Abstract Background and Aims Patients with renal diseases often have progressive decline in eGFR, with many clinical researches having outcomes of reaching renal death/renal replacement therapy. In the real life, however, patients with eGFR under 15 mL/min/1.73m2 (the same unit hereafter) lose their lives more often by cardiovascular diseases before reaching endstage renal disease. Moreover, patients with eGFR below 30 are known to have miscellaneous metabolic derangements including phosphate retention, parathyroid hormone excess, metabolic acidosis, hyperkalaemia and renal anaemia. It appears to be extremely important to recognize that, in order to avoid enfacing plethora of metabolic alterations and cardiovascular disorders that progress after renal function falls below 30, we physicians need to take a step to find ways how to keep patients' eGFR above 30 for as long time as possible. At the same time, renal clinics are fully booked already; we need to prioritize the patients who have the highest risks in reaching eGFR below 30. This simulation study was performed to draw a trajectory to clarify who are the most eligible patients for nephrologists to intervene. Method Before performing the simulations using an inverse transform sampling, a priori statistical distributions were retrieved from the real-world data, i.e., a whole-hospital-wide survey over 4 years; (a) distribution of eGFR: obtained from patients with at least 3 measurements of eGFR over 366 days or more, where a median of eGFR values was used to represent the patient's renal function, (b) distribution of eGFR slopes: similarly retrieved, which were then stratified in 6 strata with each eGFR range of [30,35)/[35,40)/[40,45)/[45,50)/[50,55)/[55,60). The simulation was performed in 4 steps. In the first step (“random number generating step”), 65,000 pairs of random numbers (U1, U2) were generated using a continuous uniform distribution, U[0,1]. In the second step (“eGFR generating step”), eGFR was generated using the clinically-retrieved distribution (a, above) from a random number U1. In the third step (“slope generating step”), a slope in eGFR decline was generated using the a priori stratified distribution (b, above) from a random number U2. In the final step (“estimation step”), renal prognosis, especially an eGFR value 10 years later, was estimated in each of the “65,000 simulated patients”, with or without an effect of nephrology consults. The primary outcome was defined as reaching the eGFR below 30 within 10 years. If this occurred, time until the eGFR 30 was also calculated. Results In the strata of simulated patients with eGFR’s [50,55) and [55,60), 35.3% and 34.9% reached their eGFR below 30, over the periods of 4.78 and 4.95 years in average. In these strata, there were marked differences in renal prognosis based on the eGFR slope, where all the patients with the fastest quartile (annual rate of decline faster than 3.09) reached the outcome, while only 4–11% of the patients with slower decline (the remaining 3/4) reached eGFR <30. The number of patients was the largest in the stratum of [55, 60). In the stratum of eGFR [45,50), 45.3% patients, including 100% of faster quartile and 24.5% of slower 3 quartiles, reached eGFR of <30. Assuming the effect of nephrology consultation as attenuating the slope by 1/3, nephrologist intervention most prominently influenced the prognosis of the patients in strata [40,45) and [45,50) in terms of proportion of patients reaching outcomes (decreases by 10.5% and 12.5%). Conclusion 1) Reaching an eGFR of <30 mL/min/1.73m2 is a newly proposed outcome in clinical research with substantial significance such as avoiding cardiovascular events and metabolic derangements that occur frequently at eGFR below 30. 2) Patients in the fastest quartile in eGFR slope, even those whose baseline eGFR 50–60, are the candidates who merit nephrology consultation.
Read full abstract
Oct 18, 2023
Biona Benny + 5
Open Access
