Appropriateness of antimicrobial therapy: a multicentre prevalence survey in the Netherlands,

Transcription

1 Surveillnce nd outbrek reports Appropriteness of ntimicrobil therpy: multicentre prevlence survey in the Netherlnds, I Willemsen 1, T vn der Kooij 2, B vn Benthem 2, J Wille 3, J Kluytmns (jnkluytmns@gmil.com) 1,4 1. Lbortory for Microbiology nd Infection Control, Amphi Hospitl, Bred, the Netherlnds 2. Netherlnds Centre for Infectious Disese Control, Ntionl Institute for Public Helth nd the Environment (Rijksinstituut voor Volksgezondheid en Milieu, RIVM), Bilthoven, the Netherlnds 3. Dutch Institute for Helthcre Improvement (CBO), Utrecht, the Netherlnds 4. Deprtment of Microbiology nd Infection Control, VU University Medicl Center (VU medisch centrum, VUmc), Amsterdm, the Netherlnds Cittion style for this rticle: Willemsen I, vn der Kooij T, vn Benthem B, Wille J, Kluytmns J. Appropriteness of ntimicrobil therpy: multicentre prevlence survey in the Netherlnds, Euro Surveill. 21;15(46):pii= Avilble online: Article published on 18 November 21 A survey ws crried out to determine the prevlence nd ppropriteness of ntimicrobil therpy (AMT) in the Netherlnds nd to identify determinnts for inpproprite AMT. Prevlence surveys of ptients hospitlised in the Netherlnds were performed three times in 28 nd 29. Ptients demogrphic, infectionrelted nd AMT-relted dt were collected from hospitl wrds. A totl of 19 hospitls prticipted, consisting of mix of university, teching nd generl hospitls, which were distributed evenly cross the country. The ppropriteness of AMT ws ssessed using stndrdised lgorithm bsed on locl AMT prescription guidelines. A totl of 7,853 ptients were included, of which 2,327 (29.6%) ptients were on AMT (rnge: %). In 372 ptients (16% of ptients on AMT), tretment ws considered inpproprite. In 265 (11.4%) ptients on AMT, ppropriteness of tretment ws not judged becuse of insufficient informtion. The percentge of ptients without judgment vried considerbly between the prticipting hospitls (rnge: %). Approprite AMT use ws significntly ssocited with ptient being in n intensive cre unit, hving centrl venous ctheter nd being given bet-lctmse-sensitive penicillins. The use of fluoroquinolones ws significntly ssocited with more frequent inpproprite use. There ws considerble nd significnt vrition between the prticipting hospitls in the mount of ntimicrobils prescribed nd the ppropriteness of their use. To improve the completeness nd relibility of such surveys, there is need for intensive trining of observers nd medicl stff in recording informtion. Introduction Point prevlence surveys re useful wys of investigting helthcre-relted events, including ntimicrobil use. The first report on ntimicrobil use mesured in prevlence surveys ws published in 1983 [1]. More recently Europen project the Europen Surveillnce of Antimicrobil Consumption (ESAC) hs stndrdised method to determine the prevlence of ntimicrobil therpy (AMT) in hospitls [2]. In previous study, performed in teching hospitl in the Netherlnds, we showed tht besides the prevlence of AMT, the ppropriteness of AMT for individul ptients could lso be determined, bsing the judgement on locl ntibiotic prescription guidelines [3]. This enbles reserchers to quntify the number of ptients who re treted even when tretment is not indicted or who re treted with drug tht is not the preferred choice. In ddition, it ws possible to identify determinnts of inpproprite use of AMT [3]. The objective of the current study ws to determine whether prevlence studies could be used in other hospitls s well nd whether they could lso be used s tool for benchmrking. The study ws coordinted by the PREZIES (Prevention of Nosocomil Infections by Surveillnce) network collbortion between prticipting hospitls, the Dutch Institute for Helthcre Improvement (CBO) nd the Ntionl Institute for Public Helth nd the Environment (Rijksinstituut voor Volksgezondheid en Milieu, RIVM), RIVM). Methods Prevlence surveys Hospitls were recruited from PREZIES. Those tht hd lredy prticipted in prevlence surveys for nosocomil infections were invited to dd AMT use to this survey. Prticiption ws on voluntry bsis. Three prevlence surveys of hospitlised ptients were performed in spring 28, utumn 28 nd spring 29. All ptients tht were present in the hospitl t :1 on the dy of the survey were included. Ptients in dy cre (including hemodilysis ptients) nd psychitric wrds were excluded. Dt from the hospitl wrds were collected by infection control prctitioners. A stndrdised cse record form, to document ptients demogrphic, infection-relted nd AMT-relted dt, ws used. The infection control prctitioners received 1

2 trining during workshops on how to collect the dt nd how to judge the ppropriteness of AMT. The following demogrphic vribles were recorded: ge, sex, medicl specility, type of wrd nd presence of infection on dmission. Nosocomil infections were recorded using the definitions from the United Sttes Centers for Disese Control nd Prevention (CDC) nd ptients hd to be symptomtic or still being treted on the dy of the survey [4,5]. Furthermore, the use of ntimicrobil gents nd dosge were noted. If more thn one ntimicrobil ws prescribed for one ptient, ll ntibiomicrobils, up to mximum of three, were registered. Antifungl nd ntivirl therpy s well s mediction for tuberculosis were excluded from the study. The min resons were tht not ll locl guidelines hd specific recommendtions for these gents nd susceptibility of pthogens to these gents ws not lwys determined by the locl microbiology lbortories. Appropriteness of ntimicrobil therpy The ppropriteness of AMT ws determined using stndrdised method developed by Gyssens et l. [6], using the following clssifictions: correct decision, incorrect decision, incorrect choice or insufficient dt. This clssifiction system obviously only tkes into ccount ptients tht re on AMT. However, using our pproch it is possible to exmine the ppropriteness of not prescribing AMT lso, s described in our previous study [3]. A correct decision ws deemed pproprite; incorrect decision nd incorrect choice were considered inpproprite (the evlution criteri re summrised in Tble 1). The ppropriteness of AMT ws judged ccording to the locl AMT prescription guidelines present in ll Tble 1 Evlution criteri for ppropriteness of ntimicrobil therpy, the Netherlnds, Ctegories nd criteri 1. Correct decision (pproprite use) 1.1 No AMT nd no infection nd no AMT needed 1.2 No AMT nd infection nd no AMT needed 1.3 AMT nd infection nd pproprite choice nd pproprite use 2. Incorrect decision (inpproprite use) 2.1 No AMT nd infection nd AMT needed 2.2 AMT nd no infection nd no prophylxis nd no AMT needed 2.3 AMT nd no infection nd prophylxis nd no AMT needed 3. Incorrect choice (inpproprite use) 3.1 Divergence from guidelines 4. Missing dt (insufficient informtion) 4.1 No AMT nd not enough dignostic informtion bout infection 4.2 Infection nd not enough dignostic informtion if AMT is needed 4.3 AMT nd not enough dignostic informtion bout infection 4.4 Infection nd not enough informtion bout AMT prticipting hospitls. These locl guidelines re bsed on the ntionl policy developed by the Dutch Working Prty on Antibiotic Policy (Stichting Werkgroep Antibiotic Beleid, SWAB) [7,8]. The infection control prctitioners ssessed the ppropriteness of AMT initilly: if they could not decide, consultnt microbiologist or infectious disese physicin mde the finl judgment. The consultnt microbiologist or infectious disese physicin lso judged ll ptients in intensive cre units, ll ptients who received AMT without hving n ctive infection (ccording to the survey), ll ptients who did not receive AMT nd did hve n ctive infection nd ll ptients who received AMT tht ws not ccording to the locl AMT guidelines. If ll the ntimicrobil gents tht ptient received were considered correct, the tretment ws considered pproprite. If one or more of the ntimicrobil gents ws considered incorrect, the tretment ws considered inpproprite. If it ws not possible to decide whether use of prticulr ntimicrobil gent ws correct due to incomplete informtion, tretment ws recorded s insufficient informtion. We did not ssess the reproducibility of the judgments. Dt nlysis, qulity control nd sttisticl nlysis Privcy of ptients is ensured by decoding ll dt, s required by the privcy regultions in the Netherlnds. Dt were entered in the PREZIES dtbse or hospitl-owned dtbse nd subsequently coded nd trnsferred to PREZIES. Dt were nlysed using SPSS version 17.. Tretment for which there ws insufficient informtion ws recorded s missing vlue. Ctegoricl vribles were nlysed by Fisher s exct test or chi-squre test where pproprite: continuous vribles were nlysed using t-test or Mnn Whitney U test where pproprite. Binry logistic regression nlysis ws performed: ll vribles with p vlue less thn.1 in univrite nlyses were entered into the multivrite model. Sttisticl significnce ws ccepted when the chnce for coincidence ws less thn 5%. Finlly, sensitivity nlysis ws performed. In this, the univrite nd multivrite nlyses were repeted: once ctegorising AMT use s pproprite for ll ptients for whom AMT use could not be judged nd once ctegorising it s inpproprite. Results A totl of 7,853 ptients were included, from 19 hospitls. They were mix of university, teching nd generl hospitls, which were distributed evenly cross the country. Of these, 13 prticipted in one of the three surveys, five prticipted in two surveys nd one prticipted in ll three. A men of 32 ptients were included per hospitl per prevlence survey (rnge: ; stndrd devition: 149). AMT: ntimicrobil therpy. 2

3 Ptient chrcteristics nd nosocomil infections Overll 3,784 (48.2%) ptients were mle, nd the men ge ws 62 yers (medin: 67 yers). On the dy of the survey 426 ptients (5.4%) hd t lest one ctive nosocomil infection. Antimicrobil therpy A totl of 2,327 ptients (29.6%) were on AMT (rnge: %). The men prevlence of AMT per prticipting hospitl is shown in Figure 1. Of the 2,327 ptients on AMT, 433 (18.6%) were treted with two ntimicrobils, nd 58 (2.5%) were treted with three or more. In totl 2,876 courses of ntimicrobil gents were dministered, of which 1,79 (59.4%) were given intrvenously (rnge: %). The first ntimicrobil gent ws considered pproprite in 1,69 (72.6%) ptients. In 149 (6.4%) ptients the first ntimicrobil gent ws considered not justified Figure 1 Men prevlence of ntimicrobil therpy per prticipting hospitl, the Netherlnds, Prevlence (%) Hospitl number The hospitls re shown in incresing order of prevlence. Verticl brs represent the 95% confidence intervls. Figure 2 Inpproprite ntimicrobil therpy nd proportion of ptients without ntimicrobil therpy judgment, by prticipting hospitl, the Netherlnds, Percentge Hospitl number b Inpproprite ntimicrobil therpy (%) Without ntimicrobil therpy judgment (%) The hospitls re shown in order of incresing proportion of inpproprite use. Due to missing informtion. b The numbering of hospitls is identicl to the hospitl numbers in Figure

4 nd therefore inpproprite. In 223 (9.6%) ptients AMT ws justified, but the choice of the gent ws not ccording to the guidelines. In 265 (11.4%) ptients no decision ws mde due to insufficient informtion. The second ntimicrobil gent ws considered pproprite in 384 ptients (78.2% of the 491 ptients treted with more thn one ntimicrobil gent), not justified in 26 (5.3%) nd justified but n incorrect choice in 39 (7.9%) ptients. In 42 (8.6%) ptients no choice ws mde due to insufficient informtion. The third ntimicrobil ws considered pproprite in 5 of the 58 ptients treted with t lest three ntimicrobil Tble 2 Use of ntimicrobil gents in prticipting hospitls, the Netherlnds, Antimicrobil gent First ntibiotic Second ntibiotic Third ntibiotic Totl n % n % n % n % bet-lctmse-sensitive penicillins bet-lctmse-resistnt penicillins Crbpenems Co-moxicillin-clvulnic cid First-genertion cephlosporins Second-genertion cephlosporins Third- nd fourth-genertion cephlosporins Co-pipercillin-tzobctm Sulfonmides nd trimethoprim Fluoroquinolones Glycopeptides 37 1,6 16 3,3 2 3,4 55 1,9 Imidzole derivtes 55 2, ,9 4 6, ,4 Lincosmides nd mcrolides 84 3,6 39 8, 1 17, ,6 Brod-spectrum penicillin 128 5,5 34 6,9 2 3, ,7 Aminoglycosides 3 1,3 4 8,2 8 13,8 78 2,7 Tetrcyclines 42 1,8 6 1,2 1 1,7 49 1,7 Other ntimicrobils 63 2,7 3 6,1 6 1,3 99 3,4 Totl 2, ,876 1 Totl number of courses of ntimicrobil gents dministered tht were recorded on the survey dys. Figure 3 Reltive risk for inpproprite use of ntimicrobil therpy by group of ntimicrobil gent, the Netherlnds, Reltive risk 1..5 bet-lctmse-sensitive penicillins Crbpenems Aminoglycosides Third- nd fourth-genertion cephlosporins Second-genertion cephlosporins β-lctmse-resistnt penicillins Co-moxy-clvulnic cid First-genertion cephlosporins Lincosomides nd mcrolides Sulfonmides nd trimethoprim Fluoroquinolones Verticl brs represent the 95% confidence intervls. Co-moxycillin-clvulnic cid s reference. 4

5 gents, not justified in two nd justified but n incorrect choice in two ptients. Four were not judged due to insufficient informtion. Judgment of the ppropriteness of AMT per ptient showed tht 372 ptients (16% of the ptients on AMT; 4.7% of the totl popultion) were treted inppropritely. Figure 2 shows the vritions in the proportion of AMT considered inpproprite in the different hospitls (rnge: %). For 265 ptients (11.4%) on AMT it ws not possible to judge ppropriteness becuse of insufficient informtion. Figure 2 shows the vritions in the proportion of ptients on AMT who could not be judged in the prticipting hospitls (rnge: %). Of the ptients who did not receive AMT (n=5,526), 945 were not judged for the ppropriteness of the decision not to tret. This ws minly due to four hospitls tht did not judge ptients who were not receiving AMT. Of the 4,581 ptients not receiving AMT who were judged, the decision not to use AMT ws considered pproprite for 4,497 (98.2%) ptients. For 22 ptients (.5% of those not on AMT who were judged), ptients did not receive AMT, lthough this ws indicted. For 62 (1.4%) ptients not on AMT it ws not possible to ssess the ppropriteness becuse of insufficient informtion. Figure 4 Reltive risk for inpproprite use of ntimicrobil therpy by medicl specilty, the Netherlnds, Reltive risk Crdio-thorcic surgery Generl surgery Lung diseses Orthopedics Internl medicine Neurology Gynecology Urology Verticl brs represent the 95% confidence intervls. Internl medicine s reference. Figure 5 Reltive risk for inpproprite use of ntimicrobil therpy, the Netherlnds, Reltive risk Presence of ESBL-producing bcteri b Admission to n intensive cre unit Arteril ctheter Centrl venous ctheter Isoltion precutions Epidurl ctheter Peripherl vsculr ctheter Nosocomil infection Urinry trct ctheter Suprpubic ctheter Verticl brs represent the 95% confidence intervls. All determinnts re dichotomous vribles, which re compred to their counterprt. b Extended-spectrum bet-lctmse-producing Grm-negtive rods. 5

6 Tble 2 shows the distribution of the use of vrious ntimicrobils. Co-moxicillin-clvulnic cid ws most commonly used, second were the fluoroquinolones nd the third were the third- nd fourth-genertion cephlosporins. Determinnts of inpproprite use of ntimicrobil therpy In the univrite nlysis, use of fluoroquinolones ws significntly ssocited with more frequent inpproprite use of AMT (reltive risk: 1.4). The use of bet-lctmse-sensitive penicillins ws significntly ssocited with more frequent pproprite use of AMT (reltive risk:.3) (Figure 3). Considering the use of AMT in the different medicl specilties, urology (p=.2) proved to be significntly ssocited with more frequent inpproprite use (Figure 4). None of the specilties ws significntly ssocited with more frequent pproprite use. Figure 5 shows tht the presence of suprpubic ctheter ws significntly ssocited with more frequent inpproprite use (reltive risk: 1.9). The following fctors were ssocited with more frequent pproprite use of AMT: hving centrl venous ctheter, peripherl vsculr ctheter or n rteril ctheter, the presence of ESBL (extended-spectrum bet-lctmseproducing bcteri, being dmitted to n intensive cre unit, being in isoltion precutions nd hving nosocomil infection. In multivrite nlyses, tking the effects of ll the bove-mentioned vribles into ccount, we found tht the hospitls themselves were importnt determinnts ssocited with pproprite or inpproprite use of AMT (Tble 3). Furthermore, incresing ge (p=.24), being in n intensive cre unit (p=.2), hving centrl venous ctheter (p=.12), peripherl vsculr ctheter (p=.5) nd nosocomil infection (p=.49) nd use of bet-lctmse-sensitive penicillins (p=.17) were significntly ssocited with pproprite use in multivrite nlyses. The presence of suprpubicl ctheter (p=.17) or the use of fluoroquinolones (p<.1) were significntly ssocited with inpproprite use of AMT. No colinerity ws found between the vribles in the multivrite model. Discussion nd conclusions The men prevlence of AMT in this study ws 29.6% (rnge: %). The most recent study tht cn be used for comprison is from the Europen Surveillnce of Antimicrobil Consumption (ESAC) which found similr prevlence of 3% (rnge: 19 59%) in 2 Europen hospitls in 26 [2]. Other prevlence studies in hospitls in the United Kingdom nd Turkey lso showed similr rtes of ntimicrobil use [9-11]. Although the overll prevlence in our study is comprble to tht of other lrge surveys, there were lrge vritions between the prticipting hospitls. This rnge of pproprite AMT use cn be explined by differences in the ptient popultions nd by differences in prescription policies between hospitls nd between individul prescribers. Of ll ptients on AMT in this study, the use ws considered inpproprite in 372 (16%; rnge: %). The ptients concerned comprised 4.7% of the totl number of ptients, which my seem reltively unimportnt. However, this mens tht nnully pproximtely 1, dys of unjustified AMT re given in hospitl with 2, ptient dys yer. Treting ptients with AMT when such tretment is not indicted is known to be ssocited with higher costs, more side effects nd more ntimicrobil resistnce [12,13]. Our study showed tht the proportion of ptients for whom AMT ws judged to be inpproprite vried between hospitls. AMT use could not be judged for 265 ptients due to insufficient informtion. Deciding on the ppropriteness of AMT use is often not esy. However, the difference between the hospitls is remrkble. The hospitls with the lowest proportion of cses tht could not be judged were hospitls with previous experience with this kind of survey. Possibly these kinds of judgments require more extensive trining. During session tht ws orgnised with the infection control prctitioners nd consultnt microbiologists to discuss the findings, it ws thought tht more trining nd discussion of difficult cses in the study group would probbly result in reduction of the number of cses tht could not be judged. Our study showed tht the prticipting hospitl is determinnt itself nd hd gret influence in the nlyses of determinnts ssocited with inpproprite or pproprite AMT use. We were unble to identify specific chrcteristics of the hospitls tht were responsible for more frequent inpproprite use. Nevertheless, use of fluoroquinolones proved to be significnt risk fctor for inpproprite use of AMT. Fluorouinolones were the second most frequently used ntimicrobils. The ESAC reported tht use of fluoroquinolones incresed most rpidly of ll groups of ntimicrobil gents, with rise of 15% or more from 2 to 25 in lmost hlf of ll prticipting countries [14]. At the sme time ntimicrobil resistnce ginst the fluoroquinolones incresed from 5% (in 21) to 14% (in 28) in Escherichi coli nd from 4% (in 25) to 8% (in 28) in Klebsiell pneumonie [15]. This highlights the importnce of undertking trgeted interventions to reduce inpproprite use of fluoroquinolones. Dt from prevlence surveys such s those described here provide support for such ction. Other determinnts ssocited with more pproprite use of AMT were vribles ssocited with the clinicl complexity of the ptients (e.g. being dmitted to n intensive cre unit nd hving centrl venous ctheter). In the Netherlnds, microbiologist or infectious disese physicin is lmost lwys consulted in the ssessment of these complicted cses. However, 6

7 the lrgest group of ptients being treted with ntibiotics comprises reltively uncomplicted cses: most of these re not monitored by the microbiologist or infectious disese physicin. A prevlence survey does include this group of ptients nd delivers informtion on the ppropriteness of use. In our experience, it is this group in which substntil improvement of the qulity of ntibiotic prescription cn be chieved. The fct tht the tretment or lck of tretment of some ptients could not be judged my hve ffected the outcome of our study. However, sensitivity nlysis showed tht this did not ffect the conclusions bout the ppropriteness of AMT use in the prticipting hospitls. We did not collect informtion on wht fctors in the hospitls with higher inpproprite use my hve contributed to this. Further studies re wrrnted, since they my offer clues for further improvement. In this study we identified those ptients who indvertently did not receive AMT (22 ptients,.3% of totl study popultion). There ws no further nlysis of the 22 ptients. In n erlier single-centre study, we found similr frction of such ptients (25 ptients,.6% of totl study popultion); further investigtion showed tht those ptients were not dversely ffected t dischrge [3]. The extent of intrvenous dministrtion of ntimicrobils (59.4%) suggests tht there is room for improvement. Intervention studies performed in the Netherlnds showed tht intrvenous dministrtion cn be reduced reltively esily by trgeted interventions [16,17]. A switch to orl therpy often results in shorter hospitl dmission. In our study, the ppropriteness of the route of dministrtion ws not ssessed In conclusion, we hve demonstrted tht it is possible to collect prevlence dt on use of AMT t ntionl level. Individul hospitl dt cn be very helpful in inititing trgeted interventions to improve AMT use [17]. However, in order to produce more relible results of such surveys, the number of ptients for whom the ppropriteness of AMT use could not be judged hs to be reduced. Therefore trining of infection control prctitioners nd consultnt microbiologists hs to be intensified nd medicl stff need to be trined in how to record informtion, in order to get n unmbiguous ssessment of use of AMT. 4. Horn TC, Gynes RP, Mrtone WJ, Jrvis WR, Emori TG. CDC definitions of nosocomil surgicl site infections, 1992: modifiction of CDC definitions of surgicl wound infections. Infect Control Hosp Epidemiol. 1992;13(1): Horn TC, Gynes RP. Surveillnce of nosocomil infections. In: Myhll CG, editor. Hospitl epidemiology nd infection control. Phildelphi: Lippincott Willims & Wilkins: 24. p Gyssens IC, vn den Broek PJ, Kullberg B, Hekster YA, vn der Meer JWM. Optimizing ntimicrobil therpy. A method for ntimicrobil drug use evlution. J Antimicrob Chemother. 1992;3: Vn Ksteren MEE, Wijnnds WJ, Stobberingh EE, Jnknegt R, Verbrugh HA, vn der Meer JW. [Optimizing ntibiotics use policy in the Netherlnds. I. The Netherlnds Antibiotics Policy Foundtion (SWAB)]. Ned Tijdschr Geneeskd. 1998;142: Dutch Working Prty on Antibiotic Policy (SWAB). Nethmp 28 Consumption of ntimicrobil gents nd ntimicrobil resistnce mong mediclly importnt bcteri in the Netherlnds. Amsterdm: SWAB; 28. Avilble from: 2EC125744F2ACAA5/$FILE/NethMp_28.pdf 9. Seton RA, Nthwni D, Burtyon P, McLughlin C, McKenzie AR, Dunds S, Ziglm H, Gourly Y, Berd K, Dougls E. Point prevlence survey of ntibiotic use in Scottish hospitls utilising the Glsgow Antimicrobil Audit Tool (GAAT). Int J Antimicrob Agents. 27;29(6): Usluer G, Ozgunes I, Leblebicioglu H, Turkish Antibiotic Utiliztion Study Group. A multicenter point-prevlence study: ntimicrobil prescription frequencies in hospitlized ptients in Turkey. Ann Clin Microbiol Antimicrob. 25;4: Ang L, Lskr R, Gry JW. A point prevlence study of infection nd ntimicrobil use t UK children s hospitl. J Hosp Infect. 28;68(4): Howrd D, Cordell R, McGown JE Jr, Pckrd RM, Scott RD 2nd, Solomon SL; Workshop Group. Mesuring the economic costs of ntimicrobil resistnce in hospitl settings: summry of the Centers for Disese Control nd Prevention-Emory Workshop. Clin Infect Dis. 21;33(9): Cosgrove SE. The reltionship between ntimicrobil resistnce nd ptient outcomes: mortlity, length of hospitl sty, nd helth cre costs. Clin Infect Dis. 26;42 Suppl 2:S vn de Snde-Bruinsm N, Grundmnn H, Verloo D, Tiemersm E, Monen J, Goossens H, et l. Antimicrobil drug use nd resistnce in Europe. Emerg Infect Dis. 28;14(11): Europen Antimicrobil Resistnce Surveillnce System (EARSS). EARSS nnul report 28. Bilthoven: Netherlnds Centre for Infectious Disese Control, Ntionl Institute for Public Helth nd the Environment (Rijksinstituut voor Volksgezondheid en Milieu, RIVM); 29. Avilble from: Net/Documents/28_EARSS_Annul_Report.pdf 16. vn Hees BC, de Ruiter E, Wiltink EH, de Jongh BM, Tersmette M. Optimizing use of ciprofloxcin: prospective intervention study. J Antimicrob Chemother. 28;61(1): Willemsen I, Cooper B, vn Buitenen C, Winters M, Andriesse G, Kluytmns J. Improving quinolone use in hospitls by using bundle of interventions in n interrupted time series nlysis. Antimicrob Agents Chemother. 21;54(9): References 1. Cooke DM, Slter AJ, Phillips I. The impct of ntibiotic policy on prescribing in London teching hospitl. A onedy prevlence survey s n indictor of ntibiotic use. J Antimicrob Chemother. 1983;11(5): Ansri F, Erntell M, Goossens H, Dvey P. The Europen surveillnce of ntimicrobil consumption (ESAC) pointprevlence survey of ntibcteril use in 2 Europen hospitls in 26. Clin Infect Dis. 29;49(1): Willemsen I, Groenhuijzen A, Bogers D, Stuurmn A, vn Keulen P, Kluytmns J. Appropriteness of ntimicrobil therpy mesured by repeted prevlence surveys. Antimicrob Agents Chemother. 27;51(3):