To determine hospital-related risk factors for surgical-site infection (SSI) following hip arthroplasty.

Prospective, multicenter cohort study based on surveillance data and data collected through a structured telephone interview. With the use of multilevel logistic regression, the independent effect of hospital-related characteristics on SSI was assessed.

Thirty-six acute care hospitals in the Dutch surveillance network for nosocomial infections (PREZIES), from 1996 to 2000.

Thirteen thousand six hundred eighty patients who underwent total or partial hip arthroplasty.

A high annual volume of operations was associated with a reduced risk of SSI (risk-adjusted risk ratio [RR] per 50 extra operations, 0.85; 95% confidence interval [CI95], 0.74–0.97). With each extra full-time–equivalent infection control staff member per 250 beds available for prevention of SSI, the risk for SSI was decreased (RR, 0.48; CI95, 0.16–1.44), although the decrease was not statistically significant. Hospital size, teaching status, university affiliation, and number of surgeons and their years of experience showed no important association with the risk of SSI.

Undergoing surgery in a hospital with a low volume of operations increases a patient's risk of SSI.

To characterize risk factors associated with pneumococcal disease and asymptomatic colonization during an outbreak of multidrug-resistant Streptococcus pneumoniae (MDRSP) among AIDS patients in a long-term–care facility (LTCF), evaluate the efficacy of antimicrobial prophylaxis in eliminating MDRSP colonization, and describe the emergence of fluoroquinolone resistance in the MDRSP outbreak strain.

Epidemiologic investigation based on chart review and characterization of SP strains by antimicrobial susceptibility testing and PFGE and prospective MDRSP surveillance.

An 80-bed AIDS-care unit in an LTCF.

Staff and residents on the unit.

From April 1995 through January 1996, 7 cases of MDRSP occurred. A nasopharyngeal (NP) swab survey of all residents (n = 65) and staff (n = 70) detected asymptomatic colonization among 6 residents (9%), but no staff. Isolates were sensitive only to rifampin, ofloxacin, and vancomycin. A 7-day course of rifampin and ofloxacin was given to eliminate colonization among residents: NP swab surveys at 1, 4, and 10 weeks after prophylaxis identified 1 or more colonized residents at each follow-up with isolates showing resistance to one or both treatment drugs. Between 1996 and 1999, an additional 6 patients were diagnosed with fluoroquinolone-resistant (FQ-R) MDRSP infection, with PFGE results demonstrating that the outbreak strain had persisted 3 years after the initial outbreak was recognized.

Chemoprophylaxis likely contributed to the development of a FQ-R outbreak strain that continued to be transmitted in the facility through 1999. Long-term control of future MDRSP outbreaks should rely primarily on vaccination and strict infection control measures.

To evaluate nosocomial infection (NI) surveillance strategies in French ICUs and to identify similar patterns defining subsets within which comparisons can be made.

A questionnaire was sent to all French ICUs, and a random sample of nonresponders was interviewed.

Three hundred ninety-five responder ICUs (69%) in France.

In 282 ICUs (71%), a dedicated ICU staff member was responsible for infection control activities. The microbiology laboratory was usually in the hospital (90%) and computerized (94%) but issued regular hospital microbiology records in only 48% of cases. Patients receiving mechanical ventilation, central venous catheterization, and urinary catheterization were 90%, 79%, and 60%, respectively. Patients were screened for carriage of mul-tidrug-resistant bacteria on admission and during the stay in 70% and 60% of ICUs, respectively, most often targeting MRSA. Quantitative cultures were used to diagnose ventilator-associated pneumonia (VAP) in 90% of ICUs, including distal specimens in 80% and bronchoscopy specimens in 60%. Quantitative central venous catheter (CVC)-segment cultures were used in 70% of ICUs. All CVCs were cultured routinely in 53% of the ICUs. Despite wide variations in infection control and surveillance strategies, multiple correspondence analysis identified 13 key points (4 structural variables and 9 variables concerning the diagnosis of VAP, the surveillance and diagnosis of catheter-related and urinary tract infections, and the mode of screening of MRSA carriers) that categorize the variability of French ICUs' approaches to NIs.

This study revealed profound differences in N1 surveillance strategies across ICUs, indicating a need for caution when using N1 surveillance data for comparisons and benchmarking.

To review the evidence evaluating perioperative intranasal mupirocin for the prevention of surgical-site infections according to type of surgical procedure.

Systematic review and meta-analysis of published clinical trials.

Studies included were either randomized clinical trial or prospective trials at a single institution that measured outcomes both before and after an institution-wide intervention (before-after trial). In all studies, intervention and control groups differed only by the use of perioperative intranasal mupirocin in the intervention group.

Patients undergoing general or nongeneral surgery (eg, cardiothoracic surgery, orthopedic surgery, and neurosurgery).

Risk of surgical-site infection following perioperative intranasal mupirocin versus usual care.

Three randomized and four before-after trials met the inclusion criteria. No reduction in surgical-site infection rate was seen in randomized general surgery trials (summary estimates: 8.4% in the mupirocin group and 8.1% in the control group; relative risk [RR], 1.04; 95% confidence interval [CI95], 0.81 to 1.33). In nongeneral surgery, the use of mupirocin was associated with a reduction in surgical-site infection in randomized trials (summary estimates: 6.0% in the mupirocin group and 7.6% in the control group; RR, 0.80; CI95, 0.58 to 1.10) and in before-after trials (summary estimates: 1.7% in the mupirocin group and 4.1% in the control group; RR, 0.40; CI95, 0.29 to 0.56).

Perioperative intranasal mupirocin appears to decrease the incidence of surgical-site infection when used as prophylaxis in nongeneral surgery. Given its low risk and low cost, use of perioperative intranasal mupirocin should be considered in these settings.

To evaluate the antibody response to influenza vaccines in nursing home residents and healthcare workers (HCWs) and its relation to residents' functional and chronic disease status during four successive seasons.

Before-after study.

Nine nursing homes during the 1998-1999 season and two during the 1999-2000, 2000-2001, and 2001-2002 seasons.

Two hundred fifty-nine residents and 79 HCWs during the 1998-1999 season; 180 and 71, respectively, during the 1999-2000 season; 162 and 71, respectively, during the 2000-2001 season; and 153 and 79, respectively, during the 2001-2002 season.

Multivariate analysis indicated that the mean fold increase in the geometric mean titers (GMTs) of hemagglutination inhibition (HI) antibodies and the response rate (the proportion of vaccinées resulting in a significant, at least fourfold increase in antibody titer) were good and no significant differences occurred for almost all strains in both residents and HCWs. The GMTs of HI antibodies and the protection rate (the proportion of participants with HI antibody titers & 40) were increased in both residents and HCWs, but were significantly lower for almost all strains in residents than in HCWs. Furthermore, multivariate analysis indicated that subdivision of residents into three groups by level of daily activities and into four groups according to underlying diseases revealed only minor differences in immune responses.

Antibody responses to the influenza vaccine were lower in residents than in HCWs. However, residents showed similar antibody responses regardless of their level of daily activity or underlying diseases.

Clostridium difficile-associated diarrhea (CDAD) is an important infection in hospital settings. Its impact on outpatient care has not been well defined.

To examine risk factors of ambulatory cancer patients with CDAD.

Case-control study.

Memorial Sloan-Kettering Cancer Center, a tertiary-care hospital.

Cases of CDAD among oncology outpatients from January 1999 through December 2000 were identified via positive C. difficile toxin assay results on stool specimens sent from clinics or the emergency department. A 1:3 matched case-control study examined exposures associated with CDAD.

Forty-eight episodes of CDAD were identified in cancer outpatients. The mean age was 51 years; 44% were female. Forty-one (85%) had received antibiotics within 60 days of diagnosis, completing courses a median of 16.5 days prior to diagnosis. Case-patients received longer courses of first-generation cephalosporins (4.8 vs 3.2 days; P = .03) and fluoroquinolones (23.6 vs 8 days; P < .01) than did control-patients. Those receiving clindamycin were 3.9-fold more likely to develop CDAD (P < .01). For each additional day of clindamycin or third-generation cephalosporin exposure, patients were 1.29- and 1.26-fold more likely to develop CDAD (P < .01 and .04, respectively). The 38 CDAD patients hospitalized during the risk period (79.2%) spent more time as inpatients than did control-patients (19.3 vs 9.7 days; P <. 001).

Antibiotic use, especially with cephalosporins and clindamycin, and prolonged hospitalization contributed to the development of CDAD. Outpatient CDAD appears to be most strongly related to inpatient exposures; reasons for the delayed development of symptoms are unknown.

Case-control studies often analyze risk factors for antibiotic resistance. Recently published articles have illustrated that randomly selected control-patients may be preferable to those with the susceptible phenotype of the organism. A possible methodologic problem with randomly selected control-patients is potential bias due to control group misclassification. This occurs if some control-patients did not have clinical cultures performed and thus might have been unidentified case-patients. If this bias exists, these studies might be expected to report lower odds ratios (ORs) because control-patients would be more like case-patients.

To analyze potential biases that might arise due to control group misclassification and potentially larger selection biases that may be introduced if control-patients are required to have at least one clinical culture.

One hundred twenty case-patients, 770 control-patients in group 1, and 510 control-patients in group 2.

Two case-control studies. Case-patients had clinical cultures positive for imipenem-resistant Pseudomonas aeruginosa. The first group of control-patients were random. The second group of control-patients were identical to those in group 1 except being required to have at least one clinical culture.

Univariate analyses showed higher ORs for case-patients versus control-patients in group 1 (imipenem [OR, 12.5], piperacillin-tazobactam [OR, 3.7], and vancomycin [OR, 4.7]) as compared with case-patients versus control-patients in group 2 (imipenem [OR, 8.0], piperacillin-tazobactam [OR, 2.5], and vancomycin [OR, 3.0]).

Requiring control-patients to have at least one clinical culture introduces a selection bias likely because it eliminates patients with less severe illness.

To quantify the occurrence of surgical-site infections (SSIs) in an Italian region and to estimate the proportion of potentially avoidable infections through benchmarking comparison.

Prospective study during 1 month based on a convenience sample of surgical patients admitted to 31 public hospitals. All of the patients undergoing an intervention included among the 44 operative procedures of the National Nosocomial Infections Surveillance (NNIS) System were enrolled. Ninety-five percent of the patients were actively observed after discharge for up to 30 days for all of the operations and for up to 1 year for operations involving implantation.

Among the 6,167 operative procedures studied, 290 infections were recorded (4.7 per 100 procedures), 206 (71%) of which were SSIs (3.3 per 100 procedures; 95% confidence interval, 2.9–3.9). One hundred thirty-five SSIs (65.5%) were superficial infections, 53 (25.7%) were deep infections, and 12 (5.8%) were organ–space infections; in 6 cases (2.9%), the type of SSI was not recorded. The frequency of SSIs observed in this study was significantly higher for several procedures than that expected when the NNIS System rates (standardized infection ratio [SIR] ranging from 1.77 to 6.42) or the Hungarian rates (SIR ranging from 1.28 to 3.04) were applied to the study population.

The high intensity of postdischarge surveillance can in part explain the differences observed. To allow for meaningful benchmarking comparison, in addition to intrinsic patient risk, data on the intensity of postdischarge surveillance should be included in published reports.

Because resident physicians (RPs) frequently have direct patient contact, those who are unimmunized against influenza potentially subject patients to unnecessary risk of infection.

To determine the rates of, knowledge regarding, and attitudes toward influenza immunization among RPs. We hypothesized that rates of and knowledge about influenza immunization did not differ between primary care (PC) and non-PC RPs.

A self-administered, anonymous questionnaire distributed to a convenience sample of 300 RPs (150 PC and 150 non-PC). The questionnaire requested influenza immunization status in the 2003-2004 and previous seasons and factors influencing respondents' decisions whether to be immunized. It included a 20-item test of knowledge about influenza immunization.

Two hundred five (68.3%) of 300 distributed questionnaires (196 that were evaluable) were returned. Response rates of PC and non-PC RPs did not differ (P = .79). The overall immunization rate of RPs in 2003-2004 was 38.3% and rates did not differ between PC (38.9%) and non-PC (37.6%) RPs. RPs most often cited “self-protection” as a reason for electing (93.3%) and “lack of time” for declining (47.1%) influenza immunization. Their ability to correctly answer questions about influenza immunization varied; their mean knowledge score was 13.7 (perfect = 20). PC and non-PC trainees did not differ by knowledge score (P = .48). However, RPs “ever vaccinated” had a higher knowledge score than those “never vaccinated” (P = .01).

RPs have low immunization rates and significant gaps in knowledge regarding influenza immunization. These problems should be addressed during their training by education on the importance, effectiveness, and safety of influenza vaccine for them and their patients.

In Victoria, Australia, from July to December 2002, 126 outbreaks of viral and suspected viral gastroenteritis were reported in healthcare institutions. Norovirus was found to account for at least 77 of the 126 outbreaks.

In October 2002, the infection control unit investigated an outbreak of acute gastroenteritis on three wards in a 500-bed, long-term-care facility in Melbourne, Victoria, Australia. Cohorting and other infection control measures were initiated.

The outbreak was controlled 32 days after the first symptoms of acute gastroenteritis were identified. Fifty-two patients and 11 staf f members were affected. Norovirus genotype 2 was detected on two of the three wards. Norovirus was not isolated in the third ward but was suspected to be the causative organism.

Outbreaks of viral gastroenteritis can cause significant morbidity in a long-term-care facility, affecting both patients and staff. In addition, the transmission of viral pathogens can be well established before there is recognition of an outbreak.

To determine whether certain hair removal policies are better than others to prevent surgical-site infections in patients undergoing clean surgery.

Publications were retrieved by a systematic search of Medline, the Cochrane Library, and EMBASE up to February 2005. Additionally, the reference lists of all identified trials were examined. All randomized trials, quasi-randomized trials, and systematic reviews or meta-analyses of randomized or quasi-randomized trials comparing hair removal policies in clean surgery were selected. Trials involving patients undergoing cranial neurosurgery were excluded. Two reviewers independently assessed trial quality and extracted data. Disagreements were resolved by discussion with a third reviewer. Data from the original publications were used to calculate the relative risk or risk difference of surgical-site infection. Data for similar outcomes were combined in the analysis, where appropriate, with the use of a random effects model.

Four trials were included in the review. No eligible systematic review or meta-analysis of randomized or quasi-randomized trials was found. The quality of the trials and how they were reported were generally unsatisfactory. Evidence regarding whether preoperative hair removal has any effect was inconclusive. When hair removal was considered necessary, evidence about the best time for removal was inconclusive. There was some evidence that hair removal by clipper is superior to removal by razor.

Because of insufficient evidence as a basis for recommendations, the practical consequences for ward management were essential when the Dutch Working Party on Infection Prevention formulated its recommendations for hair removal policies. Large randomized, controlled trials are needed to determine the optimal policy for preoperative hair removal.

Clostridium difficile is the most common cause of infectious nosocomial diarrhea and can be found in up to 30% of asymptomatic hospitalized patients. Our primary aim was to compare the clinical characteristics of hospitalized patients who received antibiotics and developed C. difficile-associated diarrhea (CDAD) with those of hospitalized patients who received antibiotics and did not develop the disease.

Case-control study comprising inpatients at a single institution.

Case-patients were defined as patients who had diarrhea and tested positive for C. difficile. Control-patients (matched 4:1 to case-patients) were defined as patients who received antibiotics for at least 5 days and did not develop CDAD.

On univariate analysis, nine variables were associated with CDAD. Only three of the variables, need for intensive care, length of stay, and macrolide antibiotic use, were found to be significant (P < .05) on logistic regression analysis. The odds ratios for status as a CDAD case were 3.68 (CI95, 1.44 to 9.40) for stay in the intensive care unit and 1.03 (CI95, 1.02 to 1.05) for each day of hospital stay. Receipt of macrolide antibiotics reduced risk significantly; the odds ratio was 0.23 (CI95, 0.19 to 0.87).

We identified need for intensive care and length of stay as important risk factors for the development of CDAD. We also identified macrolide antibiotic use as protective against its development. Patients receiving intensive care may represent a population to study for targeted prophylaxis.

There are significant limitations of the standard case-control study design for identifying risk factors for resistant organisms. The objective of this study was to develop a study design to overcome these limitations.

Theoretical analysis of different types of study designs that can be used in risk factor studies for resistant organisms.

We developed the case-case-control study design, which uses two separate case-control analyses within a single study. The first analysis compares patients infected with resistant bacteria (resistant cases) with control-patients without infection caused by the target organism, who are therefore representative of the source population; and the second analysis compares patients infected with the susceptible phenotype of the target organism (susceptible cases) with the same controlpatients without infection caused by the target organism. These two analyses provide risk models for (1) isolation of the resistant phenotype of the target organism as compared with the source population and (2) isolation of the susceptible phenotype of the organism as compared with the source population. When these two risk models are compared and contrasted, risk factors specifically associated with isolation of the resistant phenotype can be identified.

The case-case-control study design is an effective method for identifying risk factors for antimicrobial-resistant pathogens. Although the case-case-control study design has limitations, it is, in our opinion, more informative and less flawed than the standard case-control study design.

To evaluate the relationship between Staphylococcus aureus nasal and tracheal colonization and infection in medical intensive care unit (MICU) patients. The primary outcome was the incidence of S. aureus infection in colonized versus non-colonized patients.

Prospective, observational cohort study. Patients admitted to the MICU during the study period were screened for S. aureus nasal and tracheal colonization by culture and a PCR assay twice weekly. Demographic, clinical, and microbiologic data were collected in the MICU and for 30 days after discharge. PFGE and antibiotic susceptibility testing were performed on all S. aureus nasal, tracheal, and clinical isolates.

Twenty-three percent of patients (47 of 208) were nasally colonized with S. aureus. Twenty-four percent of these patients developed S. aureus infections versus 2% of noncolonized patients (P < .01). Nine of 11 patients with both nasal colonization and infection were infected by their colonizing strain. Two of 47 nasally colonized patients developed an infection with a different strain of S. aureus. Fifty-three percent of intubated patients with nasal colonization (10 of 19) had tracheal colonization with S. aureus as opposed to 4.9% of intubated, non-colonized patients (3 of 61) (P < .01). Parenteral antibiotics were ineffective at clearing nasal colonization. PCR detected S. aureus colonization (nasal and tracheal) within 6.5 hours with a sensitivity of 83% and a specificity of 99%.

The incidence of S. aureusinfection was significantly elevated in nasally colonized MICU patients. Techniques to rapidly detect colonization in this population may make targeted topical prevention strategies feasible. (Infect Control Hosp Epidemiol 2005;26:622-628)

To evaluate the secular trends in MRSA BSIs after the introduction of a nosocomial MRSA control intervention.

Before-after study.

An 850-bed community hospital with an ICU and vascular surgery, neurosurgery, bone marrow transplantation, and AIDS units. MRSA is endemic at this hospital; the prevalence of methicillin resistance among patients with S. aureus infection is greater than 50%.

Among all inpatients, MRSA BSI was identified, its origin defined, and incidence rates calculated by ward and origin.

A MRSA control program was implemented based on active surveillance cultures to identify MRSA-colonized patients, followed by isolation using contact precautions. Incidence rates of MRSA BSI during the intervention (ie, July 1, 1997, to December 31, 2001) and preintervention (ie, January 1, 1996, to June 30, 1997) periods were compared.

Sixty-nine MRSA BSIs were identified. When compared with the preintervention period, the incidence rate of MRSA BSI was reduced from 0.64 to 0.30 per 1,000 admissions (RR, 0.46; CI95, 0.25–0.87; P = .02) during the intervention period. The impact was greater in the ICU, with an 89% reduction (RR, 0.11; CI95, 0.01–0.98; P = .03), and for CVC-associated MRSA BSIs, with an 82% decrease (RR, 0.17; CI95, 0.05–0.55; P = .002). Methicillin resistance among S. aureus blood isolates decreased from 46% to 17% (RR, 0.36; CI95, 0.22–0.62; P = .0002).

A reduction in MRSA bacteremia is achievable through use of the MRSA “search and isolate” intervention even in a hospital with high rates of endemic MRSA.

Central venous catheter (CVC)-related bloodstream infections (BSIs) are known to increase rates of morbidity and mortality in both inpatients and outpatients, including hematology-oncology patients and those undergoing hemodialysis or home infusion therapy. Biofilm-associated organisms on the lumens of these catheters have reduced susceptibility to antimicrobial chemotherapy. This study tested the efficacy of tetrasodium EDTA as a catheter lock solution on biofilms of several clinically relevant microorganisms.

Biofilms of Staphylococcus epidermidis, methicillin-resistant S. aureus, Pseudomonas aeruginosa, Escherichia coli, Klebsiella pneumoniae, and Candida albicans were grown to levels of approximately 1 × 105 colony-forming units (CFU)/cm−1 on CVC segments in a model system, then subjected to the tetrasodium EDTA lock treatment.

Comparisons of biofilms before and after exposure to the 40-mg/mL−1 tetrasodium EDTA lock for 21 hours showed that the biofilm viable cell counts of all organisms tested were significantly reduced (P < .05) after exposure to the treatment.

Antimicrobial lock treatment using 40 mg/mL−1 of tetrasodium EDTA for at least 21 hours could significantly reduce or potentially eradicate CVC-associated bio-films of clinically relevant microorganisms (Infect Control Hosp Epidemiol 2005;26:515-519).

To describe the usefulness of the OraQuick Rapid HIV-1 Antibody Test (OraSure Technologies, Bethlehem, PA) in cases of occupational exposure regarding its use with source-patient sera, effects on post-exposure prophylaxis (PEP) use, potential cost savings, and effects on healthcare worker (HCW) stress reaction symptoms.

Before-and-after analysis.

A 269-bed, tertiary-care medical center with adjacent clinics.

All source-patients and HCWs experiencing an occupational exposure during the study period.

Use of the OraQuick test with patient sera was validated prior to its use for occupational exposures. Exposures from January 1 through July 10, 2003 (enzyme immunoassay [EIA] group) and July 11 through December 31, 2003 (OraQuick group) were retrospectively reviewed and the use and cost of PEP was compared for each group. Randomly selected HCWs from both groups completed a survey to assess their stress reaction symptoms.

After exclusion, there were 71 exposures in the EIA group and 79 in the OraQuick group. OraQuick results were 100% concordant with the reference standard of EIA and Western blot using patient sera. The mean number of doses ingested per course of PEP was significantly higher for HCWs in the EIA group (3.8; range, 0 to 6) compared with the OraQuick group (1.2; range, 0 to 3; P = .016). Cost analysis revealed a mean savings of $6.62 with the OraQuick test per occupational exposure. Although the survey failed to detect an overall reduction in HCW stress reaction symptoms using OraQuick for source-patient testing, 11 HCWs in the EIA group had repetitive thoughts of the exposure compared with 5 in the OraQuick group (P= .049).

Because of the reduction in ingested doses of unnecessary PEP and reduced cost of occupational exposure management with their use, rapid HJV-antibody tests should be the preferred method for source-patient testing following an occupational exposure.

To identify infection control policies and practices used by long-term-care facilities (LTCFs) in Iowa for residents with methicillin-resistant Staphylococcus aureus (MRSA) or vancomycin-resistant enterococci (VRE), and to estimate the prevalence of residents known to have these organisms.

Survey.

LTCFs in Iowa from December 2002 through March 2003.

Of the 429 LTCFs in Iowa, 331 (77%) responded to the survey. The estimated prevalence of residents known to have MRSA was 13.4 per 1,000 and that of residents known to have VRE was 2.3 per 1,000. Facilities owned by the government or those with an average of more than 86 occupied beds were more likely to have residents known to have MRSA and VRE (P = .002 and .007, respectively). Of the responding facilities, 7.3% acknowledged that they refused to accept individuals known to have MRSA and 16.9% acknowledged that they refused to accept those known to have VRE. Facilities in large communities (population, > 100,000) were least likely to deny admission to an individual known to have either MRSA or VRE (P = .05). Most facilities reported adhering to the national guidelines, but fewer than half (44.7%) of the respondents had heard of the Iowa Antibiotic Resistance Task Force's guidelines regarding residents with MRSA or VRE.

Many LTCFs in Iowa care for residents known to have MRSA or VRE, but some refuse to admit these individuals. Infection control personnel and public health officials should work together to educate LTCF staff so that residents receive proper care and resistant organisms do not spread within this setting.

To examine the extent to which the strategies recommended by the National Foundation for Infectious Diseases (NFID)-Centers for Disease Control and Prevention (CDC) co-sponsored workshop, Antimicrobial Resistance in Hospitals: Strategies to Improve Antimicrobial Use and Prevent Nosocomial Transmission of Antimicrobial-Resistant Microorganisms, have been implemented and the relationship between the degree of implementation and hospital culture, leadership, and organizational factors.

Survey.

A representative sample of U.S. hospitals stratified by teaching status, bed size, and geographic region.

Infection control professionals.

Surveyed hospitals had implemented strategies to optimize the use of antimicrobials and to detect, report, and prevent transmission of antimicrobial-resistant microorganisms. Multivariate analyses found that hospitals with a greater degree of implementation of the NFID–CDC strategic goals were more likely to have management support, education of staff, and interdisciplinary groups specifically to address these issues; they were also more likely to engage in benchmarking on broader quality of care indicators.

Most surveyed hospitals had implemented some measures to address the NFID–CDC recommendations; however, hospitals need to do much more to improve antimicrobial use and to increase their efforts to detect, report, and control the spread of antimicrobial resistance. A supportive hospital administration must foster a culture of ongoing support, education, and interdisciplinary work groups focused on this important issue to successfully accomplish these goals.