cu e o ailu e defini ion Ou come Fac o s associa ed wi h
ou come
B and e al.
(1997) [2]
1980–1991
Re ospec i e
coho ,
unicen ic
Pos -su gical knee and hip
SA-PJI
N= 33 (3 MRSA)
DAIR (n= 33)
Median 28 days in a enous
(IV) an ibio ics (be a-
lac ams, 91%; ancomycin,
9%)
Failu e: pe sis ence o
ecu ence o clinical signs
o PJI
Failu e:
59% a 1 yea
63% a 2 yea s
Mul i a iable (HR): DAIR
pe o med [2 days a e
symp om onse
Salgado e al.
(2007) [3]
1998–2004
Re ospec i e
coho ,
unicen ic
Hip and knee SA-PJI
N= 45 (12 MRSA)
DAIR (n= 20)
Pa ial/ o al 1 T (n=4)
2T(n= 15)
Resec ion a h oplas y
(n=6)
Median 42 days IV
an ibio ics (be a-lac ams o
ancomycin, 51% wi h
i ampicin). In ou DAIR,
o al i ampicin-quinolone
Failu e: elapse, ein ec ion,
dea h- ela ed, o he signs o
clinical ailu e
Failu e: 38%,
32% in MSSA,
50% in MRSA
Follow-up: median
190 days ( ange
4–2279 daus)
Mul i a iable:
MRSA, TKA, e en ion o
p os hesis
Vilchez e al.
(2010)
[14]
2000–2007
Re ospec i e
coho ,
unicen ic
Ea ly
( 2 mon hs, 15 days
o symp oms) SA-PJI,
o al and pa ial hip and
knee p os hesis
N= 53 (4 MRSA)
DAIR (n= 53)
Mean 10.6 days i an ibio ics
(cloxacillin i MSSA,
ancomycin i MRSA) and
88 days o al (le ofloxacin
plus i ampicin)
Cu e: Absence o in ec ion
symp oms, asep ic loosening
ha equi ed exchange
p os hesis
Failu e: 24.5%
Follow-up: 2 yea s
Mul i a iable (HR):
CRP [22 mg/dl
Need o 2nd DAIR
P os hesis age [25 days
In ec Dis The (2022) 11:2177–2203 2193
Table 4 con inued
Re e ences Yea s o
diagnosis and
design
Types o in ec ion and
numbe o pa ien s
Numbe o pa ien s and
managemen
Cu e o ailu e defini ion Ou come Fac o s associa ed wi h
ou come
Joulie e al.
(2011) [4]
2001–2006
P ospec i e
coho ,
unicen ic
( e e ence)
Any o al and pa ial hip
and knee SA-PJI
N= 95 e aluable (25%
MRSA)
DAIR (n= 30), 1 T
(n= 15), 2 T (n= 25),
esec ion (n= 10)
Mean 7 days IV an ibio ics
(no specified, adap ed o
suscep ibili y) and
122.3 days o al (no
specified, adap ed o
suscep ibili y, 64%
combined wi h i ampicin)
Cu e: ESR and/o CRP
no mal, non-inflamma o y
sca wi h no fis ula, no
an ibio ics
since discha ge and no
ein e en ion needed
Cu e:
DAIR, 57%;
1 T, 94%;
2 T, 86.2%;
Resec ion, 34%
Follow-up:
minimum o
12 mon hs
Mul i a iable o cu e:
Monomic obial
P os hesis emo al
Senne ille
e al.
(2011) [5]
2000–2006
Re ospec i e
coho ,
unicen ic
To al hip and knee SA-PJI
N= 98 (17.3% MRSA)
DAIR (n= 41) (pe o med i
symp oms 4 weeks)
1T(n= 14)
2T(n= 26)
Resec ion a h oplas y
(n=9)
A h odesis (n=8)
Mean 7 days IV an ibio ics
(no specified, adap ed o
suscep ibili y) and
3–6 mon hs o al
( i ampicin, n= 68;
i ampicin-fluo quinolone,
n= 39). Six DAIR wi h
supp essi e ea men
Cu e: no local o sys emic
signs o in ec ion, no need
o ein e en ion o new
an ibio ic he apy, no
in ec ion- ela ed dea hs
Cu e: 78.6%;
DAIR, 78.0%;
1 T, 100%;
2 T, 84.6%;
Resec ion, 44.4%;
A h odesis,
62.5%
Follow-up:
minimum o
2 yea s
Mul i a iable o cu e:
ASA sco e B2, use o
i ampicin-
fluo oquinolone
2194 In ec Dis The (2022) 11:2177–2203
Table 4 con inued
Re e ences Yea s o
diagnosis and
design
Types o in ec ion and
numbe o pa ien s
Numbe o pa ien s and
managemen
Cu e o ailu e defini ion Ou come Fac o s associa ed wi h
ou come
Lo a-Tamayo
e al.
(2013) [6]
2003–2010
Re ospec i e
coho ,
mul icen e
Ea ly and haema ogenous
hip ( o al and pa ial)
and knee SA-PJI
N= 345 (23% MRSA)
DAIR (n= 345)
Median days o an ibio ics:
94 in MRSA and 91 in
MSSA. Ri ampicin
combina ions in 88%: be a-
lac ams (13%) o
quinolones (75%, mainly
le ofloxacin) in MSSA, and
glycopep ides (18%, namely
ancomycin), co imoxazole
(46%), linezolid (24%) o
clindamycin (10%) in
MRSA
Failu e: in ec ion- ela ed
dea h, p os hesis
eplacemen , need o
u he deb idemen s
Failu e: 45%
(in ec ion- ela ed
dea h, 7%)
– Ea ly ailu e
(30 days), 29%
; Failu e du ing
an ibio ic he apy,
32%;
– Failu e a e
an ibio ic he apy,
39%;
Follow-up: 2 yea s
Mul i a iable (HR) o ea ly
ailu e: male, heuma oid
a h i is, bac e aemia,
polymic obial,
CRP [100 mg/L
Failu e du ing he apy:
highe age,
immunosupp essi e d ugs,
MRSA, sinus ac ,
abno mal adiog aphy,
need C2 deb idemen s,
no use o i ampicin
Failu e a e he apy:
haema ogenous in ec ion,
deb idemen delay, need
o C2 deb idemen s
Go
´mez-
Junyen
e al.
(2021)
[11]
2003–2010
Re ospec i e
coho ,
mul icen e
Hip and knee SA-PJI
N= 249 (ea ly, n= 141;
haema ogenous, n= 26;
ch onic, n= 82)
Implan emo al
(161 ini ial he apy, 88
sal age)
1T,n= 17 (6.8%)
2T,n= 188 (75.5%)
Hip esec ion, n=44
(17.7%)
Failu e: local ailu e and/o
all-cause mo ali y wi hin
60 days
Failu e: 15.6%
Local ailu e: 9.3%
Mo ali y: 12.8%
Follow-up: median
781 days,
in e qua ile
ange [IQR]
355–1375 days
Mul i a iable o ailu e:
C2 como bidi ies
In ec Dis The (2022) 11:2177–2203 2195
Table 4 con inued
Re e ences Yea s o
diagnosis and
design
Types o in ec ion and
numbe o pa ien s
Numbe o pa ien s and
managemen
Cu e o ailu e defini ion Ou come Fac o s associa ed wi h
ou come
Be z e al.
(2015)
[15]
1996–2012
Re ospec i e
coho ,
unicen ic
Hip ( o al and pa ial) PJI
N= 29 (12 MSSA, 17
MRSA, 9 s ep ococci)
DAIR (n= 38)
All exchange o mobile pa s.
An ibio ic adap ed o
suscep ibili y, wi h median
du a ion o
12 weeks (IV o a median o
14 days). Ri ampicin use: 23
(60.5%)
Pe sis ence o ecu ence o
PJI
Failu e: 18.4%
MSSA, 9.5%
MRSA, 29.4%
S ep ococci, 0%
Follow-up:
minimum o
2 yea s
Uni a ia e analysis: no
ac o s associa ed wi h
ailu e
Bouaziz e al.
(2018)
[13]
2000–2010
Re ospec i e
coho , wo
cen es
Hip and knee MSSA-PJI
N= 85 (ea ly, 33; delayed,
11; la e, 45)
DAIR (n= 62)
An ibio ic no specified,
excep o i ampicin
Ri ampicin use: ea ly 20
(62%), delayed 7 (63%), la e
26 (58%)
Failu e: need o u he
su ge y o con ol PJI (o o
ea supe in ec ion),
ampu a ion o in ec ion-
ela ed dea h
Failu e: 42%
Ea ly, 10 (30%)
Delayed, 4 (36%)
La e, 23 (51%)
Follow-up:
2.8 ±2.2 yea s
Mul i a iable (HR) o
ailu e: non-compliance
wi h su gical IDSA
guidelines (namely DAIR
pe o med i du a ion o
symp oms 3 weeks o
join age 30 days, and
s able implan wi hou
sinus ac )
Lesens e al.
(2018)
[12]
2010–2014
Re ospec i e
coho ,
mul icen e
Hip, knee, shoulde , elbow
SA-PJI
N= 137 (ea ly acu e,
n= 63; ea ly ch onic,
n= 26; la e acu e,
n= 35; la e ch onic,
n= 13)
MRSA 27 (19.7%)
DAIR (n= 137)
Mean du a ion an ibio ic:
12.6 weeks
Ri ampicin use: 85 (65%)
Ri ampicin plus
fluo quinolone: 63 (47.4%)
Supp essi e ea men : 14
(10.2%)
Failu e: p os hesis emo al,
dea h, addi ional
deb idemen o cou se o
an ibio ics, clinical and
mic obiological signs o
in ec ion
Failu e: 25%
Mo ali y: 8.8%
Follow-up: 2 yea s
Mul i a iable (HR) o
ailu e: longe AB
du a ion (p o ec i e),
i ampicin egimen
(p o ec i e), smoking,
ea ly in ec ion (p o ec i e)
2196 In ec Dis The (2022) 11:2177–2203
Table 4 con inued
Re e ences Yea s o
diagnosis and
design
Types o in ec ion and
numbe o pa ien s
Numbe o pa ien s and
managemen
Cu e o ailu e defini ion Ou come Fac o s associa ed wi h
ou come
Wou huyzen-
Bakke
e al.
(2018) [7]
2005–2015
Re ospec i e
coho , wo
cen es
Ea ly hip and knee MSSA-
PJI ea ed wi h
i ampicin plus
le ofloxacin (n= 40) o
moxifloxacin (n= 19)
N=58
DAIR (n= 58)
Du a ion an ibio ic he apy
abou 90 days. IV cloxacillin
o flucloxacillin 7–14 days
ollowed i ampicin plus
fluo quinolone
Failu e: need o e ision
su ge y
and/o supp essi e
an imic obial he apy
because o pe sis en
in ec ion, dea h- ela ed
in ec ion, ein ec ion o
elapse wi h S. au eus
Cu e: 86.4%
(87.5% in
le ofloxacin
g oup, 84.2% in
moxifloxacin
g oup)
Follow-up:
minimum 2 yea s
No pe o med
Mun
˜oz-
Gallego
e al.
(2020) [8]
2016–2017
P ospec i e
coho ,
mul icen e
Hip ( o al and pa ial) and
knee SA-PJI
N= 85 (19 MRSA)
Ea ly ( 90 days): 45
Ch onic ([90 days): 21
Haema ogenous: 19
DAIR (n= 55)
P os hesis emo al (n= 30)
Du a ion and ype o
an ibio ic no specified
Failu e: all-cause dea h wi hin
90 days, pe sis en o
elapsing signs
o s aphylococcal in ec ion,
need o sal age he apy
excep o ex a deb idemen s
in he
fi s 30 days, supp essi e
an imic obial ea men
O e all ailu e:
36.4% (DAIR,
47.2%; p os hesis
emo al, 16.6%)
Follow-up: a leas
1 yea
Uni a ia e analysis:
Global ailu e associa ed
wi h DAIR
DAIR: ailu e associa ed
wi h delay o deb idemen
and no use o i ampicin
In ec Dis The (2022) 11:2177–2203 2197

Table 4 con inued
Re e ences Yea s o
diagnosis and
design
Types o in ec ion and
numbe o pa ien s
Numbe o pa ien s and
managemen
Cu e o ailu e defini ion Ou come Fac o s associa ed wi h
ou come
ARTHR-IS
coho
2014–2016
Re ospec i e
coho ,
mul icen e
Hip ( o al and pa ial) and
knee pos su gical SA-PJI
N= 128 (28 MRSA)
DAIR (n= 99)
1T(n= 10)
2T(n= 13)
Resec ion (n=6)
Median days o an ibio ics:
16 IV and 50 o al.
Ri ampicin combina ions in
103 (80.5%); wi h
quinolones in 42 (65.8%)
Clinical and unc ional ailu e
a e he fi s p ocedu e and
a 18 mon hs
Clinical ailu e: all-cause
dea hs in fi s 2 mon hs, SA-
PJI- ela ed dea h, pe sis ence
o ecu ence o signs o
symp oms o in ec ion; need
o addi ional cou se o
an ibio ics including
supp essi e he apy, emo al
o he p os hesis i DAIR.
Func ional ailu e:
significan ly impai ed
walking due o p os hesis
loosening, Gi dles one o
a h odesis
O e all clinical and
unc ional ailu e
a e fi s su gical
p ocedu e: 32.8%
(DAIR, 33.3%;
p os hesis
emo al, 37.1%)
Conside ing only
clinical ailu e:
28.1%
Failu e a
18 mon hs:
24.2% (DAIR:
21.2%; p os hesis
emo al, 34.4%)
Conside ing only
clinical ailu e:
15.5% (DAIR:
15.5%; p os hesis
emo al: 17.2%)
Follow-up:
18 mon hs
Mul i a iable analysis:
Failu e a e fi s su gical
p ocedu e: Cha lson C2,
haemoglobin
le el 10 mg/dL,
bac e aemia, polymic obial
in ec ion and need o
pe o m addi ional
deb idemen
In DAIR: hose abo e and
BMI [30, and ime
be ween symp om onse
and DAIR; i ampicin use
was p o ec i e
A 18 mon hs: haemoglobin
le el 10 mg/dL, hip
ac u e and need o
addi ional join , su gical
p ocedu es no ela ed o
in ec ion con ol
1Tone-s age eplacemen , 2T wo-s age eplacemen , CRP C- eac i e p o ein, DAIR deb idemen , an ibio ics, and implan e en ion, ESR e y h ocy e sedimen a ion a e, HR
haza d a io, IV in a enous, MRSA me hicillin- esis an S. au eus., MSSA me hicillin-suscep ible S. au eus,PJI p os he ic join in ec ion, SA-PJI p os he ic join in ec ion p oduced
by S. au eus,TKA o al knee a h oplas y
2198 In ec Dis The (2022) 11:2177–2203
used he ASA index [5]o C2 como bidi ies o
simila easons [11]. Anaemia was p e iously
iden i ied as a p edic o in one s udy [16], bu
was no assessed a all in mos o he o he s.
Anaemia is a po en ial ma ke o nu i ional
s a us ha can inc ease issue hypoxia o e en
indica e a sys emic in lamma o y p ocess. As in
ou s udy, bac e aemia, polymic obial in ec ion
and he need o addi ional deb idemen we e
also p edic o s o TF in o he s udies [6,14].
Since DAIR is he mos equen ini ial p o-
cedu e, we also analysed p edic o s o TF in his
subg oup. Apa om he abo e a iables, obe-
si y and delay in pe o ming DAIR we e also
iden i ied as isk ac o s, while he use o
i ampicin had a p o ec i e e ec . Obesi y is a
isk ac o o PJI and was also ound o be
associa ed wi h TF in hip PJI unde going wo-
s age eplacemen [17], bu was no e en con-
side ed in mos s udies. Obesi y could be asso-
cia ed wi h wound complica ions, addi ional
deb idemen , impai ed inna e immune
esponse and changes in he pha macokine ics
o some an imic obial d ugs [18]. Delayed DAIR
ollowing onse o PJI symp oms is a known
ac o o TF ega dless o he mic oo ganism
in ol ed, bu is pa icula ly pe inen in he
case o SA-PJI [2,6,12–14]. Es ablishing a
h eshold is complex. The 21-day h eshold o
pe o ming DAIR ecommended by he guide-
lines [9] was based on one small-scale s udy
expe ience [19], and a delay o [2 days in
pa ien s wi h SA-PJI ea ed wi h be a-lac ams
was associa ed wi h inc eased TF in ano he
s udy [2] bu no when a i ampicin- luo o-
quinolone combina ion was used [6–14,16]. In
ou coho , we ound an inc eased isk o each
day o delay, suppo ing he ecommenda ion
ha deb idemen should be pe o med as ea ly
as possible.
Ri ampicin in combina ion wi h o he
an ibio ics (mainly luo oquinolones) was
epo ed o inc ease cu e a es in a small an-
domized ial [19] and in obse a ional s udies
[5,6,12,20–22]. Howe e , wo ecen me a-
analyses ound con o e sial esul s on he ole
o i ampicin: one ound no bene i in s aphy-
lococcal in ec ions [23] and he o he only a
limi ed impac [24]. The s udies included in he
me a-analysis had a conside able isk o
selec ion and immo al ime bias. In ou s udy,
on he o he hand, we ound ha i ampicin
combina ions we e associa ed wi h a p o ec i e
e ec , e en a e pe o ming sensi i i y and
landma k analyses.
Al hough p e ious s udies ha e ound highe
TF a es in PJIs caused by MRSA compa ed wi h
suscep ible s ains [3,15], ou da a did no
demons a e his associa ion, which is in line
wi h mo e ecen obse a ions [5,6]. Whe he
his is due o he use o an i-MRSA d ugs wi h
good bioa ailabili y and an i-bio ilm ac i i y,
such as linezolid, would equi e u he s udies.
Finally, we also analysed he p edic o s o TF
a e all p ocedu es had been pe o med. Apa
om o he ac o s, hip ac u e inc eased he
isk o TF, p obably e lec ing he ail y o he
pa ien s a ec ed. In hese pa ien s, he i s
su gical app oach was c ucial, since ini ial TF
was ollowed by unc ional ailu e in all cases
(da a no shown). In a p e ious mul icen e
coho s udy o pa ien s wi h hip PJIs, ac u e
was also associa ed wi h clinical ailu e and
wo se unc ional p ognosis [25]. The isk o TF
also inc eased when addi ional join su ge y no
due o pe sis en in ec ion was pe o med; his
may ha e been due o issue damage, delayed
healing o acili a ion o bac e ial
supe in ec ions.
This s udy has some limi a ions ha should
be conside ed when in e p e ing he esul s.
The sample size was oo small o in es iga e
p edic o s o o he ini ial su gical p ocedu es
and may ha e been insu icien o de ec addi-
ional p edic o s o TF; i s e ospec i e design
limi ed he a ailable a iables; esidual con-
ounding is also possible; we did no collec da a
abou e hnici y o he pa ien s; inally, some
changes in managemen may ha e occu ed
du ing he s udy pe iod. Some s eng hs include
ha i is a mul ina ional s udy, he de ini ions
o TF including clinical and unc ional aspec s
and he good p edic i e abili y o he models
de eloped.
CONCLUSIONS
In conclusion, we obse ed ha a conside able
p opo ion o SA-PJIs ailed a e ini ial su gical
In ec Dis The (2022) 11:2177–2203 2199
ea men s, al hough a subs an ial pa o hem
we e eco e ed wi h u he p ocedu es. Signi -
ican unc ional loss should be conside ed
alongside clinical ailu e, and he impo ance o
ce ain isk ac o s o TF was con i med,
including he p o ec i e ole o i ampicin-
based ea men in DAIR.
ACKNOWLEDGEMENTS
We hank all s udy pa icipan s o hei col-
labo a ion, wi hou which his manusc ip
would ha e been impossible.
O he membe s o he ARTHR-IS g oup a e:
Nienke Cupe us and Giuseppe Man e
´(Julius
Cen e o Heal h Sciences and P ima y Ca e,
Uni e si y Medical Cen e U ech , U ech , he
Ne he lands) collabo a ed in he selec ion p o-
cess o he pa icipa ing cen e s. Ana Isabel
Sua
´ ez-Ba enechea and Al a o Pascual-He -
nandez (Depa men o mic obiology, Hospi al
Uni e si a io Vi gen Maca ena), Alba Ri e a
(Depa men o Mic obiology, Hospi al de la
San a C eu i San Pau), Xa ie C usi and Ma cos
Jo da
´n (Depa men o auma ology, Hospi al
de la San a C eu i San Pau), Nicolo
`Rossi
(Depa men o Medical and Su gical Sciences,
Uni e si y o Bologna, Bologna, I aly), Tessa
an de Ke kho (Depa men o O hopaedic
Su ge y & T auma, Ma
´xima MC, Eindho en,
he Ne he lands; Depa men o O hopaedic
Su ge y & T auma, Ca ha ina Hospi al, Eind-
ho en, he Ne he lands), Juan P. Ho cajada,
Joan Go
´mez-Junyen and Albe Alie (Hospi al
del Ma , Ins i u Hospi al del Ma d’In es iga-
cions Me
`diques, Uni e si a Pompeu Fab a,
Ba celona, Spain), Mi anda an Rijen and Jan-
nie Romme (Amphia Hospi al, B eda, Ne he -
lands), Juliane Anke (Jena Uni e si y
Hospi al—Jena, Ge many), Celia Whi ehouse
and Ad ian Jones (No olk and No wich
Uni e si y Hospi al—No wich, UK), Ja ie Cobo
and Ja ie Mo eno (Hospi al Uni e si a io
Ramo
´n y Cajal—Mad id, Spain), Anne Meheu
(Cen e Hospi alie Uni e si ai e de Rennes—
Rennes, F ance), Clai e Gledel (O hopedic su -
ge y depa men , C oix Rousse Hospi al—Lyon,
F ance), Pauline Pe eau (Cen e Hospi alie
Uni e si ai e de Bo deaux—Bo deaux, F ance),
Remco J.A. an Wensen (Depa men o
O hopaedic Su ge y & T auma, Ca ha ina
Hospi al, Eindho en, he Ne he land), Gab iella
Linde ga d (No h Manches e Gene al Hospi-
al—Manches e , UK) collabo a ed in he da a
collec ion.
Funding. This wo k was suppo ed by he
Inno a i e Medicines Ini ia i e Join Unde -
aking (g an ag eemen No. 115523), COM-
BACTE-NET conso ium (Eu opean Union FP7/
2007–2013 and GlaxoSmi hKline Biologicals SA,
as EFPIA pa ne ). R.E, L.S, O. M, R. E-S, J. L–T, J.
P, J. R-B and MD. del T a e membe s o he
Spanish Ne wo k o Resea ch in In ec ious
Diseases (REIPI), suppo ed by Plan Nacional de
I?D?i 2013-2016 and Ins i u o de Salud
Ca los III, Subdi eccio
´n Gene al de Redes y
Cen os de In es igacio
´n Coope a i a, Minis e-
io de Ciencia, Inno acio
´n y Uni e sidades,
Spanish Ne wo k o Resea ch in In ec ious
Diseases (REIPI RD16/0016/0001; 0002; 0005;
0009; 0011; 0015), co- inanced by Eu opean
De elopmen Regional Fund ‘‘A way o achie e
Eu ope’’, Ope a i e P og am In elligence
G ow h 2014-2020. The s udy sponso is also
unding he jou nal’s Rapid Se ice Fee. Role o
unding sou ce: The unde s had no in luence
on he analysis and decision o published;
GlaxoSmi hKline Biologicals SA was p o ided
he oppo uni y o e iew a e sion o his
manusc ip o ac ual accu acy; au ho s a e
solely esponsible o inal con en and
in e p e a ion.
Au ho ship. All named au ho s mee he
In e na ional Commi ee o Medical Jou nal
Edi o s (ICMJE) c i e ia o au ho ship o his
a icle, ake esponsibili y o he in eg i y o
he wo k as a whole, and ha e gi en hei
app o al o his e sion o be published.
Au ho Con ibu ions. Espindola R, Vella V,
Rod ı
´guez-Ban
˜o J, Del To o MD con ibu ed in
he concep ion and design o he s udy, analysis
and in e p e a ion o da a and d a ing he
a icle. Beni o N, Mu I, Tedeschi S, Zampa ini
E, Hend iks J, So lı
´L, Mu illo O, Solde ila L,
Sca bo ough M, Sca bo ough C, Kluy mans J,
2200 In ec Dis The (2022) 11:2177–2203
Fe a i MC, Ple z M, Mcnama a I, Escude o-
Sanchez R, A ieux C, Ba aille C, Dauchy F-A,
Liu W-Y, Lo a-Tamayo J, P aena J, Us ianowski J
con ibu ed in he acquisi ion o da a and
e ising he a icle. Cinconze E, Pelleg ini M
and Bagnoli F pa icipa ed in he analysis and
in e p e a ion o da a and e ision he a icle.
All au ho s ga e hei inal app o al o he e -
sion o be submi ed.
P io P esen a ion. This s udy was p e i-
ously p esen ed as a pos e a he 32nd Eu o-
pean Cong ess o Clinical Mic obiology &
In ec ious Diseases (ECCMIC), held in Lisbon,
Po ugal, Ap il 23–26.
Disclosu es. Venanzio Vella, Elisa Cinconze,
Michele Pelleg ini and Fabio Bagnoli a e
employees o he GSK g oup o companies.
Venanzio Vella, Michele Pelleg ini and Fabio
Bagnoli hold sha es in he GSK g oup o com-
panies. Fabio Bagnoli holds pa en s pending
and issued pa en s on S aphylococcus au eus
accine o mula ions. Venanzio Vella, Elisa
Cinconze, Michele Pelleg ini and Fabio Bagnoli
decla e no o he inancial o non- inancial
ela ionships and ac i i ies. Reinaldo Espindola,
Na i idad Beni o, Isabel Mu , Sa a Tedeschi,
Eleono a Zampa ini, Johannes G.E. Hend iks,
Luisa So lı
´, Osca Mu illo, Lau a Solde ila,
Ma hew Sca bo ough, Clai e Sca bo ough, Jan
Kluy mans, Ma eo Ca lo Fe a i, Ma hias W.
Ple z, Iain Mcnama a, Rosa Escude o-Sanchez,
Ced ic A ieux, Cecile Ba aille , F e
´de
´ ic-
An oine Dauchy, Wai-Yan Liu, Jaime Lo a-
Tamayo, Julia P aena, And ew Us ianowski,
Jesu
´s Rod ı
´guez-Ban
˜o and Ma ia Dolo es Del
To o ha e no con lic s o decla e. A p esen ,
he a ilia ion o Reinaldo Espindola is In ec-
ious Diseases and Mic obiology Clinical Uni ,
Hospi al Uni e si a io Vi gen de Valme, Se ille,
Spain.
Compliance wi h E hics Guidelines. The
s udy was app o ed by he E hics Commi ees
a each si e. The need o ob ain w i en
in o med consen was wai ed due o he e o-
spec i e na u e o he s udy and anonymized
da a, excep in he case o he F ench hospi als.
The app o al o he e hics commi ee o all he
pa icipa ing cen e s, wi h he names and e -
e ences, a e e lec ed in he able S5 o he
supplemen a y ma e ial.
Da a A ailabili y. The da ase s gene a ed
and analyzed du ing he cu en s udy a e
a ailable om he co esponding au ho on
easonable eques .
Open Access. This a icle is licensed unde a
C ea i e Commons A ibu ion-NonComme -
cial 4.0 In e na ional License, which pe mi s
any non-comme cial use, sha ing, adap a ion,
dis ibu ion and ep oduc ion in any medium
o o ma , as long as you gi e app op ia e c edi
o he o iginal au ho (s) and he sou ce, p o ide
a link o he C ea i e Commons licence, and
indica e i changes we e made. The images o
o he hi d pa y ma e ial in his a icle a e
included in he a icle’s C ea i e Commons
licence, unless indica ed o he wise in a c edi
line o he ma e ial. I ma e ial is no included
in he a icle’s C ea i e Commons licence and
you in ended use is no pe mi ed by s a u o y
egula ion o exceeds he pe mi ed use, you
will need o ob ain pe mission di ec ly om he
copy igh holde . To iew a copy o his licence,
isi h p://c ea i ecommons.o g/licenses/by-
nc/4.0/.
REFERENCES
1. Tande AJ, Pa el R. P os he ic join in ec ion. Clin
Mic obiol Re . 2014;27:302–45. h ps://doi.o g/10.
1128/CMR.00111-13.
2. B and CM, Sis unk WW, Du y MC, e al. S a-
phylococcus au eus p os he ic join in ec ion ea-
ed wi h deb idemen and p os hesis e en ion.
Clin In ec Dis 1997;24:914–919. h ps://pubmed.
ncbi.nlm.nih.go /9142792/. Accessed 4 June 2021.
h ps://doi.o g/10.1093/CLINIDS/24.5.914.
3. Salgado CD, Dash S, Can ey JR, Ma culescu CE.
Highe isk o ailu e o me hicillin- esis an S a-
phylococcus au eus p os he ic join in ec ions. Clin
O hop Rela ed Res. 2007:48–53. h ps://pubmed.
ncbi.nlm.nih.go /17534195/. Accessed 3 June
2021. h ps://doi.o g/10.1097/BLO.
0b013e3181123d4e.
In ec Dis The (2022) 11:2177–2203 2201