The Impact of De-implementing Urine Dipsticks for Diagnosis of UTIs in Hospitals

Overview

The goal of this interrupted time-series analysis is to evaluate the impact of the de-implementation of urine dipsticks as a diagnostic tool for urinary tract infections (UTIs) in hospitalized patients in the North Denmark Region. The main question it aims to answer is: How does de-implementation of urine dipsticks affect the diagnosis and management of UTIs and related disorders? Specifically, does it change the following parameters: * Number and severity of UTI infections (lower and upper UTI, non-severe and severe) * Antibiotic prescription (overall, antibiotic classes, administration routes, duration, dosages) * Number of urine cultures and number of positive urine cultures * Risks of admission to intensive care units and 30-day mortality * Risk of drug toxicity * Length of hospital stay * Risk of admission to intensive care unit * 30-day risk of readmission after discharge * 6-month risks of Clostridioides difficile enterocolitis and de novo antimicrobial resistance in cultures obtained during routine clinical care. Researchers hypothesize that de-implementing urine dipsticks will lead to a reduced frequency of diagnosed cystitis, reduced antibiotic use, and fewer urine cultures without negatively affecting patient mortality or readmission risk. Researchers will compare the outcomes before and after the discontinuation of urine dipsticks across hospitals in the North Denmark Region. Furthermore, results will be compared to another Danish administrative healthcare region where dipsticks are still in use as well as urine culture data from the primary sector in the North Denmark Region. Since this is a registry-based observational study utilizing data from the electronic patient record system in the North Denmark Region, no direct contact will be made with participants.

BACKGROUND: Urinary tract infections (UTIs) are a leading cause of antibiotic treatment and hospitalizations in Europe and the United States. Urine dipsticks are a widely used point-of-care test in hospitals, with over half of admitted patients in the North Denmark Region being screened using this method. They are considered an effective tool to rule out UTIs and aim to save resources by reducing unnecessary urine cultures and guiding initial antibiotic choice when used correctly. However, for reliable results, urine dipsticks require specific preanalytical conditions, such as a clean midstream sample retained in the bladder for 4-6 hours before voiding. These criteria are rarely met in clinical practice. Moreover, the prevalence of asymptomatic bacteriuria among hospitalized patients is about 45% which requires careful interpretation and correlation with specific symptoms of UTI by the clinician. Although designed to rule out UTIs, studies suggest that urine dipsticks are often misused to rule them in, despite their low positive predictive value in hospital settings. This misuse may contribute to excessive antibiotic use. Furthermore, despite their generally high negative predictive value, a recent study in the North Denmark Region found that negative dipsticks cannot reliably exclude UTIs in symptomatic patients. Due to these limitations, urine dipsticks were de-implemented throughout the North Denmark Region in August 2023 and February 2024. In September 2024, the Danish Society of Infectious Diseases and the Danish Society of Clinical Microbiology recommended discontinuing dipstick use in all Danish hospitals. However, the impact of de-implementing such a widely used point-of-care test remains unknown, raising questions about whether the advantages of this decision outweigh its potential drawbacks. AIM This study aims to assess the impact of removing urine dipsticks in hospitals in the North Denmark Region. HYPOTHESES 1. The incidence of urine cultures will decrease after the de-implementation of the urine dipstick. 2. The incidence of the ICD-10 codes, DN300 (acute cystitis), DN308C (recurrent cystitis), DN309 (unspecific cystitis), and DA419B (urosepsis) will decrease after the de-implementation of the urine dipstick. 3. The overall use of antibiotics as well as UTI-specific antibiotics, pivmecillinam, nitrofurantoin, sulfamethizol, and trimethoprim, will decrease after the de-implementation of the urine dipstick. 4. Risk of admission into the intensive care unit will remain unchanged before and after the de-implementation of the urine dipstick. 5. The average length of hospital stay and mortality rate will remain unchanged before and after the de-implementation of the urine dipstick. 6. Risk of early readmissions, defined as a return to an inpatient unit or emergency room within 30 days after the initial discharge, will remain the same before and after the de-implementation of the urine dipstick. In a subpopulation of patients with positive urine cultures, the investigators hypothesize that: 1. The mortality rate and average length of stay will remain unchanged after the de-implementation of the urine dipstick. 2. The proportion of patients treated with antibiotics with uropathogenic coverage will remain the same before and after the de-implementation of the dipsticks (the UTI-specific antibiotics presented above and/or ampicillin/gentamicin, piperacillin/tazobactam or ciprofloxacin). METHODS Design This study will be a registry-based interrupted time-series analysis utilizing data from the Business Intelligence Portal in the North Denmark Region. Data before and after the abolishment of the dipsticks will be extracted. Setting Exposure The de-implementation of urine dipsticks in North Denmark Region in August 2023 (Aalborg) and February 2024 (Hjørring). Study period Data will be extracted from both emergency rooms and inpatient units. The data extraction will start when the Business Intelligence Portal can extract data (preferably from 2019), and end at the latest update of conducting the study. This study will include two control groups: 1. A comparable hospital cohort from the Central Denmark Region. 2. The total number of urine cultures requested by general practitioners in the North Denmark Region during the study period. STATISTICS To avoid duplicates, the primary analysis will only include each patient's first admission during the study period. Descriptive statistics: Incidence rates Since hospitalized patients are an open cohort, the incidence rates (e.g. of antibiotic use), will be calculated using the following formula: Monthly incidence rate of pivmecillinam use=((Number of unique persons treated with pivmecillinam))/(Amount of total person time in a month)=X pr.1000 patient days Confidence intervals of the incidence rates will be calculated based on the Poisson distribution, and the standard error will be calculated as: Standard error (log⁡〖rate)=1/√(number of cases)〗 The total incidence rates will be standardized using direct standardization based on the age distribution of the patient population as of September 2024 (the latest available data at the time of writing this protocol). Mortality Mortality rates will be calculated using the same methodology as presented above. In a supplementary analysis, the investigators will calculate the 30-day mortality risk: 30-day mortality risk=(Number of deaths within 30 days)/(Total number of admissions)=X % And the case-fatality rate: Case fatality=(Number of deaths in the hospital)/(Total number of admissions)=X % Charlson Severity Index The medical history extracted up to 10 years before the index date, will be extracted to calculate the Charlson Comorbidity Index using standard methodology. Comparisons The incidence and mortality rates in the North Denmark Region before and after the de-implementation of the dipstick will be compared using the principles of interrupted time series regression. This methodology offers the advantage of adjusting for seasonality. Furthermore, the rates in the North Denmark Region will be compared to those in the control groups.