Demographics

At least 141 attendings and advanced practice providers (nurse practitioners and physician assistants) and approximately 200 residents received the survey. Sixty-four providers completed our survey (Table 1). The exact response rate cannot be calculated as the total number of providers who received the survey is unknown (it is unknown how many providers received the QR code) but was a maximum of 18%. Most respondents were physicians (89%). Most physicians were out of training (73%), with the remainder being current residents. Respondents had practiced medicine for a mean of 17 years (range 1–38 yr). Adult and pediatric primary care were well represented with 29% of respondents working in the adult setting, 24% family medicine, 24% pediatrics, and 5% medicine-pediatrics. The remainder of respondents worked in adult or pediatric outpatient subspecialty care, urgent care, or another unspecified setting.

Table 1. Demographics of survey respondents. medical specialty refers to area of training and may be different for some respondents than current practice setting

Factors impacting prescribing decisions

Providers endorsed using a variety of tools and resources to support their decision-making. UpToDate® was the most utilized resource, used by 92% of respondents (Figure 1C). Providers also relied upon prior experience (43%), pharmacists (42%), or other providers (30%) (Figure 1C). Forty-one percent of providers reported use of at least one EMR-based tool such as AgileMD (a platform that provides EMR integrated clinical decision pathways), order sets, or other UChicago Medicine institutional pathways/toolkits such as those that assist with addressing antibiotic allergies (Figure 1C). Other resources used were UpToDate® Lexidrug™ (39%), AAP Guidelines (19%) and the AAP Red Book (11%) (Figure 1C). Other institutional, national, and international antibiotic prescribing guidelines including those from the University of Michigan, the Emergency Medicine Residents Association, the Infectious Diseases Society of America, the American Academy of Family Physicians, and the North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition were utilized by at least one respondent. Finally, Epocrates, Open Evidence, and Sanford Guide were all utilized by one respondent.

Figure 1. Provider perspectives on outpatient antimicrobial stewardship expectations and initiatives. (A) Provider perspectives on antibiotic stewardship. ASP, Antibiotic Stewardship Program. (B) Non-clinical factors that influence prescribing decisions. “Any factor” means a provider considers at least one non-clinical factor in making prescribing decisions. (C) Resources used by providers to make antibiotic prescribing decisions. (D) Provider perspectives on which stewardship interventions may be useful in their practice.

A variety of non-clinical factors also shape decisions about antibiotic prescribing. In fact, 78% of providers acknowledged at least one non-clinical factor impacts their antibiotic prescribing decisions (Figure 1B). Forty-one percent of providers consider a patient/family’s expectations and 51% consider social situation (Figure 1B). Providers also reported limited time (27%) and difficulty with frequently changing recommendations/guidelines (29%) as additional non-clinical factors that shape prescribing decisions (Figure 1B).

Perspectives on outpatient antimicrobial stewardship

Antibiotic stewardship is important to those surveyed. Eighty-three percent agreed that antibiotic overuse is an issue in primary care (Figure 1A). Seventy-five percent agreed that an outpatient antimicrobial stewardship program would be helpful to their practice (Figure 1A). The most popular antimicrobial stewardship interventions were related to the EMR, with 75% of respondents expressing interest in at least one (Figure 1D). Proposed EMR-based interventions in our survey included applying default dose or duration for certain antibiotics (66% “yes” responses), improving note templates to include links to guidelines or other resources (44% “yes” responses), and streamlining the process for de-labeling penicillin allergies (47% “yes” responses) (Figure 1D). Some respondents also expressed interest in audit with group feedback (34%), audit with individual feedback and anonymous comparison to peers (33%), or audit with individual feedback (14%) (Figure 1D). Examples of “other” strategies suggested by respondents included an outpatient antibiogram, patient/family educational initiatives, and incorporation of additional clinical decision supports or alerts in the EMR (Figure 1D).