We assessed the frequency and relatedness of methicillin-resistant Staphylococcus aureus (MRSA) isolates to determine whether healthcare workers, the environment, or admitted patients could be a reservoir for MRSA on a burn trauma unit (BTU). We also assessed risk factors for MRSA colonization among BTU patients.

Prospective cohort study and surveillance for MRSA carriage.

BTU of a Midwestern academic medical center.

Patients admitted to a BTU from February 2009 through January 2010 and healthcare workers on this unit during the same time period.

Samples for MRSA culture were collected on admission from the nares and wounds of all BTU patients. We also had collected culture samples from the throat, axilla, antecubital fossa, groin, and perianal area of 12 patients per month. Samples collected from healthcare workers' nares and from environmental sites were cultured quarterly. MRSA isolates were typed by pulsed-field gel electrophoresis.

Of 144 patients, 24 (17%) carried MRSA in their nares on admission. Male sex (odds ratio [OR], 5.51; 95% confidence interval [95% CI], 1.25–24.30), admission for necrotizing fasciitis (OR, 7.66; 95% CI, 1.64–35.81), and MRSA colonization of a site other than the nares (OR, 23.40; 95% CI, 6.93–79.01) were independent predictors of MRSA nasal carriage. Cultures of samples collected from 4 healthcare workers and 4 environmental cultures had positive results. Two patients were colonized with strains that were indistinguishable from strains collected from a healthcare worker or the environment.

Patients were a major reservoir for MRSA. Infection control efforts should focus on preventing transmission of MRSA from patients who are MRSA carriers to other patients on the unit.

To examine potential clinical outcomes and cost of active methicillin-resistant Staphylococcus aureus (MRSA) surveillance with and without decolonization in neonatal intensive care units (NICUs) from the perspective of healthcare providers in Hong Kong.

Decision analysis modeling.

NICU.

Hypothetical cohort of patients admitted to an NICU.

We designed a decision tree to simulate potential outcomes of active MRSA surveillance with and without decolonization in patients admitted to an NICU. Outcome measures included total direct medical cost per patient, MRSA infection rate, and MRSA-associated mortality rate. Model inputs were derived from the literature. Sensitivity analyses evaluated the impact of uncertainty in all model variables.

In the base-case analysis, active surveillance plus decolonization showed a lower expected MRSA infection rate (0.911% vs 1.759%), MRSA-associated mortality rate (0.223% vs 0.431%), and total cost per patient (USD 47,294 vs USD 48,031) compared with active surveillance alone. Sensitivity analyses showed that active surveillance plus decolonization cost less and had lower event rates if the incidence risk ratio of acquiring MRSA infections in carriers after decolonization was less than 0.997. In 10,000 Monte Carlo simulations, active surveillance plus decolonization was significantly less costly than active surveillance alone 99.9% of the time, and both the MRSA infection rate and the MRSA-associated mortality rate were significantly lower 99.9% of the time.

Active surveillance plus decolonization for patients admitted to NICUs appears to be cost saving and effective in reducing the MRSA infection rate and the MRSA-associated mortality rate if addition of decolonization to active surveillance reduces the risk of MRSA infection.

Infect Control Hosp Epidemiol 2012;33(10):1024-1030

Effective implementation of infection control programs and adherence to standard precautions are challenging in resource-limited settings. The objective of this study was to describe infection control knowledge, attitudes, and practices among healthcare workers (HCWs) in Uganda.

We conducted a survey of hospital employees who had direct contact with patients or their immediate environment. We also performed an environmental assessment of resource availability and utilization within hospital wards.

Surgical, medicine, and obstetrics wards at a national referral hospital in Kampala, Uganda.

One hundred eighty-three randomly selected HCWs.

Almost all HCWs knew to wash their hands, although nursing and support staff were less likely to perceive that HCWs' hands can be a vector of disease transmission. Hand washing was valued more as a means of self-protection than as a means to prevent patient-to-patient transmission, consistent with the prevailing belief that infection control was important for occupational safety. Sinks were not readily accessible, and soap at sinks was uncommon throughout the medicine and obstetrics wards but more commonly available in the surgery wards. Alcohol gel was rarely available.

Changing infection control practices in developing countries will require a multifaceted approach that addresses resource availability, occupational safety, and local understanding and attitudes about infection control.

Staphylococcus aureus is a cause of community- and healthcare-acquired infections and is associated with substantial morbidity, mortality, and costs. Vancomycin minimum inhibitory concentrations (MICs) among S. aureus have increased, and reduced vancomycin susceptibility (RVS) may be associated with treatment failure. We aimed to identify clinical risk factors for RVS in S. aureus bacteremia.

Case-control.

Academic tertiary care medical center and affiliated urban community hospital.

Cases were patients with RVS S. aureus isolates (defined as vancomycin E-test MIC >1.0 μg/mL). Controls were patients with non-RVS S. aureus isolates.

Of 392 subjects, 134 (34.2%) had RVS. Fifty-eight of 202 patients (28.7%) with methicillin-susceptible S. aureus (MSSA) isolates had RVS, and 76 of 190 patients (40.0%) with methicillin-resistant S. aureus (MRSA) isolates had RVS (P = .02). In unadjusted analyses, prior vancomycin use was associated with RVS (odds ratio [OR], 2.08; 95% confidence interval [CI], 1.00–4.32; P = .046). In stratified analyses, there was significant effect modification by methicillin susceptibility on the association between vancomycin use and RVS (P = .04). In multivariate analyses, after hospital of admission and prior levofloxacin use were controlled for, the association between vancomycin use and RVS was significant for patients with MSSA infection (adjusted OR, 4.02; 95% CI, 1.11–14.50) but not MRSA infection (adjusted OR, 0.87; 95% CI, 0.36–2.13).

A substantial proportion of patients with S. aureus bacteremia had RVS. The association between prior vancomycin use and RVS was significant for patients with MSSA infection but not MRSA infection, suggesting a complex relationship between the clinical and molecular epidemiology of RVS in S. aureus.

Infect Control Hosp Epidemiol 2012;33(2):160-166

To examine the effect of restricting ciprofloxacin on the resistance of nosocomial gram-negative bacilli, including Pseudomonas aeruginosa, to antipseudomonal carbapenems.

Interrupted time-series analysis.

Tertiary care teaching hospital with 11 intensive care and intermediate care units with a total of 295 beds.

All nosocomial isolates of P. aeruginosa.

Restriction of ciprofloxacin.

There was a significant decreasing trend observed in the percentage (P = .0351) and the rate (P = .0006) of isolates of P. aeruginosa that were resistant to antipseudomonal carbapenems following the restriction of ciprofloxacin. There was also a significant decreasing trend observed in the percentage (P = .0017) and the rate (P = .0001) of isolates of ciprofloxacin-resistant P. aeruginosa. The rate of cefepime-resistant P. aeruginosa isolates declined (P = .004 ) but the percentage of cefepime-resistant P. aeruginosa isolates did not change. There were no significant changes observed in the rate or the percentage of piperacillin-tazobactam-resistant P. aeruginosa isolates. There were no significant changes observed in the susceptibilities of nosocomial Enterobacteriaciae or Acinetobacter baumannii isolates that were resistant to carbapenems. Over the study period there was a significant increase in the use of carbapenems (P = .0134); the use of ciprofloxacin decreased significantly (P = .0027). There were no significant changes in the use of piperacillin-tazobactam or cefepime.

Restriction of ciprofloxacin was associated with a decreased resistance of P. aeruginosa isolates to antipseudomonal carbapenems and ciprofloxacin in our hospital's intermediate care and intensive care units. There were no changes observed in the susceptibilities of nosocomial Enterobacteriaciae or A. baumannii to carbapenems, despite increased carbapenem use. Reducing ciprofloxacin use may be a means of controlling multidrug-resistant P. aeruginosa.

Carbapenem-resistant Enterobacteriaceae (CRE) are rapidly emerging worldwide. Control group selection is critically important when analyzing predictors of antimicrobial resistance. Focusing on modifiable risk factors can optimize prevention and resource expenditures. To identify specific predictors of CRE, patients with CRE were compared with 3 control groups: (1) patients with extended-spectrum β-lactamase (ESBL)-producing Enterobacteriaceae, (2) patients with non-ESBL-containing Enterobacteriaceae, and (3) uninfected controls.

Matched multivariable analyses.

Patients possessing CRE that were isolated at Detroit Medical Center from September 1, 2008, to August 31, 2009.

Patients were matched (1:1 ratio) to the 3 sets of controls. Matching parameters included (1) bacteria type, (2) hospital/ facility, (3) unit/clinic, (4) calendar year, and (5) time at risk (ie, from admission to culture). Matched multivariable analyses were conducted between uninfected controls and patients with CRE, ESBL, and non-ESBL Enterobacteriaceae. Models were also designed comparing patients with CRE to patients with ESBL, patients with non-ESBL Enterobacteriaceae, and all 3 non-CRE groups combined.

Ninety-one unique patients with CRE were identified, and 6 matched models were constructed. Recent (less than 3 months) exposure to antibiotics was the only parameter that was consistently associated with CRE, regardless of the group to which CRE was compared, and was not independently associated with isolation of ESBL or non-ESBL Enterobacteriaceae.

Exposure to antibiotics within 3 months was an independent predictor that characterized patients with CRE isolation. As a result, antimicrobial stewardship efforts need to become a major focus of preventive Interventions. Regulatory focus regarding appropriate antimicrobial use might decrease the detrimental effects of antibiotic misuse and spread of CRE.

Systematic review and meta-analysis of randomized controlled trials and quasi-experimental studies to assess the efficacy of daily bathing with chlorhexidine (CHG) for prevention of healthcare-associated bloodstream infections (BSIs).

Medical, surgical, trauma, and combined medical-surgical intensive care units (ICUs) and long-term acute care hospitals.

Inpatients.

Data on patient population, diagnostic criteria for BSIs, form and concentration of topical CHG, incidence of BSIs, and study design were extracted.

One randomized controlled trial and 11 nonrandomized controlled trials reporting a total of 137,392 patient-days met the inclusion criteria; 291 patients in the CHG arm developed a BSI over 67,775 patient-days, compared with 557 patients in the control arm over 69,617 catheter-days. CHG bathing resulted in a reduced incidence of BSIs: the pooled odds ratio using a random-effects model was 0.44 (95% confidence interval, 0.33–0.59; P< .00001). Statistical heterogeneity was moderate, with an I2 of 58%. For the subgroup of studies that examined central line–associated BSIs, the odds ratio was 0.40 (95% confidence interval, 0.27–0.59).

Daily bathing with CHG reduced the incidence of BSIs, including central line-associated BSIs, among patients in the medical ICU. Further studies are recommended to determine the optimal frequency, method of application, and concentration of CHG as well as the comparative effectiveness of this strategy relative to other preventive measures available for reducing BSIs. Future studies should also examine the efficacy of daily CHG bathing in non-ICU populations at risk for BSI.

Infect Control Hosp Epidemiol 2012;33(3):257-267

We examined surveillance data to describe the epidemiology of tuberculosis (TB) among healthcare workers (HCWs) in the United States during the period 1995–2007.

Cross-sectional descriptive analysis of existing surveillance data.

TB cases reported to the Centers for Disease Control and Prevention from the 50 states and the District of Columbia from 1995 through 2007.

Of the 200,744 reported TB cases in persons 18 years of age or older, 6,049 (3%) occurred in individuals who were classified as HCWs. HCWs with TB were more likely than other adults with TB to be women (unadjusted odds ratio [95% confidence interval], 4.1 [3.8–4.3]), be foreign born (1.3 [1.3–1.4]), have extrapulmonary TB (1.6 [1.5–1.7]), and complete TB treatment (2.5 [2.3–2.8]).

Healthcare institutions may benefit from intensifying TB screening of HCWs upon hire, especially persons from countries with a high incidence of TB, and encouraging treatment for latent TB infection among HCWs to prevent progression to TB disease.

To evaluate the impact of postprescription review of broad-spectrum antimicrobial (study-ABX) agents on rates of antimicrobial use.

Quasi-experimental before-after study.

Five academic medical centers.

Adults receiving at least 48 hours of study-ABX.

The baseline, intervention, and follow-up periods were 6 months each in 2 units at each of 5 sites. Adults receiving at least 48 hours of study-ABX entered the cohort as case-patients. During the intervention, infectious-diseases physicians reviewed the cases after 48 hours of study-ABX. The provider was contacted with alternative recommendations if antimicrobial use was considered to be unjustified on the basis of predetermined criteria. Acceptance rates were assessed 48 hours later. The primary outcome measure was days of study-ABX per 1,000 study-patient-days in the baseline and intervention periods.

There were 1,265 patients in the baseline period and 1,163 patients in the intervention period. Study-ABX use decreased significantly during the intervention period at 2 sites: from 574.4 to 533.8 study-ABX days/1,000 patient-days (incidence rate ratio [IRR], 0.93; 95% confidence interval [CI], 0.88-0.97; P = .002) at hospital В and from 615.6 to 514.4 study-ABX days/1,000 patient-days (IRR, 0.83; 95% CI, 0.79-0.88; P < .001) at hospital D. Both had established antimicrobial stewardship programs (ASP). Study-ABX use increased at 2 sites and stayed the same at 1 site. At all institutions combined, 390 of 1,429 (27.3%) study-ABX courses were assessed as unjustified; recommendations to modify or stop therapy were accepted for 260 (66.7%) of these courses.

Postprescription review of study-ABX decreased antimicrobial utilization in some of the study hospitals and may be more effective when performed as part of an established ASP.

The impact of antimicrobial scrubs on healthcare worker (HCW) bacterial burden is unknown.

To determine die effectiveness of antimicrobial scrubs on hand and apparel bacterial burden.

Prospective, crossover trial.

Thirty HCWs randomized to study versus control scrubs in an intensive care unit.

Weekly microbiology samples were obtained from scrub abdominal area, cargo pocket, and hands. Mean log colony-forming unit (CFU) counts were calculated. Compliance with hand hygiene practices was measured. Apparel and hand mean log CFU counts were compared.

Adherence measures were 78% (910/1,173) for hand hygiene and 82% (223/273) for scrubs. Culture compliance was 67% (306/460). No differences were observed in bacterial hand burden or in HCWs with unique positive scrub cultures. No difference in vancomycin-resistant enterococci (VRE) and gram-negative rod (GNR) burden was observed. A difference in mean log mediicillin-resistant Staphylococcus aureus (MRSA) CFU count was found between study and control scrubs for leg cargo pocket (mean log CFUs, 11.84 control scrub vs 6.71 study scrub; P = .0002), abdominal area (mean log CFUs, 11.35 control scrub vs 7.54 study scrub; P = .0056), leg cargo pocket at die beginning of shift (mean log CFUs, 11.96 control scrub vs 4.87 study scrub; P = .0028), and abdominal area pocket at die end of shift (mean log CFUs, 12.14 control scrubs vs 8.22 study scrub; P = .0054).

Study scrubs were associated witfi a 4–7 mean log reduction in MRSA burden but not VRE or GNRs. A prospective trial is needed to measure die impact of antimicrobial impregnated apparel on MRSA transmission rates.

Infect Control Hosp Epidemiol 2012;33(3):268-275

To evaluate the seroprevalence and risk factors for 2009 influenza A (H1N1) virus infection among healthcare personnel.

Observational cross-sectional study.

Healthcare workers (HCWs) in an acute care hospital.

Between September 14 and October 4, 2009, before 2009 H1N1 vaccination, we collected serological samples from 461 healthy HCWs. Hemagglutination-inhibition antibody assays were conducted. To evaluate the risk factors of seropositivity for 2009 H1N1 virus, gender, age, profession, work department, usage of personal protective equipment, and seasonal influenza vaccination status data were gathered via questionnaires.

Our survey showed that doctors and nurses were at highest risk of seropositivity for the 2009 H1N1 virus (odds ratio [OR], 5.25 [95% confidence interval {CI}, 1.21–22.7]). An increased risk of seropositivity was observed among pediatric, emergency room, and internal medicine staff (adjusted OR, 1.98 [95% CI, 1.07–3.65]). Risk was also higher among HCWs who had high titers of antibodies against the seasonal H1N1 virus (adjusted OR, 1.59 [95% CI, 1.02–2.48]).

Seropositivity for the 2009 H1N1 virus was associated with occupational risk factors among HCWs.

Infect Control Hosp Epidemiol 2012;33(1):58-62

The purpose of this study was to describe patterns of infection or colonization with antibiotic-resistant gram-negative bacilli (GNB) in hospitalized children utilizing an electronic health record.

Tertiary care facility.

Pediatric patients 18 years of age or younger hospitalized from January 1, 2006, to December 31, 2008.

Children were identified who had (1) at least 1 positive culture for a multidrug-resistant (MDR) GNB, defined as a GNB with resistance to 3 or more antibiotic classes; or (2) additive drug resistance, defined as isolation of more than 1 GNB that collectively as a group demonstrated resistance to 3 or more antibiotic classes over the study period. Differences in clinical characteristics between the 2 groups were ascertained, including history of admissions and transfers, comorbid conditions, receipt of procedures, and antibiotic exposure.

Of 56,235 pediatric patients, 46 children were infected or colonized with an MDR GNB, of which 16 were resistant to 3 classes and 30 were resistant to 4 classes. Another 39 patients had positive cultures for GNB that exhibited additive drug resistance. Patients with additive drug resistance were more likely than patients with MDR GNB to have had previous admissions to a long-term facility (8 vs 2; P = .04) and had more mean admissions (7 vs 3; P <.01) and more mean antibiotic-days (P < .01 to P = .02). Six patients with additive drug resistance later had a positive culture with an MDR GNB.

An electronic health record can be used to track antibiotic class resistance in GNB isolated from hospitalized children over multiple cultures and hospitalizations.

To evaluate hospital characteristics and practices used by Thai hospitals to prevent catheter-associated urinary tract infection (CAUTI), central line-associated bloodstream infection (CLABSI), and ventilator-associated pneumonia (VAP), the 3 most common types of healthcare-associated infection (HAI) in Thailand.

Survey.

Thai hospitals with an intensive care unit and 250 or more hospital beds

Between January 1, 2010, and October 31, 2010, research nurses collected data from all eligible hospitals. The survey assessed hospital characteristics and practices to prevent CAUTI, CLABSI, and VAP. Ordinal logistic regression was used to assess relationships between hospital characteristics and use of prevention practices.

A total of 204 (80%) of 256 hospitals responded. Most hospitals (93%) reported regularly using alcohol-based hand rub. The most frequently reported prevention practice by infection was as follows: for CAUTI, condom catheters in men (47%); for CLABSI, avoiding routine central venous catheter changes (85%); and for VAP, semirecumbent positioning (84%). Hospitals with peripherally inserted central catheter insertion teams were more likely to regularly use elements of the CLABSI prevention bundle. Greater safety scores were associated with regular use of several VAP prevention practices. The only hospital characteristic associated with increased use of at least 1 prevention practice for each infection was membership in an HAI collaborative.

While reported adherence to hand hygiene was high, many of the prevention practices for CAUTI, CLABSI, and VAP were used infrequently in Thailand. Policies and interventions emphasizing specific infection prevention practices, establishing a strong institutional safety culture, and participating in collaboratives to prevent HAI may be beneficial.

The emergence of Staphylococcus aureus with reduced susceptibility to glycopeptides (glycopeptide-intermediate S. aureus [GISA] and heterogeneous GISA [h-GISA]) leads to intensive care unit (ICU) outbreaks that frequently result in ward closure. We investigated the role of hospital hygiene in the transmission and eradication of an h-GISA outbreak.

The study is a description of an original environmental investigation around a series of 12 cases.

The outbreak occurred in a 20-bed polyvalent/trauma ICU in a 2,800-bed tertiary care university hospital in France.

Specimens were obtained for surveillance and diagnostic cultures from all patients in the unit. Surface sampling was also performed. Geographic cohorting, contact isolation, emphasis on adherence to infection control practices, and environmental cleaning were implemented.

Twelve patients with h-GISA infection (n = 5) or colonization (n = 7) were identified. The mean interval between admission and h-GISA detection was 23.6 days (range, 10–89 days), with a median of 16.5 days. Environmental investigation identified an unexpected reservoir, namely, SpO2 sensors. The outbreak was controlled by a combination of measures, including eradication of this reservoir, avoiding total ward closure.

Targeted surface sampling helps to secure the environment through active investigation of various reservoirs while maintaining normal activity on the ward. In our study, this method led to the detection of an unsuspected reservoir, the eradication of which helped control the h-GISA epidemic. Further applications of this original investigative procedure should allow confirmation of its relevance and efficiency.

Infect Control Hosp Epidemiol 2012;33(2):167-174

In neonatal intensive care units (NICUs), monitoring hospital-acquired bloodstream infection (BSI) is critical to alert clinicians to variations in the incidence of infection between units and over time. We demonstrate a toolkit of monitoring techniques that account for case mix and could be implemented using routinely available clinical data. This toolkit could enable quality of care comparisons between hospitals to facilitate the sharing of improved practices.

Prospective study over 4 years.

Babies admitted to 2 tertiary London NICUs.

We derived expected numbers of BSI episodes using a Poisson regression risk model adjusting for variations in birth weight, transfers to the NICU from other hospitals, postnatal age, and days spent at each National Health Service level of care. We compared observed and expected numbers of BSI episodes using 2 monitoring techniques: standardized infection ratios (SIRs) and the sequential probability ratio test (SPRT).

Using the SIR method, observed BSI incidence increased over expected incidence in 2002 at both NICUs, but this increase did not reach statistical significance at the 1% level. Using the SPRT method, neither unit showed a clinically important increase or decrease, defined as a 30% deviation from expected incidence.

Risk-adjusted BSI monitoring can be performed using routine hospital data. NICUs could use SIRs for an annual look back at infection incidence and SPRTs for prospective, quarterly monitoring. The SIR and the SPRT methods have different strengths, and both could help clinicians improve infection control and patient care in NICUs.

To describe the epidemiology and healthcare costs of Clostridium difficile infection (CDI) identified in the outpatient setting.

Population-based, retrospective cohort study.

Kaiser Permanente Colorado and Kaiser Permanente Northwest members between June 1, 2005, and September 30, 2008.

We identified persons with incident CDI and classified CDI by whether it was identified in the outpatient or inpatient healthcare setting. We collected information about baseline variables and follow-up healthcare utilization, costs, and outcomes among patients with CDI. We compared characteristics of patients with CDI identified in the outpatient versus inpatient setting.

We identified 3,067 incident CDIs; 56% were identified in the outpatient setting. Few strong, independent predictors of diagnostic setting were identified, although a previous stay in a nonacute healthcare institution (odds ratio [OR], 1.45 [95% confidence interval (CI), 1.13-1.86]) was statistically associated with outpatient-identified CDI, as was age from 50 to 59 years (OR, 1.64 [95% CI, 1.18-2.29]), 60 to 69 years (OR, 1.37 [95% CI, 1.03-1.82]), and 70 to 79 years (OR, 1.36 [95% CI, 1.06-1.74]), when compared with persons aged 80-89 years.

We found that more than one-half of incident CDIs in this population were identified in the outpatient setting. Patients with outpatient-identified CDI were younger with fewer comorbidities, although they frequently had previous exposure to healthcare. These data suggest that practitioners should be aware of CDI and obtain appropriate diagnostic testing on outpatients with CDI symptoms.

Infect Control Hosp Epidemiol 2012;33(10):1031-1038

The main objectives of our study were to explore reasons for seasonal influenza vaccine acceptance and declination in employees of a large integrated healthcare system and to identify underlying constructs that influence acceptance versus declination. Secondary objectives were to determine whether vaccine acceptance varied by hospital location and to identify facility-level measures that explained variability.

A national health promotion survey of employees was conducted that included items on vaccination in the 2009-2010 influenza season. The survey was administered with two other institutional surveys in a stratified fashion: approximately 40% of participating employees were randomly assigned to complete the health promotion survey.

National single-payer healthcare system with 152 hospitals.

Employees of the healthcare system in 2010 who responded to the survey.

Factor analysis was used to identify underlying constructs that influenced vaccine acceptance versus declination. Mean factor scores were examined in relation to demographic characteristics and occupation. Multilevel logistic regression models were used to determine whether vaccine acceptance varied by location and to identify facility-level measures that explained variability.

Four factors were identified related to vaccine declination and were labeled as (1) “don't care,” (2) “don't want,” (3) “don't believe,” and (4) “don't know.” Significant differences in mean factor scores existed by demographic characteristics and occupation. Vaccine acceptance varied by location, and vaccination rates in the previous year were an important facility-level predictor.

Results should guide interventions that tailor messages on the basis of particular reasons for declination. Occupation-specific and culturally appropriate messaging should be considered. Continued efforts will be taken to better understand how workplace context influences vaccine acceptance.

Antimicrobial stewardship programs are being implemented in health care to reduce inappropriate antimicrobial use, adverse events, Clostridium difficile infection, and antimicrobial resistance. There is no standardized approach to evaluate the impact of these programs.

To use a structured panel process to define quality improvement metrics for evaluating antimicrobial stewardship programs in hospital settings that also have the potential to be used as part of public reporting efforts.

A multiphase modified Delphi technique.

Paper-based survey supplemented with a 1-day consensus meeting.

A 10-member expert panel from Canada and the United States was assembled to evaluate indicators for relevance, effectiveness, and the potential to aid quality improvement efforts.

There were a total of 5 final metrics selected by the panel: (1) days of therapy per 1000 patient-days; (2) number of patients with specific organisms that are drug resistant; (3) mortality related to antimicrobial-resistant organisms; (4) conservable days of therapy among patients with community-acquired pneumonia (CAP), skin and soft-tissue infections (SSTI), or sepsis and bloodstream infections (BSI); and (5) unplanned hospital readmission within 30 days after discharge from the hospital in which the most responsible diagnosis was one of CAP, SSTI, sepsis or BSI. The first and second indicators were also identified as useful for accountability purposes, such as public reporting.

We have successfully identified 2 measures for public reporting purposes and 5 measures that can be used internally in healthcare settings as quality indicators. These indicators can be implemented across diverse healthcare systems to enable ongoing evaluation of antimicrobial stewardship programs and complement efforts for improved patient safety.