Pre-dialysis hospital use and late referrals in incident dialysis patients in England: a retrospective cohort study.

BACKGROUND Timely referral to specialist kidney care can improve outcomes for patients and delay the onset of dialysis, yet late referral (LR) remains a problem in many countries. We aimed to estimate the proportion of LRs that could potentially have been detected earlier because of increases in patients' general hospital activity. METHODS A cohort of patients starting dialysis in the English NHS (National Health Service) during 2010/11 was approximated using hospital administrative data. The time between first recorded contact with a consultant nephrologist and starting dialysis was used to categorize the timeliness of referral. Monthly rates of inpatient activity prior to starting dialysis for both referral types were compared with the national average. RESULTS A cohort of 3928 patients was detected. One-third (34%) of the cohort started dialysis <90 days after their first referral to a nephrologist. Rates were higher for patients starting haemodialysis than peritoneal dialysis. The proportion of patients receiving their first dialysis as an emergency rises from 27% for those referred before 3 months to 67% for those referred on or after the day of starting dialysis. Half of the late referred patients (49%) have hospital activity rates more than double the national average (adjusted for age and sex) at 90 days before they start dialysis. CONCLUSIONS A substantial proportion of patients (49%) referred late for specialist kidney care have had regular contact with other hospital services. This could represent a missed opportunity to improve outcomes by timely management of their kidney disease.