Kidney Week

Abstract: FR-PO027

Kinetic eGFR Can Predict AKI Recovery Earlier, Especially in Patients with CKD

Session Information

• AKI: Clinical, Outcomes, Trials - I
  October 26, 2018 | Location: Exhibit Hall, San Diego Convention Center
  Abstract Time: 10:00 AM - 12:00 PM

Category: Acute Kidney Injury

• 102 AKI: Clinical, Outcomes, and Trials

Authors

• Uehara, Keita, Naha City Hospital, Naha, Okinawa, Japan

Keita Uehara,

• Uema, Takahito, Naha City Hospital, Naha, Okinawa, Japan

Takahito Uema,

• Miyara, Tadashi, Naha City Hospital, Naha, Okinawa, Japan

Tadashi Miyara,

Background

The kinetic estimated glomerular filtration rate eGFR (KeGFR) obtained via consecutive serum creatinine (sCr) values is reportedly useful for predicting renal recovery after acute kidney injury (AKI). However, no studies have determined which patients can benefit using the KeGFR model. We predicted recovery from AKI with consecutive KeGFR and sCr values, comparing the usefulness of KeGFR.

Methods

This retrospective cohort study included patients admitted to our intensive care unit (ICU) from April 2014 to March 2016 and diagnosed with AKI after ICU admission. We defined renal recovery day as the first day when increasing consecutive sCr values turn to decrease or decreasing consecutive KeGFR values turn to increase, comparing each recovery length from AKI. We examined patient characteristics when the KeGFR model was superior in predicting recovery from AKI.

Results

During the study period, 972 patients were admitted to our ICU and 74 were finally studied. On admission, median age was 75.5 (IQR: 64.0-83.0) years, median baseline sCr was 0.81 (IQR: 0.63-1.04) mg/dL, and median baseline eGFR was 65.0 (IQR: 51.1-77.6) mL/min/1.73 m². AKI recovery length was statistically shorter in KeGFR model than sCr model (median [IQR], 1.0 [1.0-2.0] vs 1.0 [1.0-2.0], p<0.01). In 17 (23.0%) patients, using the KeGFR model predicted AKI recovery earlier than when using the sCr model. Among those whose AKI recovery was predicted better than when using the sCr model, the baseline sCr was likely higher (median [IQR], 0.80 [0.61-0.93] vs 0.97 [0.79-1.42], p<0.01), baseline eGFR was likely to be lower (median [IQR], 68.0 [54.2-80.9] vs 51.6 [35.8-64.1], p=0.01), and the proportion of patients with CKD was higher (n [%], 18 [31.6] vs 13 [76.5], p<0.01). Other factors such as age, sex, mean blood pressure, sequential organ failure assessment score, infusion volume, and body weight were not associated with predicting early AKI recovery using the KeGFR model.

Conclusion

KeGFR can predict recovery from AKI earlier, especially in patients with a higher baseline sCr, lower baseline eGFR, and CKD. Using the KeGFR model, we can confirm the efficacy of our AKI therapy early, which would enable prompt and accurate management of patients with CKD.