Effects of Knowledge, Attitudes, and Practices of Primary Care Providers on Antibiotic Selection, United States Guillermo V. Sanchez, Rebecca M. Roberts, Alison P. Albert, Darcia D. Johnson, and Lauri A. Hicks Appropriate selection of antibiotic drugs is critical to not clear whether nonadherence is related to lack of famil- optimize treatment of infections and limit the spread of iarity with clinical practice guidelines or if other factors in- antibiotic resistance. To better inform public health efforts fluence antibiotic selection once a diagnosis is established. to improve prescribing of antibiotic drugs, we conducted Published qualitative studies that have examined an- in-depth interviews with 36 primary care providers in the tibiotic selection among primary care providers (PCPs) United States (physicians, nurse practitioners, and physi- are outdated and focus on non-US–based physicians; they cian assistants) to explore knowledge, attitudes, and self- reported practices regarding antibiotic drug resistance and do not include nurse practitioners or physician assistants, antibiotic drug selection for common infections. Participants who together comprise >25% of the US primary care work- were generally familiar with guideline recommendations for force (15–22). The objectives of this study are to explore antibiotic drug selection for common infections but did not US PCP knowledge, attitudes, and self-reported practices always comply with them. Reasons for nonadherence in- (KAPs) concerning antibiotic therapy, assess factors that cluded the belief that nonrecommended agents are more influence provider antibiotic choice, and provide an update likely to cure an infection, concern for patient or parent sat- on PCP attitudes regarding antibiotic resistance. isfaction, and fear of infectious complications. Providers inconsistently defined broad- and narrow-spectrum antibi- Methods otic agents. There was widespread concern for antibiotic We conducted in-depth interviews by digitally record- resistance; however, it was not commonly considered when ed telephone calls, and transcribed the recordings to text to selecting therapy. Strategies to encourage use of first-line agents are needed in addition to limiting unnecessary pre- accurately and reliably assess PCP KAPs. The qualitative scribing of antibiotic drugs. method of an open-ended interview by telephone was cho- sen to ensure candid and truthful answers from participants. We composed a screening questionnaire to recruit physi- ntibiotic prescribing guidelines establish standards of cians, nurse practitioners (NPs), and physician assistants Acare, help focus efforts on quality improvement, and (PAs) from a nationwide marketing database in the United have been shown to improve patient outcomes (1–3). Many States. We initially contacted and screened potential partic- guidelines emphasize the importance of diagnostic certain- ipants (online Technical Appendix Section A, wwwnc.cdc. ty for the management of bacterial upper respiratory tract gov/EID/article/20/12/14-0331-Techapp1.pdf) using tele- infections and promote β-lactam agents, such as amoxicil- phones, email, and fax transmission of documents. Inclu- lin or amoxicillin-clavulanate, as the preferred first-line sion criteria included self-reporting of spending >50% of therapy (4–6). Studies indicate that health care providers medical practice time in direct patient contact in a primary often do not adhere to established clinical practice guide- care setting, >30 years of age, and fluency in the English lines for the management of common infections (7–9). Pre- language. Interviewees were excluded from this study if scribing rates for second-line, broad-spectrum antibiotics they had an immediate family member who was employed among outpatients have increased, contributing to a grow- in an industry that could represent a conflict of interest, in- ing problem of antibiotic-resistant infections (10–14). It is cluding advertising or public relations, the federal govern- ment, market research, news media, or the pharmaceutical Author affiliation: Centers for Disease Control and Prevention, industry; if they had board certification in a subspecialty Atlanta, Georgia, USA outside of primary care; or if they had practiced medicine DOI: http://dx.doi.org/10.3201/eid2012.140331 >30 years at the time of recruitment. Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 20, No. 12, December 2014 2041 RESEARCH Thirty-six PCPs were selected for the study. Special- was given a clinical vignette about a patient (online Tech- ties for the 27 physician participants selected included nical Appendix, Section C) who had a diagnosis of an pediatrics (n = 9), family medicine (n = 9), and internal acute bacterial infection: acute otitis media (AOM) for medicine (n = 9). Among PAs (n = 4) and NPs (n = 5), pediatricians, acute bacterial rhinosinusitis for internists, 6 practiced in family medicine settings and 3 practiced in acute uncomplicated cystitis for family practitioners, and pediatrics. Provider specialty, years in practice, and demo- group A streptococcal pharyngitis for PAs and NPs. The graphic information are described in Table 1. participant was asked to explain his or her rationale in Interviews were conducted during May 2013. One choosing an antibiotic agent and why other PCPs might professional moderator who had >25 years of moderating choose nonrecommended antibiotics. For the purposes of experience conducted all interviews. Based on a discussion this study, antibiotics considered to be broad-spectrum guide prepared by our research team (data not shown), par- include penicillins containing β-lactamase inhibitors ticipants were first informed of the sponsoring organization (e.g., amoxicillin/clavulanate), second through fifth gen- (Centers for Disease Control and Prevention, Atlanta, GA, eration cephalosporins, macrolides, quinolones, and lin- USA), the planned use of data, and presence of listeners; comycin derivatives. Narrow-spectrum agents include they were then asked “warm-up” questions about the prac- penicillins (e.g., amoxicillin), first generation cephalospo- tice setting in which the participant worked and the patient rins, sulfonamides, and nitrofurantoin. Each participant populations whom they served. Next, each interview pro- received a cash incentive after the interview in exchange ceeded through an ordered list of open-ended questions on for participating. self-reported antibiotic prescribing practices, perceived Interviews were transcribed by project staff, and each prescribing practices of their peers, attitudes toward clini- member of the research team either listened to interview cal practice guidelines for common bacterial infections, recordings or read corresponding interview transcripts. knowledge of narrow- versus broad-spectrum antibiotic Relevant excerpts were coded into compilations of refer- agents, preferred resources and methods for medical edu- ences to specific themes and were used to identify the most cation and antibiotic treatment, and attitudes toward anti- frequent responses to discussion topics outlined in the dis- biotic resistance. cussion guide. We identified common themes using both Participants were provided a worksheet before the inductive and deductive methods that were reviewed by interview (online Technical Appendix, Section B) which all authors. Author disagreement on theme selections were asked them to rank each of 12 factors on the basis of discussed until a consensus was met. If no consensus was its perceived influence on antibiotic selection when an met by most study authors, the theme was excluded from antibiotic is indicated (e.g., illness severity, patient de- our results. We performed in-depth analyses of themes by mand for an antibiotic, or clinical practice guidelines). reading and coding transcribed responses using Nvivo 9 Worksheet answers were discussed and recorded during (QSR International, Burlington, MA). This study was re- the interview. viewed and approved for exemption status by the Human To assess compliance with clinical practice guidelines Research Protection Office of the Centers for Disease Con- and to evaluate clinical decision-making, each participant trol and Prevention. Table 1. Characteristisc of primary care providers interviewed for Results knowledge, attitudes, and practices in antibiotic drug selection, We conducted 36 interviews, each lasting ≈45 min- United States* Characteristic Physician, n = 27 NP or PA, n = 9 utes. Through analysis of provider in-depth interviews, Sex several common themes regarding antibiotic prescribing M 18 1 and antibiotic resistance were identified (Table 2). F 9 8 Race/ethnicity White 18 9 Antibiotic Selection Black 4 0 PCPs generally cited little difficulty selecting anti- Asian 3 0 biotic treatments for common infections, but indicated Hispanic 1 0 Other 1 0 that allergies, complicated medical histories, and recur- Years in practice rent infections regularly make antibiotic selection more <10 5 5 challenging. Previous experience and familiarity with an 10–20 11 3 21–30 11 1 agent were frequently cited as influential factors when Medical specialty choosing antibiotic therapy. Results from the ranking ex- Pediatrics 9 3 ercise suggest that illness severity, medical history, and Family medicine 9 6 clinical practice guidelines were important considerations Internal medicine 9 0 *NP, nurse practitioner; PA, physician assistant. across all specialties. 2042 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 20, No. 12, December