UK Renal Registry 12th Annual Report (December 2009): chapter 13: the UK Renal Registry advanced CKD study: frequency of incorrect reporting of date of start of RRT

Abstract

Background: A preliminary review of the UK Renal Registry (UKRR) pre-RRT study data revealed results suggesting that, for some patients, the date of start of renal replacement therapy (RRT), as reported to the UKRR, was incorrect and often significantly later than the true date of start. A more detailed study then aimed to validate a set of criteria to identify patients with an incorrect start date.

Methods: Pre-RRT laboratory data were electronically extracted from 8,810 incident RRT patients from 9 UK renal centres. Any patient with a low urea (<15 mmol/L) at the start of RRT or with a substantial improvement in kidney function (either a fall in urea >10 mmol/L or rise in eGFR >2 ml/ min/1.73 m) within the two months prior to RRT were considered to potentially have an incorrect date of start. In 4 selected centres, the electronic patient records of all patients flagged were reviewed to validate these criteria.

Results: Of 8,810 patients, 1,616 (18.3%) were flagged by the identification criteria as having a potentially incorrect date of start of RRT, although a single centre accounted for 41% of the total flagged cohort. Of these flagged patients, 61.7% had been assigned an incorrect date of start of haemodialysis (HD), 5.7% had evidence of acute RRT being given before the reported date of start of HD and 9.2% had evidence of starting peritoneal dialysis exchanges prior to the reported date of start. Of those flagged, 10.7% had a correct date of start of RRT.

Conclusions: Accurate reporting of RRT episodes is vital for the analysis of time dependent studies such as survival or time to transplantation. A proportion of patients starting RRT were assigned an incorrect start date. In order to improve the accuracy of this reporting the UK Renal Registry must work with renal centres and clinical staff on improving data input for the start of RRT.