Kidney Week

Abstract: FR-PO103

Acute Kidney Disease in Outpatient Setting: Epidemiological Insights from National Digitalized Big Data Practice

Session Information

• AKI: Epidemiology, Risk Factors, Prevention - I
  November 03, 2023 | Location: Exhibit Hall, Pennsylvania Convention Center
  Abstract Time: 10:00 AM - 12:00 PM

Category: Acute Kidney Injury

• 101 AKI: Epidemiology, Risk Factors, and Prevention

Authors

• Chiang, Hsiu-Yin, China Medical University Hospital, Taichung, Taiwan

Hsiu-Yin Chiang,

• Lin, Zi-Han, China Medical University Hospital, Taichung, Taiwan

Zi-Han Lin,

• Chang, David R., China Medical University Hospital, Taichung, Taiwan

David R. Chang,

• Kuo, Chin-Chi, China Medical University Hospital, Taichung, Taiwan

Chin-Chi Kuo,

Background

The burden of acute kidney disease in outpatients (AKD_OPT) is largely unknown due to health data silos and its asymptomatic nature. We established an Acute Kidney Injury Detection System (AKIDS) that integrates local electronic medical records and National Health Insurance Cloud to screen for AKD_OPT, and enable risk-based referrals in real-time.

Methods

AKIDS was integrated into the backend of outpatient HIS during 2017.12-2020.05. AKD_OPT within 180 days prior to a clinic visit is defined as the percent change between the max and min of (1) serum creatinine (SCr) >50%, or (2) eGFR >35%. AKD_OPT is considered deteriorating if the last two SCr decreased by >0.3 mg/dL; otherwise, it’s termed stable. We ensured a 1-year follow-up by including adult patients with the first AKD_OPT occurring during 2017.12-2019.06 and without ESRD or cancer. One-year composite kidney outcome (CKO) was defined as having ESRD or ≥40% drop of outpatient eGFR during follow-up. All-cause mortality was determined by National Death Registry. Multivariable Cox proportional modeling was used to estimate the risk of outcomes associated with AKD_OPT.

Results

The monthly cumulative incidence density of AKD_OPT was 6.9 to 8.1% (Figure). AKD_OPT patients had significantly higher 1-year CKO (1.8 vs 0.2%) and mortality (8.2 vs 1.4%), compared with those without. Of 79838 outpatients, 15.7% had AKD_OPT (12.7% stable; 3.0% deteriorating). Multivariable analysis found that AKD_OPT was significantly associated with increased 1-year risk of CKO by 5-fold (adjusted HR [aHR]=5.2; 95% CI=4.0-6.8) and mortality by 3-fold (2.6 [2.4-2.9]). The risk of 1-year CKO for deteriorating AKD_OPT (7.6 [5.5-10.5]) was higher than that for stable AKD_OPT (4.4 [3.3-5.8]).

Conclusion

By integrating national–local data and creating a kidney data ecosystem, we conducted the first study to fully characterize the epidemiologic profile of AKD_OPT—a obscured AKI phenotype. This study illustrates the potential of healthcare big data in transforming global approaches to AKD patterns and prevention.

Funding

• Government Support – Non-U.S.