Background: The population of incident dialysis patients is progressively ageing and dialysis outcome is particularly poor in the elderly. There is little documentation whether late referral is more frequent in the very elderly (> or =75 years) as compared with non-elderly patients and whether it contributes, at least in part, to their particularly poor outcome.
Methods: In a retrospective single center study we assessed all consecutive patients (n = 254) who had been admitted to haemodialysis between 1998 and 2001. Outcome in relation to the interval between the time of referral and start of dialysis was compared in very elderly and non-elderly patients. According to a previous analysis in our center major adverse outcome is seen in patients referred < or =8 weeks before the start of dialysis. For the present study this time interval was therefore operationally defined as 'late referral'.
Results: Expectedly 1 year after start of dialysis mortality was higher (31%) in the very elderly compared with younger patients (19%). The interval between referral and first dialysis was less in patients > or =75 years (median interval 3.5 weeks) compared with patients <75 years (median 20.5 weeks; P = 0.007). The difference in 1 year mortality between timely (>8 weeks) vs late (< or =8 weeks) referral, however, was as high in the very elderly (42% vs 16%) as in the younger patients (34% vs 9%). The relative risk of death conferred by late referral was also not significantly different in the very elderly (RR 1.80) compared with the younger (RR 2.32) patient. Using multivariate analysis timing of referral proves to be an independent factor with regard to the outcome and time of survival.
Conclusions: We conclude that late referral is more frequent in the very elderly. Although the relative risk of death conferred by late referral is similar in the very elderly and non-elderly, due to higher frequency of late referral it accounts for a large proportion of excess mortality in the very elderly.