Background <p>Chronic kidney disease (CKD) is a global public health concern, with 50–70% of the burden attributed to non-diabetic aetiology. To expand CKD research, there is a need to identify novel surrogate endpoints preceding cardiorenal outcomes that are commonly used in CKD trials. This study explored and quantified associations between intercurrent events and clinical outcomes in patients with non-diabetic CKD to inform potential surrogate endpoints.</p> Methods <p>In this retrospective observational cohort study, adults with non-diabetic, moderate-to-severe CKD (stage 3/4) were identified in the US Optum Clinformatics^® Data Mart healthcare claims database. Key outcomes were hospitalization for heart failure, kidney failure/need for dialysis, and worsening of CKD stage from baseline. Intercurrent events were defined as events observed in patient medical or pharmacy claims after the cohort inclusion date that either precluded a clinical outcome of interest or were associated with a modified risk of the respective outcome. Intercurrent events were selected a priori or by a data-driven exploratory approach. Associations between intercurrent events and clinical outcomes were explored and quantified using a Cox proportional hazards regression model.</p> Results <p>The study cohort included 504,924 patients. An outpatient heart failure diagnosis code was associated with an increased risk of consequent hospitalization for heart failure (hazard ratio [HR]: 12.92, 95% confidence interval [CI]: 12.67–13.17). CKD stage 4 diagnosis code was associated with an increased risk of kidney failure/need for dialysis (HR: 3.75, 95% CI: 3.69–3.81). Dispensation of potassium-removing resins and potassium-binding agents as an intercurrent event was associated with an increased risk of consequent worsening of CKD stage (HR: 4.83, 95% CI: 4.51–5.17). The estimated glomerular filtration rate decline in 295,174 patients with available laboratory data was associated with progressively increased risk of hospitalization for heart failure and kidney failure/need for dialysis.</p> Conclusions <p>Associations between intercurrent events and clinical outcomes in patients with non-diabetic CKD were investigated, quantified, and ranked using a large set of routinely collected data from a US claims database. Our approach may help identify novel surrogate endpoints that occur earlier in the disease course and could be leveraged as indicators of clinical outcomes in CKD research.</p>