Evaluating Approaches With the Potential to Improve Health Outcomes While Reducing Unnecessary Interventions in Infectious Disease Public Health

Abstract

We are increasingly tasked with improving value in healthcare. We wish to do more with less. More specifically, we aim to simultaneously improve health outcomes while reducing resource utilization, including unnecessary interventions and waste. This concept is apparent in the widely successful Choosing Wisely campaign which is an initiative focused on identifying and championing ‘high value’ practices in health care and translating these into meaningful change. Communicable pathogens that can be acquired and transmitted in both the community and hospital settings pose unique challenges with respect to diagnosis, therapy, and control. There are many approaches in communicable disease management which are not high value, and also interventions which do not yet exist but could improve value. Here I will discuss three approaches spanning diagnostics, therapeutics, and infection control approaches, with a focus on improving outcomes and reducing unnecessary interventions/treatment. First I present a retrospective evaluation of genome sequencing performed on isolates collected at the onset of institutional influenza outbreaks, for the purpose of evaluating the potential to reduce unnecessary outbreak/isolation practices and associated negative impacts in resident care. With this genomic epidemiologic study, I demonstrate that current CDC-based definitions appear specific for classifying influenza outbreaks, and that whole genome sequencing in its present form is unlikely to significantly reduce institution of outbreak measures including unnecessary isolation. Second, I describe a deterministic, compartmental model of the transmission and selection of extended-spectrum beta-lactamase producing Escherichia coli in the community and hospital environments. With this model, I show that although a reduction in hospital antibiotic prescriptions is most efficient, the greatest impact in both the community and hospital can be achieved through proportional reductions in antibiotic use in the community setting. Lastly, I present a retrospective evaluation of multiple genetic relatedness based-methods for predicting antibiotic susceptibility in confirmed/suspected E. coli infections. We show that rapid genetic relatedness information, derived from genome sequencing, could be used to provide informative predictions of the susceptibility of infecting E. coli isolates, with the potential to improve adequacy of empiric treatment as well as reduce the use of unnecessarily broad antibiotics.