Mackenzie Keintz, MD
Image credits: UNMC
Ambulatory antibiotic stewardship has focused mainly on respiratory infections, overlooking common conditions like urinary tract infections (UTIs), which make up a significant portion of outpatient antibiotic prescriptions. To address this, researchers developed and validated an antibiotic appropriateness metric for UTIs to improve prescribing practices, reduce unnecessary use, and combat antimicrobial resistance.
During the validation period from October 2023 to March 2024, 575 clinic visits were identified for uncomplicated UTI, pyelonephritis, or asymptomatic bacteriuria, with 98% of these visits classified as uncomplicated UTI. A total of 572 antibiotic prescriptions were reviewed, and the metric correctly identified appropriate antibiotic use in 99.8% of cases, including the correct agent, dose, and duration. Two prescriptions for diagnoses other than UTI were erroneously included, and one antibiotic duration was incorrectly captured. The metric successfully adjudicated the appropriateness of prescriptions based on the defined variables for agent, dose, and duration.
Of the 575 visits, 80 (14%) were incorrectly coded, with the majority (78) being uncomplicated UTIs that were either misclassified as complicated or lacked documented symptoms. Additionally, only 3% of urine cultures revealed pathogens resistant to first- or second-line antibiotics.
The study utilized ICD-10 codes to identify visits for UTI, pyelonephritis, and asymptomatic bacteriuria across 17 clinics within an academic health system. Appropriateness was defined based on three key components: antibiotic choice (first- or second-line agents), correct dose, and proper duration. SQL Server Management Studio was used to automate categorization of prescriptions as appropriate or inappropriate according to these criteria.
To validate the metric, six months of data were reviewed manually to verify the accuracy of both the identified cases and coding decisions. The manual review also assessed whether clinical nuances, such as allergies, renal function, or history of resistant organisms, might affect appropriateness classification.
In an interview, Mackenzie Keintz, MD, and her co-authors shared insights into the development of this metric. She emphasized the disparity between inpatient and outpatient stewardship efforts, “Ambulatory stewardship has lagged behind hospital-based stewardship programs despite the majority of antibiotic usage occurring outpatient. While progress has been made in monitoring antibiotic use for acute respiratory infections, common conditions like UTIs and skin infections have not received as much attention.”
Keintz noted that understanding current prescribing practices for UTIs was crucial: “We need to understand current prescribing practices to determine what work needs to be done to ensure appropriate antimicrobial prescribing for UTI. This was more challenging than acute respiratory infection metrics, which look at appropriate non-prescribing because antibiotics are usually required for UTI.”
ICD-10 coding, while standard, posed challenges. “There were several challenges to utilizing ICD-10 codes to identify urinary tract infections for our metric,” said Keintz. “The first was coding accuracy. In our validation, we found some cases of complicated urinary tract infection coded as simple cystitis, which led to inaccuracies in the appropriateness measure.” Additionally, she pointed out cases where UTI was coded but no symptoms were documented, which could suggest incomplete documentation or miscoding of asymptomatic bacteriuria. “There is no easy solution to these challenges,” she added. “However, we plan to implement coding education into our planned academic detailing about the measure.”
Keintz also acknowledged the limitations of ICD-10 codes in reflecting clinical nuances that might alter antibiotic decision-making, such as allergies or renal function: “ICD-10 coding is also limited in identifying clinical nuances that may alter antibiotic decision-making, such as patient allergy, renal function, or history of resistant organisms. Without in-depth chart review, these factors are difficult to identify, and such reviews are not feasible on the scale needed to create an ambulatory stewardship metric.”
To address this, the researchers reduced the metric goal for clinicians: “We chose to reduce the metric goal for clinicians to account for these nuances. The goal of ambulatory antibiotic stewardship is improvement rather than perfection.”
What You Need To Know
Researchers developed and validated an antibiotic appropriateness metric for UTIs in ambulatory settings to enhance prescribing practices and reduce unnecessary antibiotic use.
The study revealed challenges in ICD-10 coding, including misclassified diagnoses and incomplete documentation, which can impact the accuracy of the metric.
The metric has the potential for widespread adoption across institutions, helping standardize stewardship efforts and improve antibiotic prescribing in outpatient care.
Another challenge the researchers encountered was symptom-based codes, like dysuria, used by some clinicians instead of diagnostic codes like UTI. “We evaluated these codes; however, they were not specific enough to provide meaningful feedback to clinicians. Therefore, we chose to exclude symptom-based codes to ensure the metric captures only urinary tract infections to provide the most accurate data for clinician improvement.”
Looking beyond their study, Keintz envisions how the metric can benefit other institutions: “The envisioned application of this metric by other institutions involves its integration into existing ambulatory antimicrobial stewardship programs that may not have resources to develop a metric on their own. By adopting this standardized metric using readily available coding data, institutions can systematically evaluate and improve their antibiotic prescribing practices, leading to better patient outcomes and a reduction in antimicrobial resistance.”
Keintz sees the metric as a tool for ongoing improvement: “The metric serves as a tool for benchmarking and identifying areas needing improvement, thereby facilitating targeted interventions. Ultimately, widespread adoption of this metric may contribute to more consistent and appropriate antibiotic use across various healthcare settings.”
The antibiotic appropriateness metric for UTIs in ambulatory settings has been successfully validated, offering a structured approach to improve prescribing practices. While challenges like coding inaccuracies and clinical nuances remain, the metric helps support antimicrobial stewardship efforts and reduce resistance by identifying areas for improvement. Future versions may incorporate additional clinical factors to enhance its accuracy and broader applicability.
