**Background:** Bloodstream infections (BSIs) have traditionally been managed with prolonged intravenous (IV) antibiotics, but accumulating data suggest shorter courses and oral antibiotic therapy (OAT) may be appropriate in certain scenarios. No established guidelines exist regarding the role of OAT for BSIs, and practices may vary depending on clinician specialty and experience. This study assessed practice patterns regarding oral antibiotic use for treatment of bacteremia in infectious diseases clinicians (IDCs, including physicians and pharmacists and trainees) and non-infectious diseases clinicians (NIDCs).
**Methods:** An open-access, web-based REDCap survey was distributed to clinicians at a Midwestern academic medical center via e-mail and to clinicians outside the medical center using social media (Twitter and blog). The survey opened October 25, 2019, and closed January 31, 2020. It included demographic questions, generalized treatment questions evaluating which organisms and clinical syndromes influenced respondents' practice of using OAT, common antimicrobials they would consider, and 6 clinical vignettes describing hospitalized patients with resolved BSIs due to defined, pansusceptible organisms with controlled infectious syndromes. Uncomplicated BSI was defined as patients without persistent bacteremia who were clinically improving with a controlled source of infection. Physicians or pharmacists who identified infectious disease as their specialty were considered IDCs; all others were NIDCs. χ² analyses were used for categorical data, and Fisher exact test was used when >25% of cells had expected counts <5%. SAS version 9.4 was used, and P < .05 was considered significant.
**Key Results:** Of 282 survey responses, 233 (82.6%) were physicians, 46 (17.4%) were pharmacists, and 7 (2.5%) were advanced practice providers. IDCs represented 195 (69.2%) of respondents, with 87 (30.9%) NIDCs. About half (49.1%) responded via e-mail, 50.9% via social media, and 94% originated in the United States. The Twitter post received 7,831 impressions and 359 engagements.
For clinical vignettes, IDCs and NIDCs reported similar practices for routine OAT use in E. coli UTI (83.5% vs 79.1%; P = .585), S. pneumoniae pneumonia (68.6% vs 75.9%; P = .316), and streptococcal cellulitis (56.9% vs 60.9%; P = .781). However, fewer IDCs than NIDCs selected OAT for MRSA gluteal abscess (11.9% vs 25.6%; P = .012) and MSSA septic arthritis (13.9% vs 20.9%; P = .219). IDCs more frequently chose to always repeat blood cultures in MRSA gluteal abscess (99.0% vs 83.9%; P < .001) and MSSA septic arthritis (98.4% vs 83.1%; P < .001). For MSSA septic arthritis, more IDCs reported treating for >14 days (83.3% vs 46.5%; P < .001). For MRSA BSI, IDCs more often chose linezolid (65.8% vs 26.7%), while NIDCs more often chose TMP-SMX (51.1% vs 20.6%; P < .001).
For organism-specific OAT use (independent of syndrome), IDCs were more likely to select routine OAT for gram-negative anaerobes (84.6% vs 59.8%; P < .0001), Klebsiella spp (84.5% vs 69.0%; P < .009), Proteus spp (83.6% vs 71.3%; P < .027), and other Enterobacterales (79.5% vs 60.9%; P < .004). Both groups reported high routine OAT use for E. coli (88.2% vs 81.6%; P = .145). The lowest routine OAT use was for S. aureus (11.3% IDCs vs 29.9% NIDCs; P = .0006). For complicated syndromes, IDCs more often reported routine OAT use for prosthetic joint infection (21.5% vs 8.1%; P < .05) and vertebral osteomyelitis (18.5% vs 5.8%; P < .05), but less often for meningitis (2.6% vs 13.8%; P < .05). Only 41.6% of IDCs would select OAT for endocarditis routinely or in special circumstances.
**Clinical Implications:** This study reveals significant practice variation between IDCs and NIDCs regarding OAT use for BSIs, with both groups showing gaps between current evidence and reported practice. IDCs demonstrated greater alignment with evidence for gram-negative BSIs but were more conservative than evidence may warrant for streptococcal BSI and endocarditis. NIDCs were more likely to consider OAT for S. aureus BSI despite limited supporting data and current guideline recommendations favoring IV therapy. These findings highlight opportunities for targeted education of NIDCs regarding appropriate OAT candidates and for IDCs to better incorporate emerging evidence (e.g., from the POET and OVIVA trials) into practice. The development of standardized guidelines for OAT use in BSI management could help reduce variability and improve patient outcomes.