27 Apr UCI Develops AI Model To Determine Best Antibiotic Match for an Individual Patient

MedicalResearch.com Interview with:

Dr. Gohil

Shruti K. Gohil, MD
Assistant Professor, Infectious Diseases, Department of Medicine
Associate Medical Director, Epidemiology & Infection Prevention, Infectious Diseases
UCI School of Medicine

MedicalResearch.com: What is the background for this study?

• Antibiotic resistance, which occurs when germs like bacteria and fungi mutate to defeat the drugs designed to kill them, is a major public health threat.
• Data show that 40-50% of patients hospitalized with pneumonia receive broad spectrum antibiotics when they do not need them.
• Helping clinicians tailor antibiotic prescriptions to individual patients can improve patient outcomes by preserving healthy bacteria in the body and reducing the risk of future antibiotic resistance.

MedicalResearch.com: What are the main findings?

1. • The two newly published studies, involved over 220,000 patients with pneumonia or UTI in 59 HCA Healthcare hospitals. In half of the hospitals, clinicians were given algorithm driven computerized alerts with information about the best antibiotic match for an individual patient at the moment antibiotics were being prescribed.
   • Results showed:

1. 1. • 1. INSPIRE Pneumonia Trial: Compared with routine stewardship, the group using CPOE prompts had a 28% reduction in empiric extended-spectrum days-of-therapy (rate ratio 0.72, 95% CI:0.66-0.78, p<0.001).
        2. INSPIRE UTI Trial: Compared with routine stewardship, the group using CPOE prompts had a 17% reduction in empiric extended-spectrum days-of-therapy (rate-ratio 0.83, 95% CI 0.77-0.89 p<0.001).
        3. There was no difference in safety outcomes of ICU transfers or length of hospitalization between the two groups in either study

MedicalResearch.com: Can the algorithm take into account the patient’s risk of antibiotic resistance so that the right antibiotics can be started until cultures are known? 

1. • If a provider orders an extended-spectrum antibiotic to treat a patient for pneumonia or UTI in the first 3 days of hospitalization (empiric period), then the algorithm assesses the patient’s likelihood of having an infection due to an multidrug-resistant organism (MDRO).
   • If the risk of an MDRO infection is estimated to be low (<10%), then the provider receives a prompt informing them of their patient’s low risk and recommends a standard-spectrum antibiotic instead. 

MedicalResearch.com: How readily can the algorithm be incorporated into hospital records ie EPIC?

• • We defined the risk threshold to determine high vs low risk at 10% and found that most of the rules that inform the algorithms are remarkably simple, with only a few risk factors needed for programming.
  • Programming effort would need to focus on capturing these risk factors, most of which should be available in most hospital record systems.
    • For example, the top two risk factors that separated low from high risk from low risk patients was the patient’s history of MDRO and hospital MDRO prevalence.  These factors can be identified through existing data within the medical record.

MedicalResearch.com: What recommendations do you have for future research as a results of this study?

• We plan to evaluate the impact of antibiotic choices and risk for future development of MDROs, C. difficile, and other adverse outcomes.
• We also plan to evaluate how the prompts worked to change physician prescribing pattern.
• Finally we plan to evaluate the cost effectiveness of the intervention.

MedicalResearch.com: Is there anything else you would like to add? Any disclosures?

• In addition to showing the effectiveness of a prompt-based antibiotic stewardship intervention, this study also confirms that the vast majority of patients hospitalized for pneumonia and UTI do not grow MDROs and do not warrant use of extended-spectrum antibiotics.
• No disclosures

Citations:

• Gohil SK, Septimus E, Kleinman K, et al. Stewardship Prompts to Improve Antibiotic Selection for Pneumonia: The INSPIRE Randomized Clinical Trial. Published online April 19, 2024. doi:10.1001/jama.2024.6248
• Stewardship Prompts to Improve Antibiotic Selection for Urinary Tract Infection Shruti K. Gohil, MD, MPH; Edward Septimus, MD; Ken Kleinman, ScD; Neha Varma, MPH; Taliser R. Avery, MS; Lauren Heim, MPH; Risa Rahm, PharmD; William S. Cooper, PharmD; Mandelin Cooper, PharmD; Laura E. McLean, MEd; Naoise G. Nickolay, RPh; Robert A. Weinstein, MD; L. Hayley Burgess, PharmD; Micaela H. Coady, MS; Edward Rosen, BA; Selsebil Sljivo, MPH; Kenneth E. Sands, MD, MPH; Julia Moody, MS; Justin Vigeant, BA; Syma Rashid, MD; Rebecca F. Gilbert, BA; Kim N. Smith, MBA; Brandon Carver, BA; Russell E. Poland, PhD; Jason Hickok, MBA; S. G. Sturdevant, PhD; Michael S. Calderwood, MD, MPH; Anastasiia Weiland, MD; David W. Kubiak, PharmD; Sujan Reddy, MD, MSc; Melinda M. Neuhauser, PharmD, MPH; Arjun Srinivasan, MD; John A. Jernigan, MD, MS; Mary K. Hayden, MD; Abinav Gowda, BS; Katyuska Eibensteiner, BA; Robert Wolf, BS; Jonathan B. Perlin, MD, PhD; Richard Platt, MD, MSc; Susan S. Huang, MD, MPH
  JAMA

The information on MedicalResearch.com is provided for educational purposes only, and is in no way intended to diagnose, cure, or treat any medical or other condition.

Some links may be sponsored. Products are not endorsed.

Always seek the advice of your physician or other qualified health and ask your doctor any questions you may have regarding a medical condition. In addition to all other limitations and disclaimers in this agreement, service provider and its third party providers disclaim any liability or loss in connection with the content provided on this website.

Last Updated on April 27, 2024 by Marie Benz MD FAAD