eFigure 1. Flow Diagram Depicting Patients Treated for Community-Acquired Pneumonia (CAP) Who Were Included in (N = 219) or Excluded From (N = 211) the Analysis
eFigure 2. Flow Diagram Depicting Patients Treated for Present-on-Admission Urinary Tract Infection (UTI) Who Were Included in (N = 452) or Excluded From (N = 394) the Analysis
eFigure 3. Flow Diagram Depicting Fluoroquinolone (FQ) Patients Included in (N = 550) and Excluded From (N = 518) the Analysis
eFigure 4. Flow Diagram Depicting Patients Receiving Intravenous Vancomycin Treatment (VANC) Who Were Included in (N = 403) or Excluded From (N = 709) Analysis
eFigure 5. Community-Acquired Pneumonia (CAP) Analysis Pathway
eFigure 6. Present-on-Admission Urinary Tract Infection (UTI) Analysis Pathway
eFigure 7. Fluoroquinolone Treatment (FQ) Analysis Pathway
eFigure 8. Intravenous Vancomycin Treatment (VANC) Analysis Pathway
eTable. Summary of Antimicrobial Prescribing Quality Across AQUA Events
Customize your JAMA Network experience by selecting one or more topics from the list below.
Identify all potential conflicts of interest that might be relevant to your comment.
Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.
Err on the side of full disclosure.
If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.
Not all submitted comments are published. Please see our commenting policy for details.
Magill SS, O’Leary E, Ray SM, et al. Assessment of the Appropriateness of Antimicrobial Use in US Hospitals. JAMA Netw Open. 2021;4(3):e212007. doi:10.1001/jamanetworkopen.2021.2007
What percentage of hospital antimicrobial use in the US deviates from recommended practices, such as treatment selection or duration, on the basis of medical record documentation?
In this cross-sectional study of 1566 patients at 192 hospitals, antimicrobial use deviated from recommended practices for 55.9% of patients who received antimicrobials for community-acquired pneumonia or urinary tract infection present at admission or who received fluoroquinolone or intravenous vancomycin treatment.
The findings suggest that standardized assessments of hospital antimicrobial prescribing quality can be used to estimate the appropriateness of antimicrobial use in large groups of hospitals.
Hospital antimicrobial consumption data are widely available; however, large-scale assessments of the quality of antimicrobial use in US hospitals are limited.
To evaluate the appropriateness of antimicrobial use for hospitalized patients treated for community-acquired pneumonia (CAP) or urinary tract infection (UTI) present at admission or for patients who had received fluoroquinolone or intravenous vancomycin treatment.
Design, Setting, and Participants
This cross-sectional study included data from a prevalence survey of hospitalized patients in 10 Emerging Infections Program sites. Random samples of inpatients on hospital survey dates from May 1 to September 30, 2015, were identified. Medical record data were collected for eligible patients with 1 or more of 4 treatment events (CAP, UTI, fluoroquinolone treatment, or vancomycin treatment), which were selected on the basis of common infection types reported and antimicrobials given to patients in the prevalence survey. Data were analyzed from August 1, 2017, to May 31, 2020.
Antimicrobial treatment for CAP or UTI or with fluoroquinolones or vancomycin.
Main Outcomes and Measures
The percentage of antimicrobial use that was supported by medical record data (including infection signs and symptoms, microbiology test results, and antimicrobial treatment duration) or for which some aspect of use was unsupported. Unsupported antimicrobial use was defined as (1) use of antimicrobials to which the pathogen was not susceptible, use in the absence of documented infection signs or symptoms, or use without supporting microbiologic data; (2) use of antimicrobials that deviated from recommended guidelines; or (3) use that exceeded the recommended duration.
Of 12 299 patients, 1566 patients (12.7%) in 192 hospitals were included; the median age was 67 years (interquartile range, 53-79 years), and 864 (55.2%) were female. A total of 219 patients (14.0%) were included in the CAP analysis, 452 (28.9%) in the UTI analysis, 550 (35.1%) in the fluoroquinolone analysis, and 403 (25.7%) in the vancomycin analysis; 58 patients (3.7%) were included in both fluoroquinolone and vancomycin analyses. Overall, treatment was unsupported for 876 of 1566 patients (55.9%; 95% CI, 53.5%-58.4%): 110 of 403 (27.3%) who received vancomycin, 256 of 550 (46.5%) who received fluoroquinolones, 347 of 452 (76.8%) with a diagnosis of UTI, and 174 of 219 (79.5%) with a diagnosis of CAP. Among patients with unsupported treatment, common reasons included excessive duration (103 of 174 patients with CAP [59.2%]) and lack of documented infection signs or symptoms (174 of 347 patients with UTI [50.1%]).
Conclusions and Relevance
The findings suggest that standardized assessments of hospital antimicrobial prescribing quality can be used to estimate the appropriateness of antimicrobial use in large groups of hospitals. These assessments, performed over time, may inform evaluations of the effects of antimicrobial stewardship initiatives nationally.
Optimizing antimicrobial use is critical to slowing the spread of resistant pathogens. In 2014, the US Centers for Disease Control and Prevention (CDC) called for acute care hospitals to implement antimicrobial stewardship programs with the goal of improving antimicrobial use to optimize infection cure rates and minimize harms.1 In 2014 and 2015, the White House released the US National Strategy and Action Plan for Combating Antibiotic-Resistant Bacteria, which established antibiotic stewardship outcomes to accomplish by 2020, including a 20% reduction in inappropriate inpatient antibiotic use for monitored conditions and medications.2,3 National initiatives have bolstered stewardship efforts in recent years, and data from the CDC’s National Healthcare Safety Network have shown increases in the percentage of hospitals with comprehensive antimicrobial stewardship programs.1
Efforts to evaluate antimicrobial stewardship programs’ effect on hospital antimicrobial use typically focus on volume rather than prescribing quality4-6; it is not clear whether the volume of antimicrobial use correlates with appropriateness.7 Prescribing decisions for hospitalized patients are associated with many factors, including comorbidities, allergies, adverse effects, and drug interactions. In addition, the lack of current national treatment guidelines for some infections makes evaluating the appropriateness of US hospital antimicrobial use challenging. Hospital antimicrobial stewards often perform intensive, small-scale medication use evaluations to answer specific questions about appropriateness. Larger-scale evaluations are more difficult to conduct.
We developed and implemented a multicenter objective data collection as part of a hospital prevalence survey of health care–associated infections and antimicrobial use conducted by the CDC’s Emerging Infections Program in 2015. We used these data to assess the appropriateness of antimicrobial use for selected prescribing events in a large group of hospitals and to establish a baseline to which data from subsequent surveys could be compared for estimation of the association of national antimicrobial stewardship interventions with the appropriateness of antimicrobial use at these hospitals.
This study used data collected by the Emerging Infections Program, which conducted cross-sectional prevalence surveys of health care–associated infections and antimicrobial use in 2011 and 2015 at selected hospitals in 10 states (California, Colorado, Connecticut, Georgia, Maryland, Minnesota, New Mexico, New York, Oregon, and Tennessee); methods and results have been published previously.8-11 Each hospital selected a survey date between May 1 and September 30, 2015. Patients were randomly selected from the census on the morning of the survey date.8-11 The human subjects advisor in the CDC’s National Center for Emerging and Zoonotic Infectious Diseases determined that the survey was a nonresearch public health activity. Emerging Infections Program sites and hospitals determined that the survey was a nonresearch activity or approved it with an informed consent waiver. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.
Data collected in the 2011 survey identified 4 common antimicrobial prescribing events for assessment in the 2015 survey: 2 infection-based events, including treatment of community-acquired pneumonia (CAP) or treatment of urinary tract infection (UTI) present at admission, and 2 antimicrobial-based events, including treatment with fluoroquinolones (FQ) or treatment with intravenous vancomycin (VANC). Medical record abstraction forms were developed for each event to compose the antimicrobial quality assessment (AQUA) component of the 2015 survey. Emerging Infections Program staff, who were not required to be clinicians or antimicrobial stewards, reviewed medical records retrospectively to collect information on comorbidities, health care exposures, antimicrobial allergies, illness severity, infections during the hospitalization, microbiology and other test results, and treatment.
Patients were eligible for multiple AQUA data collections based on the antimicrobials given to the patient on the survey date or the previous day and the reported rationale for use. Patients were eligible for CAP data collection if they were 1 year or older, did not have certain underlying conditions or exposures, and were receiving antimicrobials for treatment of CAP on the survey date or the previous day. Patients were excluded from CAP data collection for the following reasons: (1) a stay in a nursing home, long-term care facility, or long-term acute care hospital before admission to the survey hospital; (2) hospitalization for 2 or more days in the 90 days before admission (other than the current admission); (3) receipt of intravenous antimicrobials, cancer chemotherapy, or wound care in the 30 days before admission; (4) requirement for long-term hemodialysis or mechanical ventilation at home; (5) diagnosis of cystic fibrosis or AIDS or another acquired or congenital immunodeficiency; (6) history of solid organ or hematopoietic stem cell transplant; and (7) treatment with high-dose corticosteroids or other immunosuppressive agents for more than 30 days. Patients were eligible for UTI data collection if they were 1 year or older and receiving antimicrobials on the survey date or the previous day for treatment of UTI present at the time of admission to the survey hospital. Patients were eligible for FQ data collection if they were 18 years or older and receiving FQ treatment on the survey date or the previous day, and patients were eligible for VANC data collection if they were 1 year or older and receiving VANC treatment on the survey date or the previous day.
Data were analyzed from August 1, 2017, to May 31, 2020. Final data downloaded from the prevalence survey data system on November 16, 2017, were analyzed using SAS, version 9.4 (SAS Institute Inc) or OpenEpi, version 3.01.12 AQUA analysis pathways were developed by CDC staff13-15 and refined from 2018 to 2020 with input from an antimicrobial stewardship expert group convened by The Pew Charitable Trusts.
Among patients eligible for AQUA data collection, a subset was eligible for analysis in which the finalized pathways were used. For the CAP pathway, the analysis included patients who (1) were 18 years or older, (2) had radiographic evidence of pneumonia during the first 5 hospital days, (3) had signs or symptoms of pneumonia during the first 2 hospital days, (4) received inpatient pneumonia treatment for 3 or more calendar days, and (5) did not have other infections reported during their hospitalizations. For the UTI pathway, analysis included patients who (1) were not pregnant and did not have neutropenia or a history of transplant, (2) received inpatient UTI treatment for 1 or more calendar days, and (3) did not have other infections reported. For the FQ pathway, analysis included patients with only 1 infection type who received FQ for 1 or more days. For the VANC pathway, analysis included patients with only 1 infection type who received VANC for more than 3 days (eMethods and eFigures 1-4 in the Supplement). We excluded patients from the VANC or FQ pathway if they were included in the CAP or UTI pathway.
Analysis pathways categorized the quality of antimicrobial use for each patient’s AQUA event(s) (eMethods in the Supplement). Because categorizations were based on data collected using standardized forms rather than clinical judgment at the time of medical record review, we used the terms supported and unsupported as proxies for appropriate and inappropriate or unnecessary use. Antimicrobial use was supported if there was medical record evidence that (1) treatment was clinically indicated for the infection for which the patient had a reported diagnosis, (2) antimicrobial selection was consistent with available guidelines or microbiology data, and (3) duration was consistent with recommendations in available guidelines.7,16-20 In cases involving more severe or complicated infections (eg, sterile site infections, sepsis, or infections due to selected pathogens such as mycobacteria), duration was not considered in the determination of whether prescribing was supported. Antimicrobial use for which some aspect was unsupported by medical record data (hereafter referred to as unsupported use) was defined as (1) use of antimicrobials to which the pathogen was not susceptible, use in the absence of signs or symptoms of infection, or use without supporting microbiologic data; (2) use of antimicrobials that deviated from guidelines; or (3) use that exceeded recommended duration.
Percentages of supported vs unsupported use and 95% CIs were calculated for each analysis pathway and across all pathways. For patients included in both the FQ and the VANC analysis pathways, discordant determinations were resolved to a single determination by 1 of the authors (S.S.M.).
Of 12 299 patients, 1566 patients (12.7%) in 192 hospitals were included in AQUA analyses; the median age was 67 years (interquartile range, 53-79 years), and 864 (55.2%) were female. Characteristics of patients included in each analysis pathway are shown in Table 1. Of 12 299 patients included in the survey, 6084 (49.5%) received antimicrobial medications on the survey date or day before, and 4476 of these patients (73.6%) received antimicrobial treatment for infection, as reported previously.11 Among 4476 patients receiving antimicrobial treatment for infection, 2680 (59.9%) were eligible for 1 or more AQUA data collections: 430 (9.6% of all patients receiving antimicrobial treatment for infection) in the CAP analysis pathway, 846 (18.9%) in the UTI pathway, 1112 (24.8%) in the VANC pathway, and 1068 (23.9%) in the FQ pathway (eFigures 1-4 in the Supplement). Of the 2680 patients eligible for AQUA data collection, 1566 were included in the analyses: 219 (14.0%) in the CAP pathway, 452 (28.9%) in the UTI pathway, 550 (35.1%) in the FQ pathway, and 403 (25.7%) in the VANC pathway.
Among 219 patients in the CAP pathway (Table 2 and eFigure 5 in the Supplement), antimicrobial prescribing was categorized as supported for 45 (20.5%; 95% CI, 15.6%-26.3%) and unsupported for 174 (79.5%; 95% CI, 73.7%-84.4%). Most patients with unsupported CAP treatment were treated for 8 or more days (103 of 174 [59.2%]) or received antimicrobials on inpatient treatment day 3 that were inconsistent with current guidelines (68 of 174 [39.1%]).
Among 452 patients in the UTI pathway (Table 3 and eFigure 6 in the Supplement), antimicrobial treatment was categorized as supported for 105 (23.2%; 95% CI, 19.5%-27.3%) and unsupported for 347 (76.8%; 95% CI, 72.7%-80.5%). Unsupported antimicrobial use was most commonly attributed to lack of documented signs or symptoms of UTI (174 of 347 [50.1%]), continued treatment without qualifying microbiologic evidence of infection (95 of 347 [27.4%]), or excessive treatment duration (74 of 347 [21.3%]).
Among 550 patients in the FQ pathway (Table 4 and eFigure 7 in the Supplement), antimicrobial prescribing was categorized as supported for 294 (53.5%; 95% CI, 49.3%-57.6%) and unsupported for 256 (46.5%; 95% CI, 42.4%-50.7%), most commonly because of FQ treatment for 8 or more days in patients with lower respiratory tract, abdominal, or gastrointestinal infections without supporting microbiologic data (161 of 256 [62.9%]).
Among 403 patients in the VANC pathway (Table 5 and eFigure 8 in the Supplement), antimicrobial use was categorized as supported for 293 (72.7%; 95% CI, 68.2%-76.9%) and unsupported for 110 (27.3%; 95% CI, 23.1%-31.8%). Unsupported treatment was commonly attributed to continuation of VANC in patients who did not appear to require it (54 of 110 [49.1%]), for example, patients without susceptible or likely susceptible pathogens identified from microbiologic testing or patients with cultures positive for pathogens susceptible to penicillin, ampicillin, or oxacillin and without a severe or unspecified penicillin allergy.
After exclusion of patients in the CAP or UTI pathway from the VANC and FQ pathways, 58 patients (3.7%) remained in multiple analysis pathways (VANC and FQ). Determinations in the 2 pathways were concordant for 32 of 58 patients (55.2%): 22 with supported and 10 with unsupported treatment. Discordant determinations (eg, unsupported for VANC and supported for FQ) were observed for 26 patients; after data for these patients were reviewed, 1 had an overall determination of supported treatment and 25, unsupported treatment. After discordant determinations were resolved, antimicrobial prescribing was determined to be supported for 690 of 1566 patients (44.1%; 95% CI, 41.6%-46.5%) and unsupported for 876 of 1566 patients (55.9%; 95% CI, 53.5%-58.4%) (eTable in the Supplement).
Among patients included in a multicenter hospital prevalence survey of health care–associated infections and antimicrobial use, a substantial percentage of CAP, UTI, FQ, and VANC treatment was unsupported by medical record data collected using a standardized approach (55.9% overall and as high as 79.5% for CAP). Common reasons for unsupported use included long duration, antimicrobial selection that deviated from guidelines, absence of documented signs or symptoms of infection, and lack of microbiologic evidence of infection.
Few recent, large studies have addressed inpatient antimicrobial prescribing quality.21-24 Comparison of our results with the results of these other studies is difficult because different approaches to data collection and different definitions of inappropriate or unnecessary prescribing of antimicrobials were used. In the other studies, antimicrobial prophylaxis and treatment were included and antimicrobial stewardship program personnel or other medical professionals collected the data and made determinations about antimicrobial prescribing quality.21-24 These studies also focused their assessments on antimicrobial prescriptions rather than infection syndromes. We focused solely on antimicrobials used to treat infections rather than including prophylaxis; did not require data collectors to have clinical or stewardship expertise; and used analysis pathways to categorize prescribing quality for 2 antimicrobial-based and 2 infection-based events.
Other studies have used terms such as inappropriate and suboptimal to describe prescribing quality but defined them in different ways. The use of multiple different definitions of appropriate and inappropriate prescribing is a particular challenge for hospital antimicrobial stewardship.7 Tribble et al22 considered suboptimal antimicrobial use to be inappropriate or appropriate with modification required; reasons included pathogen-drug mismatch, duplicate treatment (eg, 2 antimicrobials to cover anaerobes), unnecessary intravenous antimicrobial administration, overly broad coverage, and reasons classified as other. In contrast, the Australian Hospital National Antimicrobial Prescribing Survey defines inappropriate antimicrobial prescribing as being either suboptimal or inadequate.24 Suboptimal prescribing includes overly broad coverage, duplicate treatment, excessively long treatment, and failure to de-escalate on the basis of microbiology test results; inadequate prescribing includes antimicrobial use when antimicrobials are not needed and prescribing in which the antimicrobial selection, dose, route, or duration is deemed unlikely to treat the pathogen or likely pathogen.24 We opted to use the terms supported and unsupported as proxies for appropriate and inappropriate or unnecessary use because we did not require that data collection be performed by clinicians, and determinations were made through analysis pathways rather than by antimicrobial stewards using their clinical expertise and judgment to evaluate individual patient records.
We observed that the percentages of unsupported use were higher for infection-based events than for antimicrobial-based events. This finding may have been associated in part with our inclusion of more specific criteria in the infection-based analysis pathways according to treatment guidelines from professional societies, which tend to focus on types of infections. Although US infectious diseases and pharmacy professional societies have issued a guideline on therapeutic monitoring of VANC use for serious infections caused by methicillin-resistant Staphylococcus aureus,25 few national guidelines have focused on appropriate therapeutic uses of specific antimicrobials. In addition, it was not feasible to include specific criteria to cover aspects of prescribing for all possible infection types in the antimicrobial-based pathways. The larger percentages of supported FQ and VANC use compared with antimicrobial use for treatment of CAP and UTI may have been attributable to this exclusion. We believe that for the approach that we used, the infection-based assessments were more practical for implementation on a large scale and identified more opportunities for improving use.
One example of an opportunity for improvement suggested by our analysis is excessive treatment duration, which was the most common reason for unsupported CAP treatment and has been reported in multiple other studies.26-28 We calculated total treatment duration, including days of inpatient therapy plus the planned duration of postdischarge treatment. Current CAP guidelines recommend treatment for a minimum of 5 days, even if the patient has reached clinical stability before 5 days, stating that “as most patients will achieve clinical stability within the first 48 to 72 hours, a total duration of therapy of 5 days will be appropriate for most patients.”17 Exceptions are noted for CAP caused by methicillin-resistant S aureus or Pseudomonas aeruginosa, for which the recommended duration of treatment is 7 days.17 In our analysis, among 142 patients with CAP for whom duration of therapy was assessed, 103 (72.5%) were treated for at least 8 days. Among hospitalized veterans with uncomplicated pneumonia in 2013, 93.1% of patients with CAP received treatment for longer than the recommended duration.26 Among patients with CAP who were hospitalized in 2017 and 2018 in a Michigan Hospital Medicine Safety Consortium study, 71.3% received treatment for longer than the recommended duration.27 Given the harm associated with excessive treatment, studies are needed to establish effective approaches to reducing treatment duration, particularly after discharge.27,28
Absence of signs or symptoms of infection was another common reason for unsupported antimicrobial use among patients receiving UTI treatment. Recent updated guidelines29 have addressed the problem of inappropriate treatment of asymptomatic bacteriuria. Despite efforts to discourage treatment of asymptomatic bacteriuria, a large percentage of patients receiving UTI treatment in our analysis—approximately 38%—lacked documented signs or symptoms of infection. This is higher than the percentage observed in a similar analysis performed in 2011,30 in which approximately 23% of patients without a catheter who were being treated for UTI did not have documented signs or symptoms of infection. Results of a Veterans Health Administration study showed that among hospitalized patients with positive urine culture results in 2013 and 2014, 72% with asymptomatic bacteriuria received antibiotics.31 Interventions that incorporate elements such as education and clinical decision support have been shown to be associated with reductions in antimicrobial use for asymptomatic bacteriuria.32-34
This study has limitations. The numbers of hospitals and patients included in our analysis were limited and from just 10 states; consequently, the results may not be generalizable. We assessed antimicrobial treatment only and not surgical or medical prophylaxis; data on surgical prophylaxis from the Emerging Infections Program hospital prevalence survey have been published.11 Because of the complexity of evaluating inpatient antimicrobial use, we included only selected patients who were treated for a single infection type. Therefore, only 35.0% of patients receiving antimicrobial treatment during hospitalization were assessed, which is a limitation of an approach that does not use antimicrobial stewards to review and interpret data from individual patient records. Determining the appropriateness of antimicrobial use for the remaining 65% of patients, many of whom may have received antimicrobials for complicated infections, may be challenging with the use of our approach. In a small percentage of patients included in both the FQ and the VANC analysis pathways, discordant determinations had to be resolved by 1 of the authors (S.S.M.). Further refinement of the data collection and analysis pathways may reduce this need in future assessments. In addition, our assessment was based solely on medical record documentation. Incomplete documentation or failure to collect certain data, such as all antimicrobials received by patients during hospitalization in the FQ or VANC pathways, could have affected our results. We were not able to validate the results obtained using the analysis pathways with reviews of a subset of patient records by infectious diseases specialists or pharmacists. In addition, we did not assess risk factors for unsupported antimicrobial use.
The findings suggest that standardized assessments of hospital antimicrobial prescribing quality can be used to estimate the appropriateness of antimicrobial use in large groups of hospitals. National assessments of prescribing quality to complement data on the volume of antimicrobial use in hospitals and improve prescribing practices may ultimately depend on the ability to access and analyze electronic health record data across hundreds or thousands of health care facilities. Until such approaches are feasible, the AQUA assessment may be repeated over time as part of intermittent prevalence surveys of health care–associated infections and antimicrobial use to describe changes in prescribing quality and estimate the effects of national antimicrobial stewardship initiatives.
Accepted for Publication: January 24, 2021.
Published: March 18, 2021. doi:10.1001/jamanetworkopen.2021.2007
Correction: This article was corrected on April 16, 2021, to fix multiple rounding errors in the Abstract Results, Results, and Table 1; incorrect sentence order in the Results section; indentation errors in Tables 4 and 5; typos in the Conflict of Interest Disclosures; and data errors in the Supplement.
Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2021 Magill SS et al. JAMA Network Open.
Corresponding Author: Shelley S. Magill, MD, PhD, Division of Healthcare Quality Promotion, Centers for Disease Control and Prevention, 1600 Clifton Rd, HB16-3, Atlanta, GA 30329 (firstname.lastname@example.org).
Author Contributions: Dr Magill and Ms O’Leary had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Magill, Kainer, Bamberg, Johnston, Lynfield, Nadle, Thompson, Pierce, Maloney, Wilson, Dumyati, Edwards, Neuhauser.
Acquisition, analysis, or interpretation of data: Magill, O'Leary, Ray, Kainer, Evans, Johnston, Janelle, Oyewumi, Lynfield, Rainbow, Warnke, Nadle, Thompson, Sharmin, Pierce, Zhang, Ocampo, Maloney, Greissman, Wilson, Dumyati, Edwards, Chea, Neuhauser.
Drafting of the manuscript: Magill, O'Leary.
Critical revision of the manuscript for important intellectual content: Magill, O'Leary, Ray, Kainer, Evans, Bamberg, Johnston, Janelle, Oyewumi, Lynfield, Rainbow, Warnke, Nadle, Thompson, Sharmin, Pierce, Zhang, Ocampo, Maloney, Greissman, Wilson, Dumyati, Edwards, Chea, Neuhauser.
Statistical analysis: Magill, O'Leary, Edwards.
Obtained funding: Magill, Bamberg, Maloney, Dumyati.
Administrative, technical, or material support: Magill, Ray, Kainer, Evans, Johnston, Janelle, Oyewumi, Rainbow, Warnke, Nadle, Thompson, Sharmin, Zhang, Ocampo, Maloney, Greissman, Neuhauser.
Supervision: Magill, Ray, Kainer, Bamberg, Lynfield, Nadle, Thompson, Pierce, Maloney, Wilson, Dumyati, Edwards.
Conflict of Interest Disclosures: Dr Kainer reported receiving nonfinancial support from the Council of State and Territorial Epidemiologists, the Society for Healthcare Epidemiology of America (SHEA), the American Society for Microbiology, and the Public Health Association of Australia; receiving personal fees from Infectious Disease Consulting Corporation; and receiving personal fees and nonfinancial support from the Infectious Disease Consulting Corporation (IDCC), WebMD, and Pfizer outside the submitted work. Dr Bamberg reported grants from the Centers for Disease Control and Prevention (CDC) to the Colorado Department of Public Health and Environment during the conduct of the study. Ms Johnston reported receiving grants from CDC to the Colorado Department of Public Health and Environment during the conduct of the study. Ms Janelle reported receiving grants from CDC during the conduct of the study. Dr Lynfield reported receiving grants from CDC Emerging Infections Program Cooperative Agreement during the conduct of the study and being co-editor and receiving royalties for a book from the American Academy of Pediatrics, which were donated to the Minnesota Department of Health, outside the submitted work. Ms Warnke reported receiving grants from the CDC during the conduct of the study. Ms Nadle reported receiving grants from the CDC Emerging Infections Program Cooperative Agreement outside the submitted work. Dr Thompson reported receiving grants from CDC Emerging Infections Program during the conduct of the study. Dr Pierce reported receiving grants from CDC Emerging Infection Program and the CDC Epidemiology and Laboratory Capacity during the conduct of the study and personal fees from SHEA outside the submitted work. Ms Zhang reported receiving grants from the CDC during the conduct of the study. Ms Maloney reported receiving grants from the CDC Emerging Infections Program Cooperative Agreement during the conduct of the study and receiving a SHEA Education & Research Foundation Public Health Scholarship. Dr Wilson reported receiving grants from the CDC to the Maryland Department of Health. Dr Dumyati reported receiving grants from the CDC and personal fees from Roche Molecular Diagnostics during the conduct of the study and from Seres Therapeutics outside the submitted work. No other disclosures were reported.
Funding/Support: The Emerging Infections Program Hospital Prevalence Survey of Healthcare-associated Infections and Antimicrobial Use was supported by the CDC through the Emerging Infections Program Cooperative Agreement. The process of refining antimicrobial quality assessment analysis pathways was supported in part by The Pew Charitable Trusts through the CDC Foundation. The Pew Charitable Trusts sponsored in-person and telephone expert meetings.
Role of the Funder/Sponsor: CDC staff members contributed to the design and conduct of the study; management, analysis, and interpretation of the data; preparation, review or approval of the manuscript; and decision to submit the manuscript for publication. Staff of The Pew Charitable Trusts contributed to the interpretation of the data and review of the manuscript.
Group Members: Emerging Infections Program Hospital Prevalence Survey Team: California Emerging Infections Program, Oakland, CA: Karen Click; Linda Frank, RN, BSN, PHN; Deborah Godine, RN; Brittany Martin, MPH; Erin Parker, MPH; and Lauren Pasutti, MPH. Colorado Department of Public Health and Environment, Denver: Sarabeth Friedman, CNM, MSN; Annika Jones, MPH; and Tabetha Kosmicki, MPH, CIC. Connecticut Emerging Infections Program, New Haven and Hartford, CT: James Fisher, MPH; Amber Maslar, MPA; James Meek, MPH; and Richard Melchreit, MD. Division of Healthcare Quality Promotion, CDC, Atlanta, GA: Farzana Badrun, MD, MS (Eagle Medical Services); Lauren Epstein, MD, MPH; Ryan Fagan, MD, MPH; Anthony Fiore, MD, MPH; Nicole R. Gualandi, RN, MSN/MPH; and Arjun Srinivasan, MD. Georgia Emerging Infections Program, Decatur: Scott K. Fridkin, MD; Susan L. Morabit, MSN, RN, PHCNS-BC, CIC; and Lewis A. Perry, DrPH, MPH, RN. Maryland Department of Health, Baltimore: Rebecca Perlmutter, MPH, CIC, and Elisabeth Vaeth, MPH. Minnesota Department of Health, St Paul: Annastasia Gross, MPH, MT (ASCP); Jane Harper, MS, BSN, CIC; Brittany Pattee, MPH; and Nabeelah Rahmathullah, MBBS, MPH. New Mexico Department of Health, Santa Fe: Joan Baumbach, MD, MS, MPH, and Marla M. Sievers, MPH. New York Emerging Infections Program and University of Rochester Medical Center, Rochester: Cathleen Concannon, MPH; Christina Felsen, MPH; and Anita Gellert, RN. Oregon Health Authority, Portland: Monika Samper, RN. Tennessee Department of Health, Nashville: Raphaelle H. Beard, MPH; Patricia Lawson, RN, MS, MPH; Daniel B. Muleta, MD, MPH; and Vicky P. Reed, RN.
Disclaimer: The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the CDC or the position or policy of the Department of Veterans Affairs (VA).
Additional Contributions: The following members of The Pew Charitable Trusts Hospital Antimicrobial Stewardship Expert Panel made contributions to data interpretation: David Hyun, MD (The Pew Charitable Trusts); Susan Davis, PharmD (Wayne State University); Elizabeth Dodds Ashley, PharmD, MHS (Duke University); Scott K. Fridkin, MD (Emory University); Matthew Bidwell Goetz, MD (VA Greater Los Angeles Healthcare System and David Geffen School of Medicine at UCLA); Kalpana Gupta, MD, MPH (VA Boston Healthcare System and Boston University School of Medicine); Timothy Jenkins, MD, MSc (Denver Health, University of Colorado); Makoto M. Jones, MD, MS (VA Salt Lake City Health Care System and University of Utah School of Medicine); Holly D. Maples, PharmD (University of Arkansas for Medical Sciences); Larissa May, MD, MSPH, MSHS (University of California Davis); Joshua Metlay, MD, PhD (Massachusetts General Hospital and Harvard Medical School); Talene A. Metjian, PharmD (Children’s Hospital of Philadelphia); Marc J. Meyer, RPh, BPharm, CIC, FAPIC (Southwest Health System, Cortez, CO); Jason Newland, MD, MEd (Washington University, St Louis, MO); Pranita D. Tamma, MD, MHS (Johns Hopkins University School of Medicine); Barbara W. Trautner, MD, PhD (Michael E. DeBakey VA Medical Center and Baylor College of Medicine); and Valerie Vaughn, MD, MSc (University of Michigan and VA Ann Arbor Healthcare System). Members were not compensated for their contributions, but The Pew Charitable Trusts supported their travel to an expert meeting.