Effect of a Diagnostic Algorithm for Urinary Tract Infection in General Practice

STUDY DESIGN AND SETTING Cluster randomized controlled single blinded study of a diagnostic aid in general practice RECRUITMENT OF GENERAL PRACTICES Recruitment of general practices was done through online advertisement in email newsletters for general practice, invitation by post of 200 practices and invitation of 44 general practices already participating in a medical audit project regarding UTI (ref til Gloria). Only practices in the Capital Region of Denmark could participate.

RANDOMIZATION AND BLINDING Randomization was performed using an online random number generator and an employee not part of the investigation team sent out the diagnostic algorithm to the intervention group. The practices were asked not to reveal their allocation to the investigators in case of contact. Allocation was concealed to the investigators until after analysis of the primary outcomes.

INTERVENTION AND CONTROLS The practices in the intervention group received a laminated diagnostic algorithm by post and got access to a smart-phone integrated web-page, which could calculate post-test probabilities based on pre-test probabilities and accuracy of diagnostic tests .The diagnostic algorithm consists of three parts, one about urine dipstick, one about point-of-care microscopy and one about urine culture . The practices were instructed, use of the algorithm was voluntary. Control practices did not receive a diagnostic algorithm and did not have access to the smart-phone integrated web-page.

RECRUITMENT OF PATIENTS The practices were told to register diagnostics and treatment on the first 20-40 patients presenting in general practice with symptoms of UTI regardless of age, sex and comorbidity. For each patient, the practice sent a urine sample to the microbiological department as reference.

INCLUSION AND EXCLUSION CRITERIA Inclusion criteria were all patients with symptoms of UTI who had not formerly participated in the study and where urinalysis was performed. The only exclusion criterion was acute admission to hospital.

DATA COLLECTION AND MANAGEMENT On the patient-level, the study was observational and anonymous. The practices registered clinical history, diagnostics and treatment using a case-report form where data on all patients fit into one sheet of paper. On day one (the day of consultation), clinical history, diagnostics, diagnosis and treatment were registered. On day two (the day after the consultation), result of the point-of-care urine culture, if such was performed, and the subsequent diagnosis and treatment were registered. The result of the reference urine culture was registered when the result was reported back to the practice on day 4-6. The intervention group could use the algorithm as they saw fit.

REFERENCE CULTURE Urine for the microbiological department was incubated in a standardized boric-acid container (Urine-Monovette®, Sarstedt) and collected by a transport service from the microbiological departments. At the microbiological laboratories (Herlev and Hvidovre), urine sample were analyzed on Inoqul A™ Bi-plate (CHROMagar and blood agar) with 10 μL on each half of the agar. The susceptibility pattern was determined on Mueller Hinton agars with disks containing mecillinam, trimethoprim, nitrofurantoin and sulfamethizol. Significant growth was defined as growth of ≥103 cfu/mL for E. coli and S. saprophyticus, ≥104 cfu/mL for other typical uropathogens and ≥105 cfu/ml for possible uropathogens in accordance with European consensus[21]. Plates with significant growth of more than two uropathogens were labeled as mixed cultures (inconclusive). The result was sent to practices within 4-5 workdays electronically and practices registered the result of the culture and susceptibility against mecillinam, trimethoprim, nitrofurantoin and sulfamethizol on the case-report.

ETHICS AND PATIENT SAFETY The study was presented to the ethical committee of Copenhagen and did not require ethical approval since the study was purely observational on the patient level and did not interfere with patient-treatment. Registration of patient data was done anonymously and did not require approval from the Danish data protection agency.