Review articles.

Review rtiles Systemti literture nlysis nd review of trgeted preventive mesures to limit helthre-ssoited infetions y metiillin-resistnt Stphyloous ureus R Kök1, K Beker 2, B Cookson 3, J E vn Gemert-Pijnen 4, S Hrrth 5, J Kluytmns 6, M Mielke 7, G Peters 2, R L Skov 8, M J Struelens 9, E Tonelli 10, W Witte 11, A W Friedrih (lex.friedrih@umg.nl) 12 1. Institute of Hygiene, University Hospitl Münster, Münster, Germny 2. Institute of Medil Miroiology, University Hospitl Münster, Münster, Germny 3. Division of lnfetion nd lmmunity, University College London, London, United Kingdom 4. Fulty of Behviourl Sienes, University of Twente, Enshede, the Netherlnds 5. Infetion Control Progrm, University of Genev Hospitls nd Medil Shool, Genev, Switzerlnd 6. Deprtment of Medil Miroiology nd Infetion Control, VU University Medil Centre, Amsterdm nd Amphi Hospitl Molengrht, Bred, the Netherlnds 7. Roert Koh Institute, Deprtment for Infetious Diseses, Berlin, Germny 8. Deprtment for Miroiology nd Infetion Control for Miroiologil Surveillne nd Reserh, Sttens Serum Institut, Copenhgen, Denmrk 9. Europen Centre for Disese Prevention nd Control, Stokholm, Sweden 10. Division of Infetious Diseses, Deprtment of Internl Mediine I, University Hospitl Tüingen, Tüingen, Germny 11. Roert Koh Institute, Referene Centre for Stphylooi, Wernigerode, Germny 12. Deprtment of Medil Miroiology, University of Groningen, University Medil Center Groningen, Groningen, The Netherlnds Cittion style for this rtile: Kök R, Beker K, Cookson B, vn Gemert-Pijnen JE, Hrrth S, Kluytmns J, Mielke M, Peters G, Skov RL, Struelens MJ, Tonelli E, Witte W, Friedrih AW. Systemti literture nlysis nd review of trgeted preventive mesures to limit helthre-ssoited infetions y metiillin-resistnt Stphyloous ureus. Euro Surveill. 2014;19(29):pii=20860. Aville online: http://www.eurosurveillne.org/viewartile.spx?artileid=20860 Artile sumitted on 20 April 2013 / pulished on 25 July 2014 Metiillin-resistnt Stphyloous ureus (MRSA) is mjor use of helthre-ssoited infetions in Europe. Mny exmples hve demonstrted tht the spred of MRSA within helthre settings n e redued y trgeted infetion ontrol mesures. The im of this systemti literture nlysis nd review ws to summrise the evidene for the use of teril ultures for tive surveillne the enefit of rpid sreening tests, s well s the use of deolonistion therpies nd different types of isoltion mesures. We inluded 83 studies pulished etween 2000 nd 2012. Although the studies reported good evidene supporting the role of tive surveillne followed y deolonistion therpy, the effetiveness of singleroom isoltion ws mostly shown in non-ontrolled studies, whih should inspire further reserh regrding this issue. Overll, this review highlighted tht when plnning the implementtion of preventive interventions, there is need to onsider the prevlene of MRSA, the inidene of infetions, the ompeting effet of stndrd ontrol mesures (e.g. hnd hygiene) nd the likelihood of trnsmission in the respetive settings of implementtion. Bkground Metiillin-resistnt Stphyloous ureus (MRSA) is mjor use of helthre-ssoited infetions in Europe. In 2008, the Europen Centre for Disese Prevention nd Control (ECDC) estimted tht totl numer of 171,200 nosoomil MRSA infetions re quired nnully in the Memer Sttes of the Europen Union (EU), nd in Ielnd nd Norwy, resulting in 5,400 ttriutle exess deths, more thn 1 million exess dys of hospitlistion nd EUR 380 million exess in-hospitl osts [1]. The urden of MRSA infetions ws lso shown in n nlysis of dt on helthre-ssoited infetions olleted prospetively from Europen intensive re units (ICU) etween 2005 nd 2008, where 1.7% of ll ptients developed S. ureus pneumoni or loodstrem infetions. A men of 35% of these infetions were used y MRSA. Moreover, the hzrd rtio for mortlity ws 5.6-times higher (95% onfidene intervl (CI): 3.4 9.4) for ptients with MRSA loodstrem infetion thn for ptients without S. ureus teremi [2]. Among the proposed methods to prevent MRSA, mny (e.g. hnd hygiene nd trnsmission-sed preutions) hve een used for generl infetion ontrol, nd their effetiveness hs een reviewed extensively [3,4]. However, there is n ongoing disussion out the evidene for the effetiveness of severl more speifi prevention methods whih, nevertheless, hve een inluded in stndrds for the prevention nd ontrol of MRSA in mjority of Europen ountries [5]. Therefore, the sope of this review ws to nlyse systemtilly reent literture (pulished fter 2000) with respet to the following questions relted to MRSA prevention nd ontrol: www.eurosurveillne.org 1

1. Does sreening of ptients efore or on dmission redue the inidene of MRSA infetion or trnsmission? How do PCR-sed rpid tests for the diret detetion of MRSA from sreening speimens influene the inidene of MRSA olonistion or infetion ompred with ulture-sed methods? 2. Does the deolonistion of nsl MRSA or S. ureus rrige using mupiroin nsl ointment, lone or in omintion with other gents, redue olonistion or the development of infetions? 3. Does isoltion in single rooms of ptients olonised or infeted with MRSA prevent the spred of MRSA etter thn the use of trnsmission-sed preutions (hnd hygiene, gloves, prons) lone? Wht is the effet of pre-emptive isoltion of risk ptients for MRSA rrige (until sreening results re ville)? Methods A systemti literture nlysis nd review ws performed following the Preferred Reporting Items for Systemti Reviews nd Met-Anlyses (PRISMA) guidelines [6]. To identify relevnt pulitions, PuMed, EMBASE nd Sopus were serhed for rtiles pulished etween 1 Jnury 2000 nd 31 Otoer 2012 in English lnguge. The serh terms were: MRSA AND (prevention OR ontrol OR prophylxis OR preventive mesures OR preventive therpy OR preventive tretment OR preution OR sreening OR tive surveillne OR deoloniztion OR mupiroin OR surveillne ulture* OR hromogeni OR PCR OR polymerse hin retion OR rpid test OR isoltion OR hygiene OR effiien* OR effetive*) AND (helthre OR hospitl OR nursing home OR long-term re filit*); the serh terms were dpted for serh in EMBASE: MRSA AND deoloniztion, MRSA AND isoltion, MRSA AND sreening. Titles nd strts were sreened independently y two reviewers (RK nd AWF). Studies with outomes mesuring the inidene of MRSA olonistion or infetion were inluded. Exlusion riteri were: Studies tht did not report on the effets of the preventive mesures on infetion or trnsmission; studies performed in settings other thn hospitls, long-term re filities nd nursing homes; se series, outrek reports nd Figure Flow digrm for the seletion of studies on preventive mesures ginst to limit helthre-ssoited infetions y metiillin-resistnt Stphyloous ureus, pulished 2000 2012 (n=9,340) 9,340 reords identified through dtse serhing nd sreened 9,189 reords exluded 1 51 full -text rtiles ssessed for eligiility 82 rtiles exluded (outrek reports, se series, reviews, studies with outome prmeters not meeting inlusion rit eri) 14 rtiles dded fter serh of literture lists 83 originl rtiles inluded A ssignment to one of the study questions Sreening D eolonistion Isoltion 2 www.eurosurveillne.org

Tle 1 Studies on the effetiveness of the use of tive surveillne (sreening) for metiillin-resistnt Stphyloous ureus, pulished 2000 2012 (n=41) MRSA ; Culture-sed tests Cmus [9]; 4.8 9%; 2002 03; Frne; MICU; RCT. Cherny [10]; ; 2002 06; Germny; ICU nd surgery; CS (interrupted time series). Clny [12]; 3.7%; 2003 04; Ellingson [15]; ; 1999 2008; Hospitl-wide; CS (interrupted time series). Eveillrd [42]; 4.7 12.1%; 2003; Frne; Hospitl-wide; CS. Gould [47]; 6 16%; 1999 2003; United Kingdom; CS (interrupted time series). Turnround time (PCR/ Design Sreening followed y Outome Result ulture) 48 h 48 h Intervention: sreening of high-risk ptients (nose, perineum, wounds, spirtes) t dmission, weekly therefter nd t dishrge; Control: sme methods s in the intervention group, ut the sreening results were not reported. Intervention: sreening (nose, throt, wounds) of ll ptients; Control: seletive sreening of ontt ptients or ptients with history of MRSA rrige. Intervention: nsl sreening of ll ptients t dmission nd weekly therefter; Intervention: sreening (nose, wounds) of ll ptients t dmission nd t dishrge + ehviourl hnge strtegies, hnd hygiene, environmentl disinfetion; Intervention: sreening of ll ptients dmitted to ICUs (nose, xill, retl) nd of high-risk ptients dmitted to other wrds; prospetive dt quisition without historil or prospetive ontrol group. Intervention: sreening (nose, throt, groin, xill) of ll ptients t dmission; Control: phse without ny or non-ompulsory Gloves, gowns, msk (lso pre-emptively), deolonistion Privte rooms, gowns, gloves, deolonistion Privte rooms, gowns, gloves Privte rooms, gowns, gloves Contt preutions similr to guidelines from the United Sttes Centers for Disese Control nd Prevention Privte rooms, rriernursing (unspeified), deolonistion A, I I I C/I I C/I, B MRSA quisition in the intervention group vs the ontrol group: 6.5% vs 5.3%; p=0.58; Proportion of ptients who quired MRSA infetion ws identil: 1.6% (n=4) vs 1.6% (n=4); p>0.99; Rte of ICU-quired infetion ws identil: 16.5% vs 16.5%, p=0.98. Chnge in the level of infetions: -0.163 MRSA infeted ptients/1,000 pd (95% CI: -0.276 to -0.05); Slope: -0.01 MRSA-infeted ptients/1,000 pd (95% CI: 0.018 0.003). Derese of MRSA infetions (6.1 vs 4.1 infetions/1,000 ensus-dys; p=0.01) nd of nosoomil (>72 h fter dmission) MRSA infetions (4.5 vs 2.8 infetions/1,000 ensus-dys; p=0.01). Inidene of MRSA olonistion or infetion deresed y 21.8% (95% CI: 8.8 33.7) from 2.40 ses/1,000 pd to 1.88/1,000 pd t risk. Inidene of MRSA from linil speimens/100 dys of hospitlistion for MRSA rriers identified t dmission of ws 3.1% when the progrmme ws ompletely implemented, ompred with 10.4% when no sreening ws performed (p<0.001). By time series regression nlysis, the proportion of ptients with MRSA (infetion nd olonistion) deresed from 15% to 5% (95% CI: 3.5 19.3; p=0.005); no signifint effet on MRSA teremi rtes. CI: onfidene intervl; CS: omprtive study; ICU: intensive re unit; MICU: medil ICU; MRSA: metiillin-resistnt Stphyloous ureus; MSICU: medil/surgil ICU; : not ville; OR: odds rtio; PICU: peditri ICU; pd: ptient-dys; RCT: rndomised ontrolled tril; RR: reltive risk; SICU: surgil ICU; SSI: surgil-site infetions; MRSA prevlene in the study setting per 100 ptients dmitted (exept stted differently). Turnround time of the sreening test result (strtified y PCR-sed test vs ulture-sed test, if oth were ompred in the respetive study). Outome mesures: A=MRSA quisition/trnsmission; B=MRSA teremi; C/I=ses of olonistion or (ll/unspeified types of) infetion; I=ses of severl or unspeified types of infetion; W/ SSI=wound infetions/surgil-site infetions.

Tle 1B Studies on the effetiveness of the use of tive surveillne (sreening) for metiillin-resistnt Stphyloous ureus, pulished 2000 2012 (n=41) MRSA ; Culture-sed tests Holzmnn-Pzgl [19]; 2.7 8.3%; 2007 09; PICU; Lwes [38]; 3.1%; 2006 10; United Kingdom; Hospitl-wide; CS without ontrol (times series nlysis). Hung [21]; 12%; 1996 2004; CS (interrupted time series). Huskins [20]; 9.5 12.6%; 2005 06; RCT. Kelly [40]; 1.13 1.63%; 2005 07; Irelnd; Orthopedi surgery; Turnround time (PCR/ Design Sreening followed y Outome Result ulture) 48 h mostly <24 h 48 h 5.2 ± 1.4 d Intervention: nsl sreening of ll ptients t dmission nd weekly therefter; Intervention: nsl sreening of ll ptients t dmission; isoltion filities nd deolonistion; hnd-hygiene mpign; Compred to: no ontrol group; oservtion over time. Intervention: mpigns for theter plement, hnd hygiene, nsl sreening of ll ptients t dmission nd weekly therefter introdued step y step; Intervention: nsl sreening of ll ptients t dmission, weekly therefter; Control: ontrol ICUs where sreening ws performed s in intervention ICUs ut without reporting of the results. Intervention: period 1: pre-dmission sreening (nose, xille, groin) of ll eletive orthopedi ptients; period 2: seprtion (dmission to nother hospitl) of trum ptients from eletive ptients; Gloves, gowns C/I Privte rooms, deolonistion Contt isoltion preutions (unspeified) Gloves, gowns C/I Deolonistion prior to dmission B B C/I Yerly MRSA inidene density deresed from 2006 to 2009 (6.88 vs 1.45/1,000 pd; p<0.01); nd from 2007 to 2009 (7.32 vs 1.45/1,000 pd; p<0.01). Redution of MRSA teremi (0.26/1,000 ute oupied ed dys (AOBD) vs 0.07/1,000 AOBD; p<0.001). In multivrite time-series nlysis, introdution of sreening resulted in redution of MRSA teremi, hospitl-ssoited inidene density nd 30-dy mortlity fter MRSA teremi (p<0.001). MRSA sreening ws ssoited with 67% derese in the inidene density of MRSA teremi in ICUs (p<0.002), 39% derese in non-icus, nd 53% derese hospitl-wide. Inidene of events of olonistion or infetion with MRSA/1,000 pd did not differ signifintly etween intervention nd ontrol ICUs fter djustment for the seline inidene (40.4 vs 35.6; p=0.35). Inidene of MRSA infetions delined from 0.49% in the ontrol phse to 0.35% (p=0.108) in period 1, nd to 0.23% (p=0.05) in period 2. MRSA olonistion deteted rose from 1.13% (ontrol phse) to 1.63% (period 1) nd 1.59% (period 2) (p=0.002). CI: onfidene intervl; CS: omprtive study; ICU: intensive re unit; MICU: medil ICU; MRSA: metiillin-resistnt Stphyloous ureus; MSICU: medil/surgil ICU; : not ville; OR: odds rtio; PICU: peditri ICU; pd: ptient-dys; RCT: rndomised ontrolled tril; RR: reltive risk; SICU: surgil ICU; SSI: surgil-site infetions; MRSA prevlene in the study setting per 100 ptients dmitted (exept stted differently). Turnround time of the sreening test result (strtified y PCR-sed test vs ulture-sed test, if oth were ompred in the respetive study). Outome mesures: A=MRSA quisition/trnsmission; B=MRSA teremi; C/I=ses of olonistion or (ll/unspeified types of) infetion; I=ses of severl or unspeified types of infetion; W/ SSI=wound infetions/surgil-site infetions.

Tle 1 Studies on the effetiveness of the use of tive surveillne (sreening) for metiillin-resistnt Stphyloous ureus, pulished 2000 2012 (n=41) MRSA ; Culture-sed tests Luet [43]; 6.5%; 1995 2001; Frne; CS. Mlde [37]; 3.2 6.7%; 1996 2004; United Kingdom; Vsulr surgery; Pn [28]; ; 1996 2001; Itly; Hospitl-wide; Reilly [27]; 3.9%; 2008 09; United Kingdom; Hospitl-wide; CS (efore-nd fter). Rodriguez-Bno [30]; 9%; 1995 2008; Spin; Hospitl-wide; CS (interrupted time series). Turnround time (PCR/ Design Sreening followed y Outome Result ulture) Intervention: in period 1 nd 2, nsl sreening of ll ptients t dmission nd weekly therefter; Control: prospetive dt quisition without ontrol group, in period 2 promotion of hnd hygiene. Intervention: nsl sreening of ll ptients t dmission or for eletive dmissions 1 3 weeks prior to dmission; Intervention: nsl sreening of high-risk ptients on high-risk wrds t dmission nd in different intervls therefter. Control: phse without ny or with non-ompulsory Intervention: nsl sreening of ll ptients t dmission; Intervention: phse 2 sreening of ll ptients (nose nd vrious speimens) t dmission nd weekly therefter nd helthre workers; phse 3 sreening of ptients dmitted from other filities; Control: phse without ny or with non-ompulsory Privte rooms, gloves, gowns Deolonistion W/SSI Gloves, deolonistion, gowns (only for infeted ptients) Privte rooms, other preutions unspeified, deolonistion Privte rooms, ontt preutions, deolonistion A B C/I C/I Inidene of MRSA quisition/100 exposed ptients (per 1,000 pd) deresed from 7% (5.43) in period 1 to 2.8% (2.39) in period 2. Period 2 ws n independent protetive ftor influening the inidene of MRSA quisition (OR vs period 1: 0.49; p<0.0001). MRSA wound infetions mong MRSA-positive eletive dmissions redued from 20/36 (56%) to 15/67 (22%) (p=0.002); mong MRSA-positive emergeny dmissions from 35/56 (63%) to 53/121 (44%) (p=0.042). Mjor lim mputtion rtes mong MRSA-positive dmissions redued from 10/36 (18%) to 6/67 (9%) (p=0.026). Inidene rte of MRSA teremi deresed y 42% from 0.64 to 0.37/1,000 dmissions (RR 0.57; 95% CI: 0.35 0.92; p=0.03). This effet ws mostly due to redution of teremi ses relted to entrl venous theters. MRSA infetions (7.5/1,000 pd) redued signifintly over the study period (p=0.0209); dmission prevlene deresed from 5.5% to 3.5% (p< 0.0001). MRSA olonistion nd infetion rtes (0.56 ses/1,000 pd; 95% CI: 0.49 0.62) deresed signifintly to 0.28 ses/1,000 pd (95% CI: 0.17 0.40) in phse 2 nd to 0.07/1,000 pd (95% CI 0.06 0.08) in phse 3. CI: onfidene intervl; CS: omprtive study; ICU: intensive re unit; MICU: medil ICU; MRSA: metiillin-resistnt Stphyloous ureus; MSICU: medil/surgil ICU; : not ville; OR: odds rtio; PICU: peditri ICU; pd: ptient-dys; RCT: rndomised ontrolled tril; RR: reltive risk; SICU: surgil ICU; SSI: surgil-site infetions; MRSA prevlene in the study setting per 100 ptients dmitted (exept stted differently). Turnround time of the sreening test result (strtified y PCR-sed test vs ulture-sed test, if oth were ompred in the respetive study). Outome mesures: A=MRSA quisition/trnsmission; B=MRSA teremi; C/I=ses of olonistion or (ll/unspeified types of) infetion; I=ses of severl or unspeified types of infetion; W/ SSI=wound infetions/surgil-site infetions.

Tle 1d Studies on the effetiveness of the use of tive surveillne (sreening) for metiillin-resistnt Stphyloous ureus, pulished 2000 2012 (n=41) MRSA ; Culture-sed tests Shitrit [31]; 1.6 5.6%; 2002 04; Isrel; Geritri wrd; Souweine [46]; ; 1994 06; Frne; Thompson [32]; 8.1%; 1996 2008; United Kingdom; Tomi [45]; ; 1998 2002; Sloveni; Trohé [44]; 4.2%; 1995 2000; Frne; ICU; CS. Wng [33]; 17.6 26.5%; 2005 06; Tiwn; Turnround time (PCR/ Design Sreening followed y Outome Result ulture) 3d Intervention: sreening (nose, sputum for intuted, perineum, wounds) of high-risk ptients t dmission nd in different intervls therefter; Intervention: sreening (nose, retum) of ll ptients t dmission, weekly therefter nd t dishrge; Intervention: sreening (nose, throt) of ll ptients t dmission nd weekly therefter; Intervention: sreening (nose, throt, wounds nd devies) of high-risk ptients t dmission; Intervention: nsl sreening of ll ptients t dmission, weekly therefter nd t dishrge; prospetive dt quisition without historil or prospetive ontrol group. Intervention: sreening (nres, throt/sputum, xille, inguinl re, wounds) of ll ptients t dmission, every 3 dys therefter nd t dishrge; Control: s in intervention phse ut results were not reported. Privte rooms, gowns, gloves, deolonistion Gloves, gowns, deolonistion Privte rooms, gowns, gloves, deolonistion Privte rooms, gowns, gloves, deolonistion (All ptients in privte rooms), gloves, gowns, deolonistion Privte rooms, gloves, gowns B I A, B C/I A A, I Men numer of MRSA teremi ses per month deresed from 3.6 ses to 1.8 ses fter the intervention (p<0.001). Numer of ptients infeted y MRSA (inluding ses of teremi, pneumoni, urinry trt infetion, theter infetion, wound infetion) deresed from 5.2% to 1.7% (p=0.018). MRSA quisition/1,000 ed-dys deresed from 49.0 (95% CI: 34.4 63.6) to 28.3 (95% CI: 21.7 34.9), 19.3 (95% CI: 16.3 22.3) nd 11.8 (95% CI: 7.3 16.3), respetively; MRSA teremi ses/1,000 ed-dys deresed from 7.6 (95% CI: 4.7 10.5) to 3.7 (95% CI: 2.6 4.8) nd 0.4 (95% CI: 0 2.9). MRSA ses inresed from 4.5 to 8.0/1,000 dmissions fter implementtion of sreening (p=0.02); the proportion of quired MRSA ses deresed from 50% in 1999 to 6% in 2002 (p=0.001). The overll MRSA quisition rte ws 7.9 ses/1,000 pd (p=); it delined in the first three yers fter the implementtion of sreening ut inresed gin, when the dmission prevlene inresed. The inidene of quiring MRSA during ICU sty did not differ signifintly during intervention nd ontrol phses in two prtiipting hospitls (9.6% vs 9.98%; p=0.94; 13.92% vs 13.52%; p=0.81). The inidene of MRSA infetion did not differ either (p=0.719; p=0.932). CI: onfidene intervl; CS: omprtive study; ICU: intensive re unit; MICU: medil ICU; MRSA: metiillin-resistnt Stphyloous ureus; MSICU: medil/surgil ICU; : not ville; OR: odds rtio; PICU: peditri ICU; pd: ptient-dys; RCT: rndomised ontrolled tril; RR: reltive risk; SICU: surgil ICU; SSI: surgil-site infetions; MRSA prevlene in the study setting per 100 ptients dmitted (exept stted differently). Turnround time of the sreening test result (strtified y PCR-sed test vs ulture-sed test, if oth were ompred in the respetive study). Outome mesures: A=MRSA quisition/trnsmission; B=MRSA teremi; C/I=ses of olonistion or (ll/unspeified types of) infetion; I=ses of severl or unspeified types of infetion; W/ SSI=wound infetions/surgil-site infetions.

Tle 1e Studies on the effetiveness of the use of tive surveillne (sreening) for metiillin-resistnt Stphyloous ureus, pulished 2000 2012 (n=41) MRSA ; Culture-sed tests Wrren [34]; 7.2 11.4%; 2002 04; SICU; Wernitz [36]; 20.6%; 1999 2002; Germny; Hospitl-wide; West [35]; 5.3 9.7%; 2001 03; Hospitl-wide; PCR-sed tests Aldey [7]; 6.8 7.3%; 2006 07; United Kingdom; Medil/surgil wrd; Awd [8]; 18%; 2005 08; Hospitl-wide; Turnround time (PCR/ Design Sreening followed y Outome Result ulture) 72 h 19.3 22.7 h / 42.2 51.8 h Intervention: nsl sreening of ll ptients t dmission, weekly therefter nd t dishrge; Intervention: sreening (nose, throt, skin, devies, wounds) of high-risk groups t dmission; Intervention: nsl MRSA sreening of risk ptients t dmission nd weekly therefter; Intervention: phse 1: rpid test on surgil wrd (nres, xille, groin) for ll ptients t dmission nd dishrge; ulture-sed sreening (nres, xille, groin, throt) on medil wrd (4 months) for ll ptients t dmission nd dishrge; Control: phse 2: swith of wrds nd tests. Intervention: multiple mesures (nsl sreening of ll ptients t dmission/trnsfer nd dishrge; ontt isoltion of MRSA infeted or olonised ptients, hnd hygiene mpign, ulturl trnsformtion mpign; Privte rooms, gloves, gowns Privte rooms, gowns, gloves, deolonistion Contt isoltion, gowns, gloves Privte rooms (not for ll); ontt preutions (unspeified) Contt isoltion (unspeified) A I I C/I A, B, I, W/SSI MRSA dmission prevlene inresed (7.2% vs 11.4%; p=0.001); MRSA quisition rte onstnt (7.0 vs 5.5 MRSA ses/1,000 pd; p=0.29). The stndrdised infetion rtio ws 0.52 (95% CI: 0.37 0.71), inditing tht 48% of the expeted hospitlquired MRSA infetions were prevented. Men numer of nosoomil MRSA infetions deresed y 39% from 0.76 to 0.45/1,000 pd (p=0.05) in one, nd y 21% from 0.72 to 0.57 /1,000 pd (p=0.35) in nother hospitl. Hospitl-quired MRSA inidene (ses of olonistion nd infetion) on surgil wrd not redued: 20 (phse 1) vs 22.1/1,000 ed-dys (phse 2) (p=0.69); hospitlquired MRSA inidene rte in medil wrd inresed in rpid test phse: 11.8 (phse 1) vs 20.3/1,000 eddys (phse 2) (p=0.03). MRSA trnsmission deresed from 5.8 to 3.0/1,000 ed-dys (p=0.05); overll MRSA nosoomil infetions deresed from 2.0 to 1.0/1,000 ed dys (p=0.016); overll SSI deresed (p<0.05); nosoomil MRSA loodstrem infetions deresed from 2.9 to 2.5/1,000 ed-dys (p> 0.05). CI: onfidene intervl; CS: omprtive study; ICU: intensive re unit; MICU: medil ICU; MRSA: metiillin-resistnt Stphyloous ureus; MSICU: medil/surgil ICU; : not ville; OR: odds rtio; PICU: peditri ICU; pd: ptient-dys; RCT: rndomised ontrolled tril; RR: reltive risk; SICU: surgil ICU; SSI: surgil-site infetions; MRSA prevlene in the study setting per 100 ptients dmitted (exept stted differently). Turnround time of the sreening test result (strtified y PCR-sed test vs ulture-sed test, if oth were ompred in the respetive study). Outome mesures: A=MRSA quisition/trnsmission; B=MRSA teremi; C/I=ses of olonistion or (ll/unspeified types of) infetion; I=ses of severl or unspeified types of infetion; W/ SSI=wound infetions/surgil-site infetions.

Tle 1f Studies on the effetiveness of the use of tive surveillne (sreening) for metiillin-resistnt Stphyloous ureus, pulished 2000 2012 (n=41) MRSA ; PCR-sed tests Chowers [11]; 2.7 3.7%; 2003 08; Isrel; Hospitl-wide; CS (interrupted time series). Conterno [13]; 2%; 2000 5; Cnd; ICU; Medil/surgil wrd; CS (interrupted time series). Cunninghm [14]; 7.0%; 2005 06; Unired Kingdom; Hrrth [16]; 6.7%; 2003 05; Switzerlnd; Turnround time (PCR/ Design Sreening followed y Outome Result ulture) 24 h / 2 4 d 1.6 d / 3.8 d <1 d / 3 d 22 h / 93 h Intervention: period 1: high-risk ptients sreened t dmission (smple unspeified) + ompline monitoring; period 2: ompline monitoring with sreening/ ontt isoltion disontinued; period 3: PCR-sed sreening of high-risk ptients introdued (smple unspeified); period 4: monitoring re-introdued nd deolonistion disontinued; Control: period 0 without ny or with nonompulsory sreening (sreening of ontt ptients only). Intervention: dmission sreening of high-risk ptients (nose, retum, skin lesions, theter exit sites) using PCR-sed test; Control: dmission sreening of high-risk ptients using ulture-sed test. Intervention: PCR-sed nsl sreening of ll ptients t dmission nd dishrge; Control: sreening with onventionl ultures of ll ptients t dmission. Phse 1: sreening (nose, perineum) of high-risk ptients (ulture-sed); phse 2: universl sreening (PCR-sed) of ll ptients; phse 3: sme s phse 2 ut generl preemptive isoltion. Contt isoltion (unspeified), deolonistion Privte rooms, gloves, gowns; disontinued if PCR not onfirmed y ulture Privte room (if ville), stndrd infetion ontrol preutions, deolonistion Gowns, gloves, msks, deolonistion B C/I A I Period 0 vs period 1: verge numer of teremi ses per 1,000 pd ws redued y ftor 0.55 (95% CI: 0.36 0.83); period 0 vs period 4: verge numer of teremi ses per 1,000 pd deresed y ftor of 0.27 (95% CI: 0.14 0.58); period 1 vs period 4: verge numer of teremi ses per 1,000 pd redued y ftor 0.51 (95% CI: 0.27 0.88) (p=0.02). Insignifint derese of 0.14 nosoomil (deteted >48 h fter dmission) MRSA ses/1,000 pd per month (95% CI: 0.18 0.46) fter the introdution of PCR detetion (p=0.39). Inidene of MRSA trnsmission 13.89 vs 4.9/1,000 pd during ulture nd PCR-phse (RR redution: 0.65; 95% CI: 0.28 1.07). Redution in medil ICU-quired MRSA infetions (RR: 0.3; 95% CI: 0.1 0.7); no effet in surgil ICU (RR: 1.0; 95% CI: 0.6 1.7). CI: onfidene intervl; CS: omprtive study; ICU: intensive re unit; MICU: medil ICU; MRSA: metiillin-resistnt Stphyloous ureus; MSICU: medil/surgil ICU; : not ville; OR: odds rtio; PICU: peditri ICU; pd: ptient-dys; RCT: rndomised ontrolled tril; RR: reltive risk; SICU: surgil ICU; SSI: surgil-site infetions; MRSA prevlene in the study setting per 100 ptients dmitted (exept stted differently). Turnround time of the sreening test result (strtified y PCR-sed test vs ulture-sed test, if oth were ompred in the respetive study). Outome mesures: A=MRSA quisition/trnsmission; B=MRSA teremi; C/I=ses of olonistion or (ll/unspeified types of) infetion; I=ses of severl or unspeified types of infetion; W/ SSI=wound infetions/surgil-site infetions.

Tle 1g Studies on the effetiveness of the use of tive surveillne (sreening) for metiillin-resistnt Stphyloous ureus, pulished 2000 2012 (n=41) MRSA ; PCR-sed tests Hrrth [17]; 5.1%; 2004 06; Switzerlnd; Surgil wrds; Prospetive ohort study. Hrdy [18]; 3.6%; 2005 07; United Kingdom; Surgil wrds; Prospetive ohort study. Jeyrthnm [23]; 6.7%; 2006 07; United Kingdom; Medil/surgil wrd; RCT. Jog [24]; 2.5%; 2004 06; United Kingdom; Crdi surgery; Kjoonegrd [41]; 11.6%; 2009 10; MICU/SICU; Turnround time (PCR/ Design Sreening followed y Outome Result ulture) 22.5 h 0.9 d / 3.3 d 22 h / 46 h Intervention: nsl PCR-sed sreening of ll ptients dmitted to intervention wrds; sreening (swith of intervention nd ontrol wrds fter 9 months). Intervention: Nsl PCR-sed sreening of ll ptients dmitted to wrds ssigned to intervention group; Control: ontrol wrds with ulture-sed sreening; swith of wrds in intervention nd ontrol groups fter 8-month period. Intervention: ll ptients t 10 wrds rndomised to perform rpid or onventionl sreening (nose, xill, groin, skin reks) t dmission nd dishrge; fter wshout period the wrds swed the sreening methods; Control: ptients sreened using onventionl ultures. Intervention: nsl sreening of ptients dmitted for rdi surgery; Intervention: nsl (nd initilly perinel) sreening of ll ICU ptients t dmission; Privte rooms (if ville), gowns, gloves, msks, deolonistion Privte rooms, gowns, gloves, deolonistion Privte rooms, gowns, gloves, deolonistion Privte rooms, stndrd preutions, deolonistion A, I, W/SSI A A W/SSI Contt preutions I Intervention vs ontrol phse: nosoomil (>48 h fter dmission) MRSA infetion rte 1.11 vs 0.91/1,000 pd (djusted inidene rte rtio: 1.20; 95% CI: 0.85 1.69); quisition rte 1.69 vs 1.59/1,000 pd (inidene rte rtio: 1.1; 95% CI: 0.8 1.4); MRSA SSI rte 1.14 vs 0.99/100 surgil interventions (inidene rte rtio: 1.2; 95% CI: 0.8 1.7). Rpid sreening redued MRSA quisition y 1.49 times (95% CI: 1.115 2.003; p=0.007). No hnge in djusted quisition rte (djusted OR: 0.91, 95% CI: 0.61 1.34; p=0.63); MRSA wound infetions in the ontrol rm vs the rpid-test rm (OR: 0.91; 95% CI: 0.48 1.7; p=0.77). Overll SSI rte (ll orgnisms) 3.3% in ontrol vs 2.2% in intervention phse; signifint redution of MRSA SSIs (1.15% vs 0.26%; p<0.05; RR: 0.77; 95% CI: 0.056 0.95). Inrese of helthre-ssoited MRSA infetions fter introdution of sreening (0.8/1,000 dmissions vs 1.6/1,000 dmissions; p=0.037). CI: onfidene intervl; CS: omprtive study; ICU: intensive re unit; MICU: medil ICU; MRSA: metiillin-resistnt Stphyloous ureus; MSICU: medil/surgil ICU; : not ville; OR: odds rtio; PICU: peditri ICU; pd: ptient-dys; RCT: rndomised ontrolled tril; RR: reltive risk; SICU: surgil ICU; SSI: surgil-site infetions; MRSA prevlene in the study setting per 100 ptients dmitted (exept stted differently). Turnround time of the sreening test result (strtified y PCR-sed test vs ulture-sed test, if oth were ompred in the respetive study). Outome mesures: A=MRSA quisition/trnsmission; B=MRSA teremi; C/I=ses of olonistion or (ll/unspeified types of) infetion; I=ses of severl or unspeified types of infetion; W/ SSI=wound infetions/surgil-site infetions.

Tle 1h Studies on the effetiveness of the use of tive surveillne (sreening) for metiillin-resistnt Stphyloous ureus, pulished 2000 2012 (n=41) MRSA ; PCR-sed tests Kurup [22]; 13%; 2007 08; Singpore; Leonhrdt [25]; 1.8 4%; 2009 10; Hospitl-wide; Mrtinez-Cpolino [26]; 13 23%; 2007 8; Prvez [39]; 10.8%; 2008; Hospitl-wide; Roisek [29]; 6.3 8.3%; 2003 07; Turnround time (PCR/ Design Sreening followed y Outome Result ulture) 24 h in 90% of ll ses <24 h / 18 28 h Phse 2: 2.5 d (in-house PCR); phse 3: 0.67 d (ommeril PCR) Intervention: nsl sreening of ll ptients t dmission to the ICU, weekly therefter nd t dishrge; Intervention: nsl sreening of ll dult ptients t dmission or efore in one intervention hospitl; Control: phse with trgeted sreening of highrisk ptients. Intervention: nsl sreening of ll ptients t dmission nd weekly therefter; Intervention: nsl sreening of ll ptients t dmission; Intervention: nsl sreening of ll ptients in the ICU (phse 2) nd generl hospitl-wide sreening nd retesting upon ICU dmission (phse 3); Control: ptients without sreening in phse 1. Privte rooms, gowns, gloves, deolonistion Privte rooms, gowns, gloves, msk, deolonistion Privte rooms, gowns, gloves I I I, B Contt isoltion W/SSI Privte rooms, gowns, gloves, deolonistion I No sttistilly signifint differene in MRSA infetion rte in oth ICUs omined (2.7 to 2.4/1,000 pd; p=0.48). Non-signifint deline in hospitl-quired MRSA infetions of 0.12 perentge points (p=0.34) during the intervention period. Derese in MRSA ventiltor-ssoited pneumoni from 0.95 to 0.17/1,000 pd nd 0.47 to 0.0/1,000 pd in Hospitl 1 nd 2, respetively; derese of MRSA loodstrem infetions from 0.22 to 0.13/1,00 pd nd 0.93 to 0.31/1,000 pd in Hospitl 1 nd 2, respetively; derese of overll hospitl-wide MRSA infetions only in Hospitl 2 (0.63 vs 0.31/1,000 pd); sttistil nlysis. No hnge in the MRSA SSI rte (22/3,862 (0.56%) vs 30/4,076 (0.73%); p=0.362). Aggregte hospitl-ssoited MRSA disese prevlene density hnged y -36.2% (95% CI: -65.4% to 9.8%; p=0.17) from seline to phse 2, nd y -69.6% (95% CI: -89.2% to -19.6%; p=0.03) from seline to phse 3. CI: onfidene intervl; CS: omprtive study; ICU: intensive re unit; MICU: medil ICU; MRSA: metiillin-resistnt Stphyloous ureus; MSICU: medil/surgil ICU; : not ville; OR: odds rtio; PICU: peditri ICU; pd: ptient-dys; RCT: rndomised ontrolled tril; RR: reltive risk; SICU: surgil ICU; SSI: surgil-site infetions; MRSA prevlene in the study setting per 100 ptients dmitted (exept stted differently). Turnround time of the sreening test result (strtified y PCR-sed test vs ulture-sed test, if oth were ompred in the respetive study). Outome mesures: A=MRSA quisition/trnsmission; B=MRSA teremi; C/I=ses of olonistion or (ll/unspeified types of) infetion; I=ses of severl or unspeified types of infetion; W/ SSI=wound infetions/surgil-site infetions.

reviews (the literture lists of the reviews were mnully sreened for dditionl relevnt pulitions). Dt were extrted y AWF nd RK independently using stndrdised form. The study designs were ssigned ording to modified study design sheme pulished y the Centre for Reviews nd Dissemintion t the University of York, United Kingdom, in the NHS eonomi evlution dtse hndook from 2007. Forml ssessment of the qulity of studies ws not performed. Due to the different study outomes inluded, forml met-nlysis ws onsidered inpproprite. Heterogeneity in methodology nd outome mesures lso prevented quntittive ssessment of pulition is. Results The literture serh identified 9,340 rtiles, 151 of whih were retrieved s full texts fter review of titles nd strts. Of these, 69 rtiles fulfilled the riteri for inlusion nd further 14 rtiles were dded fter serh through the literture lists of exluded review rtiles (Figure). Overll, 83 rtiles were inluded in the review [7-89]. Sreening We identified 41 studies tht investigted the question whether sreening for MRSA rriers efore or on dmission hd n impt on MRSA quisition or infetion rtes (Tle 1) [7-47]. Culture-sed sreening Twenty-five studies used ulture-sed sreening pprohes, inluding two rndomised ontrolled trils (RCTs) nd 23 omprtive studies mostly using efore-nd-fter design [9,10,12,15,19-21,27,28,30-38,40,42-47]. Of these 25 studies, seven used unspeified ulture-sed tehniques [12,21,27,28,37,40,46], eight used MRSA hromogeni medi (t lest prtilly) [19,31-34,38,45,47] nd the others used mnnitol slt, oxillin slt or lood grs. An estimte for the turnround times (TAT) of sreening results ws only reported in eight of the 25 studies (1 d 5.2 d) [10,12,19-21,33,34,38]. Overll, 19 of the 23 omprtive studies inluded reported trends of deresing rtes of MRSA infetion or olonistion [10,12,15,19,21,27,28,30-32,35-38,40,42,43,45,46], two reported miguous results [44,47], nd two reported no redution of MRSA infetions or trnsmission [33,34]. The two RCTs found no redution of MRSA infetions or trnsmission [9,20]. PCR-sed sreening Sixteen studies used PCR-sed sreening tehniques in their intervention phses, inluding one RCT, two prospetive ohort studies nd 13 omprtive studies [7,8,11,13,14,16-18,22-26,29,39,41]. The TAT of the PCR sreening result ws reported in 11 of 16 studies (0.67 d 1.5 d) [7,11,13,14,16-18,23,25,26,29]. Overll, seven of 16 studies doumented positive effets on the ourrene of MRSA infetions or trnsmissions fter implementtion of sreening [8,11,14,18,24,26,29]. One study reported miguous results [16]. Among the studies reporting derese of infetion or trnsmission, five ompred the intervention group (PCR-sed sreening) to ontrol group without tive surveillne, with non-ompulsory tive surveillne or with sreening of limited risk groups [8,11,24,26,29], nd two ompred with ontrol group where routine ulture-sed sreening ws performed [14,18]. Among the eight studies whih ould not doument deresing trends in MRSA infetions or trnsmission following the implementtion of sreening, three ompred PCRsed sreening with ulture-sed sreening [7,13,23], four ompred the intervention to ontrol periods without ny tive surveillne of MRSA [17,22,39,41], nd one ompred the intervention with seline period where PCR-sed sreening of seleted risk ptients ws performed [25]. Sreening (PCR-sed nd ulture-sed) vs no sreening strtified y outome mesure In eight of nine studies (89%) using this outome prmeter, MRSA teremi rtes deresed fter implementtion of sreening [8,11,21,26,28,31,32,38,47]. Inidene of MRSA quisition or trnsmission deresed in three of eight studies (38%) ssessing this outome prmeter [8,9,17,32-34,43,44]. Three of five studies (60%) using wound infetion nd surgilsite infetions (SSI) s n outome prmeter showed deresing SSI rtes fter implementtion of sreening [8,17,24,37,39]. A derese of MRSA ws oserved in 20 of 23 studies (87%) using ll or unspeified MRSA infetions or ses of olonistion/infetion s their outome prmeters [8-10,12,15-17,19,20,22,25-27, 29,30,35,36,40-42,45-47]; mong these studies, one found derese only in medil ICUs [16]. PCR-sed vs ulture-sed sreening Five investigtions ompred PCR-sed to ulturesed sreening [7,13,14,18,23]. All five doumented tht the TAT ws redued when ompred to ulturesed pprohes (Tle 1). However, three studies found no differene in MRSA quisition or infetion rtes [7,13,23]. In ontrst, one efore-nd-fter study found redution in the inidene of MRSA trnsmission fter introdution of the PCR-sed test whih lmost rehed sttistil signifine, nd one ohort study reported redution in MRSA quisition rtes [14,18]. Deolonistion A totl of 11 RCTs, 23 omprtive studies nd one prospetive ohort study evluted the effetiveness of mupiroin-sed nsl deontmintion regimens for the prevention of S. ureus infetions (Tle 2) [48-82]. Of ll 11 RCTs, six demonstrted signifintly deresing infetion trends fter implementtion of deolonistion [48,51,52,72,73,75]; for one of these, this ws only oserved when seletive digestive deontmintion ws dded to nsl deolonistion [52], nd for one RCT, the effet ws only nlysed for Grm-positive infetions (whih were mostly MRSA) [75]. Strtified y www.eurosurveillne.org 11

types on infetions prevented, the RCTs showed tht deolonistion deresed deep S. ureus SSI [48], overll S. ureus infetions [48,51,73], overll infetion rtes [52], Grm-positive pneumoni [75] nd S. ureus exit-site infetions [72]. Among the 24 non-rndomised studies identified, 19 reported evidene tht the use of mupiroin ws effetive in reduing infetion. Of the seven studies performed in ICUs, six (86%) demonstrted n effet; speifilly, derese in pneumoni nd hospitlquired S. ureus infetion [59], in the overll infetion rtes in ICUs [50,70], in MRSA SSI nd loodstrem infetions (BSI) in ICUs [55], nd in the overll numer of MRSA infetions in ICUs [80,81]. Non-ontrolled studies implementing deolonistion in non-icu settings led to derese in overll nd peristoml MRSA infetions [57,76], in the inidene of S.ureus/MRSA SSI in surgil units [55,58,64,65,71,77,79], in overll S. ureus/mrsa infetions in gstrointestinl surgery nd orthopedis [49,82], nd in the totl rte of SSI or wound infetions [53,60,67]. Strtified y different implementtion settings, four of five studies doumented suess mong ptients undergoing rdiothori surgery [53,65,66,71,77], four of six in orthopedi deprtments [49,60,61,63,64,79], nd six of seven in other or mixed surgil deprtments [54,55,58,67,73,75,82]. Moreover, seven of eight studies performed in ICU settings [50,52,55,59,68,70,80,81], two of two performed in hemodilysis units [51,72], two of five performed in different non-surgil deprtments [56,57,69,76,78], nd one of three studies performed hospitl-wide or in oth medil nd surgil deprtments [48,62,74], demonstrted suessful effets of mupiroin-tretment. Strtified y different ustive orgnisms, eight studies showed tht mupiroin-tretment led to derese in the overll inidene of infetions due to ll orgnisms [49,53,60,64,65,67,70,77]. In the sme studies, this effet ws prtilly non-signifint for S. ureus/ MRSA infetions in prtiulr [53,60,67,70]. Four studies reported derese in infetions used y methiillin-sensitive S. ureus (MSSA) [48,51,55,65]. Twelve investigtions reveled redution in MRSA infetions [49,50,55,57,58,64,76,77,79-82], six showed deresing trends for S. ureus (MRSA nd/or MSSA) infetions [50,59,71-73,82] nd one reported redution of pneumoni used y Grm-positive teri (mostly MRSA) [75]. Mny of the studies identified in this review used mupiroin-only regimens [51,55,59,60,63,67,70-73,75,78,82]. Others omined nsl mupiroin with other topil gents to support deolonistion, inluding hlorhexidine [48,50,53,56-58,61,62,64-68,74,81], trilosn [49,76,79], extr-nsl use of mupiroin [69,77,80], seletive digestive deontmintion [52], povidone-iodine [49], nd systemi ntiiotis [54]. Isoltion Fousing on the physil isoltion of ptients in seprte single or ohort rooms, we identified one ohort study nd seven omprtive studies reporting on the effetiveness of this mesure (Tle 3) [16,83-89]. Five studies were performed in ICU settings [16,83-85,88], one in vsulr surgery wrd, one in dieti food unit, nd one hospitl-wide [86,87,89]. In two of these studies, nurse ohorting ws performed in ddition to single-room isoltion [83,86]. Overll, one ohort nd three omprtive studies reported on enefiil effets of single-room isoltion (not performed pre-emptively) on MRSA olonistion or infetion [85,86,88] nd on quisition rtes [84]. Two omprtive studies did not find redution of trnsmission [83] or MRSA prevlene [87]. Three studies ssessed the role of pre-emptive isoltion mesures pending the results of sreening [16,86,89]. In one efore-nd-fter study, pre-emptive isoltion preutions led to redution of the MRSA quisition rte (0.21% vs 0.07%; p=0.04) [89]. In retrospetive omprtive study pling ll dmitted ptients in pre-emptive isoltion, the numer of nosoomil MRSA isoltes ws redued (p=0.005). However, simultneous introdution of ohort isoltion fility with dedited stff mkes the effets of this mesure indistinguishle from the effets of preemptive isoltion [86]. The third ws study tht evluted the effets of simultneous implementtion of pre-emptive isoltion nd rpid sreening test on the inidene of MRSA infetions in two ICUs [16] resulting in signifint redution of ICU-quired infetions in medil ut not in surgil ICU. Disussion Improving the rtionl use of ntiiotis nd the implementtion of hnd hygiene re lerly ornerstones of MRSA prevention nd ontrol [90-92]. Moreover, enhmrking nd puli reporting systems hve reently een demonstrted to suessfully support infetion ontrol mesures [93]. However, the effetiveness of sreening, deolonistion nd isoltion for MRSA prevention when implemented routinely in settings with endemi MRSA, remins ontroversil. For exmple, it is deted to wht extent miroiologil, strin-speifi ftors hve ontriuted to the deresing MRSA trends [94,95]. Therefore, the present review imed to fous on three importnt mesures nd to summrise the urrent evidene for their impt on MRSA prevention. Sreening The strtegy of sreening is sed on the finding tht miroiologil ultures performed for linil resons fil to detet previously unknown MRSA rriers t dmission in 69 to 85% of ptients [96,97]. Tehnilly, sreening n e performed y ulture-sed methods (sreening sw streked onto non-seletive or hromogeni medi) or PCR-sed tests. 12 www.eurosurveillne.org

Tle 2 Studies on the effetiveness of Stphyloous ureus deolonistion using mupiroin-sed regimens, pulished 2000 2012 (n=35) Bode [48]; 2005 07; Netherlnds; Surgery nd internl mediine; Rndomised pleoontrolled tril. Boelert [51]; ; Belgium; Hemodilysis; Rndomised pleoontrolled tril. Cmus [52]; 1996 08; Frne; MICU; Rndomised pleoontrolled tril. Cimohowski [53]; 1995 99; Crdiothori surgery; Prospetive omprtive (efore-nd-fter). Cordov [54]; ; Dermtology (Mohs surgery); Retrospetive omprtive (efore-nd-fter). Tretment regimen Tretment of Mupiroin 2xd nd hlorhexidine gluonte (40 mg/ml) sop for 5 dys; further ourses of sme tretment for ptients stying >3 weeks. Mupiroin 3xd for 2 weeks; susequently 3x per week for 9 months. Group 1: mupiroin 3xd for 5 dys; gin 5 dys if nsl S. ureus; hlorhexidine gluonte (4%) totlody wshing 2xd (until 24 h fter extution; mx 90 dys); Group 2: sme s group 1 plus seletive digestive deontmintion Mupiroin the night nd morning efore surgery, efore surgery, then 2xd for 5 dys; hlorhexidine shower efore surgery. Mupiroin 2xd for 5 7 dys nd orl trimethoprim-sulfmethoxzole for 5 7 dys S. ureus rriers only S. ureus rriers only All ptients irrespetive of rrige All ptients irrespetive of rrige rriers Effets of tretment strtified y pthogen All orgnisms MRSA+ MSSA MRSA MSSA n.s NS Effet of tretment Redution of hospitl-quired MSSA infetion (3.4% vs 7.7%; RR: 0.42; 95% CI: 0.23 0.75), deep MSSA SSI (RR: 0.21; 95% CI: 0.07 0.62) ut not of superfiil MSSA SSI 0.45 (0.18 1.11) nd MSSA lower respirtory infetions 0.82 (RR: 0.82; 95% CI: 0.12 5.78). Redution of MSSA infetions (1/104 ptientmonths vs 6/147 ptient-months; p<0.05). Group 1: numer of quired infetions did not differ (OR: 0.98; 95% CI: 0.6 1.58; p=0.92). Group 2: numer of quired infetions inl. VAP, UTI, theter-relted infetions differed (OR: 0.42; 95% CI: 0.25 0.73; p=0.002). Redution of overll SSI (0.9 vs 2.7%; p=0.005), ut not S. ureus SSI (4/854 vs 11/992; p>0.05). MRSA SSI: 0.3% in historil ohort (12/3,633) vs 0% in tretment group (0/962); sttistil nlysis ; Fisher's ext test performed y the uthors of this review: p=0.08). Types of infetions nlysed seprtely, SSI, VAP/LRTI, (BSI, UTI ssessed, ut smll numers), VAP, UTI, BSI Sternl SSI SSI BSI: loodstrem infetions; CI: onfidene intervl; diverse: diverse or ll types of infetions ICU: intensive re unit; LRTI: lower respirtory trt infetions; MICU: medil intensive re unit; MRSA: metiillin-resistnt Stphyloous ureus; MSSA: metiillin-sensitive Stphyloous ureus; : no dt ville; NS: not signifint; redution; inrese; OR: odds rtio; pd: ptient-dys; RR: reltive risk; SSI: wound infetions or surgil-site infetion, VAP: ventiltor-ssoited pneumoni; UTI: urinry trt infetions; Mupiroin refers to mupiroin nsl ointment unless speified otherwise. Chlorhexidine nd trilon ody wshes, 1xd or 2xd or 3xd refers to pplition 1x, 2x or 3x per dy. Only when seletive digestive deontmintion ws dded to mupiroin-tretment. MSSA nd ogulse-negtive stphylooi. d Grm-positive infetions (mostly MRSA).

Tle 2 Studies on the effetiveness of Stphyloous ureus deolonistion using mupiroin-sed regimens, pulished 2000 2012 (n=35) Dupeyron [56]; 1999 2001; Frne; Digestive disese unit; Prospetive omprtive (efore-nd-fter). Dupeyron [57]; 2000 04; Frne; Gstroenterology; Prospetive omprtive (interrupted-time-series). Frser [59]; 2006 07; MICU; Retrospetive omprtive (efore-nd-fter). Gernt-vn der Sluis [60]; 1992 06; The Netherlnds; Orthopedi wrds; Prospetive omprtive (efore-nd-fter). Hdley [61]; 2007 09; Orthopedi wrds; Retrospetive omprtive (efore-nd-fter). Tretment regimen Tretment of Mupiroin 3xd for 5 dys; hlorhexidine (4%) every seond dy during mupiroin tretment; further tretment ourses in se of filure. Mupiroin 3xd for 5 dys; hlorhexidine (4%) every seond dy during mupiroin tretment; further ourses in se of filure. Mupiroin (5 doses) Mupiroin thrie efore surgery Mupiroin (2%) for 5 dys (dose unspeified); hlorhexidine one preopertively. rriers rriers S. ureus rriers only All ptients irrespetive of rrige All ptients irrespetive of rrige Effets of tretment strtified y pthogen All orgnisms MRSA+ MSSA MRSA MSSA NS NS NS NS Effet of tretment Overll MRSA infetions: 1.41/1,000 pd in ontrol period nd 1.46/1,000 pd in intervention period (sttistil nlysis ). Redution of overll MRSA infetions (1.41/1,000 pd in the yer efore intervention to 1.40, 0.74, 0.59/1,000 pd in different periods therefter, p=0.002). Redution of S. ureus VAP (p=0.03; RR 0.12; 95% CI: 0.01 0.83), overll S. ureus infetions (p=0.03; RR 0.37; 95% CI: 0.14 0.90), ut NS effet on S. ureus BSI (p=0.28). Redution of overll SSI (p=0.02), ut NS redution of S. ureus SSI (p=0.3). No redution of SSI rte (1.28% in the tretment vs 1.45% in the ontrol group; p=0.809) nd no redution of MRSA SSI (0.24% vs 0.30%; NS). Types of infetions nlysed seprtely, BSI, VAP/LRTI SSI SSI BSI: loodstrem infetions; CI: onfidene intervl; diverse: diverse or ll types of infetions ICU: intensive re unit; LRTI: lower respirtory trt infetions; MICU: medil intensive re unit; MRSA: metiillin-resistnt Stphyloous ureus; MSSA: metiillin-sensitive Stphyloous ureus; : no dt ville; NS: not signifint; redution; inrese; OR: odds rtio; pd: ptient-dys; RR: reltive risk; SSI: wound infetions or surgil-site infetion, VAP: ventiltor-ssoited pneumoni; UTI: urinry trt infetions; Mupiroin refers to mupiroin nsl ointment unless speified otherwise. Chlorhexidine nd trilon ody wshes, 1xd or 2xd or 3xd refers to pplition 1x, 2x or 3x per dy. Only when seletive digestive deontmintion ws dded to mupiroin-tretment. MSSA nd ogulse-negtive stphylooi. d Grm-positive infetions (mostly MRSA).

Tle 2 Studies on the effetiveness of Stphyloous ureus deolonistion using mupiroin-sed regimens, pulished 2000 2012 (n=35) Hrrth [62]; 1995 07; Switzerlnd; Hospitl-wide; Rndomised pleoontrolled tril. Hung [80]; 2003 06; Tiwn; Neontl ICU; Prospetive omprtive study (efore-nd fter). Klmeijer [63]; 1997 09; The Netherlnds; Orthopedi wrds; Rndomised pleoontrolled tril. Keshtgr [55]; 2000 06; United Kingdom; ICU nd surgery; Prospetive omprtive study (efore-nd fter). Kim [64]; 2005 07; Orthopedi wrds; Prospetive omprtive (efore-nd-fter). Kluytmns [65]; 1989 92; The Netherlnds; Crdiothori surgery; Retrospetive omprtive (efore-nd-fter). Tretment regimen Tretment of Mupiroin 2xd for 5 dys; hlorhexidine for 5 dys. Mupiroin 2xd for 5 dys Mupiroin 2xd until dy of surgery (t lest 2 doses efore surgery). Mupiroin 3xd for 5 dys; hlorhexidine (use unspeified exept for hirwsh on dys 1, 3, 5). Mupiroin 2xd for 5 dys; hlorhexidine 1xd for 5 dys (3 dys for MSSA). Mupiroin 2xd for 5 dys; hlorhexidine efore surgery. rriers rriers All ptients irrespetive of rrige rriers S. ureus rriers only All ptients irrespetive of rrige Effets of tretment strtified y pthogen All orgnisms MRSA+ MSSA MRSA MSSA NS NS NS Effet of tretment No redution of overll MRSA infetions (3/48 vs 7/50; p=0.32). Redution of overll MRSA infetions in the group of neontes treted (92/783 vs 5/450; OR: 11.85; 95% CI: 4.6-33.3). No signifint redution of endogenous S. ureus SSI (0.3% in tretment vs 1.7% ontrol group; RR: 0.19; 95% CI: 0.02-1.62). Redution of MRSA BSI y 38.5% (p<0.001), MSSA BSI y 30.4% (p<0.001), MRSA SSI y 12.7% (p=0.021); ut inrese of MSSA SSI y 12.7% (p=0.006). Redution of overll SSI (p=0.0093), MRSA SSI (0.19% vs 0.06%, p=0.0315), MSSA SSI (0.26% vs 0.13%, p=0.0937). 2.8%; p<0.001) nd of S. ureus/ogulsenegtive Redution of the overll rte of SSI (7.3% vs stphylool SSI (p=0.0032). Types of infetions nlysed seprtely SSI BSI, SSI SSI SSI BSI: loodstrem infetions; CI: onfidene intervl; diverse: diverse or ll types of infetions ICU: intensive re unit; LRTI: lower respirtory trt infetions; MICU: medil intensive re unit; MRSA: metiillin-resistnt Stphyloous ureus; MSSA: metiillin-sensitive Stphyloous ureus; : no dt ville; NS: not signifint; redution; inrese; OR: odds rtio; pd: ptient-dys; RR: reltive risk; SSI: wound infetions or surgil-site infetion, VAP: ventiltor-ssoited pneumoni; UTI: urinry trt infetions; Mupiroin refers to mupiroin nsl ointment unless speified otherwise. Chlorhexidine nd trilon ody wshes, 1xd or 2xd or 3xd refers to pplition 1x, 2x or 3x per dy. Only when seletive digestive deontmintion ws dded to mupiroin-tretment. MSSA nd ogulse-negtive stphylooi. d Grm-positive infetions (mostly MRSA).