[From the first symptoms to terminal renal failure: need for a nephrologic follow-up]

As patients with chronic renal failure are frequently referred late to nephrologists, we decided to quantify the magnitude of late referral and its consequences.

We studied retrospectively an inception cohort of 62 patients starting dialysis (either hemodialysis or continuous ambulatory peritoneal dialysis) during 1993 with a 4-year follow-up.

The mean delay between either first symptoms of renal disease, or first evidence of renal failure and nephrologist referral was 10 years and 3 years 56 days, respectively. About 47% of the patients were referred less than 6 months before starting dialysis, and 27.5% less than 1 month. Blood pressure levels were higher in patients referred less than 6, 3 and 1 month (P < 0.05), as was creatinine concentration in patients referred less than 1 month (P < 0.05). In contrast, plasma calcium was lower for referral less than 6 months (P < 0.05) and 3 months (P < 0.005), as was bicarbonate concentration for referral less than 3 and 1 month (P < 0.05). Initial hospitalisation stay was prolonged (x1.5) for late referral less than 3 months (56.4 +/- 39 days vs 35.9 +/- 33.6 days, P < 0.05) as was 6 months hospitalisation length for referral less than 3 months (x1.6) (52.9 +/- 40.6 days vs 33.2 +/- 28.7 days, P < 0.05) and less than 1 month (x1.8) (61 +/- 45 days vs 33.9 +/- 28.7 days, P < 0.05) and < 1 month (x1.8) (61 +/- 45 days vs 33.9 +/- 28.7 days, P < 0.05). Only 44.1% of patients started hemodialysis with a functioning arteriovenous fistula, and patients requiring temporary access had a 4.4-fold longer initial (60.1 +/- 41.7 days vs 13.6 +/- 11.6 days, P < 0.005) and 6-month (59.6 +/- 39 days vs 13.6 (9.1, P < 0.005) hospitalisation stay. The four-year mortality rate was unaffected by the delayed referral but strongly and independently predicted by age, diabetes and hypoalbuminemia.

Early nephrologic referral and timely initiated dialysis decrease morbidity at the start of dialysis and both hospitalisation length and costs.