Application of a Prediction Model for Directing Antibiotic Use in the Treatment of Urinary Tract Infection in an Ambulatory Setting

Urinary tract infection (UTI) is when bacteria enter the urinary system and cause an infection. UTIs cause symptoms including burning when peeing, a feeling of an increased urge to pee, and cloudy or strong-smelling urine. Sometimes, severe UTIs can also cause fever, abdominal pain, and/or lower back pain. In the emergency department (ED), healthcare providers rely on symptoms, along with a urine analysis and a urine culture to diagnose a UTI. A urine analysis involves taking a sample of urine and analyzing different factors like color, acidity, presence of blood cells, presence of bacteria. An abnormal urine analysis increases the likelihood that patients might have a UTI, but it does not confirm it. A positive urine analysis will lead to provider's sending a sample of urine for a urine culture. A urine culture is used to grow whatever bacteria is in the collected urine. If growth is seen on the culture, then this confirms a patient has a UTI. This also specifies which bacteria grew on the culture. The lab can also take it a step further and do an antibiotic test to check which antibiotic the bacteria is sensitive to. When a urine analysis comes back abnormal in an ER setting, patients are prescribed an antibiotic before the culture and antibiotic sensitivity tests come back. If a patients condition is not critical, they will be discharged home before the culture results come back. If the culture comes back positive, the pharmacists will evaluate the culture and antibiotic sensitivity tests, then call patients to inform them whether they are taking a suitable antibiotic. This study aims to decrease the unnecessary use of antibiotics because this contributes to antibiotic resistance which is considered a global public health issue. Antibiotic resistance occurs when bacteria develop the ability to withstand certain antibiotics that used to be effective against them, which makes it difficult to treat the infection. One of the factors that increase the risk of antibiotic resistance is the overuse of antibiotics. In this study, investigators will be incorporating a prediction model and a negative callback system to decrease unnecessary antibiotic use.