Antimicrobial stewardship programs (ASPs) strive to improve patient outcomes while minimizing unintended consequences of antimicrobials. A common strategy used by ASPs is audit-and-feedback: prospectively auditing inpatients receiving antimicrobials and providing real-time feedback to primary prescribers.Reference Pope, Dellit, Owens and Hooton 1 Prior research on audit-and-feedback has focused largely on the technical intervention itself.Reference Camins, King and Wells 2 – Reference Palmay, Elligsen and Walker 4 The success of any intervention, however, also depends upon socio-adaptive changes in a hospital’s culture.Reference Saint and Stock 5 Understanding how the design and delivery of audit-and-feedback influences its ultimate effectiveness is essential to successfully implementing stewardship.
In this study, we sought to understand how ASPs within the Society for Healthcare Epidemiology of America Research Network (SRN) have designed and implemented their audit-and-feedback processes.
METHODS
An 11-item electronic survey was emailed to SRN members. 6 A primary email was sent on October 15, 2015, and a reminder on October 29, 2015. Any acute care hospital that participated in the SRN was eligible. The responder needed to be a pharmacist or physician who was actively involved in the hospital’s ASP, and the ASP needed to be engaged in some form of prospective audit-and-feedback.
Respondents were asked questions about their ASP’s approach to prospective audit–and-feedback (Online Supplementary Material). Responses were recorded as dichotomous options, open-ended free text, and Likert scales from 1 to 5. The SRN’s database provided basic information on each hospital. Data were entered into an electronic spreadsheet (Excel; Microsoft) and were analyzed using SAS, version 9.4 (SAS Institute).
The project was deemed nonhuman subjects research by the University of Iowa Institutional Review Board.
RESULTS
The survey was sent to 215 SRN members. Sixty-one SRN members from unique institutions responded and were eligible for participation. Of 58 respondents who reported their professional title, 48 (82.8%) were physicians and 10 (17.2%) were pharmacists. The median (interquartile range) hospital size was 452 (210–650) beds. Nine facilities (14.8%) were outside the United States, 35 (57.4%) were teaching hospitals, and 28 (45.9%) were tertiary care facilities.
These 61 ASPs primarily addressed antimicrobial use in adult patients (42 [68.9%]). Ten ASPs (16.4%) addressed antimicrobial use in both adults and pediatrics, and 9 (14.8%) focused on pediatrics alone.
Of 60 responses about the performance of antimicrobial audits, 36 (60.0%) reported that a majority of their antimicrobial audits were performed by pharmacists, 18 (30.0%) by both physicians and pharmacists, 5 (8.3%) by physicians, and 1 (1.7%) by a microbiologist. Of 59 responses about which ASP member provides the majority of their feedback, 30 (50.8%) said pharmacists, 21 (35.6%) said both pharmacists and physicians, and 8 (13.6%) said physicians alone.
Data on feedback strategies were available for 57 programs. In 22 programs (38.6%), a majority of recommendations were communicated to the primary prescriber via the telephone. Seven programs (12.3%) delivered a majority of their recommendations through face-to-face conversations with the primary prescriber. Four programs (7.0%) asked members of the primary prescriber’s team to pass on a majority of their recommendations, and 2 (3.5%) primarily used electronic messages. Twenty-two programs (38.6%) did not identify a single strategy used for more than 50% of cases. In all, 11 programs (19.3%) exclusively provided verbal feedback to the primary prescriber. Only 9 programs (15.8%) documented their recommendations in the medical record for more than half the cases.
Among the 34 programs with complete information on hospital beds, the median (interquartile range) number of recommendations per week was 9 (5–19) for every 100 hospital-beds covered by the ASP. Eight programs (23.5%)—all with more than 300 hospital-beds—estimated that they made at least 100 recommendations per week. Table 1 provides the number of audit-and-feedback recommendations per week based on the number of hospital beds covered by the stewardship program. Respondents estimated that their recommendations were accepted at a median (interquartile range) frequency of 85% (80%–90%).
TABLE 1 Estimated Number of Audit-and-Feedback Recommendations Made per Week Based on the Number of Hospital Beds Covered by the ASP

NOTE. ASP, antimicrobial stewardship program; IQR, interquartile range.
The strategy believed to be most effective for changing the behavior of resistant prescribers was having the stewardship physician directly engage him/her in conversation. This strategy was perceived to be effective or very effective by 34 (57.6%) of 59 respondents. Table 2 summarizes the perceived effectiveness of other strategies to address clinicians resistant to recommendations.
TABLE 2 Perceived Effectiveness of Strategies to Change the Antimicrobial-Prescribing Behavior of Resistant Prescribers From 59 Respondents

Fifty-eight participants responded to the open-ended question, “What is the biggest barrier your ASP faces when trying to improve antimicrobial use at your hospital?” Responses were organized into 7 themes. The most common barrier was lack of resources, including limited time and personnel (25 [43.1%]). Additional barriers included physician attitudes about antimicrobials (14 [24.1%]), the need for more rigorous informatics support (8 [13.8%]), prescribers resistant to feedback (5 [8.6%]), lack of high-quality data on appropriate antimicrobial use (2 [3.4%]), communication (2 [3.4%]), and lack of stewardship metrics (2 [3.4%]).
DISCUSSION
Randomized controlled trials have demonstrated that prospective audit-and-feedback can reduce unnecessary antimicrobial use.Reference Camins, King and Wells 2 , Reference Lesprit, Landelle and Brun-Buisson 3 The optimal implementation of these processes, however, has not been well defined. Our survey demonstrates variability in how ASPs have structured their audit-and-feedback processes, including differences in who performs audits, who provides feedback to prescribers, and how the feedback is communicated.
Resource limitations, which were identified as a major barrier to stewardship, may account for some variability in ASP implementation. Hospital size may also be influencing each program’s approach.
Despite limited evidence to support any one implementation model, certain approaches to audit-and-feedback may be more likely to succeed than others. For example, prescribers may be more receptive to feedback if the recommendation is coming directly from a physician colleague and if the prescriber is able to discuss his/her concerns with changing/stopping antimicrobials.Reference Aagaard, Gonzales and Camargo 7 In contrast, our survey shows pharmacists alone are providing a majority of the feedback. Furthermore, although seemingly more effective, verbal conversations are not universally utilized.Reference Arora, Manjarrez, Dressler, Basaviah, Halasyamani and Kripalani 8
ASPs trying to promote a collaborative approach to antimicrobial use may encounter physicians who are not interested in changing their prescribing behavior. Research in other fields has shown that certain personality types resist organizational change through either active or passive means.Reference Saint, Kowalski, Banaszak-Holl, Forman, Damschroder and Krein 9 Although direct physician-to-physician conversations were identified by more than half of respondents as an effective or very effective strategy for addressing resistant prescribers, this perception was not shared by all. In addition, a large proportion of respondents indicated they had never used several proposed strategies, which suggests there are opportunities to address resistant prescribers in a different manner.
On the basis of this survey, the lack of protected time and salary support was perceived to be a major barrier to the development of more comprehensive programs. However, it is unclear whether more resources would lead to more effective stewardship at participating hospitals. A study of ASPs at 44 academic medical centers found that the amount of ASP resources did not predict facility-level antimicrobial use, but the greater use of stewardship strategies was associated with less antimicrobial use.Reference Pakyz, Moczygemba, Wang, Stevens and Edmond 10
Our study has some limitations. Survey respondents were all members of the SRN and may be different from ASP teams at non-SRN hospitals. In addition, it is unclear how many SRN members who met our inclusion criteria did not respond to our survey. Our response rate was slightly lower than prior SRN studies. 6 Although physicians were the primary respondents to our survey, pharmacists were the primary actors of stewardship. As a result, estimates of recommendations and acceptance rates may be inaccurate. Further, it is possible that respondents had different opinions of what constituted a recommendation, thereby skewing our results. Finally, our study was not designed to ascertain factors associated with more recommendations and higher acceptance rates.
In conclusion, our survey identified variability in how ASPs have implemented audit-and-feedback processes. Given this variability and the overwhelming need to expand the reach of antimicrobial stewardship, additional research is needed to identify the most effective strategies for auditing antimicrobial use and providing feedback to prescribers.
ACKNOWLEDGMENTS
Financial support. None reported.
Potential conflicts of interest. All authors report no conflicts of interest relevant to this article.
SUPPLEMENTARY MATERIAL
To view supplementary material for this article, please visit http://dx.doi.org/10.1017/ice.2016.57