Antimicrobial Resistant E. Coli Before and After California Senate Bill 27 (ares): A Natural Experiment

ABSTRACTOveruse of antibiotics contributes to antimicrobial resistance (AMR) and is a growing threat to human healthworldwide. In the U.S., industrial food animal production (IFAP) is a powerful driver of antibiotic use. Seventypercent of antibiotics are sold for use in livestock, including tetracycline, gentamicin, and third-generationcephalosporins (3GC). In an attempt to reduce AMR and population health impacts, beginning January 1,2018, California's Senate Bill 27 (SB27) will ban all non-therapeutic uses of antimicrobials in livestock (includ-ing growth promotion and routine disease prevention) and require a veterinarian's prescription for therapeuticindications. We will quantify the effect of this legislation on AMR E. coliE. coli is an ideal pathogen for evaluating changes in antimicrobial use andAMR related to livestock production because (1) AMR profiles of E. coli circulating in livestock and humansoverlap; and (2) extraintestinal pathogenic E. coli are a leading cause of human UTI, which cost Americans$1-2 billion annually. We plan to use data ranging from the molecular to the population level in a natural exper-iment to address three specific aims. In Aim 1, we will use whole genome sequencing and antimicrobial sus-ceptibility testing to characterize resistance genes, phenotypes, and E. coli populations isolated from retailchicken meat and UTI. We hypothesize that contaminated retail meat products contribute to the burden ofAMR E. coli UTI in humans. In Aim 2, we will leverage longitudinal (2015-2021) electronic health recordson retail chicken meat and human E.coli urinary tract infections (UTIs).(EHRs) on more than 7 million outpatients with 100,000 E. coli UTIs per year from two major California healthsystems to estimate longitudinal changes in the proportion of tetracycline-, gentamicin-, and 3GC-resistant E.coli isolated from retail chicken meat and laboratory-confirmed E. coli UTI before and after the implementationof SB27. We will also employ synthetic control statistical analyses to estimate the causal effect of SB27 onAMR trends in retail chicken meat and E. coli UTI. We anticipate that the enactment of SB27 will be associatedwith reductions in AMR of E. coli isolated from retail chicken meat and human UTI cases. In Aim 3, we willconduct spatial analyses using patient addresses from EHRs and IFAP location data. We hypothesize that pri-or to SB27, residential proximity to IFAP will be a risk factor for tetracycline-, gentamicin-, and 3GC-resistant E.coli UTI; after SB27, this risk will be reduced. The approach is innovative because it uses a quasi-experimentaldesign built around a first-in-the U.S. policy-mandated reduction in livestock antimicrobial use. Results of thiswork will make a significant contribution to the evidence base connecting food systems and UTI and will de-termine if this state-level legislative intervention affects AMR trends in retail meat and human UTIs, informingdecision-making about implementation of similar policies across the U.S.