#4381 incidence and frequencies of hyperkalemia in patients with chronic kidney disease: a retrospective database study from us clinical care

Abstract Background and Aims Hyperkalemia (HK) is associated with significant risks for premature mortality, adverse clinical outcomes, and with a potentially more rapid decline of renal function. Reported numbers on the epidemiology of HK in patients diagnosed with chronic kidney disease (CKD) vary significantly and are unclear in real-world patients across different CKD stages, eGFR and UACR strata. We aimed to evaluate the occurrence of moderate to severe hyperkalemia episodes among a large, representative sample of patients with CKD from the US between 2009 and 2020. Method Using Optum Electronic Health Records data, we included patients with CKD between January 2009 and December 2020, who had at least two qualifying estimated glomerular filtration rate (eGFR) (15-60 ml/min/1.73 m2) and/or UACR values (≥30 mg/g) between 90 and 365 days apart. The second qualifying value was considered confirmatory and set as the index date. Patients had to have 365 days of baseline activity prior to index, be ≥18 years of age, and not show any diagnoses, procedures, or lab values of kidney failure or hemodialysis or kidney transplant during baseline. A HK episode was defined in two ways, (1.) as either a combination of two elevated inpatient or outpatient serum potassium values (sK+) ≥5.5 mmol/l, not longer than 7 days apart, or (2.) a combination of one elevated sK+ and the initiation of pharmacotherapy (e.g., i.v. calcium or insulin-glucose, nebulized albuterol, potassium binders) or a diagnostic code for HK, not longer than 3 days apart. We calculated relative frequencies and incidences of HK in the overall CKD study population and specific subgroups of interest. Results 1,771,900 patients met our selection criteria for CKD, with advanced stages 3 and 4 predominantly represented (85.7% and 10.2%, respectively). The cohort consisted of 57.7% females, 83.8% Caucasian and 9.8% African American. Most common baseline comorbidities were hypertension (68.5%), hyperlipidemia (55.1%) and T2D (34.2%). 69.3% of patients were prescribed antihypertensives, 45.7% statins and 45.6% antiarrhythmics. 99.1% of patients had at least one baseline potassium measurement with values averaging 4.3 mmol/L (median, 4.3; IQR, 4.0-4.55). During an average follow up 3.9 years, 220,339 (12.4%) patients experienced at least one episode of hyperkalemia. Of those, 69.3%, 17.5%, and 13.2% showed one, two, and three or more HK episodes, respectively. Across all patients with CKD, the mean incidence rate was 3.37 (95% CI, 3.36-3.38) cases per 100 patient years (PYs). HK incidence correlated with lower eGFR and increased UACR values, with rates of 1.32 (1.25-1.39), 2.48 (2.40-2.55), 3.00 (2.99-3.01) and 8.80 (8.71-8.88) cases/100 PYs for patients with CKD stage 1, stage 2, stage 3 and stage 4, respectively. Highest incidence rates (13.81; 12.96-14.70) were found in patients with UACR values ≥3500, irrespective of their eGFR value. In addition, across disease-related subgroups, significantly higher incidence rates were found in patients co-diagnosed with T2D (5.43; 95% CI, 5.40-5.47) and HF (8.70; 8.62-8.77), and sMRA users (7.66; 7.57-7.76) at baseline. Conclusion Our contemporary findings demonstrate that HK is common in patients with CKD undergoing routine clinical care in the US, and it is notable in patients with reduced eGFR and elevated UACR. In addition, HK was more predominant in patients with T2D, heart failure or sMRA use, emphasizing a need for more routine sK+ monitoring in patients with these risk factors. Further research is needed to assess additional intrinsic risks, clinical consequences and management approaches of HK in patients with CKD to further inform clinical practice.