MO474: Expectation and Acceptance of a Clinical Decision Support Software by Nephrologist End-Users: The Ckdnapp Survey

Abstract

Abstract BACKGROUND AND AIMS Chronic kidney disease (CKD) is a major public health problem. CKD constitutes a complex disease due to differing underlying disease etiologies in each patient, which can in turn lead to many complications, comorbidities and polypharmacy. Monitoring disease progression and personalized treatment efforts are crucial for optimal long-term patient outcomes. In order to achieve this, physicians need to integrate different levels of data, e.g. clinical/demographic parameters, biomarkers and drug information, with medical knowledge. Clinical decision support systems (CDSS) can tackle these issues and improve patient management. ‘CKDNapp’ (CKD Nephrologist App), a CDSS application for nephrologists, based on mathematical models using machine-learning techniques, is currently being developed (https://ckdn.app). CKDNapp is intended to become a tool for daily clinical use. The nephrologists’ attitude towards any CDSS and CKDNapp in particular is of prime importance for its successful implementation into the daily medical routine. This survey investigates nephrologists’ experiences with CDSS in general and their expectations towards a reliable and useful application supporting their daily medical routine. METHOD CKDNapp survey is ongoing and has been conducted by telephone or as a do-it-yourself online interview in the form of a 38-item questionnaire. The answers of nephrologists from all regions across Germany were collected and analyzed using the Electronic Data Capture System, RedCap [1]. CKDNapp survey is divided into four modules: (1) experiences with CDSS, (2) expectations of a helpful CDSS, (3) evaluation of the planned contents of CKDNapp and (4) ethical aspects of CDSS (in collaboration with the BMBF-funded DESIREE project; https://www.desiree-forschung.de/desiree/index.php). All questions were based on a literature search for questionnaire items on CDSS [2, 3]. Response formats include the Likert scale or multiple choice. Descriptive statistical analyses of all questions were calculated. RESULTS In total, 44 participants took the survey, and completeness of answers ranged from 85% to 100%. Participants were aged 51–60 years old, male (64%) and had been working in nephrology outpatient clinics for a median of 12 years. Nephrologists treated a median of 35 patients/day. A total of 85% of participants reported never or rarely use a CDSS in patient care. The most frequently given reason for this was a lack of knowledge about CDSS. Nevertheless, 79% of participants believed CDSS to be helpful in the management of patients with CKD and 71% would be willing to use a CDSS given the chance to do so. When rating the importance of planned CKDNapp features, prediction of CKD progression (97%, Figure 1) and in-silico simulations of disease progression when changing, e.g. lifestyle or medication (97%) were most important, followed by the need for integration of available CKD guidelines (95%), prediction of acute kidney injury (95%), prediction of mortality risk (80%), and easy access to patient information (76%). The spectrum of answers to ethical aspects of CDSS (utility of CDSS for experienced versus inexperienced nephrologists, aspects of machine learning Fig. 2, discrimination of minority groups, etc.) was diverse. CONCLUSION This survey provides insights into experience with and expectations of outpatient nephrologists on CDSS in general and CKDNapp in particular. Despite the current lack of knowledge on CDSS, the willingness to integrate CDSS into daily patient care and the evaluation of planned CKDNapp features was high.