Outpatient visit intervals in chronic kidney disease: adherence to Kidney Disease: Improving Global Outcomes recommendations and determinants of variation

Vollständiger Abstract

Worum geht es in dieser Arbeit?

<h4>Background</h4>Chronic kidney disease (CKD) is a major global health problem owing to its progressive nature, complications, and substantial healthcare burden. The Kidney Disease: Improving Global Outcomes (KDIGO) guidelines recommend outpatient visit intervals based on CKD severity; however, adherence to these recommendations and the factors influencing the visit interval remain unclear.<h4>Methods</h4>This retrospective observational study included 13,621 patients with CKD treated at a tertiary hospital in South Korea. Using electronic medical records, we used linear regression and Shapley Additive exPlanations (SHAP), a method for interpreting model prediction, to assess the influence of clinical, demographic, and healthcare utilization factors on outpatient visit intervals. Key variables included the CKD stage, comorbidities, age, distance to the hospital, and physician assignment.<h4>Results</h4>Visit intervals shortened as CKD severity increased, although variability was substantial. Adherence to KDIGO-recommended intervals was higher in earlier CKD stages, but declined in intermediate and advanced stages, except for stage G5-A3, where 90% of visits met guidelines. Shorter intervals were associated with advanced stage, older age, greater healthcare use, and shorter distance to the hospital. SHAP analysis confirmed these predictors, identifying CKD stage as the dominant factor and revealing notable physician-level variation.<h4>Conclusion</h4>CKD stage was the primary determinant of outpatient follow-up intervals, but adherence to KDIGO recommendations varied across stages. Visit schedules were also affected by age, comorbidities, healthcare utilization, accessibility, and physician practice. These findings underscore the need for tailored follow-up strategies that reflect both disease severity and clinical context.

Abstract: PubMed · Datensatz