Unnecessary and inappropriate antibiotic use leads to antibiotic resistance and one of the most common and deadly healthcare-associated infections: Clostridioides difficile infection (CDI).Reference Davies, Longshaw and Davis1,Reference Lessa, Mu and Bamberg2 Clostridioides difficile accounts for nearly half a million infections and 15,000 deaths annually in the United States (US) alone.3 The US government has used a combination of financial incentives, public reporting, and regulatory oversight to try to reduce healthcare-associated infections. Although these policies have led to decreases in most healthcare-associated infections, CDI and deaths from CDI have not decreased.Reference Magill, O’Leary and Janelle4,Reference Dubberke, Rohde and Saint5
Successful strategies for reducing CDI include antibiotic stewardship and infection prevention.Reference Gerding, Muto and Owens6 For example, England implemented a national CDI prevention campaign in 2007 emphasizing both antibiotic stewardship and infection prevention. That program reduced CDI and related mortality by at least 60%.Reference Wilcox, Shetty and Fawley7 Although the US has adopted many infection prevention measures, it has been slower to adopt antibiotic stewardship. In a 2013 national survey, we found that although infection prevention practices for CDI were nearly universal in US hospitals, only 52% had an antibiotic stewardship program.Reference Saint, Fowler and Krein8 The same year, an analysis of the National Healthcare Safety Network Annual Hospital Survey found that only 39% of US hospitals met all of the Centers for Disease Control and Prevention (CDC) “core elements” for antibiotic stewardship programs.Reference Pollack, van Santen, Weiner, Dudeck, Edwards and Srinivasan9 This deficit in antibiotic stewardship is concerning given mounting data suggesting that antibiotic stewardship plays a larger role than infection prevention in reducing CDI.Reference Dingle, Didelot and Quan10
To address the continued problem of antibiotic overuse and CDI, the major hospital accrediting body in the US, The Joint Commission, made antibiotic stewardship programs a condition for hospital accreditation as of January 1, 2017. Specifically, The Joint Commission began requiring hospitals to have multidisciplinary stewardship teams including the following, when available: (1) an infectious diseases (ID) physician, (2) an infection preventionist, (3) a pharmacist, and (4) a practitioner (not defined; presumed in this analysis to include any physician).Reference Barlam, Cosgrove and Abbo11 Concurrently, the US is facing a shortage of available ID specialists, especially those with antibiotic stewardship expertise.Reference Chandrasekar, Havlichek and Johnson12 Whether all US hospitals, especially smaller hospitals that may have limited resources and staffing, have been able to meet The Joint Commission goal is not known. This concern led the CDC to recommend that critical access hospitals (<25 beds) include a pharmacist and a “physician leader” on their stewardship team.13
Furthermore, since our 2013 survey, the concept of “diagnostic stewardship”—or strategies and policies to improve appropriate use of microbiological testing—has expanded as a key method to improve antibiotic use.Reference Morgan, Malani and Diekema14 Similarly, the adoption of many methods to prevent, treat, and improve testing for CDI, has not been well characterized. In 2017, we conducted a random survey of US hospitals to determine the following: (1) the presence and composition of antibiotic stewardship programs; (2) current methods of CDI prevention, treatment, and testing; (3) diagnostic stewardship practices; and (4) whether antibiotic stewardship program composition, CDI prevention strategies, and diagnostic stewardship vary by hospital bed size.
Methods
Data collection
The current study was part of an ongoing survey in which, every 4 years, we ask infection preventionists across the US what practices their hospitals are using to prevent common healthcare-associated infections.Reference Saint, Fowler and Krein8,Reference Krein, Kowalski, Hofer and Saint15 –Reference Krein, Hofer and Kowalski17 Survey methods have been previously described.Reference Saint, Kowalski and Kaufman16 Briefly, we randomly sampled 900 medical and surgical hospitals with an intensive care unit across the US. Three hospitals were excluded from the study due to closure or status change. Surveys were mailed in May 2017, with subsequent reminders to nonresponders. Hospitals that employed >1 infection preventionist were asked to have the lead infection preventionist serve as the primary respondent, though we encouraged consulting with others to complete the questionnaire. Hospital characteristics, including bed size and The Joint Commission accreditation status, were obtained by linking to the 2013 American Hospital Association database. Institutional review board approval as an exempt study was obtained from the University of Michigan.
Survey measures
Similar to prior surveys,Reference Krein, Kowalski, Hofer and Saint15–Reference Krein, Hofer and Kowalski17 participants were queried regarding hospital characteristics and details of their infection prevention programs. In addition, participants were asked how frequently certain CDI prevention practices were used in their facility (responses: 1 [never] to 5 [always]).Reference Saint, Fowler and Krein8 We defined responses of 4 or 5 (ie, “almost always” or “always”) as regular use of the respective CDI prevention practices. Hospitals were also asked, “Has your hospital implemented a urine culture stewardship initiative? (Yes/No)” and “Does your hospital have an antibiotic stewardship program? (Yes/No).” If participants answered yes to having an antibiotic stewardship program, they were asked to indicate who of the following was on their antibiotic stewardship team: (1) infection preventionist, (2) ID physician, (3) hospitalist (primarily inpatient general internist), (4) other physician, (5) pharmacist (with ID training), (6) pharmacist (without ID training), (7) nurse, or (8) other.
Data analysis
To create nationally representative estimates, survey responses are shown as weighted proportions, with sampling weights based on the inverse probability of selection and response by bed size. To evaluate whether antibiotic stewardship program composition and CDI strategies varied by bed size, we used weighted logistic regression with hospital bed size as a continuous variable. Odds ratios are reported for every 10-bed increase. For ease of visualization, responses were separated into 3 commonly used bed-size categories: <50 beds (small), 50–250 beds (medium), >250 beds (large). P < .05 was considered statistically significant. We used SAS version 9.4 software (SAS Institute, Cary, NC) for all analyses.
Role of the funding source
The funders of the study had no role in the study design; data collection, analysis, or interpretation; writing of the report; or in the decision to submit the paper for publication. The corresponding author had full access to all study data and had final responsibility for the decision to submit for publication.
Results
The survey response rate was 59% (530 of 897). Two responding hospitals (0.2%) could not be linked to bed size data and were excluded, leaving 528 hospitals in the final analysis. Hospital characteristics are shown in Table 1.
Note. CI, confidence interval.
a Data on medical school affiliation, presence of a hospital epidemiologist, and full-time equivalents for infection preventionists were obtained from our survey data. The remaining data were obtained by linking respondents to the 2013 American Hospital Association database.
Antibiotic stewardship team composition
Nearly all (95%) hospitals reported having an antibiotic stewardship program. Most stewardship teams had a pharmacist (99%), a physician (95%, including ID physician [69%], hospitalist [48%], or other physician [44%]), and/or an infection preventionist (91%). Team members with ID training were less common: 52% of antibiotic stewardship programs had an ID-trained pharmacist, and 69% had an ID physician. Although most hospitals (78%) had either an ID physician or ID pharmacist, only 43% had both. Less than half of hospitals (41%) met The Joint Commission accreditation standard (ID physician, infection preventionist, pharmacist, practitioner); however, most hospitals (95%) met the minimum standards set by the CDC for critical access hospitals (pharmacist, physician).
As bed size increased, hospitals were more likely to have antibiotic stewardship team members (eg, hospitalists, non-ID physicians, nurses) with ID training (Fig. 1). In contrast, the presence of generalists on antibiotic stewardship team did not vary by bed size. Larger hospitals were more likely to meet both The Joint Commission recommendation for multidisciplinary stewardship programs and the CDC minimum recommendations for critical-access hospitals (Table 2).
Note. ID, infectious diseases; CDC, Centers for Disease Control and Prevention. Includes hospitals (N = 493) that reported having an antibiotic stewardship program.
a Survey responses are shown as weighted proportions with sampling weights based on the inverse probability of selection and response by bed size.
b Hospital bed size as reported by the American Hospital Association. P < .05 was considered significant.
c “Practitioner” is terminology used by The Joint Commission. For this analysis, ID physicians, hospitalists, and other physicians were included as practitioners.
CDI prevention, treatment, and testing practices
The reported use of guideline-recommendedReference McDonald, Gerding and Johnson18 CDI prevention practices, which have been shown to reduce CDI, was high (>90%). Furthermore, CDI prevention was perceived as “important” or “very important” to hospital leadership in 89% of hospitals (Table 3). In contrast, use of innovative, “high-tech” methods for CDI prevention and availability of fecal microbiota transplant varied, with larger hospitals more likely to report their use. Across hospitals, methods of testing for CDI varied; approximately half of hospitals (56%) used polymerase chain reaction (PCR) and 25% used a combination of testing methods. Larger hospitals were more likely to use PCR and less likely to use toxin or glutamate dehydrogenase antigen enzyme immunoassay (EIA) testing methods (Table 4).
Note. CDI, Clostridioides difficile infection.
a Survey responses are shown as weighted proportions with sampling weights based on the inverse probability of selection and response by bed size.
b Regular use is defined by a score of 4 or 5 on Likert scale.
c Question used a Likert scale from 1 (“Not at all important”) to 5 (“Very important”).
Note. ATP, adenosine triphosphate; CDI, Clostridioides difficile infection; C. difficile, Clostridioides difficile; EIA, enzyme immunoassay; PCR, polymerase chain reaction; UV, ultraviolet.
a Survey responses are shown as weighted proportions with sampling weights based on the inverse probability of selection and response by bed size.
b Regular use is defined by a score of 4 or 5 on Likert scale.
c Hospital bed size as reported by the American Hospital Association. P < .05 was considered significant.
Diagnostic stewardship practices
Hospitals commonly (91%) reported rejecting formed stool that was submitted for CDI testing; however, the use of urine culture stewardship was uncommon (33%). Both diagnostic stewardship practices increased with higher bed size (Table 4).
Discussion
In this large national survey of US hospitals, most hospitals had an antibiotic stewardship program and used evidence-based CDI prevention practices. Notably, the number of hospitals reporting antibiotic stewardship programs has nearly doubled in 4 years,Reference Saint, Fowler and Krein8 coinciding with new US hospital accreditation standards.19 Nevertheless, few hospitals met accreditation standards for multidisciplinary teams and ID expertise was limited, particularly as hospital size decreased. Despite national interest in CDI prevention, practices varied across hospitals, with less use of novel CDI practices and diagnostic stewardship strategies, especially at small hospitals.
National society guidelines in the US have long recommended antibiotic stewardship as a key tool for CDI prevention.Reference Cohen, Gerding and Johnson20 However, the use of antibiotic stewardship programs has been limited.Reference Saint, Fowler and Krein8 Our survey revealed that antibiotic stewardship programs have become nearly universal following new national accreditation requirements. The Joint Commission is the top hospital accreditation body (accrediting 75% of hospitals in our survey), and this uptake in stewardship is likely related to the 2017 standard. Furthermore, although hospitals struggled to meet the specific 2017 Joint Commission recommendation for multidisciplinary antibiotic stewardship programs, nearly all hospitals had multidisciplinary stewardship teams, often including a pharmacist, a physician, and an infection preventionist.
Despite the critical role of pharmacists in antibiotic stewardship, only half of US hospitals reported having pharmacists with ID training. Specialized pharmacy training programs are a more recent development and training spots are in short supply,Reference Gauthier, Worley and Laboy21 potentially limiting access to ID pharmacists, especially at smaller hospitals. Instead, clinical pharmacists without ID training often develop local expertise to improve antibiotic prescribing or obtaining additional training in stewardship through national organizations. Similarly, data suggest that antibiotic stewardship programs are best led by ID physicians with additional stewardship training.Reference Barlam, Cosgrove and Abbo11,Reference Ostrowsky, Banerjee and Bonomo22 In our study, two-thirds of hospitals had ID physicians on their stewardship teams. Lack of ID-trained leaders at small hospitals may limit antibiotic stewardship: in 2015, only 31% of hospitals with ≤50 beds met all CDC core elements.Reference O’Leary, van Santen, Webb, Pollock, Edwards and Srinivasan23 Although systematic changes to attract trainees to the field of ID and to antibiotic stewardship may help, this is a long-term solution.Reference Luther, Shnekendorf and Abbo24,Reference Walensky, Del Rio and Armstrong25 Other options to help distribute ID expertise across hospitals include access to expertise through quality collaborativesReference Vaughn, Gandhi, Conlon, Chopra, Malani and Flanders26 or “tele-stewardship” in which antibiotic use data are collected remotely and ID physicians are available via a stewardship “hotline.”Reference Stenehjem, Hersh and Buckel27 The Infectious Diseases Society of America suggests telestewardship as a way to provide cost-effective subspecialty care to resource-limited populations.Reference Young, Abdel-Massih and Herchline28,Reference Siddiqui, Herchline and Kahlon29 Unfortunately, many systems are prevented from using telestewardship due to medico-legal barriers and lack of financial reimbursement.Reference Bishop, Kong, Schulz, Thursky and Buising30
Even with these strategies, the numbers of ID physicians specializing in antibiotic stewardship for all US acute-care hospitals, outpatient clinics, and long-term care facilities are insufficient. To account for this shortage, the CDC instead recommends that each critical access hospital (<25 beds) have, at minimum, a pharmacist and physician on its stewardship team.13 In our survey, most small hospitals were able to meet this recommendation by having a non-ID physician, such as a hospitalist, on their stewardship team. Hospitalists can play a role in antibiotic stewardship because they prescribe most antibiotics for hospitalized patients, they can improve frontline provider buy-in, and they are often engaged in complementary quality improvement efforts.Reference Vaughn and Flanders31–Reference Mack, Rohde and Jacobsen33 Similarly, the role of nurses in antibiotic and diagnostic stewardship is underappreciated but growing.Reference Carter, Greendyke and Furuya34 A stewardship model that relies on pharmacy and physician generalists and nurses could apply not only to small hospitals, but to outpatient and long-term care settings where availability of ID experts are limited.
In addition to antibiotic stewardship, infection prevention strategies are critical in preventing CDI and are often multidisciplinary. Unlike stewardship, infection prevention has been a focus of US policies for decades; thus, it is not surprising that CDI prevention was considered important by hospital leadership and that evidence-based CDI prevention practices were common. The significance of variation in “high-tech” CDI prevention strategies is unclear because evidence for many of the practices is mixed and the cost is often quite high. Thus, small hospitals may be appropriately delaying purchasing expensive new technology until more evidence supports their use.
The lack of diagnostic stewardship strategies, especially at smaller hospitals, is concerning. For example, PCR techniques alone are only recommended to diagnose patients with a high pretest probability of CDI; even then, a multistep algorithm including toxin is generally preferred.Reference McDonald, Gerding and Johnson18 Thus, hospitals should implement diagnostic testing algorithms to improve CDI diagnostic accuracy. Ideally, such algorithms would include an assessment of symptoms and alternative causes to determine when testing is necessary. One high-yield method is automatically rejecting testing for CDI in patients with formed stools.Reference Morgan, Malani and Diekema14 Similarly, urine culture stewardship is key to reducing inappropriate antibiotic use for asymptomatic bacteriuria.Reference Brown, Daneman and Schwartz35 Up to 40% of hospitalized patients treated with antibiotics for presumed urinary tract infection have asymptomatic bacteriuria.Reference Petty, Vaughn and Flanders36 Antibiotic use in this group increases adverse events and prolongs hospitalization without improving outcomes.Reference Petty, Vaughn and Flanders36
Our study has several limitations. First, we relied on self-reporting. Although infection preventionists are most likely to know their hospital’s current practices related to CDI, their responses may not represent true practice. Second, while our sampling strategy was designed to obtain a nationally representative sample, it is possible that participating hospitals differed from nonparticipating hospitals. For example, hospitals with less developed stewardship and infection prevention teams may have been less likely to respond. Third, because we did not collect data on availability of ID specialists, we were unable to determine whether lack of ID involvement in stewardship at small hospitals reflected limited access or limited interest by available ID specialists. Finally, due to the inherent difficulties with survey methodology we were unable to assess the implementation of CDI and diagnostic stewardship strategies.
Herein, we have provided a snapshot of antibiotic stewardship program team composition, CDI strategies, and diagnostic stewardship in US hospitals in the period immediately following the 2017 Joint Commission standard for antibiotic stewardship. Although nearly all hospitals now have an antibiotic stewardship program, team compositions differ by hospital size, and most hospitals do not meet ideal recommendations for multidisciplinary teams. Specifically, smaller hospitals appear to have limited ID expertise on their stewardship teams and to struggle with deploying diagnostic stewardship strategies. In addition to implementing diagnostic stewardship, preventing CDI at small hospitals may be enhanced by emphasizing the role of generalists, such as clinical pharmacists, nurses, and hospitalists, in antibiotic stewardship.
Acknowledgments
The findings and conclusions in this manuscript are those of the authors and do not necessarily represent the official position of the Department of Veterans Affairs.
Financial support
This work was supported by the Blue Cross Blue Shield of Michigan Foundation (grant no. 2413.II) and a US Department of Veterans’ Affairs (VA) National Center for Patient Safety funded Patient Safety Center of Inquiry. Dr. Krein was also supported by a VA Health Services Research and Development Service Research Career Scientist award (grant no. RCS 11-222).
Conflicts of interest
Dr. Flanders has received royalties from Wiley Publishing and expert witness testimony and grant support from Blue Cross Blue Shield of Michigan and the Agency for Healthcare Research and Quality.