#565 I will present the results of a novel automated method for screening chronic kidney disease (CKD) in primary care, leveraging artificial intelligence and adherence to established clinical guidelines

Abstract Background and Aims Chronic kidney disease (CKD) is a significant global health concern due to its high prevalence and associated complications. Often undiagnosed until advanced stages, early detection is crucial for effective management. The KDIGO and National Scientific Societies guidelines provide evidence on which patients should be screened. Primary Care Centers (PCCs) can play a vital role in identifying undiagnosed CKD, but they require tools to assist in this process. Artificial intelligence (AI) is transforming clinical laboratories, and one of its applications is Clinical Decision Support Systems (CDSS). These digital solutions are designed to assist healthcare professionals in making clinical decisions and can be instrumental in opportunistic screening to facilitate early CKD detection and management. The objectives were to explore a new method to screening for CKD in PCC patients using a CDSS, and to present the results of a pilot study. Method A cross-sectional intervention study was conducted from August 1st, 2023, to December 31, 2024, in patients visiting PCCs for any reason and requiring laboratory tests. The study was approved by the Hospital Ethics Research Committee. The pilot study was carried out at the San Juan PCC, serving a population of 26,129 inhabitants (52% female, 19% aged over 65 years). The PCC employs 15 doctors and 24 nurses and handles 199,816 primary care consultations annually. Our laboratory has implemented a CDSS as a rule engine (AlinIQ CDS v8.2; Abbott, Chicago, IL) to manage pathways and actions. The CDSS allows identification of requests including a specific test in current and recent previous requests. In addition, the CDSS integrates with clinical databases, such as electronic medical record, to get information on all current and previous patient diagnoses. If certain conditions are detected in its rules engine, CDSS is capable of generating alerts or adding tests. Figure 1 illustrates the implemented workflow. The CDSS identified patients eligible for potential CKD screening based on inclusion criteria that implemented a Spanish National Scientific Societies consensus document: patients from the local health catchment area and aged ≥60 years or having specific comorbidities (hypertension, diabetes, or coronary heart disease). Exclusion criteria included age <18 years, previous CKD diagnosis, pregnancy, or prior screening within one year. The CDSS offered clinicians, via real-time pop-ups, the option to assess CKD risk and add urine albumin:creatinine ratio (UACR) and estimated glomerular filtration rate (eGFR) tests. Laboratory reports provided CKD risk assessments and management recommendations aligned with KDIGO guidelines. The reagent costs for UACR were 0.26€ for strip analysis and 0.62€ for quantification. Costs were divided by the number of patients classified as KDIGO low and high risk and further divided by high-risk patients to evaluate two key performance indicators: cost per patient screened and cost per patient identified as high CKD risk. Results During the study period, 18,377 requests were analyzed, and 1,515 patients were flagged for CKD screening. This represented 8.2% of requests and 5.8% of the population. A total of 1,318 patients (87.0%) accepted and provided samples (mean age 62.4 ± 11.7 years; 53.0% female). The CDSS classified 1,222 patients (92,7%) as low CKD risk. Ninety-six patients (7.3%) were identified as moderate risk. Of these, eleven patients were directly classified as CKD, while 85 required confirmation of values, with 43 ultimately confirmed. Overall, fifty-four patients (4.1%) were identified with CKD (47 were recommended for managed in PCC, while 7 follow-up in nephrology). The cost analysis showed an expense of 0.51€ per screened patient. The total reagent expenditure was 671.28€. The cost per identified CKD patient was 12.4€. Conclusion Opportunistic CKD screening is feasible, cost-effective, and well-received. The acceptance rates and significant identification of high-risk patients support CKD detection and management. Future research should expand this approach and evaluate long-term impacts.