infection control programmes to control antimicrobial resistance

WHO/CDS/CSR/DRS/2001.7ORIGINAL: ENGLISHDISTRIBUTION: GENERAL

World Health Organization

Infection controlprogrammesto controlantimicrobialresistanceLindsay E. Nicolle

Copies can be obtained from the CDS Information Resource CentreWorld Health Organization, 1211 Geneva 27, Switzerland

fax: +41 22 791 42 85 email: [email protected]

WHO/CDS/CSR/DRS/2001.7ORIGINAL: ENGLISHDISTRIBUTION: GENERAL

Infection controlprogrammesto containantimicrobialresistanceLindsay E. NicolleDepartment of Internal MedicineUniversity of Manitoba, Canada

World Health Organization

A BAC

KGRO

UND D

OCUM

ENT F

OR

THE W

HO GL

OBAL

STRA

TEGY

FOR C

ONTA

INME

NT OF

ANTIM

ICROB

IAL

RESIS

TANC

E

Acknowledgement

The World Health Organization wishes to acknowledge the support of the United States Agency for Inter-national Development (USAID) in the production of this document.

World Health Organization 2001

This document is not a formal publication of the World Health Organization (WHO), and all rights are reserved by the Organiza-tion. The document may, however, be freely reviewed, abstracted, reproduced and translated, in part or in whole, but not for sale orfor use in conjunction with commercial purposes.

The views expressed in documents by named authors are solely the responsibility of those authors.

The designations employed and the presentation of the material in this document, including tables and maps, do not imply theexpression of any opinion whatsoever on the part of the secretariat of the World Health Organization concerning the legal status ofany country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. Dotted lines onmaps represent approximate border lines for which there may not yet be full agreement.

The mention of specific companies or of certain manufacturers products does not imply that they are endorsed or recommended byWHO in preference to others of a similar nature that are not mentioned. Errors and omissions excepted, the names of proprietaryproducts are distinguished by initial capital letters.

Designed by minimum graphicsPrinted in Switzerland

Contents

iii

WHO/CDS/CSR/DRS/2001.4 DRUG RESISTANC IN MALARIA

Executive Summary 1

1. Antimicrobial resistance in health care facilities 3

2. Containing antimicrobial resistance in health care facilities 4

3. Infection control programmes 53.1 Effectiveness of infection control programmes 53.2 Impact on antimicrobial resistance 53.3 Impact on MRSA and VRE 53.4 Summary 6

4. Outbreak management 74.1 Elements of outbreak management 74.2 Literature review 74.3 Limitations of published reports 74.4 Control of outbreaks caused by resistant organisms 84.5 Spontaneous disappearance of resistant strains 8

5. Handwashing/hand hygiene 95.1 Recommendations 95.2 Hand antiseptics 9

6. Isolation and other barrier practices 106.1 Practices 106.2 Enterobacteriaceae 106.3 MRSA 106.4 VRE 116.5 Summary 116.6 Standard barrier precautions 116.7 Multiply-resistant Mycobacterium tuberculosis 12

7. Environmental cleaning and sterilization 13

8. Antimicrobial interventions 148.1 Promotion of antimicrobial resistance 148.2 Systemic prophylaxis 148.3 Topical prophylaxis 148.4 Mupirocin for S. aureus 148.5 Antimicrobial-impregnated medical devices 15

9. Resources for infection control 169.1 Cost-effectiveness 169.2 Prioritization of infection control resources 16

10. Resource-poor countries 1710.1 Infection control 1710.2 Outbreaks with resistant organisms 1710.3 Endemic antimicrobial resistance 17

iv

11. Research agenda 19

12. Discussion 20

13. Conclusions 22

14. Bibliography 23

Annex. Control of outbreaks of nosocomial antimicrobial resistant organisms 29

INFECTION CONTROL PROGRAMMES TO CONTAIN ANTIMICROBIAL RESISTANCE WHO/CDS/CSR/DRS/2001.7

1WHO/CDS/CSR/DRS/2001.7 INFECTION CONTROL PROGRAMMES TO CONTAIN ANTIMICROBIAL RESISTANCE

Executive Summary

cially in high-risk units, decreases transmissionof some resistant organisms. Barrier practicescannot, however, by themselves, fully preventnor ultimately contain the progression of resist-ance.

5. The current worldwide epidemics of MRSA andVRE have progressed despite intense nationaland local infection control measures to identifyand contain the spread of these organisms.

6. The appropriate use of prophylactic anti-microbials prevents some nosocomial infections.However, any prophylactic antimicrobial use,especially in high-risk patients, contributes toantimicrobial pressure and to the emergence ofantimicrobial-resistant organisms. In this case,appropriate infection control activity mayactually promote antimicrobial resistance.

7. In some facilities, reallocation of infection con-trol resources to comply with recommendationsfor control of colonization of MRSA and VREhas compromised other infection control func-tions, potentially increasing the frequency ofnosocomial infection.

8. Overall, infection control programmes havesome efficacy in containing antimicrobial resist-ance, particularly when an outbreak with aresistant strain is identified. Other infection con-trol practices decrease transmission of bothresistant and susceptible organisms amongpatients, and intensification of these practicesto limit transmission of resistant organisms likelyhas some short-term efficacy in decreasing en-demic resistance. Ultimately, however, limitingantimicrobial resistance rests primarily withantimicrobial use rather than infection control.

9. If the infection control responsibility is expandedto incorporate control of transmission and colo-nization with resistant organisms, rather thandecreasing infections, additional resources tosupport the increased activity must be allocated.

1. Health care facilities, particularly acute carefacilities, are important sites for the developmentof antimicrobial resistance. The intensity of an-timicrobial use together with populations highlysusceptible to infection create an environmentwhich facilitates both the emergence and trans-mission of resistant organisms.

2. Optimal infection control programmes in healthcare facilities decrease the frequency of nosoco-mial infection. Such programmes have beenidentified as important components of anycomprehensive strategy for the control of anti-microbial resistance, primarily through limitingtransmission of resistant organisms amongpatients. The successful containment of antimi-crobial resistance in acute care facilities, how-ever, also requires adequate clinical microbiologylaboratory support and a strong antimicrobialuse programme.

3. Infection control interventions are effective incontrolling some outbreaks of colonization andinfection with antimicrobial-resistant organismsin health care facilities. In fact, antimicrobialresistance frequently is a phenotypic marker forthe outbreak organism which facilitates identi-fication and early initiation of interventions tocontrol the outbreak. The application of infec-tion control measures, however, is not uniformlysuccessful in limiting outbreaks with resistantorganisms.

4. Barrier practices including patient isolation andthe use of gloves, gowns, or masks, are widelyrecommended for the control of endemic anti-microbial resistance. The effectiveness of thesepractices is controversial, and studies evaluatingtheir efficacy are contradictory. The extent towhich these practices decrease resistance likelyvaries with other determinants, such as preva-lence of antimicrobial-resistant organisms in thefacility, characteristics of the patient population,staffing ratio and expertise, and patient volumes.New implementation of these practices, espe-


2

10. Further systematic evaluation of infection con-trol interventions in containing endemic anti-microbial resistance in acute care facilities, aswell as other health care settings, is needed.This should include studies of the naturalhistory and impacts of antimicrobial-resistantorganisms in facilities as well as effectiveness,feasibility, and costs of specific infectioncontrol interventions.


1. Antimicrobial resistance inhealth care facilities

munity infections in Brazil (9). The flow of resist-ant organisms is not, however, unidirectional. Re-sistant strains have emerged in the community, suchas penicillin-resistant Streptococcus pneumoniae (10)and resistant Salmonella spp (11,12), and been in-troduced into acute care facilities with resultingnosocomial infections.

Antimicrobial resistance is a predictable outcomeof antimicrobial use. The rapidity with which re-sistance emerges and its extent are proportional tothe intensity of antimicrobial use (1). Resistancefirst emerges in populations with a high frequencyof infection, due to either underlying patientstatus or interventions compromising host defences,resulting in a high rate of antimicrobial use. Wherepatients at risk are in close proximity, the transmis-sion of organisms between patients will be facili-tated, and the opportunity for a single strain todisseminate widely is enhanced. All these featuresare present in health care facilities, particularly acutecare facilities and areas such as intensive care units(2). Thus, health care facilities, particularly thosewhich are large and care for the most complexpatients, are a focal point in the emergence of anti-microbial resistance.

Resistant organisms have repeatedly been firstdescribed in high-risk patients of acute care facili-ties (Table I). Some organisms, such as resistantfungi in neutropenic patients (3), or resistant Pseu-domonas aeruginosa in intensive care unit patients(4), are a risk only for selected hospitalized patients.These organisms contribute to morbidity and costin restricted patient groups, with little impact inless immunocompromised hospitalized patients orin the community. In other instances, such as ex-tended spectrum -lactamase producing Enterobac-teriaceae (5) or vancomycin-resistant enterococci(6), strains spread among patients and may causeinfection in patients with less acuity in the acutecare hospital, in chronic care or long-term carefacilities, or in community health care. An organ-ism of particular concern is Staphylococcus aureus,an important human pathogen in both the hospi-tal and the community. Resistance, first to penicil-lin in the 1950s (7), and now methicillin (8), hasemerged and became widespread in hospitals andsubsequently spread to the community. In anotherexample, hospitals were found to be the source ofstrains of resistant Salmonella which caused com-

TABLE I. ANTIMICROBIAL-RESISTANT ORGANISMS OFCONCERN IN HEALTH CARE FACILITIES

Organism Resistant to

Staphylococcus aureus methicillinvancomycin

Staphylococcus epidermidis vancomycin

Enterococci aminoglycoside (high level)ampicillinvancomycin

Streptococcus pneumoniae penicillin

Enterobacteriaceae aminoglycosidesthird-generation cephalosporinsmonobactamsceftazidime

Pseudomonas aeruginosa fluoroquinolonesextended-spectrum penicillinsfluoroquinolonesaminoglycosidesceftazidimecarbapenems

Acinetobacter spp aminoglycosidesceftazidimecarbapenems

Mycobacterium tuberculosis isoniazidrifampinstreptomycinethambutolpyrazinamide

Candida spp amphotercin bazoles

Herpes simplex acycolvir

Cytomegalovirus foscarnet


4

2. Containing antimicrobial resistance inhealth care facilities

prevention. This will include discussion of specificcomponents of infection control activity, and willfocus primarily on acute care facilities. Importantquestions requiring further investigation to clarifyand quantify the impact of effective infection con-trol strategies in containing antimicrobial resistancewill also be identified.

In health care facilities, the infection controlprogramme is one of three essential overlappingprogrammes with activities which address the prob-lem of antimicrobial resistance (1,2). The clinicalmicrobiology laboratory provides isolation and sus-ceptibility testing of organisms from clinical speci-mens, and surveillance data to summarize theprevalence of antimicrobial resistance in a facility.The third essential activity is the antimicrobial useprogramme which makes recommendations forantimicrobials for the hospital formulary consid-ering the impacts of antimicrobial resistance bothfor the individual infected patient as well as theenvironment, and monitors antimicrobial use andappropriateness. These three programmes mustfunction cooperatively to support the goal of con-tainment of antimicrobial resistance, but also haveresponsibility for service delivery beyond antimi-crobial resistance. While acknowledging that inte-gration of infection control, the laboratory, andantimicrobial use programmes are essential toaddress antimicrobial resistance, this discussion willfocus specifically on the infection control pro-gramme. Activities of the other two programmeswill only be addressed where they are directly rel-evant to infection control functions.

Multifaceted proposals to address the problem ofantimicrobial resistance have uniformly stated thatoptimal infection control programmes in healthcare facilities are an essential component. TheWHO Global Strategy for Containment of Anti-microbial Resistance recommends that hospitalmanagement establish infection control pro-grammes with responsibility for effective manage-ment of antimicrobial resistance in hospitals andensure that all hospitals have access to such a pro-gramme(1). Recommendations for strengtheninginfection control programmes and activity are alsoincluded in the United States Public HealthAction Plan to Combat Antimicrobial Resistance(13), in Controlling Antimicrobial Resistance: AnIntegrated Action Plan for Canadians (14), and theEuropean Commission Opinion of the ScientificSteering Committee on Antimicrobial Resistance(15). Despite this consensus on the importance ofinfection control activities in health care facilities,there has been limited evaluation of the evidenceto support the effectiveness of infection control incontaining antimicrobial resistance and preventingadverse clinical outcomes attributable to antimi-crobial resistance.

This discussion will review the evidence support-ing the efficacy of infection control programmesand activities in containing the prevalence of anti-microbial resistance and limiting infections withresistant organisms acquired in health care facili-ties. The question to be addressed is what, if any, isthe additional benefit of an infection control pro-gramme in the containment of antibiotic resistancebeyond the benefit inherent in overall infection


3. Infection control programmes

fections least likely to be preventable through in-fection control activity are those which occur inthe most highly immunocompromised patients,where patient vulnerability overwhelms preventiveefforts. Such patients would include, for example,those receiving allogeneic bone marrow transplantsor with over 50% body surface area burns. Theseare also the patients most likely to require prophy-lactic or therapeutic antimicrobials. Thus, infec-tion control programmes may be less effective indecreasing infections for these patients at greatestrisk for antimicrobial-resistant organisms. Withintensification of infection control activity, theproportion of antimicrobial-resistant infectionsmay, then, increase.

The goal of infection control programmes is todecrease the incidence of infections in patients andstaff (16). Considerations relevant to antimicrobial-resistant organisms, however, extend beyond thisgoal by including patient colonization with resist-ant organisms as well as infection. The outcome ofcolonization is seen to be an important measure ofthe total burden of antimicrobial resistance in apopulation, as well as predicting the future burdenof infection with these organisms. The effective-ness of an optimal infection control programme indecreasing the total burden of colonization ofpatients with resistant organisms has also not beenadequately evaluated.

3.3 Impact on MRSA and VRE

While the overall impact on antimicrobial-resist-ant organisms in a facility is not known, some ob-servations relevant to specific organisms have beenreported. For methicillin-resistant Staphylococcusaureus (MRSA), intense and comprehensive infec-tion control programmes, including screening ofstaff and patients, strict isolation or cohorting, anddecolonization therapy of patients and staff wereinitially recommended for control in some coun-tries. These interventions are both expensive andburdensome. In both the United States (18, 19) andthe United Kingdom (20, 21), and elsewhere (22)

3.1 Effectiveness of infectioncontrol programmes

The essential features and appropriate resources foran optimal infection control programme have beenidentified (16) (Table II). The SENIC study showedthat effective infection control programmes whichincluded surveillance, control activities, and appro-priate personnel and leadership decreased the fre-quency of endemic nosocomial infections by 30%to 50% (17). The specific role of such programmesin containing antimicrobial resistance has not beenreported. An assumption would be that such a pro-gramme would decrease antimicrobial-resistant in-fections proportional to the overall decrease innosocomial infections. The absolute number ofinfections with resistant organisms would, then,decrease but the proportionate amount of antimi-crobial-resistant infections would remain stable.

TABLE II. ACTIVITIES OF AN OPTIMAL INFECTIONCONTROL PROGRAMME

surveillance of nosocomial infections outbreak investigation and control policy development, review and compliance monitoring

isolation practiceshand hygienesterilization/disinfection of equipment and supplieshousekeepinglaundryfood

employee health relevant to infections education of staff, patients, visitors

3.2 Impact on antimicrobial resistance

The impact of infection control on antimicrobial-resistant infections, however, is not necessarilystraightforward. If the overall frequency of infec-tion in a facility is decreased, then antimicrobialuse may also be decreased. This would mean lessantimicrobial pressure in the institutional environ-ment, leading to a decrease in the prevalence ofantimicrobial-resistant organisms and proportion-ately fewer infections with resistant organisms withintensified infection control activity. However, in-


6

recommendations have subsequently been adjustedto promote less intense infection control measures.This followed from an apparent failure of the ini-tial recommendations to limit the increase inendemic MRSA (22, 23), and repeated reportswhere decreased intensity of infection control in-terventions was not followed by increased rates ofnosocomial transmission or infection with MRSAin the facility (2428).

A relevant report with a unique perspective isthat by Meers and Leong (29). They describe anexperience with MRSA in a newly opened teach-ing hospital where no infection control programmeto control MRSA was ever implemented. Theorganism was first isolated in patient specimensshortly after the hospital opened, and the preva-lence of MRSA in S. aureus isolates increased overthe next year. For the subsequent three years theprevalence of MRSA remained stable, and was re-sponsible for about 50% of nosocomial S. aureusinfections. The authors argue the nosocomialS. aureus infection rate was similar to that reportedfrom other facilities, and the only negative impactof MRSA in the facility was the increased cost ofantimicrobials to treat infected patients. They sug-gest their experience did not support intense andcostly infection control interventions to controlendemic MRSA in an acute care facility.

Similar to the MRSA experience, the recom-mendation for and widespread implementation of

comprehensive guidelines to control endemic van-comycin-resistant enterococci (VRE), includingintense infection control interventions (30), havenot prevented progression of the American VREepidemic nationally (31), or in individual institu-tions (3236). Whether these intense infectioncontrol programmes delayed the progression ofendemicity with these organisms cannot be assessed.

3.4 Summary

Overall, while optimal infection control pro-grammes should be expected to decrease theoccurrence of infections and, possibly, colonizationwith resistant organisms, the effect of such pro-grammes and the duration of any effect are notknown. In fact, the efficacy of an infection controlprogramme in limiting endemic colonization orinfection with resistant organisms has not beenunequivocally demonstrated. The available evi-dence, primarily the experience with MRSA andVRE, suggests infection control programmes havelimited efficacy in preventing endemic infection orcolonization with antimicrobial-resistant organismsbecoming established in a facility. If programmesdo decrease the prevalence of resistance, the dura-tion of this effect, especially in the face of an in-creasing prevalence of resistant organisms in otherfacilities or in the community is also unclear (37,38).


4. Outbreak management

publication is encouraged, or if there is somethingunique about the outbreak. This might include anoutbreak with a new strain, such as an antimicro-bial-resistant organism, identification of a vectornot previously described, or the use of a new epi-demiological typing method. Outbreaks success-fully controlled are also more likely to be reportedthan those for which control measures were noteffective, or were only partially effective.

Another limitation in assessing the effectivenessof outbreak investigation and control measures isthat reports are descriptive rather than compara-tive. Interventions are applied universally withinthe outbreak population and there is no controlgroup, randomization, or blinded assessment. Thisintroduces bias in evaluating the effectiveness ofthe interventions. There are obvious reasons forthese limitations, including ethical considerationsand the need for immediate and complete contain-ment, but the potential bias must be recognized.With no control population, apparent containmentof the outbreak may simply reflect the natural his-tory of the outbreak strain in the population (39).In addition, the impact of any single interventioncan seldom be ascertained as multiple, usuallysimultaneous, interventions are invariably initiated.Many reports also provide a limited duration offollow-up, and where control or eradication isreported, the durability of the effect is not known.

One example illustrating the difficulty in evalu-ating the contribution of outbreak control in con-tainment is the report of Bernards et al. (40). Thisdescribes three patients transferred to three differ-ent Dutch hospitals who were colonized or infectedwith both MRSA and antimicrobial-resistantAcinetobacter baumannii. All patients were imme-diately placed in strict isolation, following Dutchinfection control guidelines. There was no trans-mission of MRSA in any facility. Two of the threefacilities experienced outbreaks with the importedAcinetobacter strainsin one facility the outbreakresolved spontaneously with no investigation orcontrol measures, and in the other facility the strainwas eradicated after intense outbreak investigation

4.1 Elements of outbreak management

Effective outbreak management is an essentialfunction of an infection control programme (16).Specific activities include identification, coordina-tion of response, case-finding, description of theextent and temporal course, input into case man-agement, analysis of exposure and patient variablesto identify risks, and introduction and evaluationof specific control measures to terminate the out-break. Mobilization of resources beyond infectioncontrol is usually necessary, and the clinical micro-biology laboratory and antimicrobial use pro-gramme are two key components. In outbreakcontrol the activities of these two programmes arean integral part of the infection control response.

4.2 Literature review

The largest body of evidence which supports a spe-cific role for infection control in containing anti-microbial resistance is in outbreak control. Manyreports which describe hospital outbreaks of anti-microbial-resistant organisms are summarized in theannex to this report. These were identified througha Medline search using the key words outbreak andresistant, supplemented by complete review of in-fection control journalsthe Journal of HospitalInfection, Infection Control and Hospital Epidemiol-ogy, and the American Journal of Infection Control.Only reports in English and those published after1970 have been included. The focus is largely acutecare facilities, and only includes those reports wheresufficient information was provided to assess theimpact of control measures. The interventionsinstituted, as far as could be ascertained frominformation in the published reports, are also sum-marized.

4.3 Limitations of published reports

In evaluating this body of information, it must beappreciated there is likely substantial publicationbias. Outbreaks are more likely to be reported ifthey have occurred in an academic centre where


8

and control. The authors report the one facilitywhich did not experience an outbreak was the onlyone with isolation facilities appropriate to preventairborne spread, and suggests this explains theabsence of transmission in that facility. However,measures for the control of airborne transmissionwere not instituted for at least one of the two out-breaks, and this resolved. From this report, it ispossible to conclude that isolation precautions areeffective, or that they are ineffective, that outbreakcontrol is necessary and effective, or is not neces-sary, and that respiratory isolation is essential, or itis not necessary!

4.4 Control of outbreaks caused byresistant organisms

Even given these limitations, a summary assessmentof the reports in the annex would be that outbreakmanagement is effective in limiting the spread ofand preventing infections caused by antimicrobial-resistant organisms in the acute care hospital. Manyof these outbreaks were, in fact, identified becausethe outbreak strain had a unique antimicrobial sus-ceptibility. They may not have been recognized orcontrolled as promptly if the organisms were notphenotypically unique because of the resistancemarker. The effectiveness of interventions is mostconvincing for those outbreaks where a unique en-vironmental reservoir or staff member who was acarrier were identified and there was abrupt andcomplete termination of transmission with eradi-cation of the reservoir. The many reports describ-ing eradication of a new MRSA strain from aninstitution without endemic MRSA are also con-vincing evidence for the effectiveness of outbreakmanagement. Finally, repeated reports of termina-tion of multiply-resistant tuberculosis outbreaksamong hospitalized HIV patients and staff follow-ing the introduction of interventions to controlairborne transmission document that appropriateinfection control interventions are effective for con-taining outbreaks with this organism.

Despite this assessment, a theme emerging frommany reports is that initial interventions were not

effective. These initial interventions usuallyincluded intensification of hand hygiene, barrierprecautions or isolation, and staff education, all ofwhich would be reinforcing usual infection con-trol practice. Further control measures introducedafter failure of these initial interventions usuallyincluded extraordinary measures, such as cohortingof patients and staff, ward closure, treatment ofcolonization of patients or staff, or antimicrobialrestriction. These practices would not normally besustained once the outbreak is contained. In manyreports, interventions decreased the frequency ofinfection or colonization, but could not eradicatethe strain. In some reports, despite intense, multi-faceted control interventions, the outbreak was notcontained, and the outbreak strain became endemicin the institution.

4.5 Spontaneous disappearance ofresistant strains

Several reports also describe the emergence, dissemi-nation, and subsequent spontaneous decrease ordisappearance of an outbreak strain in the absenceof control efforts. The disappearance of certainMRSA phage types from Europe in the 1970s and1980s may be one example (41, 42). Spontaneousdecline or disappearance has also been reported inAmerican facilities for gentamicin-resistant Entero-bacteriaceae (43, 44) and Pseudomonas aeruginosa(45), and even aminoglycoside-resistant transmis-sible elements (46). In some reports, an apparentenvironmental reservoir was eliminated withoutspecific control interventions (4749). Factorswhich explain the apparently spontaneous declineor disappearance of resistant strains are not known.An important variable, of course, is antimicrobialpractice. The natural history of dissemination,persistence, and replacement of antimicrobial-resistant strains requires further study. These ob-servations of apparently spontaneous resolution ordecline in resistant organisms causing outbreaks,however, suggests the impact of control measuresmay be overestimated in some reports.


5. Handwashing/hand hygiene

5.2 Hand antiseptics

Antimicrobial-resistant organisms do not have ahigher frequency of resistance to agents used forhand disinfection when compared to susceptiblestrains of the same species (55, 56). However,bacteria with intrinsic resistance to some anti-microbials, such as Providencia stuartii orPseudomonas aeruginosa, may also have intrinsicresistance to some antiseptics. Thus, where handantiseptics are used for decontamination, a differ-ential effect for some resistant organisms could beobserved. Onesko and Wienke (57) reported, inan uncontrolled study, that introduction of aniodine lotion soap to replace natural soap in twohigh-prevalence MRSA wards, one an intensive careunit, led to an 80% decrease in nosocomial MRSA.The effect was non-specific, however, as there wasa general decrease in other organisms, both suscep-tible and resistant. Webster et al. (58) reportederadication of MRSA from a neonatal intensive careunit when a triclosan disinfectant replacedchlorhexidine gluconate 4% for handwashing. Thiswas accompanied by an increase in Pseudomonasaeruginosa infections, an organism with intrinsicresistance to triclosan (59). Pittet et al. (60) haverecently reported that use of an alcohol-based 0.5%chlorhexidine gluconate handrub solution was fol-lowed by increased compliance with hand hygieneand a significant decrease over the subsequent fouryears of all nosocomial infections and MRSA trans-mission rates. Further comparative studies will benecessary to characterize the impact of specific hand-washing practices in containing antimicrobialresistance.

5.1 Recommendations

Many antimicrobial-resistant organisms, includingMRSA and VRE, are primarily transmitted betweenpatients on the hands of staff. Appropriate handdecontamination should be effective in decreasingthe transmission of these organisms, as well asstrains that are not antimicrobial-resistant. Therewould not, however, be any unique benefit forresistant organisms. Good practice will limit trans-mission of all organisms carried on the hands ofstaff, and some of these will be antimicrobial-resistant.

Hand decontamination for staff of health carefacilities participating in direct patient care may bewith soap and water, or an antiseptic or anti-microbial solution. Handwashing is effective in pre-venting nosocomial infections (50), as first dem-onstrated by the classic studies of Semmelweiss (51).The evidence to support specific practices in handdecontamination, however, including frequency,products, and methods is limited as reports are gen-erally non-comparative, observational, or comprisedof in vitro studies. The recently published evidence-based Guidelines for Preventing Hospital-acquiredInfections from the United Kingdom documentsthis when seven recommendations for hand hygieneare made, all with a level of evidence of category3expert opinion with limited scientific evidence(52). Despite this limited evidence, guidelines uni-formly recommend an antiseptic handwashingagent be used when caring for patients infected orcolonized with antimicrobial-resistant organisms(5254).


10

6. Isolation and other barrier practices

care unit with intensified handwashing, single -useequipment, and glove use.

Alford and Hall (43) report a 15-year experi-ence with gentamicin-resistant Enterobacteriaceaein a Veterans hospital in the United States.Consecutive outbreaks and endemic infection withSerratia marcesens, Klebsiella pneumoniae, andEnterobacter cloacae emerged following acquisitionof gentamicin resistance by these organisms. Pseu-domonas aeruginosa resistant to gentamicin alsoprogressively increased in prevalence. All interven-tions, including recommended barrier precautions,failed to limit the emergence and nosocomial trans-mission of these organisms.

6.3 MRSA

Thompson et al. (67) reported the introduction ofscreening and barrier precautions led to a decreasein MRSA prevalence and incidence in a UnitedStates hospital in the subsequent 12 months. In apaediatric intensive care unit, Casseron-Zerbib etal. (68) reported a 90% decrease in MRSA carriagefollowing the introduction of an MRSA contain-ment programme which included increased screen-ing of patients and intensified handwashing,isolation, and other barrier methods. There wasa significant decrease in the overall nosocomial in-fection rate entirely attributable to a decrease inMRSA infection, although there was no decline inthe nosocomial infection rate in transplant patientsin the unit. In a Swiss hospital with endemic MRSA,introduction of infection control measures includ-ing barrier precautions was followed by a 50%decline in bacteraemia and MRSA isolation over asubsequent five-year period (69). Schmitz et al. (70)also reported the new implementation of barrierpractices and intensified environmental cleaning inan intensive care unit with a high rate of endemicMRSA was followed by a decline in transmissionof this organism. There are also, however, manyreports where decreasing the intensity of barrierprecautions for patients infected or colonized withMRSA was not followed by increased rates of

6.1 Practices

The use of physical barriers and spatial separationin managing patients with an increased likelihoodof transmitting infectious agents to other patientsor staff members is a key infection control func-tion. These practices have been variously designatedisolation, infection control precautions (61), bodysubstance isolation (62), barrier precautions, stand-ard precautions (63), or routine precautions (64).The interventions usually include identification ofpatients and patient care activities at risk for trans-mission of organisms, geographical separation withisolation or cohorting, use of gloves, gowns andother protective equipment by staff to prevent con-tamination, and ensuring compliance with thesepractices by staff, patients, and visitors. These pre-cautions have been recommended for several dec-ades for limiting transmission of resistant organismswithin acute care facilities (61). Their effectivenessis not well measured, however, and remains con-troversial in the endemic, rather than the outbreaksituation. Available studies usually lack concurrentcontrols and include multiple simultaneous inter-ventions so the specific role of barrier precautionsis seldom defined.

6.2 Enterobacteriaceae

Lucet et al. (65) reported a four-year prospectiveobservational study of nosocomial acquisition ofextended spectrum -lactamase (ESBL)-producingEnterobacteriaceae in a Paris hospital following theintroduction of patient screening and barrier pre-cautions. In the first year, the incidence of nosoco-mial acquisition of these organisms did not decrease.After refinement and re-enforcement of the use ofbarrier precautions, there was a subsequent decreasefrom 0.56 to 0.06 cases/100 admissions over thesubsequent three years. Concurrent decreases in theincidence of MRSA and Acinetobacter baumanniinosocomial transmission were also observed.Similarily, Soulier et al. (66) reported a 40%decrease in ESBL-producing Enterobacteriaceaecolonization in a gastrointestinal surgical intensive

11

WHO/CDS/CSR/DRS/2001.7 INFECTION CONTROL PROGRAMMES TO CONTAIN ANTIMICROBIAL RESISTANCE

nosocomial transmission or infection (2228). Inaddition, Goetz and Muder (39) report a contin-ued increase in MRSA in their institution despite,initially, strict isolation and, subsequently, bodysubstance isolation for all patients. They also ob-served a periodic variation in the number of MRSApatients which, in the short term, could have beeninterpreted, as effectiveness of the infection con-trol interventions, but was not sustained.

Souweine et al. (71), in a 10-bed intensive careunit in another French facility, examined retrospec-tively the impact of introduction of infection con-trol measures including education, surveillancecultures, antiseptic handwashing, gown and gloves,and mupirocin use for patients with MRSA on colo-nization and infection with antimicrobial-resistantorganisms. The overall rate of colonization orinfection with MRSA, ESBL-producing Klebsiellapneumoniae or multi-resistant Enterobacter aerogenesdecreased from 15% to 6.8%. The decrease forMRSA was significant between the two periodsprior to and following the intervention, but wasnot significant for the other two organisms. How-ever, the length of stay was also one-third lower inthe post-intervention period, at least partially dueto promotion of prompt discharge, so some of theobserved effect may have been explained by theshorter lengths of stay.

6.4 VRE

In another report from an intensive care unit, VREacquisition was similar whether gowns with glovesor gloves alone were used for direct patient care(72). Whether handwashing without gloves wouldhave been as effective as gloves was not tested.Bonten et al. (73) reported that nosocomial VREacquisition in another intensive care unit correlatedwith colonization pressure, but not with compli-ance with infection control precautions ofhandwashing and gloving. Brooks et al. (74) stud-ied three different intervention approaches to con-tain VRE on three different wards. These includedenhanced environmental decontamination on oneunit, intensive continuing re-education on infec-tion control policies and precautions on a secondunit, and replacement of disposable oral and rectalthermometers by tympanic thermometers for alltemperatures on a third unit. Thermometer replace-ment was most effective, and the decrease was sus-tained at 9 months. There was a short-term decreasein VRE transmission with enhanced environmen-tal sanitation, but this was not sustained at

9 months. There was no decrease and, in fact, anincrease at 9 months, for the infection controlinterventions. Montecalvo et al. (75) reported thatenhanced infection control which includedcohorting of patients and staff, gown and glove usefor patients of unknown status, and monitoringcompliance was followed by a decrease in infectionand colonization with VRE on an adult oncologyunit. However, antimicrobial use was also morehighly controlled and significantly decreased dur-ing the period of intensive infection control inter-ventions, so it is unclear to what extent theextraordinary infection control interventions, orantimicrobial use restriction, contributed to theobserved decline.

6.5 Summary

The conflicting observations of the impact of bar-rier practices in these reports may be explained bydifferences in study design, target organisms,patient populations, the specific interventions ini-tiated, compliance, level of endemicity of theresistant organism in the facility, concurrent altera-tions in antimicrobial use, or the normal variationin the natural history of a strain in the populationobserved. The reports which observed a positiveimpact usually instituted concurrent controls in ad-dition to barrier precautions. Thus, current evidenceto document the effectiveness of patient isolationand other barrier precautions in the containmentof resistant organisms in the non-outbreak situa-tion is conflicting, constrained by the limitationsof reported studies, and not compelling. If theseinterventions are effective, the impact is most likelyto be observed in selected high-risk patient groups,such as those in intensive care units.

6.6 Standard barrier precautions

Within the past decade, a practice for routinepatient care which encourages rigorous handhygiene and consistent use of gloves and other per-sonal protective equipment whenever contamina-tion is anticipated in the care of any patient hasbeen promoted and widely implemented (52, 63,64). One rationale for the development of this ap-proach was the inability to consistently identifypatients colonized with resistant strains. If thisstandard of practice is rigorously adhered to, moreintense barrier precautions for patients identifiedas colonized or infected with a resistant organismwhich may be transmitted by contact should not


12

provide additional benefit. This is perhaps con-firmed, for MRSA, by the many reports wheredeintensification of barrier precautions have notbeen associated with an increase in the endemicrate of resistant organisms (2428). Reports whichsuggest barrier precautions are effective in contain-ing endemic bacterial resistance are, generally, fromfacilities which did not follow current recommen-dations for standard practice (65, 67). The role ofspecific additional barrier practices for patients colo-nized or infected with resistant organisms in thecontext of current practice recommendations hasnot been determined.

6.7 Multiply-resistantMycobacterium tuberculosis

Another isolation practice is the use of respiratoryisolation to prevent transmission of organisms bythe airborne route. The continued effectiveness ofprecautions for the prevention of nosocomial trans-mission of multiply-drug-resistant tuberculosisafter initial outbreak control is convincing evidencethat these infection control precautions for airbornetransmission are effective (76, 77). They are, ofcourse, equally effective in preventing transmissionof drug-susceptible tuberculosis. It was the uniquecircumstance of exposure of highly susceptible HIVpatients and resistant tuberculosis strains whichmade it apparent that previous practice was notadequate for nosocomial tuberculosis control.

13


7. Environmental cleaning and sterilization

tration of these organisms was present in the stand-ing water for cut plants. Removal of plants andwater, together with daily dry mopping, was tem-porally followed by decreased isolation of these re-sistant organisms in nosocomial infection inpatients on the ward. As summarized in the annexto this report, identification of an environmentalsource, and disinfection or sterilization interven-tions to control that source have repeatedly beeneffective in control of outbreaks with resistant or-ganisms.

For both MRSA and VRE the physical environ-ment has been proposed to be an important sourcefor acquisition of these resistant organisms bypatients. This is based on repeated observations ofsubstantial contamination of rooms, furniture, andequipment of patients colonized or infected withthese organisms (80, 81). More intense house clean-ing, including stronger disinfectants and increasedfrequency of cleaning have been suggested to con-trol endemic transmission (81). However, theevidence for a significant unique role for the envi-ronment beyond transmission of organisms on thehands of staff, if there is compliance with recom-mended normal cleaning practice, is not convinc-ing (21, 74, 82).

Sterilization and disinfection of patient care equip-ment will be equally effective for antimicrobial-resistant and susceptible organisms. As noted in theannex, many outbreaks with resistant organismshave been attributed to inadequate cleaning or dis-infection of equipment. Once again, however, theisolation of a resistant organism may have facili-tated early identification of the problem.

The role of the hospital environment in acqui-sition of endemic nosocomial infection remainscontroversial. For selected organisms, such as fun-gal infection with hospital construction andLegionella infection with water systems, a directassociation between the environment and infectionis accepted (78). Even without compelling evidencethat the environment is a major factor in acquisi-tion of other nosocomial infections, it is acceptedthat a certain standard of cleanliness and safety isrequired for hospital environmental surfaces, linenhandling, water and food supply, and waste dis-posal (52). These practices should, of course, limittransmission of both antimicrobial-resistant andsusceptible organisms.

There is one report of a decrease in endemicnosocomial infections with aminoglycolide-resistant Gram-negative organisms with environ-mental interventions alone (79). A high concen-


14

8. Antimicrobial Interventions

tory infections, improved survival has not beenconvincingly proven, and this strategy remainscontroversial (9294). A consistent theme, how-ever, is the emergence of organisms resistant toantimicrobials used for the prophylactic regimen,contributing to a high prevalence of resistantorganisms in the intensive care unit (9598).

8.3 Topical prophylaxis

The use of topical prophylactic antimicrobials inacute care facilities has also promoted widespreadantimicrobial resistance. Topical gentamicin usedfor the prophylaxis of burn wound infection in the1970s and 1980s was followed by the widespreademergence of aminoglycoside- resistant Gram-negative organisms which caused large andsustained outbreaks in many burn units (44, 45,99). In another example, the topical use of an anti-biotic ointment at the central line insertion sitedecreases the risk of line infection, but increasesthe frequency of infection with Candida spp, a moreresistant pathogen (100).

8.4 Mupirocin for S. aureus

An evolving problem, to a large extent directlyattributable to infection control intervention, is theemergence of mupirocin resistance in Staphylococ-cus aureus (101). Initial reports of efficacy of topi-cal mupirocin for eradication of nasal carriage ofMRSA (102) led to recommendations for and wide-spread use of topical mupirocin for decolonizationof patients and staff in controlling both outbreakand endemic MRSA. In some units, mupirocin wasused for all patients irrespective of whether theyhad documented MRSA colonization (103, 104).This enthusiastic application of widespreaddecolonization therapy has been followed bydevelopment of a high prevalence of mupirocin-resistant MRSA in reports from different parts ofthe world (105108). In one Canadian teachinghospital, mupirocin resistance among MRSAincreased from 2.7% to 65% over a three-year

8.1 Promotion of antimicrobialresistance

The use of prophylactic antimicrobials to preventinfection is an important infection control inter-vention. However, widespread use of any antimi-crobial will ultimately result in emergence oforganisms resistant to that agent. Infection controlactivity, then, may promote antimicrobial resist-ance. There is a trade-off between the potentialdecreased requirement for antimicrobials becauseinfections have been prevented, and increasedresistance because of antibiotic pressure fromprophylactic use. Where the balance liesbenefi-cial or detrimental, will vary over time and withthe perspectivethat of the individual patient orof the wider community.

8.2 Systemic prophylaxis

One of the most widely supported uses of prophy-laxis is in surgery. For selected surgical procedures,preoperative prophylaxis will decrease post-operative infections (83, 84). However, even withappropriate surgical prophylaxis, some infectionswill occur post-operatively, and these infections aremore likely to be caused by organisms resistant tothe antimicrobial used for prophylaxis (8588). Thenormal host flora is also altered to a higher preva-lence of resistant strains following surgical prophy-laxis (88, 89).

In another example, infections which followsystemic antimicrobial prophylaxis for bacterial,fungal, or viral infections in patients with prolongedchemotherapy-induced neutropenia are with organ-isms resistant to the prophylactic agent. The serialemergence of bacteria resistant to antimicrobialsused for prophylaxis has been the impetus for acontinuing evolution of antibacterial and antifun-gal prophylaxis in the neutropenic population (90).

A third example is selective gut decontamina-tion for intensive care unit patients, where topicaland systemic antimicrobials are given to preventnosocomial intensive care unit infection (91). Whilesome studies report a benefit in decreasing respira-

15


period when used as an adjunct to infection con-trol measures for a continuing MRSA outbreak(107). The increasing use of topical nasal mupirocinfor prophylaxis of S. aureus infection in high-riskpopulations, particularly dialysis patients (109, 110)and patients undergoing clean surgical operations(111), would also be expected, ultimately, to leadto increasing mupirocin resistance among bothmethicillin-susceptible and resistant S. aureus.

8.5 Antimicrobial-impregnatedmedical devices

The introduction of antimicrobial-impregnatedmedical devices to decrease the frequency ofdevice-related nosocomial infection is a relatedissue. For short-term central vascular catheters clini-cal trials suggest a benefit in decreasing line infec-tions with antimicrobial-impregnated lines (112).It is argued that devices incorporating antisepticsubstances, such as the silver sulfadiazine cuff (100,113) or the chlorhexidine-silver sulfadiazine coated

catheter (114), are less likely to promote emergenceof resistance than antibiotic-impregnated catheters.However, widespread use of these devices in highly-susceptible intensive care unit patients would stillprovide optimal conditions for the emergence ofresistant strains. Coagulase-negative staphylococciare a particular concern as they are importantdevice-associated pathogens and have repeatedlydemonstrated a facility to acquire resistance.Antimicrobials incorporated into other devices,such as the indwelling bladder catheter, have beenless convincing in decreasing infection, and are notyet widely used in practice (115). However, thereis intense continuing investigation and develop-ment of a variety of medical devices which incor-porate antimicrobial substances for control ofnosocomial infections. Further careful evaluationof these devices will be essential to determine therelative benefits of infection reduction comparedwith the future risks of antimicrobial resistance.


16

9. Resources for infection control

strategies in developed countries, have added a sub-stantial burden to infection control programmes(20,32,39). In most facilities, the increaseddemands of infection control activity to manageantimicrobial-resistant organisms have not beenaccompanied by additional infection control re-sources. Other important activities, such as surveil-lance, have initially temporarily and sometimesindefinitely, been restricted or lapsed (20). Increasedattention to patient care practice to prevent trans-mission of organisms should, in fact, have a posi-tive impact on all nosocomial infections. However,the disarray in infection control activity occasionedby inordinate demands for antimicrobial resistancecontrol may result in a less effective programme inother areas, with a potential increase in nosocomialinfections. The intense focus on resistant organ-isms, much of which addresses colonization ratherthan infection, may undermine effective infectioncontrol. Infection control programmes in somefacilities have successfully obtained increased re-sources to accommodate the increased activity forantimicrobial resistance containment (68, 89), butthis has certainly not been the case universally. Ineffect, a redirection of the agenda and resources ofinfection control from preventing infections to con-taining antimicrobial resistance has a potential toincrease all nosocomial infections, hence increas-ing antimicrobial use and, one assumes, antimicro-bial resistance.

9.1 Cost-effectiveness

There is limited information addressing the cost-effectiveness of infection control interventions incontaining antimicrobial resistance. The fewrelevant publications are compromised by meth-odological problems. The topic is reviewed in de-tail in another report developed for the GlobalForum for Health Research, Cost-effectivenessanalysis: Interventions against antimicrobial resist-ance (116). This report also describes the com-plexities of assessing the costs of antimicrobialresistance and containment, particularly with re-spect to estimating the impact of current practiceinto the future.

9.2 Prioritization of infectioncontrol resources

There is another aspect of the economic impact ofantimicrobial resistance with respect to infectioncontrol. For all health care facilities, infection con-trol resources are limited relative to the burden ofinfections and potential infection control activity.Thus, an infection control programme must alwaysprioritize activity, and reallocation of resources fora new or expanding problem will redistribute re-sources from other potentially effective programmecomponents. The continuing global increase inMRSA and VRE, in particular, and attempts tocomply with recommended comprehensive control

17


10. Resource-poor countries

for the limited use of molecular typing methodsfor characterizing outbreak strains. Molecular typ-ing methods were only reported to have been usedfor the VRE outbreak in an oncology unit in SouthAfrica (131) and a Tunisian outbreak of Salmonellawien (127). This reflects the limited access to clini-cal microbiology support in many areas.

Several reports provide little information describ-ing control measures instituted to limit theoutbreak (121, 127, 132, 133, 139). Clinical pres-entations and microbiological observations aredescribed rather than epidemiological investigationand intervention. In some cases, control measureswere not attempted because of lack of resources.Interventions instituted in other outbreaks werelimited, and of a lower intensity than those usuallyapplied in facilities in developed countries. Inaddition, some of the interventions, such as fumi-gation, would not be considered useful (128, 130).Several outbreaks were not contained, despite con-trol efforts (122, 125, 128, 138). The resistant strainwas, however, eradicated from some institutions(126, 131, 137), particularly where an environmen-tal source (130, 134, 136) or staff carrier (123, 124,129, 135) was identified. The limitations inherentwith publication bias, however, must also beacknowledged for these reports.

10.3 Endemic antimicrobial resistance

A high prevalence of endemic antimicrobial-resist-ant organisms in acute care facilities in developingcountries has been repeatedly reported (140143).In addition, patients transferred from institutionsin developing countries have been the source forintroduction of a resistant strain into acute carefacilities in a developed country, with subsequentoutbreaks due to the resistant organism in the re-ceiving facility (144146). Thus, endemic antimi-crobial resistance is common in health care facilitiesin developing countries. There is little information,however, which describes the origin, patient risks,or impact of antimicrobial resistance. The effec-

10.1 Infection control

The interactions of infection control and antimi-crobial resistance containment in resource-poorcountries need special consideration. Some of theresistant organisms of concern and the impact ofantimicrobial resistance on nosocomial infectionsare similar to the experience in developed coun-tries. However, other organisms, modes of acquisi-tion, or approaches to control may be unique. Manyacute care facilities in these countries have noeffective infection control activity. In other areas,such as South (117) and Central America (118),South-East Asia (119), and Eastern Europe (120),substantial progress in developing infection con-trol programmes has occurred. There are still,however, limitations in resources and expertise forinfection control. A particular deficit which com-promises infection control function is limitedaccess to adequate clinical laboratory support.

10.2 Outbreaks with resistant organisms

Information relevant to infection control and anti-microbial resistance in developing countries isprimarily found in descriptions of outbreaks attrib-uted to resistant organisms (121139). Some pre-liminary summary observations of these reports canbe made. First, while the organisms currently ofconcern in developed countriesMRSA, VRE, andmultidrug-resistant Gram-negative bacteriaareobserved, over half of these reports describe out-breaks caused by Salmonella spp, Shigella spp, orVibrio cholerae. These organisms are unusual causesof nosocomial infection in acute care facilities indeveloped countries today. Secondly, with the ex-ception of the reports from a burn unit (125) andan oncology unit (131), these outbreaks all occurredin neonatal or paediatric units. This likely reflectsboth the patient population and distribution ofresources for care of different patient groups.Finally, in contrast to the reports from developedcountries summarized in the annex, the outbreaksreported from developing countries are remarkable


18

tiveness of infection control measures in these set-tings in limiting the spread of endemic resistantorganisms or preventing infections caused by theseorganisms is not known.

The current situation in developing countries,with a high prevalence of resistant organisms inhealth care facilities but rudimentary infection con-trol, may be a potential opportunity. The intro-duction and monitoring of the impact of infection

control interventions in facilities could permit anevaluation of the effect of infection controlprogrammes and specific activities of these pro-grammes. This is not feasible in developed coun-tries where a relatively higher level of infectioncontrol practice is already in place, and the impactof infection control specifically with respect to con-tainment of antimicrobial resistance is difficult toevaluate.

19


11. Research agenda

fection in the use of prophylactic antimicrobials bedetermined? What are the relative benefits ofinfection control activity and a stringent antimi-crobial use programme? What infection controlmeasures are appropriate for health care deliveredin the community or long-term care facilities?

There are also many organism-specific questions.Is it more effective to focus control efforts on allS. aureus, rather than methicillin-resistant S. aureus?In which patient populations might efforts to con-trol VRE be of value? What is the basic microbiol-ogy of organism transmission? What conditionsenhance organism transmission and how does thisvary for different organisms? Is the expansion ofESBL-producing Enterobacteriaceae within a hos-pital population any different than that observedwith susceptible Enterobacteriaceae?

Any serious agenda to contain antimicrobialresistance in health care settings must begin toaddress the large knowledge gap with respect to therole of infection control. Otherwise, infectioncontrol programmes will continue to consumeresources and require disruption in patient care inthe pursuit of antimicrobial-resistance containment,but in the absence of evidence that these activitiesare essential. This situation is not, ultimately, sus-tainable. With such large deficits in understand-ing, it is not realistic to expect to limit the currentprogression of resistance emergence and transmis-sion in health care facilities.

Greater in-depth knowledge of the role of an in-fection control programme in containing antimi-crobial resistance in health care facilities is needed.This will also require addressing some basic ques-tions about antimicrobial-resistant strains in healthcare facilities. For instance, what is the natural his-tory of antimicrobial-resistant strains in patients andthe health care environment, and how does anti-microbial therapy modify this? What factors deter-mine spontaneous decline or disappearance of astrain? The burden of illness attributable to resist-ant organisms, rather than simply the number oforganisms, must be measured. Additional studiesdescribing morbidity and mortality directly due toantimicrobial resistance are essential for estimationof the benefits of resistance containment. Theglobal and organism-specific costs of resistance alsomust be measured. Valid models to support pre-dictions of future costs resulting from loss of effi-cacy of current antimicrobials are necessary.

The unique contributions of specific infectioncontrol interventions to contain resistance must bedocumented, as the health care system will alwaysfunction within constrained resources. Whichinfection control interventions do not provide abenefit, and in which settings? What is the impactof different handwashing agents? When are gloves,or gowns, essential? What is the optimal screeningstrategy to identify colonized patients? How shouldthe balance between resistance promotion in thelong term and short-term benefit of decreased in-


20

12. Discussion

negative organisms in the 1970s led to the recog-nition of the importance of equipment and care ofpatients with indwelling catheters, to prevent trans-mission, leading to current recommendations forcare with dedicated equipment, and handwashingand glove use. Similarly, the inherent uncertaintyof identifying patients with resistant organisms ledto current recommendations for a higher standardof handwashing and glove use for all patient care.These practices will decrease all nosocomial infec-tions. Antimicrobial resistance is certainly notdesirable, but clearly has been beneficial for infec-tion control practice.

Evaluating the role of barrier infection controlpractices in containing endemic antimicrobial re-sistance is problematic. These practices appear tohave some efficacy when they are introduced intohigh-risk units in facilities where barrier practiceshave not previously been used. Even in facilitieswith a high standard of infection control practice,however, these interventions are not sufficient toultimately prevent the emergence and expansionof resistant organisms. Theoretically, to be effec-tive, barrier interventions must completely inter-rupt all transfer of microorganisms among patients.If this is achievable, it would only be with the in-stitution of extreme and highly costly measures,including dedicated staff to fully isolate onepatient from another. This would require substan-tial investment in personnel, building infrastruc-ture, and equipment. Such commitment does notseem feasible or appropriate given the current limi-tations in knowledge of the effectiveness of infec-tion control interventions. Where, and with whichpatient groups, is the appropriate trade-off betweenintensity of barrier practice and prevention of trans-mission of resistant organisms so that containmentis feasible and effective? To address this questionrequires further knowledge of the impacts of anti-microbial resistance, including an estimate of thefuture loss from decreased antimicrobial efficacyoccasioned by current failure to control resistance.

The contribution of infection control to anti-microbial pressure in health care facilities, and the

Despite the universal acceptance of infectioncontrol as a key element for containment of anti-microbial resistance in acute care facilities, theinteractions of infection control and antimicrobialresistance in these settings is complex, and not wellstudied.

Infection control activity is effective in control-ling outbreaks of infection caused by antimi-crobial-resistant strains. However, an adequateinfection control response is not universally norconsistently effective. In many facilities, outbreakstrains have become endemic, despite vigorous andappropriate control measures. The variables whichdetermine success or failure of outbreak controlhave not been systematically analysed, although itappears that when a point source can be identifiedcontrol is likely to be achieved. The example oftuberculosis is also evidence that when specificadministrative and engineering interventions canbe instituted control may also be achieved.

Antimicrobial resistance may also be seen as hav-ing a positive impact for infection control. As aphenotypic marker it may facilitate identificationof an outbreak strain. The introduction and trans-mission of a new strain of methicillin-susceptibleS. aureus into a facility is likely to go unnoticed. Amethicillin-resistant strain introduced into a facil-ity without endemic MRSA will be identified asunusual as soon as it is isolated from a clinical speci-men. Control measures initiated to limit the trans-mission of the resistant organism will also decreasetransmission of other organisms and, possibly, de-crease nosocomial infections globally in patientssubject to the interventions. Similarly, contami-nated equipment in an intensive care unit may notbe recognized if there are a small number of infec-tions with an endemic organism. If the organism isresistant, and this is an unusual phenotype for theunit, it will be identified early and investigation toidentify environmental contamination undertakenexpeditiously, limiting further infections. Antimi-crobial resistance also has likely been beneficial byleading to improvements in standards of patientcare. Repeated outbreaks of resistant Gram-

21


emergence of antimicrobial resistance through theuse of prophylactic topical or systemic antimi-crobials should be acknowledged. It is usuallyassumed that appropriate prophylactic antimicro-bial use will decrease infections and lower overallantimicrobial usea benefit for containing anti-microbial resistance. But resistance follows fromantimicrobial useappropriate or inappropriateand in this respect the goals of infection controland resistance containment may be divergent.Perhaps the way to frame this problem is toacknowledge that some resolution of the compet-ing priorities of direct patient care, resistance con-tainment, and infection control is necessary, thatthere is always a threshold at which appropriateantimicrobial use becomes inappropriate use, andthis may vary over time.

The goal of an infection control programme isto limit nosocomial infections in patients and staff.The required components for an optimal pro-gramme have been determined, and the efficacy ofthese programmes is well documented. For patientsafety, appropriate resources must be made avail-able to support infection control programmes. Aneffective infection control programme should alsoreduce infections with antimicrobial-resistantorganisms within a global reduction of all nosoco-mial infections. If infection control programmes

are assigned an additional role of limiting the trans-mission of organisms between patients to decreasecolonization as well as infection, then necessaryresources to perform this expanded function shouldbe identified and provided. The redirection of re-sources from effective infection control activity toantimicrobial resistance control is likely counter-productive, as the overall burden of nosocomialinfection may increase. The main focus of infec-tion control must remain on infection reduction.

Within a health care facility, the major determi-nant of antimicrobial resistance is antimicrobial use.Antimicrobial use leads to the initial emergence ofresistance, and is the major determinant of persist-ence of endemic resistance in a facility. In a sense,infection control programmes and activities attemptto limit the damage created by antimicrobial usepractices over which they have little control. Thepre-eminent importance of antimicrobial use strat-egies in containing resistance must be acknowl-edged. Infection control should not be heldaccountable for containment of antimicrobial re-sistance in the absence of aggressive antimicrobialrestriction and optimal use promotion in a facility.The development, implementation, and monitor-ing of an antimicrobial use programme, and theimportance of this activity, must be reinforced inany discussion of containment of resistance.


22

13. Conclusions

fection control activity or expansion of responsi-bility to include containment of colonization withresistant organisms will benefit either the goal ofdecreasing nosocomial infections or decreasing en-demic antimicrobial resistance cannot be estimatedwith information currently available. Promotinginfection control activity to contain antimicrobialresistance in the absence of effective, highly restric-tive, antimicrobial use programmes would appear,ultimately, to be futile. Infection control pro-grammes in health care facilities should be sup-ported and reinforced in their prime roletheprevention of infection, regardless of the presenceor absence of antimicrobial resistance. This shouldlead to optimal patient outcomes, and limit the pro-gression of resistance to the extent that infectioncontrol activity may have an impact.

Despite a consensus that institutional infectioncontrol programmes are important for containingantimicrobial resistance, the interactions betweeninfection control activity and antimicrobial resist-ance are not straightforward. Optimal infectioncontrol programmes, whose goal is to minimizenosocomial infections, may decrease the prevalenceof resistance and infections caused by resistantorganisms, but may also contribute to the emer-gence of antimicrobial resistance, and may be moreeffective in outbreak management because resist-ance facilitates identification of unusual organismsin the hospital.

The overarching benefit of infection controlprogrammes in decreasing nosocomial infections,some of which may be with resistant organisms, isclear. The extent to which an intensification of in-

23


14. Bibliography

14. Health Canada. Controlling Antimicrobial Resist-ance: An Integrated Action Plan for Canadians. CanComm Dis Rep (CCDR) 1997;Suppl 23S7:132.

15. Opinion of the Scientific Steering Committee on Anti-microbial Resistance. European CommissionDGXXIV 1999. www.europa.eu.int/comm/dg24/health/sc/ssc/out50_en.html

16. Scheckler WE et al. Requirements for infrastructureand essential activities of infection control and epi-demiology in hospitals: a consensus panel report.Society of Healthcare Epidemiology of America.Infect Control Hosp Epidemiol 1998;19:114124.

17. Haley RW et al. The efficacy of infection surveil-lance and control programs in preventing nosoco-mial infections in US hospitals. Amer J Epidemiol1985;121:182205.

18. Boyce JM et al. Methicillin-resistant Staphylococcusaureus: a briefing for acute care hospitals and nurs-ing facilities. Infect Control Hosp Epidemiol 1994;15:105113.

19. Mulligan ME et al. Methicillin-resistant Staphyloco-ccus aureus: a consensus review of the microbiology,pathogenesis, and epidemiology with implicationsfor prevention and management. Amer J Med1993;94:313328.

20. Humphreys H, Duckworth G. Methicillin-resistantStaphylococcus aureus: a reappraisal of control meas-ures in the light of changing circumstances. InfectControl Hosp Epidemiol 1997;36:167170.

21. Working Party Report. Revised guidelines for thecontrol of methicillin-resistant Staphylococcus aureusinfection in hospitals. Jour Hosp Infect 1998;39:253290.

22. Adeyemi-Doro FAB et al. Living with methicillin-resistant Staphylococcus aureus: A 7-year experiencewith endemic MRSA in a university hospital. InfectControl Hosp Epidemiol 1997;18:765767.

23. Austin DJ, Anderson RM. Transmission dynamicsof epidemic methicillin-resistant Staphylococcusaureus and vancomycin-resistant Enterococci in Eng-land and Wales. J Infect Dis 1999;179:883891.

24. Harstein AJ et al. Control of methicillin-resistantStaphylococcus aureus in a hospital and an intensivecare unit. Infect Control Hosp Epidemiol 1995;16:405411.

1. World Health Organization. Global Strategy for Con-tainment of Antimicrobial Resistance. World HealthOrganization 2001, WHO/CDS/CSR/DRS 2001.2.

2. Goldman DA et al. Strategies to prevent and con-trol the emergence and spread of antimicrobial-resistant microorganisms in hospitals. JAMA 1996;275:234240.

3. Abbas J et al. Candida krusei fungemia: An escalat-ing serious infection in immunocompromisedpatients. Arch Intern Med 2000;160:26592664.

4. Bonten MJ et al. Characteristics of polyclonal ende-micity of Pseudomonas aeruginosa colonization inintensive care units: Implications for infection con-trol. Amer J Respir Crit Care Med 1999;160:12121219.

5. Wiener J et al. Multiple antibiotic-resistant Kleb-siella and Escherichia coli in nursing homes. JAMA1999;281:517523.

6. Bonilla HF et al. Colonization with vancomycin-resistant Enterococcus faecium: Comparison of a longterm care unit with an acute care hospital. InfectControl Hosp Epidemiol 1997;18:333339.

7. Finland M. Changing patterns of susceptibility ofcommon bacterial pathogens to antimicrobial agents.Ann Intern Med 1972;76:10091036.

8. Moreno F et al. Methicillin-resistant Staphylococcusaureus as a community organism. Clin Infect Dis1995;21:13081312.

9. Riley LW et al. The significance of hospitals asreservoirs for endemic multiresistant Salmonellatyphimurium causing infection in urban Brazilianchildren. J Infect Dis 1984;150:236241.

10. Millar MR et al. Outbreak of infection withpenicillin-resistant Streptococcus pneumoniae in a hos-pital for the elderly. J Hosp Infect 1994;27:99104.

11. Lamb VA et al. Outbreak of Salmonella typhymuriumgastroenteritis due to an imported strain resistant toampicillin, chloramphenicol, and trimethoprim-sulfamethoxazole in a nursery. J Clin Microbiol1984;20:10761079.

12. Robins-Browne RM et al. A hospital outbreak ofmultiresistant Salmonella typhimurium belonging tophage type 193. J Infect Dis 1983;147:210216.

13. US Public Health Action Plan to Combat Antimicro-bial Resistance (Part 1: Domestic Issues) 2001. http://www.cdc.gov/drugresistance/actionplan/index.htm


24

25. Cohen SH, Morita MM, Bradford M. A seven yearexperience with methicillin-resistant Staphylococcusaureus. Amer J Med 1991;31(suppl 3B):233237.

26. Guigit M et al. Effectiveness of simple measures tocontrol an outbreak of nosocomial methicillin-resistant Staphylococcus aureus infections in anintensive care unit. Infect Control Hosp Epidemiol1990;11:2326.

27. Ribner BS, Landry MN, Gholson GL. Strict versusmodified isolation for prevention of nosocomialtransmission of methicillin-resistant Staphylococcusaureus. Infect Control 1986;7:317320.

28. Fazal Ba et al. Trends in the prevalence of methicil-lin-resistant Staphylococcus aureus with discontinua-tion of an isolation policy. Infect Control HospEpidemiol 1996;17:372374.

29. Meers PD, Leong KY. The impact of methicillin andaminoglycoside-resistant Staphylococcus on thepattern of hospital-acquired infection in an acutehospital. J Hosp Infect 1990;16:231239.

30. Centers for Disease Control and Prevention. Rec-ommendations for preventing the spread of vanco-mycin resistance: recommendations of the HospitalInfection Control Practices Advisory Committee(HICPAC) MMWR 1995;44(RR-12):112.

31. Fridkin SK, Gaynes RP. Antimicrobial resistance inintensive care units. Clin Chest Med 1999; 20:303316.

32. Morris J et al. Enterococci resistant to multiple an-timicrobial agents, including vancomycin: establish-ment of endemicity in a university medical center.Ann Intern Med 1995;123:250259.

33. Pegues D et al. Emergence and dissemination of ahighly vancomycin-resistant vanA strain of Entero-coccus faecium at a large teaching hospital. J ClinMicrobiol 1997;35:15651570.

34. Goetz AM et al. Infection and colonization withvancomycin-resistant Enterococcus faecium in anacute care Veterans Affairs Medical Center: a 2-yearsurvey. Amer J Infect Control 1998; 26:558562.

35. Lai KK et al. Failure to eradicate vancomycin resist-ant enterococci in a university hospital and the costof barrier precautions. Infect Control Hosp Epidemiol1998;19:647652.

36. Quale J et al. Experience with a hospital-wide out-break of vancomycin-resistant enterococci. Am JInfect Control 1996;24:372379.

37. Merrer J et al. Colonization pressure and risk ofacquisition of methicillin-resistant Staphylococcusaureus in a medical intensive care unit. InfectControl Hosp Epidemiol 2000;21:718723.

38. Warshawsky B et al. Hospital and community basedsurveillance of methicillin-resistant Staphylococcusaureus: Previous hospitalization is the major riskfactor. Infect Control Hosp Epidemiol 2000;21:724727.

39. Goetz AM, Muder RR. The problem of methicillin-resistant Staphylococcus aureus: A critical appraisal ofthe efficacy of infection control procedures with asuggested approach for infection control programs.Amer J Infect Control 1992;20:8084.

40. Bernards AT et al. Methicillin-resistant Staphylococ-cus aureus and Acinetobacter baumannii: An unex-pected difference in epidemiologic behaviour. Am JInfect Control 1998;26:544551.

41. Kayser FH. Methicillin-resistant staphylococci196575. Lancet 1975;2:650653.

42. Rosdahl VT, Knudsen AM. The decline of methi-cillin resistance among Danish staphylococcusstrains. Infect Control Hosp Epidemiol 1991;21:8388.

43. Alford AH, Hall A. Epidemiology of infectionscaused by gentamicin-resistant Enterobacteriaceaeand Pseudomonas aeruginosa over 15 years at theNashville Veterans Administration Medical Center.Rev Infect Dis 1987;9:10791086.

44. Bridges K et al. Gentamicin and silver resistant pseu-domonas in a burns unit. Br Med J 1979;1:446479.

45. Snelling CF, et al. Resistance of gram-negativebacilli to gentamicin. J Infect Dis 1971;124 (Suppl)S264272.

46. Mayer KH et al. Molecular evolution, species distri-bution and clinical consequences of an endemicaminoglycoside resistance plasmid. Antimicrob AgentsChemother 1986;29:628633.

47. Agerton T et al I. Transmission of a highly drugresistant strain of Mycobacterium tuberculosis. Com-munity outbreak and nosocomial transmission via acontaminated bronchoscope. JAMA 1997;278:10731077.

48. Alvarez M et al. Noscomial outbreak caused byScedosporium prolificans (inflatum): four fatal casesin leukemic patients. J Clin Microbiol 1995;33:32903295.

49. Echols RM et al. Multidrug resistant Serratiamarcesens bacteriuria related to urologic instrumen-tation. South Med J 1984;77:173177.

50. Larson EA. A causal link between handwashing andrisk of infection? Examination of the evidence.Infect Control Hosp Epidemiol 1988;9:2836.

51. Newsom SWB. Pioneers in infection control: IgnazPhilipp Semmelweiss. J Hosp Infect 1993;23:175187.

52. Pratt RJ et al, the epic guideline development team.Guidelines for preventing hospital-acquired infec-tions. J Hosp Infect 2001;47(Suppl):S3S4.

53. Larson EL. APIC guideline for handwashing andhand antisepsis in health care settings. Amer J InfectControl 1995;23:251269.

25


54. Health Canada. Handwashing, cleaning, disinfec-tion, and sterilization in health care. Can Comm DisRep (CCDR) Supplement, Vol 24S4, July 1998.

55. Russell AD. Bacterial resistance to disinfectants:present knowledge and future problems. J Hosp In-fect 1999;43(Suppl):S5768.

56. Jones R. Bacterial resistance and topical antimicro-bial wash products. Am J Infect Control 1999;27:351363.

57. Onesko KM, Wienke EC. The analysis of theimpact of a mild low-iodine lotion soap on the re-duction of nosocomial methicillin-resistant Staphy-lococcus aureus: a new opportunity for surveillanceobjectives. Infect Control 1987;8:284288.

58. Webster J, Faoagali JL, Cartwright D. Eliminationof methicillin-resistant Staphylococcus aureus from aneonatal intensive care unit after handwashing withtriclosan. J Paediatr Child Health 1994;30:5964.

59. Chuanchuen R et al. Cross-resistance betweentriclosan and antibiotics in Pseudomonas aeruginosais mediated by multidrug efflux pumps: Exposureof a susceptible mutant strain to triclosan selects nfxBmutants overexpressing MexCD-OprJ. AntimicrobAgents Chemother 2001;45:428432.

60. Pittet D et al. Infection Control Programme. Effec-tiveness of a hospital-wide programme to improvecompliance with hand hygiene. Lancet 2000;356:13071312.

61. CDC Isolation techniques for use in hospitals 2nd Ed.Washington D.C. US Government Printing Office.1975 (DHEW Publication no. (CDC) 768314.

62. Lynch P et al. Rethinking the role of isolation prac-tices in the prevention of nosocomial infections. AnnIntern Med 1987;107:243246.

63. Garner JS. The Hospital Infection Control PracticesAdvisory Committee: Guidelines for isolation pre-cautions in hospitals. Infect Control Hosp Epidemiol1996;17:5380.

64. Health Canada. Routine practices and additionalprecautions for preventing transmission of infectionin health care. Can Comm Dis Rep (CCDR) Supple-ment, 25S4, July 1999.

65. Lucet J-C et al. Control of a prolonged outbreak ofextended-spectrum beta-lactamase producing En-terobacteriaceae in a university hospital. Clin InfectDis 1999; 29:14111418.

66. Soulier A et al. Decreased transmission of Entero-bacteriaceae with extended-spectrum beta-lactamasesin an intensive are unit by nursing reorganization.J Hosp Infect 1995;31:8997.

67. Thompson RL, Cabezudo I, Wenzel RP. Epidemi-ology of nosocomial infections caused by methicil-lin-resistant Staphylococcus aureus. Ann Intern Med1982;97:309317.

68. Casseron-Zerbib M et al. A control programme formethicillin-resistant Staphylococcus aureus contain-ment in a paediatric intensive care unit: evaluationand impact on infections caused by other organisms.J Hosp Infect 1998;40:225235.

69. Harbarth S et al. Effect of delayed infection controlmeasures on a hospital outbreak of methicillin-resistant Staphlyococcus aureus. J Hosp Infect 2000;46:4349.

70. Schmitz FJ et al. Methicillin-resistant Staphylococ-cus aureus strains in the greater Dusseldorf area. EurJ Epidemiol 1997;13:709717.

71. Souweine B et al. Role of infection control meas-ures in limiting morbidity associated with multi-resistant organisms in critically ill patients. J HospInfect 2000;45:107116.

72. Slaughter S et al. A comparison of the effect of uni-versal use of gloves and gowns with that of glove usealone on the acquisition of vancomycin-resistantenterococci in a medical intensive care unit. AnnIntern Med 1996;125:448456.

73. Bonten MJ et al. The role of colonization pressurein the spread of vancomycin-resistant enterococci:an important infection control variable. Arch InternMed 1998;158:11271132.

74. Brooks S et al. Reduction in vancomycin-resistantenterococcus and Clostridium difficile infectionsfollowing change in tympanic thermometers. InfectControl Hosp Epidemiol 1998;19:333336.

75. Montecalvo MA et al. Infection-control measuresreduce transmission of vancomycin-resistant ente-rococci in an endemic setting. Ann Intern Med1999;131:269272.

76. Wenger MN et al. Control of nosocomial transmis-sion of multidrug-resistant Mycobacterium tubercu-losis among healthcare workers and HIV-infectedpatients. Lancet 1995;345:235240.

77. Maloney SA et al. Efficacy of control measures inpreventing nosocomial transmission of multidrug-resistant tuberculosis to patients and health careworkers. Ann Intern Med 1995;122:9095.

78. CDC Guideline for prevention of nosocomial pneu-monia. Respir Care 1994;39:11911236.

79. Taolin D, Merz PM. Flower vases in hospitals asreservoirs of pathogens. Lancet 1973;2:12791281.

80. Barg NL. Environmental contamination withStaphylococcus aureus and outbreaks: the cause or theeffect? Infect Control Hosp Epidemiol 1993;14:14711476.

81. Weber DJ, Rutala WA. Role of environmentalcontamination in the transmission of VRE. InfectControl Hosp Epidemiol 1997;18:306309.

82. Hayden MK. Insights into the epidemiology andcontrol of infection with vancomycin-resistantenterococci. Clin Infect Dis 2000;31:10581065.


26

83. Dellinger EP et al. Quality standard for antimicro-bial prophylaxis in surgical procedures. Infect Con-trol Hosp Epidemiol 1994;15:182188.

84. Antibiotic prophylaxis in surgery: Summary of aSwedish-Norwegian Consensus conference. ScandJ Infect Dis 1998;30:547557.

85. Ross CB et al. Ceftriaxone versus cefazolin inperipheral arterial operations: a randomized, pro-spective trial. South Med J 1997;90:1622.

86. Andersen BM et al. Multiply beta-lactam resistantEnterobacter cloacae infections linked to the environ-mental flora in a unit for cardiothoracic and vascu-lar surgery. Scand J Infect Dis 1989;21:181191.

87. Ratto GB et al. Long-term antimicrobial prophy-laxis in lung cancer surgery: correlation betweenmicrobiological findings and empyema develop-ment. Lung Cancer 1994;11:345352.

88. Flynn DM et al. Patients endogenous flora as thesource of nosocomial Enterobacter in cardiac sur-gery. J Infect Dis 1987;156:363368.

89. Terpstra S et al. Rapid emergence of resistantcoagulase-negative staphylococci on the skin afterantibiotic prophylaxis. J Hosp Infect 1999;43:195202.

90. Hughes WT et al. 1997 guidelines for the use ofantimicrobial agents in neutropenic patients withunexplained fever. Infectious Diseases Society ofAmerica. Clin Infect Dis 1997;25: 551573.

91. First European Consensus Conference in IntensiveCare Medicine. Selective decontamination in inten-sive care unit patients. Intensive Care Med 1992;18:182188.

92. DAmico R et al. Effectiveness of antibiotic prophy-laxis in critically ill adult patients: a systematicreview of randomized trials. Br Med J 1998;316:12751285.

93. Webb CH. Selective decontamination of the diges-tive tract, SDD: a commentary. J Hosp Infect 2000;46: 106109.

94. Kollef MH, Fraser VJ. Antibiotic resistance in theintensive care unit.

infection control programmes to control antimicrobial resistance

Documents

health care facilities

antimicrobial interventions

control of outbreaks

resistant organisms

barrier practices

malariaexecutive summary

topical prophylaxis

systemic prophylaxis