. The median Hb of prevalent patients on HD was g/l.. 59% of HD patients and 55% of PD patients had Hb

Transcription

1 UK Renal Registry 17th Annual Report: Chapter 7 Haemoglobin, Ferritin and Erythropoietin amongst UK Adult Dialysis Patients in 13: National and -specific Analyses Julie Gilg a, Rebecca Evans a, Anirudh Rao a, Andrew J Williams b a UK Renal Registry, Bristol, UK; b Morriston Hospital, Swansea, UK Key Words Anaemia. Chronic kidney disease. Dialysis. End stage renal disease. Epidemiology. Erythropoietin. Erythropoietin stimulating agent. European Best Practice Guidelines. Ferritin. Haemodialysis. Haemoglobin. NICE. Peritoneal dialysis. Renal Association Summary In the UK in 13:. The median haemoglobin (Hb) of patients at the time of starting dialysis was g/l with % of patients having a Hb 5 g/l.. The median Hb in patients starting haemodialysis (HD) was 97 g/l (IQR 88 6) and in patients starting peritoneal dialysis (PD) was 9 g/l (IQR ).. At start of dialysis, 53% of patients presenting early had Hb 5 g/l whilst only 36% of patients presenting late had Hb 5 g/l.. The median Hb of prevalent patients on HD was 112 g/l with an IQR of 3 1 g/l.. The median Hb of prevalent patients on PD was 113 g/l with an IQR of g/l.. For both HD and PD patients, 83% had Hb 5 g/l.. 59% of HD patients and 55% of PD patients had Hb 5 and 41 g/l.. The median ferritin in HD patients was 424 mg/l (IQR 2 616) and 95% of HD patients had a ferritin 5 mg/l. In England, Wales and Northern Ireland in 13:. The median ferritin in PD patients was 285 mg/l (IQR ) with 88% of PD patients having a ferritin 5 mg/l.. The median erythropoietin stimulating agent (ESA) dose was higher for HD than PD patients (7,333 vs. 4, IU/week). 151

2 The UK Renal Registry The Seventeenth Annual Report Introduction This chapter describes analyses of the UK Renal Registry (UKRR) data relating to the management of anaemia in dialysis patients during 13. The diagnosis and management of anaemia in chronic kidney disease and the standards to be achieved have been detailed in the Kidney Disease Improving Global Outcomes (KDIGO), Kidney Disease Outcomes Quality Initiative (KDOQI), European Best Practice Guidelines (EBPG) and UK Renal Association guidelines [1 4]. The health economics of anaemia therapy using ESAs has also been subject to a NICE systematic review which concluded that treating to a target haemoglobin (Hb) 1 1 g/l is cost effective in HD patients [5]. This chapter reports on the analyses of data items collected by the UKRR largely in the context of the 5th edition of the UK Renal Association s Anaemia in CKD guidelines and recommendations which was published at the end of [4]. Table 7.1 lists the audit measures from these guidelines along with explanations for why some of the measures were not reported on. Methods The incident and prevalent renal replacement therapy (RRT) cohorts for 13 were analysed. The UKRR extracted quarterly data electronically from renal centres in England, Wales and Northern Ireland; data from Scotland were provided by the Scottish Renal Registry. Haemoglobin levels are given in g/l as the majority of UK laboratories have now switched to reporting using these units rather than g/dl. For the analyses of Hb for incident patients, those patients commencing RRT on PD or HD were included whilst those receiving a pre-emptive transplant were excluded. Hb measurements from after starting dialysis but still within the same quarter of the year were used. Therefore, depending on when in the quarter a patient started RRT the Hb data could be from zero to days later. The haemoglobin values the UKRR receives should be the closest available measurement to the end of the quarter. Patients who died within the first days on treatment were excluded. Results are also shown with the cohort subdivided into early and late presenters (date first seen by a nephrologist, or more days and less than days before starting dialysis respectively). For the analyses of prevalent patients, those patients receiving dialysis on 31st December 13 were included if they had been on the same modality of dialysis in the same centre for at least three months. In order to improve completeness the last available Table 7.1. Summary of recommended Renal Association audit measures relevant to anaemia management RA audit measure Included in UKRR annual report? Reason for non-inclusion 1. Proportion of CKD patients with egfr, ml/min by 4 variable MDRD method with an annual Hb level 2. Proportion of patients starting an ESA without prior measurement of serum ferritin and/or TSAT 3. Proportion of patients on renal replacement therapy with Hb level, who are not prescribed an ESA 4. Each renal unit should audit the type, route and frequency of administration and weekly dose of ESA prescribed No No Yes UKRR reports the completeness of these data items UKRR does not currently collect CKD data UKRR does not know when all patients start ESA treatment UKRR does not collect TSAT data 5. The proportion of CKD stage 4 5 patients with Hb 12 g/dl No UKRR does not currently collect CKD data 6. The proportion of patients treated with an ESA with Hb.12 g/dl Yes 7. Each renal unit should monitor ESA dose adjustments No UKRR does not collect this data 8. Proportion of patients with serum ferritin levels, ng/ml at start of treatment with ESA 9. Proportion of pre-dialysis and PD patients receiving iron therapy; type: oral vs. parenteral No No UKRR does not know when all patients start ESA treatment UKRR does not currently collect CKD data/poor data completeness. Proportion of HD patients receiving IV iron No Poor data completeness 11. Prevalence of resistance to ESA among renal replacement therapy Yes patients 12. Proportion of HD patients who received a blood transfusion within the past year No Data held at NHS Blood and Transplant 152

3 Chapter 7 Anaemia management in UK dialysis patients measurement for each patient from the last two quarters was used for Hb and from the last three quarters for ferritin. Scotland was excluded from the analysis for ferritin for PD patients as this data was not included in its return. The completeness of data items were analysed at both centre and country level. As in previous years, all patients were included in analyses but centres with less than % completeness were excluded from the caterpillar and funnel plots showing centre level results. s providing relevant data from less than patients were also excluded from the plots. The number preceding the centre name in the caterpillar plots is the percentage of patients who have data missing. Summary statistics including minimum, maximum, interquartile ranges (IQR), averages (mean and median) and standard deviations were calculated. The median values and the IQRs are shown using caterpillar plots. The percentages achieving standards were also calculated. These are displayed using caterpillar plots with the percentages meeting the targets and 95% confidence intervals (CIs) shown. Funnel plots show the distribution of the percentages meeting the targets and also whether any of the centres are significantly different from the average. Longitudinal analysis was performed to show overall changes in achievement of standards from 1998 to 13. Erythropoietin data from the last quarter of 13 were used to define which patients were receiving ESAs. Scotland was excluded from this analysis as data regarding ESA was not included in its return. Each individual was defined as being on ESA if a drug type and/or a dose was present in the data. s reporting fewer than % of HD patients or fewer than 45% of PD patients being treated with ESAs were considered to have incomplete data and were excluded from further analysis. It is recognised that these exclusion criteria are relatively arbitrary but they are in part based upon the frequency distribution graph of centres ESA use as it appears in the data. The percentage of patients on ESAs was calculated from these data and incomplete data returns risk seriously impacting on any conclusions drawn. For analyses of ESA dose, values are presented as weekly erythropoietin dose. Doses of less than 1 IU/week (likely to be darbepoietin) were harmonised with erythropoietin data by multiplying by. No adjustments were made with respect to route of administration. Patients who were not receiving ESAs were not included in analyses of dose (rather than being included with dose = ). Until two years ago, UKRR annual reports only used the dose from the final quarter of the year. Now, starting with the cohort of patients receiving ESAs in the final quarter and having a dose value present for that quarter, any further dose values available from the earlier three quarters of the year were used (provided the patient was on the same treatment and receiving the same drug in those quarters). The average (mean) of the available values was then used in analyses rather than the dose in the final quarter. The ESA data were collected electronically from renal IT systems but in contrast to laboratory linked variables the ESA data required manual data entry. The reliability depended upon the data source, whether the entry was linked to the prescription or whether the prescriptions were provided by the primary care physician. In the latter case, doses may not be as reliably updated as the link between data entry and prescription is indirect. Results Anaemia management in incident dialysis patients Haemoglobin in incident dialysis patients The Hb at the time of starting RRT gives the only indication of concordance with current anaemia management recommendations in the pre-dialysis (CKD 5 not yet on dialysis) group. The percentage of data returned and outcome Hb are listed in table 7.2. Results are not shown for two centres (Carshalton, London Guys) because data completeness was less than %, results are not shown for Dumfries & Galloway as there were less than patients with data. The median Hb of patients at the time of starting dialysis in the UK was g/l. The median starting Hb by centre is shown in figure 7.1. The percentage of patients having a Hb 5 g/l has fallen over the last several years to % from 55% in the 9 cohort. The percentage starting with a Hb 5 g/l by centre is given in figure 7.2. The variation in the proportion of patients starting dialysis with Hb 5 g/l between centres remained high (27 %). Using only centres with time of presentation data, the median Hb in the late presenters was 94 g/l with only 36% of patients having a Hb 5 g/l compared with a median Hb of 1 g/l and 53% of patients having a Hb 5 g/l in the early presenters group. In both groups there was a large amount of variation between centres in the percentage of patients having a Hb5 g/l (7 66% in the late presenters and 29 93% in the early presenters). The lower median Hb in late presenters may reflect inadequate pre-dialysis care with limited anaemia management, anaemia of multisystem disease or inter-current illness. Hb of patients at the time of starting HD was 97 g/l (IQR 88 6 g/l) and in those starting PD it was 9 g/l (IQR g/l). When starting dialysis, 43% of HD patients had a Hb 5 g/l, compared with 74% of PD patients. Incident dialysis patients from 12 were followed for one year and the median haemoglobin (and percentage with a Hb 5 g/l) of survivors on the same treatment at the same centre after a year was calculated for each quarter. Only patients who had Hb data for each of the four time points were included in this analysis. This was sub-analysed by modality and length of pre-rrt care (figures 7.3, 7.4). Hb was higher in the second quarter on dialysis than during the quarter at start of dialysis reflecting the benefits of treatment administered. Over 79% of incident patients surviving to a year had Hb 153

4 The UK Renal Registry The Seventeenth Annual Report Table 7.2. Haemoglobin data for incident patients starting haemodialysis or peritoneal dialysis during 13, both overall and by presentation time All incident patients Early presenters only (5 days) Late presenters only (, days) % data return N with data Hb g/l 5 g/l Hb g/l 5 g/l Hb g/l 5 g/l England B Heart B QEH Basldn Bradfd Brightn Bristol Camb Carlis Carsh Chelms Colchr Covnt Derby Donc Dorset Dudley Exeter Glouc Hull Ipswi Kent L Barts L Guys L Kings L Rfree L St.G L West Leeds Leic Liv Ain Liv Roy M RI Middlbr Newc Norwch Nottm Oxford Plymth Ports Prestn Redng Salford Sheff Shrew Stevng Sthend Stoke Sund Truro Wirral Wolve York

5 Chapter 7 Anaemia management in UK dialysis patients Table 7.2. Continued All incident patients Early presenters only (5 days) Late presenters only (, days) % data return N with data Hb g/l 5 g/l Hb g/l 5 g/l Hb g/l 5 g/l N Ireland Antrim Belfast Newry Ulster West NI Scotland Abrdn Airdrie D & Gall 88 7 Dundee Edinb Glasgw Inverns Klmarnk Krkcldy Wales Bangor Cardff Clwyd Swanse Wrexm England 92 4, N Ireland Scotland Wales UK 92 5, Blank cells: centres excluded from analyses due to poor data completeness or low patient numbers or because presentation time data not available 1 1 N = 5,376 Upper quartile Hb Lower quartile Haemoglobin g/l 1 Carlis 4 West NI Newry 9 Wrexm 14 Chelms Bristol 32 Airdrie 32 L West 11 Inverns 1 Cardff Ports 7 Wirral 3 Brightn 19 Camb 7 Stoke Shrew Donc 4 Ulster 1 L Rfree Dorset Derby Glouc 1 Newc Oxford 3 Krkcldy 23 Hull 6 Ipswi 1 Redng Plymth Swanse 2 Edinb 5 Sund 2 M RI 1 L Barts 2 Liv Ain 8 B QEH 1 Stevng 6 Belfast Liv Roy 1 Middlbr Leic 5 Covnt Norwch Prestn 1 Kent 31 Klmarnk 18 Glasgw 4 Wolve B Heart Dudley Abrdn 12 Dundee 2 Salford 12 Bradfd 14 York Basldn 16 L St.G L Kings 2 Leeds 21 Colchr 4 Antrim 8 England 4 N Ireland 14 Scotland 2 Wales 8 UK Fig haemoglobin for incident dialysis patients at start of dialysis treatment in

6 The UK Renal Registry The Seventeenth Annual Report N = 5,376 Bristol 4 West NI Carlis Newry 9 Wrexm 11 Inverns 32 L West Shrew 1 Cardff 32 Airdrie 7 Stoke 7 Wirral Ports 14 Chelms Donc 3 Brightn 2 Edinb Derby 3 Krkcldy Dorset 19 Camb 1 L Rfree 1 Redng Glouc 1 Newc Oxford 6 Ipswi Swanse Plymth 23 Hull 4 Ulster 5 Sund 2 Liv Ain 1 L Barts 2 M RI Liv Roy 8 B QEH 5 Covnt 6 Belfast Abrdn 12 Dundee 1 Middlbr Prestn 31 Klmarnk 12 Bradfd 1 Stevng 18 Glasgw Norwch 1 Kent Leic Dudley 2 Salford B Heart 16 L St.G 4 Wolve L Kings 4 Antrim Basldn 14 York 2 Leeds 21 Colchr 8 England 4 N Ireland 14 Scotland 2 Wales 8 UK Fig Percentage of incident dialysis patients with Hb 5 g/l at start of dialysis treatment in 13 Upper 95% Cl % with Hb > g/l Lower 95% Cl 5 g/l regardless of the modality or the length of pre-rrt care. The annual distribution of Hb in incident dialysis patients is shown in figure 7.5. Since 6, the proportion of incident patients with Hb 51 g/l has fallen from 17% to % and the proportion of patients with Hb, g/l continues to gradually increase over the years from % to %. In the 13 cohort with presentation time data, 64% of patients in the late presentation group had Hb, g/l compared with 47% in the early presentation group. ESA by time on dialysis in early vs. late presenters Incident dialysis patients from 12 were followed for one year and the percentages receiving an ESA were calculated for each quarter for survivors on the same treatment at the same centre after a year. This was subanalysed by modality and length of pre-rrt care Haemoglobin g/l PD early PD late HD early HD late Start 3 months 6 months 12 months Time since commencing dialysis Fig haemoglobin, by time on dialysis and length of pre-rrt care, for incident dialysis patients in 12 Percentage of incident patients PD early PD late HD early HD late Start 3 months 6 months 12 months Time since commencing dialysis Fig Percentage of incident dialysis patients in 12 with Hb 5 g/l, by time on dialysis and by length of pre-rrt care Percentage of incident patients < >1 (Hb, g/l) Year of start Fig Distribution of haemoglobin in incident dialysis patients by year of start 156

7 Chapter 7 Anaemia management in UK dialysis patients Percentage on ESA PD early PD late HD early HD late Start 3 months 6 months 12 months Time since commencing dialysis Fig Percentage of incident dialysis patients in 12 on ESA, by time on dialysis and by length of pre-rrt care (figure 7.6). For HD patients at the start of treatment there was a difference between early and late presenters in the percentage of patients receiving an ESA. This difference was greatly reduced by three months after starting and had disappeared within one year of starting dialysis. For PD patients there was a similar difference between the early and late group and this difference persisted over the first year after starting dialysis. However, caution is advised in interpreting this figure as the number (23) of patients in the PD late group was relatively small. Anaemia management in prevalent dialysis patients Compliance with data returns for haemoglobin and serum ferritin and percentages on ESA are shown for the 71 renal centres in the UK in table 7.3 for both HD and PD patients. Completeness of data returns was generally good for Hb and ferritin. The percentages on ESA are shown as they appear in the data received by the UKRR. For some centres, the ESA data was completely missing and for others it appears to be partially complete (i.e. very low percentages of patients appearing to be on Table 7.3. Percentage completeness of data returns for haemoglobin and serum ferritin and percentages on ESA for prevalent HD and PD patients in 13 HD PD N Hb Ferritin % on ESA N Hb Ferritin % on ESA England B Heart B QEH Basldn Bradfd Brightn Bristol Camb Carlis Carsh Chelms Colchr Covnt Derby Donc Dorset Dudley Exeter Glouc Hull Ipswi Kent L Barts L Guys L Kings L Rfree L St.G L West 1, Leeds Leic

8 The UK Renal Registry The Seventeenth Annual Report Table 7.3. Continued HD PD N Hb Ferritin % on ESA N Hb Ferritin % on ESA Liv Ain Liv Roy M RI Middlbr Newc Norwch Nottm Oxford Plymth Ports Prestn Redng Salford Sheff Shrew Stevng Sthend Stoke Sund Truro Wirral Wolve York N Ireland Antrim Belfast Newry Ulster West NI Scotland Abrdn Airdrie D & Gall Dundee Edinb Glasgw Inverns Klmarnk Krkcldy Wales Bangor Cardff Clwyd Swanse Wrexm England 18, , N Ireland Scotland 1, Wales 1, UK 22, , Blank cells: centres with no PD patients or because data was not available Percentages on ESA are shown, but it is believed that there were data problems for those centres with apparently less than % of HD patients or 45% of PD patients on ESA The country level averages for the % on ESA are based only on those centres whose % was above the limits mentioned above These overall averages are for E,W & NI (not UK) 158

9 Chapter 7 Anaemia management in UK dialysis patients ESAs). It is believed that there were problems with data entry and/or data transfer for those centres where the percentage on ESA was less than % for HD patients or 45% for PD patients. These centres have been excluded from further analyses of ESA use. Summary statistics for haemoglobin, serum ferritin and ESA are shown for the 71 renal centres in the UK in tables 7.4 for HD and 7.5 for PD patients respectively. Haemoglobin in prevalent haemodialysis patients The median Hb of patients on HD in the UK was 112 g/l with an IQR of 3 1 g/l and 83% of HD patients had a Hb 5 g/l (table 7.4). The median Hb by centre is shown in figure 7.7. Compliance with the target range of Hb 5 and 41 g/l (figure 7.8) continues to increase year on year, 53% in, 56% in 11, 57% in 12 and 59% in 13. The percentages of HD patients with Hb below g/l and above 1 g/l, as well as the percentages meeting the target, are shown by centre in figure 7.9. Funnel plots are shown for the minimum (Hb 5 g/l) and target range (Hb 5 and 41 g/l) in figures 7. and 7.11 respectively. Many centres complied well with respect to both the minimum and target range Hb standards. Some centres complied well with the percentage with Hb 5 g/l (figure 7.) but had a poor compliance with percentage of patients with Hb 5 and 41 g/l (figure 7.11). Table 7.4 can be used in conjunction with figures 7. and 7.11 to identify centres. Haemoglobin in prevalent peritoneal dialysis patients Overall, 83% of patients on PD had a Hb 5 g/l (table 7.5). The median Hb of patients on PD in the UK in 13 was 113 g/l with an IQR of g/l. The median Hb by centre is shown in figure The Table 7.4. Summary statistics for haemoglobin, serum ferritin and ESA for prevalent HD patients in 13 N with Hb data Hb g/l 5 g/l 1 g/l ferritin mg/l % ferritin 5 mg/l % ferritin. and mg/l %on ESA ESA dose (IU/week) % with Hb 5 g/l and not on ESA England B Heart , 21 B QEH , 11 Basldn , 6 Bradfd , 4 Brightn Bristol , 9 Camb Carlis , 24 Carsh Chelms , 7 Colchr Covnt , Derby Donc ,2 Dorset , 5 Dudley Exeter ,2 8 Glouc Hull , Ipswi , 19 Kent ,2 7 L Barts L Guys L Kings , 7 L Rfree L St.G L West 1, Leeds , 12 Leic , 2 159

10 The UK Renal Registry The Seventeenth Annual Report Table 7.4. Continued N with Hb data Hb g/l 5 g/l 1 g/l ferritin mg/l % ferritin 5 mg/l % ferritin. and mg/l %on ESA ESA dose (IU/week) % with Hb 5 g/l and not on ESA Liv Ain Liv Roy M RI Middlbr , 21 Newc , 32 Norwch , 9 Nottm ,7 11 Oxford , 8 Plymth Ports Prestn Redng , Salford Sheff , 11 Shrew ,2 9 Stevng Sthend , 5 Stoke Sund ,423 9 Truro Wirral Wolve , 17 York ,2 N Ireland Antrim ,375 5 Belfast ,2 8 Newry , 8 Ulster , 3 West NI , 6 Scotland Abrdn Airdrie D & Gall Dundee Edinb Glasgw Inverns Klmarnk Krkcldy Wales Bangor ,2 18 Cardff Clwyd Swanse , 14 Wrexm , England 17, , 11 N Ireland , 6 Scotland 1, Wales 1, , 14 UK 21, , Blank cells: centres excluded from analyses due to poor data completeness or low patient numbers or because the data item was not available ESA data only shown for those centres for which the % on ESA was % or more For ESA, these overall averages are for E,W & NI (not UK) 1

11 Chapter 7 Anaemia management in UK dialysis patients Table 7.5. Summary statistics for haemoglobin, serum ferritin and ESA for prevalent PD patients in 13 N with Hb data Hb g/l 5 g/l 1 g/l ferritin mg/l % ferritin 5 mg/l % ferritin. and mg/l %on ESA ESA dose (IU/week) % with Hb 5 g/l and not on ESA England B Heart , 34 B QEH , Basldn , 33 Bradfd ,7 35 Brightn Bristol , 37 Camb ,4 44 Carlis , 22 Carsh Chelms , Colchr n/a Covnt , 46 Derby Donc , Dorset ,9 28 Dudley Exeter , 24 Glouc Hull , 46 Ipswi , 25 Kent , 42 L Barts L Guys L Kings , L Rfree L St.G L West Leeds ,2 21 Leic , 25 Liv Ain Liv Roy M RI Middlbr , 36 Newc Norwch , 23 Nottm , Oxford , 16 Plymth Ports Prestn Redng Salford Sheff , 38 Shrew , 38 Stevng Sthend Stoke Sund 8 Truro Wirral Wolve , 36 York ,

12 The UK Renal Registry The Seventeenth Annual Report Table 7.5. Continued N with Hb data Hb g/l 5 g/l 1 g/l ferritin mg/l % ferritin 5 mg/l % ferritin. and mg/l %on ESA ESA dose (IU/week) % with Hb 5 g/l and not on ESA N Ireland Antrim , Belfast , 12 Newry , 12 Ulster 4 West NI , 21 Scotland Abrdn Airdrie D & Gall 112 Dundee Edinb Glasgw Inverns Klmarnk Krkcldy Wales Bangor Cardff Clwyd Swanse , 26 Wrexm ,4 42 England 2, , 31 N Ireland , 14 Scotland Wales , UK 3, , 31 Blank cells: centres excluded from analyses due to poor data completeness or low patient numbers or because the data item was not available n/a no PD patients ESA data only shown for those centres for which the % on ESA was 45% or more For ferritin and for ESA these overall averages are for E,W & NI (not UK) N = 21,2 Upper quartile Hb Lower quartile Haemoglobin g/l Carlis Edinb 1 Chelms 1 Krkcldy Norwch Derby Ports 1 L West Sund Donc 1 Liv Roy Redng Bradfd Dorset 7 Colchr Glouc 15 Stoke Wolve Shrew 1 Liv Ain 1 Glasgw 1 Antrim 1 Wirral 2 Middlbr Hull Plymth 7 M RI Klmarnk 5 Newry Belfast Bristol Leic Oxford Cardff Inverns 1 Airdrie Ulster 1 L Rfree 1 Stevng 5 Carsh York Ipswi 1 L St.G B Heart 5 Dudley Swanse Newc Kent 49 Camb 1 Prestn L Barts B QEH 1 Dundee 3 West NI 2 Brightn Covnt Wrexm Leeds D&Gall 1 Basldn L Kings 12 Salford 2 Abrdn 5 England 1 N Ireland 1 Scotland Wales 5 UK Fig haemoglobin in patients treated with HD by centre in

13 Chapter 7 Anaemia management in UK dialysis patients N = 21,2 Upper 95% Cl % with Hb > and <1 g/l Lower 95% Cl Inverns Bristol L Kings Ulster D&Gall 5 Carsh 1 Dundee Leeds 3 West NI 1 Airdrie B QEH 1 L West York Covnt 1 L Rfree Swanse 49 Camb 1 Antrim 1 Stevng 1 Prestn Dorset 1 Chelms 5 Dudley Glouc Plymth B Heart 5 Newry 2 Brightn L Barts Kent Derby 1 Wirral 1 L St.G Shrew Sund Ipswi 2 Abrdn 7 M RI Hull Cardff Wrexm 1 Liv Ain 12 Salford Bradfd Oxford 1 Krkcldy Belfast Newc 1 Basldn 2 Middlbr 1 Glasgw Klmarnk 15 Stoke Leic Norwch Edinb Carlis Wolve Ports Redng Donc 1 Liv Roy 7 Colchr 5 England 1 N Ireland 1 Scotland Wales 5 UK Fig Percentage of HD patients with Hb 5 and 41 g/l by centre in 13 compliance with Hb 5 and 41 g/l is shown in figure In 13, 55% of prevalent PD patients had a Hb within the target range. The distribution of Hb in PD patients by centre is shown in figure The funnel plots for percentage with Hb 5 g/l and for the percentage of patients with Hb 5 and 41 g/l are shown in figures 7.15 and 7.16 respectively. Table 7.5 can be used in conjunction with figures 7.15 and 7.16 to identify centres in the funnel plot. Relationship between Hb in incident and prevalent dialysis patients in 13 The relationship between the percentage of incident and prevalent dialysis (HD and PD) patients with a Hb 5 g/l is shown in figure As expected, all centres had a higher percentage of prevalent patients achieving a Hb 5 g/l than that for incident patients. Overall in the UK, 83% of prevalent patients, compared with % of incident patients, had a Hb 5 g/l in 13. Compliance with the current minimum standard (Hb 5 g/l) is shown by year ( ) for incident and prevalent patients (all dialysis patients) in figure The decline in achieving this standard appears to be levelling off. Ferritin in prevalent haemodialysis patients The median and IQR for serum ferritin for patients treated with HD are shown in figure The percentages with serum ferritin 5 mg/l,. mg/l to Hb >1 g/l Hb 1 g/l Hb < g/l Exeter Inverns Sthend Bristol L Kings Ulster D&Gall Carsh Dundee Leeds West NI Bangor Airdrie Nottm Truro B QEH L West York Covnt L Rfree Swanse Camb Antrim Stevng Prestn Dorset Chelms Dudley Glouc Plymth B Heart Newry Brightn L Barts Kent Derby Wirral L St.G Shrew Sund Ipswi Abrdn M RI Hull Cardff Wrexm Liv Ain Salford Bradfd Oxford Krkcldy Belfast Newc Basldn Middlbr Glasgw Klmarnk Sheff Stoke Clwyd Leic Norwch Edinb Carlis Ports Wolve Redng Donc Liv Roy Colchr England N Ireland Scotland Wales UK Fig Distribution of haemoglobin in patients treated with HD by centre in

14 The UK Renal Registry The Seventeenth Annual Report 95 Dotted lines show 99.9% limits Solid lines show 95% limits Dotted lines show 99.9% limits Solid lines show 95% limits , 1, Number of patients with data in centre Fig. 7.. Funnel plot of percentage of HD patients with Hb 5 g/l by centre in 13 1, 1, Number of patients with data in centre Fig Funnel plot of percentage of HD patients with Hb 5 and 41 g/l by centre in N = 3,133 Upper quartile Hb Lower quartile Haemoglobin g/l 1 1 Redng 8 Inverns York Middlbr Brightn 3 Stevng Carlis Norwch Donc Chelms Dorset Bristol 1 Derby 1 Stoke Plymth 1 Covnt Prestn 6 Dundee 11 Newc Antrim West NI Ports Airdrie 1 M RI Hull 3 Leic 2 L St.G 1 L Barts Swanse 3 Klmarnk 6 Krkcldy 1 Oxford Bradfd Wrexm Wolve Shrew 3 Salford Liv Roy B QEH Abrdn 9 D&Gall Ipswi Cardff 5 Camb Kent Glasgw Belfast 1 L Kings Leeds L Rfree 6 Carsh Basldn B Heart Dudley Edinb Glouc Newry L West Liv Ain 22 Wirral 2 England N Ireland 3 Scotland Wales 2 UK Fig haemoglobin in patients treated with PD by centre in 13 Upper 95% Cl % with Hb > and <1 g/l N = 3,133 Lower 95% Cl Newry Glasgw 6 Krkcldy 6 Dundee Antrim Bristol Ipswi 5 Camb Airdrie Belfast West NI Swanse 3 Stevng Middlbr 3 Leic Basldn Dudley 1 L Kings Kent Edinb 9 D&Gall 8 Inverns Glouc Leeds Brightn Abrdn Donc Carlis Wolve 11 Newc Ports Chelms 1 Oxford 2 L St.G B Heart 1 Stoke Cardff Hull Norwch B QEH 3 Salford Redng L West Prestn Liv Ain 1 Derby 1 Covnt 1 L Barts Dorset 22 Wirral Wrexm 3 Klmarnk Liv Roy L Rfree Shrew 6 Carsh 1 M RI Bradfd Plymth York 2 England N Ireland 3 Scotland Wales 2 UK Fig Percentage of PD patients with Hb 5 and 41 g/l by centre in

15 Chapter 7 Anaemia management in UK dialysis patients Hb >1 g/l Hb 1 g/l Hb < g/l Newry Glasgw Krkcldy Dundee Exeter Antrim Sthend Bristol Ipswi Camb Airdrie Belfast West NI Swanse Stevng Middlbr Leic Basldn Dudley L Kings Kent Nottm Edinb D&Gall Inverns Glouc Leeds Brightn Clwyd Abrdn Donc Carlis Wolve Newc Ports Truro Chelms Oxford L St.G B Heart Stoke Cardff Hull Norwch B QEH Salford Redng L West Prestn Liv Ain Derby Covnt L Barts Dorset Wirral Wrexm Klmarnk Liv Roy L Rfree Shrew Carsh Sheff M RI Bradfd Bangor Plymth York England N Ireland Scotland Wales UK Fig Distribution of haemoglobin in patients treated with PD by centre in Dotted lines show 99.9% limits Solid lines show 95% limits Dotted lines show 99.9% limits Solid lines show 95% limits Number of patients with data in centre Fig Funnel plot of percentage of PD patients with Hb 5 g/l by centre in Number of patients with data in centre Fig Funnel plot of percentage of PD patients with Hb 5 g/l and 41 g/l by centre in 13 Incident dialysis patients Prevalent dialysis patients Exeter Carlis Bristol Chelms Bangor L West Glouc Edinb Clwyd Antrim Newry Norwch Shrew Inverns Krkcldy Dorset Ports Glasgw Sund Derby Airdrie Redng Ulster Middlbr Dudley Hull Stoke Swanse Kent Stevng M RI West NI Camb Truro Sheff Oxford Wolve York Plymth Donc Ipswi Dundee Newc Wirral Carsh Leeds L Rfree Belfast B Heart Liv Roy Colchr Bradfd Nottm Cardff Leic Prestn Brightn B QEH Wrexm L St.G Liv Ain Sthend Klmarnk L Barts Covnt L Kings Salford Basldn Abrdn England N Ireland Scotland Wales UK Fig Percentage of incident and prevalent dialysis patients with Hb 5 g/l by centre in

16 The UK Renal Registry The Seventeenth Annual Report Upper 95% Cl % with Hb > g/l Lower 95% Cl Incident patients Year Prevalent patients Fig Percentage of incident and prevalent dialysis patients ( ) with Hb 5 g/l mg/l, and mg/l are shown in figures 7., 7.21 and 7.22 respectively. Most centres achieved greater than % compliance with a serum ferritin 5 mg/l for HD patients. The HD population had a median ferritin value of 424 mg/l, IQR Twenty centres had greater than % ( %) of their patients having ferritin mg/l (figure 7.22). Twelve of these had values over 25%. The serum ferritin correlated poorly with median Hb achieved and ESA dose (table 7.4). Ferritin in prevalent peritoneal dialysis patients The median and IQR for serum ferritin for patients treated with PD are shown in figure The percentages with serum ferritin 5 mg/l,. mg/l and mg/l, and mg/l are shown in figures 7.24, 7.25 and 7.26 respectively. The PD population had a lower median ferritin value (285 mg/l, IQR ) than the HD population. Twenty-nine centres reported less than % of PD patients being compliant with serum ferritin 5 mg/l, although this appeared to have little bearing on their achieved median Hb or median ESA dose when compared with other centres (table 7.5). Erythropoietin stimulating agents in prevalent haemodialysis patients As shown in previous reports there was substantial variation in the average dose of ESA prescription used. 1, 1, 1, N =,722 Upper quartile ferritin Lower quartile Ferritin µg/l 2 Plymth 1 Antrim 5 D&Gall 26 L Guys 4 Abrdn 1 Chelms Ulster Ipswi 2 Stevng 5 Airdrie 38 West NI Prestn 1 L Kings 1 Bristol 9 Colchr 1 Bradfd 18 Brightn 1 Sund 8 Krkcldy 1 Liv Ain 1 Dorset 1 Ports 1 L Rfree Redng 1 Norwch Liv Roy Wolve 7 Edinb Leeds 1 Kent 1 Antrim 1 Wirral Derby 1 L Barts 1 Belfast York Newc 11 Glasgw 27 Wrexm 1 Shrew 2 L St.G 1 Donc Hull M RI Swanse Covnt 2 Glouc 2 L West 7 Carsh 1 B QEH 1 Basldn Leic 1 B Heart 21 Dundee 24 Camb 6 Dudley Stoke 12 Klmarnk Oxford Cardff 35 Inverns 5 England 18 N Ireland 11 Scotland 3 Wales 6 UK Fig ferritin in patients treated with HD by centre in

17 Chapter 7 Anaemia management in UK dialysis patients Upper 95% Cl % with ferritin > µg/l Lower 95% Cl N =,722 Ulster 5 D&Gall 5 Airdrie 1 Chelms 9 Colchr 1 Bradfd 18 Brightn 27 Wrexm 2 Plymth 1 Donc Redng 35 Inverns 1 Bristol 1 Antrim 4 Abrdn 2 L St.G 1 L Kings 2 Stevng 1 Dorset Covnt 1 Sund Ipswi 26 L Guys Leic York 1 Ports 2 L West 1 B Heart 1 Shrew Leeds 1 Wirral 2 Middlbr Hull 1 Kent 1 L Barts Derby Wolve Newc 7 Carsh Cardff M RI Prestn 1 L Rfree 1 Belfast 1 Norwch 1 B QEH 38 West NI 1 Liv Ain 11 Glasgw 7 Edinb Oxford Stoke 6 Dudley 1 Basldn 24 Camb Swanse 2 Glouc Liv Roy 8 Krkcldy 21 Dundee 12 Klmarnk 5 England 18 N Ireland 11 Scotland 3 Wales 6 UK Fig. 7.. Percentage of HD patients with ferritin 5 mg/l by centre in 13 N =,722 Upper 95% Cl % with ferritin > and < µg/l Lower 95% Cl 35 Inverns 7 Carsh 1 Basldn 1 B QEH 2 L West York Hull 6 Dudley Covnt Cardff Leic 1 B Heart 24 Camb 27 Wrexm Oxford 2 L St.G M RI 1 Donc Stoke 1 Wirral 1 L Barts 1 Shrew 21 Dundee 12 Klmarnk Newc Redng 1 Antrim 2 Glouc Derby Swanse Leeds Wolve 11 Glasgw 1 Dorset 1 Bradfd 1 Kent 1 Ports 1 Belfast 1 Norwch 9 Colchr 1 L Rfree 18 Brightn 1 Sund 5 Airdrie 7 Edinb Liv Roy 1 L Kings 1 Liv Ain 1 Bristol Prestn 2 Stevng Ipswi 4 Abrdn 5 D&Gall 38 West NI Ulster 26 L Guys 2 Middlbr 8 Krkcldy 1 Chelms 2 Plymth 5 England 18 N Ireland 11 Scotland 3 Wales 6 UK Fig Percentage of HD patients with ferritin. mg/l and mg/l by centre in 13 Upper 95% Cl N =,722 % with ferritin > µg/l Lower 95% Cl 6 Dudley York Cardff 35 Inverns 1 B QEH Leic 7 Carsh 1 Basldn 1 Donc 21 Dundee Hull 2 L West Covnt Oxford Stoke M RI 24 Camb 2 Glouc 12 Klmarnk 1 B Heart 2 L St.G Swanse 1 L Barts 1 Shrew 1 Ports Newc Wolve 1 Wirral Derby Leeds 1 Belfast 11 Glasgw 18 Brightn 1 Dorset 27 Wrexm Redng 9 Colchr 1 Antrim 1 L Rfree 1 Liv Ain 1 Bradfd 1 Chelms 1 Bristol 1 Norwch 1 Kent Liv Roy Ulster 7 Edinb 1 L Kings 1 Sund 2 Stevng Ipswi 38 West NI Prestn 5 D&Gall 5 Airdrie 8 Krkcldy 4 Abrdn 26 L Guys 2 Middlbr 2 Plymth 5 England 18 N Ireland 11 Scotland 3 Wales 6 UK Fig Percentage of HD patients with ferritin mg/l by centre in

18 The UK Renal Registry The Seventeenth Annual Report 1, N = 2,771 Upper quartile ferritin Lower quartile Ferritin µg/l Wirral 19 L Rfree 4 Carlis 3 Derby Redng Ipswi Brightn 2 Stoke 14 Newc 3 Donc 4 Ports 4 Swanse Prestn 5 Bristol Antrim Middlbr Leeds West NI Liv Ain 1 Hull 4 Leic 21 Plymth 1 B QEH Liv Roy Dorset Newry Camb 5 Kent B Heart L West Covnt 47 Wrexm Shrew Bradfd 1 Oxford Belfast 1 L Kings L St.G 8 L Barts 16 Stevng York 36 L Guys 4 M RI Wolve 6 Carsh Norwch Chelms 19 Glouc 4 Dudley Basldn Cardff 7 England N Ireland 15 Wales 8 E, W & NI Fig ferritin in patients treated with PD by centre in 13 Upper 95% Cl % with ferritin > µg/l N = 2,771 Lower 95% Cl Antrim Wirral Middlbr 47 Wrexm 4 Carlis 19 L Rfree Dorset 3 Derby 4 Ports 1 Hull 3 Donc 5 Bristol Redng Camb Brightn Prestn 4 Swanse Leeds 14 Newc West NI Belfast Liv Ain 4 Leic B Heart 2 Stoke 5 Kent 8 L Barts L St.G L West 1 B QEH Newry York 1 L Kings 1 Oxford 21 Plymth Liv Roy 4 M RI Wolve Shrew Covnt Chelms 6 Carsh Ipswi Bradfd 36 L Guys 16 Stevng Basldn Norwch 19 Glouc 4 Dudley Cardff 7 England N Ireland 15 Wales 8 E, W & NI Fig Percentage of PD patients with ferritin 5 mg/l by centre in 13 Upper 95% Cl % with ferritin > and < µg/l N = 2,771 Lower 95% Cl Camb 47 Wrexm Chelms Belfast Dorset 1 Hull York L St.G L West 4 M RI 6 Carsh Antrim 1 L Kings Wolve Middlbr 5 Bristol 8 L Barts Leeds Newry 4 Leic 3 Donc 4 Swanse B Heart 1 Oxford 36 L Guys 1 B QEH Liv Ain Covnt 4 Ports Basldn 5 Kent Shrew 19 Glouc Prestn Liv Roy 16 Stevng Bradfd West NI Redng Cardff 4 Carlis 2 Stoke Brightn Norwch 3 Derby 4 Dudley 14 Newc 21 Plymth 19 L Rfree Wirral Ipswi 7 England N Ireland 15 Wales 8 E, W & NI Fig Percentage of PD patients with ferritin. mg/l and mg/l by centre in

19 Chapter 7 Anaemia management in UK dialysis patients Upper 95% Cl N = 2,771 % with ferritin > µg/l Lower 95% Cl 16 Stevng Camb 19 Glouc L St.G 47 Wrexm 4 M RI Chelms Cardff 6 Carsh 4 Dudley Dorset Wolve Covnt 1 B QEH Basldn 1 Oxford Bradfd 1 L Kings York 36 L Guys 5 Kent 1 Hull B Heart L West Newry 4 Swanse Antrim 8 L Barts 4 Ports 3 Donc West NI 5 Bristol 2 Stoke Belfast Shrew Liv Roy 4 Leic Norwch Middlbr Leeds Redng Prestn Liv Ain 3 Derby Brightn 14 Newc 21 Plymth 4 Carlis 19 L Rfree Wirral Ipswi 7 England N Ireland 15 Wales 13 UK Fig Percentage of PD patients with ferritin mg/l by centre in 13 The median dose for prevalent HD patients in England, Wales and Northern Ireland was 7,333 IU/week. The median dose varied from 4, IU/week (Carlisle) to 14, IU/week (Reading) with a median Hb for these centres of 1 g/l (Carlisle) and 115 g/l (Reading) (table 7.4). The 13 median dose was similar to that for 12 (7,248 IU/week). Erythropoietin stimulating agents in prevalent peritoneal dialysis patients For prevalent PD patients the median dose was substantially lower than for HD patients. The median dose was 4, IU/week with a range of 2, to 8, (table 7.5). The 13 median dose is similar to that for 12 (4,2 IU/week). ESA prescription and association with achieved haemoglobin For HD patients, centre level median Hb is plotted against median ESA dose in figure 7.27 and compliance with the RA standards for Hb 5 g/l and 41 g/l is plotted against median ESA dose in figure For these figures, Hb data was only used for those patients who were receiving an ESA and had dose data available. There was no strong relationship in either figure. It is known that not all patients treated with dialysis who have a Hb above 1 g/l are receiving ESA. It has been suggested that it may be inappropriate to include those patients not receiving ESA within the group not meeting this RA target. There are two reasons: firstly, the high Hb remains outside the control of the clinician, and secondly, the recent trials suggesting that it may be 117 Hb g/l Compliance with Hb 1 g/l , 5, 7, 9, 11, 13, 15, ESA dose (IU/week) Fig Hb versus median ESA dose in HD patients on ESA, by centre in , 5, 7, 9, 11, 13, 15, ESA dose (IU/week) Fig Compliance with Hb 1 g/l versus median ESA dose in HD patients on ESA, by centre in

20 The UK Renal Registry The Seventeenth Annual Report Hb >1 g/l not on ESA Hb >1 g/l on ESA Hb 1 g/l Hb < g/l Exeter Sthend Bristol L Kings Ulster Leeds West NI Bangor Nottm B QEH York Covnt Swanse Antrim Prestn Dorset Chelms Glouc B Heart Newry Kent Shrew Sund Ipswi Hull Wrexm Bradfd Oxford Belfast Newc Basldn Middlbr Sheff Clwyd Leic Norwch Carlis Wolve Redng Donc England N Ireland Wales E, W & NI Fig Distribution of haemoglobin in patients treated with HD and the proportion of patients with Hb.1 g/l receiving ESA by centre in 13 detrimental to achieve a high Hb in renal patients were based only upon patients treated with ESAs [6,7]. Figures 7.29 and 7. show the percentages of HD and PD patients in each centre whose Hb lies above, within or below the RA guidelines of 1 g/l. These charts also show the proportion of patients with a Hb above the upper limit who were receiving, or were not receiving an ESA. These analyses are restricted to the centres with acceptable ESA returns as stipulated above. These figures show that 23% of HD patients had a Hb.1 g/l. Most of these patients (77%) were on ESAs. Whereas for PD, 28% of patients had a Hb.1 g/l, but only about 47% of these were on ESAs. ESA prescription: age and modality associations The proportion of patients on an ESA was higher for HD (88%) than PD (68%) and this difference was present and similar across all age groups (figure 7.31). The Hb >1 g/l not on ESA Hb >1 g/l on ESA Hb 1 g/l Hb < g/l Newry Exeter Antrim Sthend Bristol Ipswi Camb Belfast West NI Swanse Middlbr Leic Basldn L Kings Kent Nottm Glouc Leeds Donc Carlis Wolve Chelms Oxford B Heart Hull Norwch B QEH Prestn Covnt Dorset Wrexm Shrew Sheff Bradfd York England N Ireland Wales E, W & NI Fig. 7.. Distribution of haemoglobin in patients treated with PD and the proportion of patients with Hb.1 g/l receiving ESA by centre in 13 1

21 Chapter 7 Anaemia management in UK dialysis patients (95% CIs) HD PD (95% CIs) HD PD Age range (years) Fig Percentage of dialysis patients on ESA, by age group and treatment modality (13) Age range (years) Fig Percentage of whole cohort (13) who are not on ESA and have Hb 5 g/l, by age group and treatment modality proportion of patients who had a Hb 5 g/l without requiring ESA (by age group and modality) is shown in figure ESAs and time on renal replacement therapy The percentage of patients on ESA by time on RRT and dialysis modality is shown in figure This is a cross-sectional analysis at the final quarter of 13. Patients who had previously changed RRT modality were included in this analysis. The proportion of PD patients requiring ESA rises with duration of RRT from 63% after 3 12 months, to % after or more years. This almost certainly reflects loss of residual renal on ESA (95% CIs) HD PD 3 months 1 2 years 2 3 years 3 5 years 5 years > years to <1 year Time on RRT Fig on ESA by time on RRT (13) function. For at least the first years on RRT, a greater percentage of HD patients were receiving ESA treatment than patients on PD for any given duration on RRT. Resistance to ESA therapy Figure 7.34 shows the frequency distribution of weekly ESA dose adjusted for weight by treatment modality. Data in the literature on prevalence of ESA resistance in the ERF population is very sparse. RA guidelines define resistance to ESA therapy as failure to reach the target Hb level despite SC epoetin dose > IU/kg/ week (4 IU/kg/week IV epoetin) or darbepoetin dose 1 < < <1 ESA dose (IU/kg/week) Fig Frequency distribution of mean weekly ESA dose corrected for weight in 13 1 < <4 4 or more HD PD 171

22 The UK Renal Registry The Seventeenth Annual Report Upper 95% Cl % with Hb > g/l Lower 95% Cl Haemodialysis Year Peritoneal dialysis Fig Percentage of prevalent HD and PD patients ( ) with Hb 5 g/l >1.5 mcg/kg/week. For the purposes of this analysis the centres were restricted to those with good completeness for weight (over 75%) and ESA dose data (35 centres for HD and 18 centres for PD). As per the above definition and assuming that HD patients largely receive ESA intravenously and PD patients receive ESA subcutaneously, the prevalence of high doses of ESA was.7% (n = 57) and 1.1% (n = 5) for HD and PD patients respectively. For these patients the dose range for HD was IU/kg/week and for PD IU/kg/week. For patients on HD with high ESA doses, 44% (n = 25) had Hb, g/l and 39% were within 1 g/l. For patients on PD with high ESA doses, % (n = 2) had Hb, g/l and % were within 1 g/l. The percentage of patients with ESA resistance, defined by those failing to reach Hb 5 g/l are.3% for HD and.5% for PD. Caution needs to be applied when interpreting these results as the numbers for the above calculations are small. Success with guideline compliance Compliance with current minimum standards by year (1998 to 13) is shown in figure 7.35 for prevalent patients (by treatment modality). Figure 7.36 shows the percentage of anaemic patients (Hb, g/l) receiving an ESA. A minority of patients had a Hb, g/l and were not receiving ESA therapy. Across the age groups this was between 3 9% for HD patients and 6 19% for PD patients. There are several potential explanations for this. Treatment with ESA may have been stopped in patients who were unresponsive or avoided in those with malignancy. Others may have been on ESA treatment but not had it recorded. Table 7.6 shows that the percentage of all patients treated with an ESA and having Hb.1 g/l ranged between 3 29% for HD and between 26% for PD. For HD, there was a small percentage of patients having ferritin levels, mg/l and being on an ESA ( %). The percentages were somewhat higher for PD ( 31%). Table 7.7 shows the percentage completeness for drug type, dose, route and frequency of administration for centres reporting ESA data. The completeness was generally good for drug type and dose but patchy for frequency and route of administration. (95% CIs) HD PD Age range (years) Fig with Hb, g/l who were on ESA, by age group and treatment modality (13) 172