Urinary Protein-to-Creatinine Ratios during Hospitalization Are Not Invalidated by AKI

Background: Experts have warned that urine protein to creatinine ratio (UPCR) measurement is invalid during acute kidney injury (AKI) because creatinine levels in the blood and urine are rapidly changing. The theoretical concern is that the UPCR will be falsely elevated when the serum creatinine is rising since the urine creatinine in the denominator of the UPCR will be lower during this period. While some anecdotal experiences seem to corroborate this theoretical concern, the issue has not been systematically investigated using real-world clinical data. Methods: We examined adults hospitalized with stage 2 or 3 AKI (defined as peak:nadir serum creatinine during hospitalization ≥2) at the University of California, San Francisco 1/1/2014 to 12/31/2021. Between a pair of UPCRs measured in the same patient, we used linear regression to look for associations between the difference in UPCR values and the difference in serum creatinine slopes at the time of UPCR collection. We also calculated test characteristics of inpatient UPCR cutoffs to predict outpatient UPCR cutoffs. Results: At least one UPCR was measured during hospitalization in 19% of the AKI patients in our cohort. Among 329 patients with multiple UPCRs measured during AKI hospitalization (median difference in serum creatinine slopes 0.37 mg/dL per day), the UPCR value with the steepest upward serum creatinine at the time of measurement was similar to the UPCR value with the steepest downward serum creatinine (median difference in UPCRs 0.06 g/g, 95% CI -0.26 to 0.50 g/g). There was no association between the difference in serum creatinine slopes when the UPCRs were collected and the difference in UPCR values (UPCR 0.05 g/g higher per mg/dL/day serum creatinine slope, 95% CI -0.36 to 0.47, p=0.80). This remained true in a subset analysis of those with serum creatinine slope differences > 1 mg/dL per day. An inpatient UPCR ≤ 3.0 g/g ruled out outpatient nephrotic-range proteinuria (> 3.0 g/g) more than 90% of the time (inpatient UPCR ≤ 0.15 g/g ruled out outpatient proteinuria >0.15 g/g > 79% of the time). Conclusion: UPCRs were commonly measured during an episode of AKI and were not systematically affected by the serum creatinine slope at the time of UPCR collection. UPCRs measured during an episode of AKI remain informative despite non-steady state serum creatinine at the time of collection and should not be wholly disregarded. Funding: NIDDK Support