Comorbidities in UK patients at the start of renal replacement therapy (Chapter 6)
UK Renal Registry, Southmead Hospital, Bristol BS10 5NB
Correspondence and offprint requests to: Charlie Tomson, UK Renal Registry, Southmead Hospital, Bristol BS10 5NB. Email: Charlie.tomson{at}nbt.nhs.uk
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Comorbidity returns have continued to improve, albeit slowly, with centres running Mediqal software having the highest rates of completeness.
Diabetes as a primary renal diagnosis accounted for 20% of those starting RRT, but a further 7% had diabetes present as a comorbid condition. The incidence of smoking remained high at 17% of diabetic patients, which was similar to that found in non-diabetics.
Twelve percent of the patients starting RRT had a previous myocardial infarction (MI) and 31% of those aged over 65 years starting RRT had ischaemic heart disease (IHD).
Patients starting on peritoneal dialysis (PD) were on average 9 years younger than those on haemodialysis (HD) and had fewer comorbidities present.
Patients starting RRT without any comorbidity present had a lower median estimated glomerular filtration rate (eGFR) than those with comorbid conditions.
Patients with a previous MI or coronary artery bypass grafting (CABG), started RRT with slightly higher mean haemoglobin than those without comorbid conditions or other comorbid conditions.
On univariate survival analysis, diabetes was not associated with an increased risk of death amongst patients aged over 65 years, possibly due to its close association with other comorbidities in this age group.
In the multivariate survival analysis, the presence of ischaemic/neuropathic ulcers was the predictor of worst survival, followed by malignancy, previous MI and age per 10 year increment.
Smoking was less common in both South Asian and black patients than whites (7 vs 17%) starting RRT. Twenty-three percent of both South Asian and white patients started RRT with IHD compared with only 12% of Black patients.
Keywords: chronic kidney disease; co-morbidity; dialysis; end stage renal disease; epidemiology; haemodialysis; peritoneal dialysis; renal replacement therapy
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Description of the extent of comorbidity amongst patients starting treatment for established renal failure is important for a number of reasons.
- Patients with significant comorbidity may require more in-patient and out-patient care, and their treatment is therefore likely to cost more; information on comorbidity may therefore help commissioners and providers to plan services.
- Marked national and international variations in the take-on rate for renal replacement therapy (RRT) may partly be explained by differing policies and attitudes relating to provision of RRT to patients with significant comorbidity. These differences may result from differences in referral, differences in acceptance for RRT or both. Study of the outcomes of RRT amongst patients with and without comorbidity may help explain and reduce these variations.
- Comorbidity may influence survival amongst patients on RRT and may affect survival differently depending on the modality of RRT. Differences in survival rates between patients on different modalities of RRT and differences in survival rates between different renal units, cannot therefore be fully understood unless data on comorbidity are collected and analysed.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Clinical staff in each renal unit are responsible for recording (in yes/no format), on their renal unit IT system, the presence or absence of 14 comorbid conditions and on current tobacco smoking (Table 6.1) in each patient starting RRT. Definitions of each of these conditions are given in web Appendix B Definitions, Statistical Methodology and Analysis Criteria (www.renalreg.org). Analyses are restricted to incident patients. Many other national Registries only collect data on patients who have survived the first 90 days of RRT and for the purposes of comparisons with their results, some analyses are restricted to patients surviving the first 90 days of RRT. Complete data on comorbidity for a given patient is considered to have been provided if there is at least one yes/no answer to one of the 14 questions. For some analyses, comorbidities have been collapsed into broader categories.
- ‘Ischaemic heart disease’ is defined as the presence of a ‘yes’ to a history either of angina; MI in past 3 months; MI >3 months; or coronary artery bypass grafting (CABG)/angioplasty (or more than one of these).
- ‘Peripheral vascular disease’ is defined as the presence of a ‘yes’ to a history either of claudication; ischaemic or neuropathic ulcers; non-coronary angioplasty, vascular graft or aneurysm; or amputation for peripheral vascular disease.
- ‘Vascular disease’ is defined as the presence of cerebrovascular disease or any of the data items that comprise ‘peripheral vascular disease’.
|
Data on completeness of co-morbidity returns from each renal unit may differ from those in previous reports because some renal units have provided additional data on comorbidity of previous years’ incident cohorts since original submission.
(Since 2004, the presence or absence of a clinical diagnosis of heart failure was also recordable. However, very few renal units are able to collect or submit this data item and it is not included in any of the analyses reported here).
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Completeness of comorbidity returns from each participating renal unit
Table 6.2 shows that completeness of data returns still varies markedly from renal unit to renal unit with some units continuing to provide data on 100% of patients and others providing no data. There is no relationship between the size of the renal unit and the completeness of data returns. After excluding renal units that returned no data at all, the average completeness of data returns from units ranged from 1% to 100% (mean 63.6%) for 2005, a moderate improvement on a mean of 48.1% in 2000. Amongst all incident patients, data on comorbidity was available on 39% of patients starting in 2000 and on only 43% in 2005 (Table 6.3).
|
|
An analysis of completeness of data returns by the type of renal unit IT system showed no pattern other than very high returns from all centres using the Mediqal system (nine centres: completeness 93.3–100%). As stated above, a return was considered to be ‘complete’ if there was at least one answer to the 14 questions on the comorbidity screen. However, most records that contained at least one answer contained answers to most or all of the other questions; in 2005, of entries that contained at least one entry on comorbidity, 1.34% contained 11 answers, 1.21% contained 12 answers, 7.28% contained 13 answers and 89.95% contained answers to all 14 questions.
Frequency of each comorbidity condition
Table 6.4 gives the frequency of each comorbidity (as a proportion of the total number of incident patients for whom data was available for that item) for patients aged <65 years and 
65 years as well as the total frequency of each comorbidity in the incident population.
|
The denominator for each percentage reported is the number of patients for whom a yes/no answer was provided for that comorbidity.
Frequency of multiple comorbidity
Just under 50% of patients for whom comorbidity data were available starting RRT in 2000–2005 were reported as having no comorbidity present. More than one comorbidity was reported as present in 27% (Table 6.5).
|
Frequency of comorbidity by age band
Figures 6.1 and 6.2 illustrate the rising frequency of comorbidity with increasing age up to 74 years in incident patients; the lower rate of reported comorbidity amongst patients over 75 years may reflect a ‘healthy survivor effect’ or decisions made by nephrologists and/or by patients aged >75 years with cardiovascular comorbidity not to embark on RRT. Smoking is less commonly reported amongst patients starting RRT aged 55 years or older. Ischaemic heart disease, cerebrovascular disease and peripheral vascular disease are all more frequent amongst older as compared with younger patients.
|
|
Frequency of comorbidity amongst patients with diabetes
Diabetes was recorded as the primary renal disease in 20.2% of all patients starting RRT 2000–2005. Table 6.6 compares comorbidity amongst patients with diabetes and patients without diabetes (as cause or comorbidity), showing higher rates of ischaemic heart disease, cerebrovascular disease and peripheral vascular disease amongst diabetic patients.
|
Age and comorbidity in patients starting haemodialysis compared with those starting PD
Figure 6.3 illustrates the younger age profile of patients being treated with PD 90 days after the start of RRT, compared with those starting HD. The median age of patients on PD at day 90 was 58.3 years, compared with 66.9 years for those on HD (P < 0.001, Kruskal–Wallis).
|
Table 6.7 compares the prevalence of each comorbidity in patients on HD and PD at day 0 of starting RRT, showing significantly higher prevalence (at a higher age) amongst HD of all comorbid conditions other than MI more than 3 months ago, CABG, smoking and non-coronary angioplasty. These data probably reflect a perception amongst UK nephrologists, nurses and their patients, that PD is in general more suitable for younger and fitter patients.
|
The percentages are out of the total population of patients on that modality at 90 days, with data for that comorbidity.
Frequency of comorbidity by ethnic origin
For Registry returns, data on ethnic origin was retrieved from fields within renal unit IT systems that were completed by physicians or nurses. These were supplied either as ‘old’ Patient Administration System (PAS) codes (white = 0, black Caribbean = 1, black African = 2, black/other/non-mixed origin = 3, Indian = 4, Pakistani = 5, Bangladeshi = 6, Chinese = 7) or as ‘new’ PAS codes (see web Appendix B www.renalreg.org). For the purposes of analysis, ‘new’ PAS codes are collapsed into the ‘old’ PAS categories, and further collapsed into white (0), black (1,2 or 3), Asian (4, 5 or 6) and Chinese (7).
Figure 6.4 illustrates the presence or absence of comorbidity by ethnic origin, showing a lower prevalence of comorbid conditions amongst patients of black or Asian origin compared with those of white origin. Figures 6.5, 6.6 and 6.7 show that the lower prevalence of comorbidity amongst patients of black or Asian origin is not attributable to younger age amongst these groups, as the prevalence of comorbidity is lower even in the 18–34 year age group than in the white population. Table 6.8 shows the prevalence of major comorbidities in each group; smoking was more common in the white population and ischaemic heart disease and peripheral vascular disease less common in the black population. Table 6.9 gives details of the age structure of each major ethnic group at the start of RRT. Figure 6.8 illustrates the lower prevalence of diabetes amongst ‘white’ patients starting RRT compared with that in other ethnic groups.
|
|
|
|
|
|
|
Renal function at the time of starting RRT and comorbidity
Using the abbreviated four variable MDRD calculation, the eGFR of patients starting RRT was calculated and is shown in Table 6.10. Data from patients with no available creatinine measurement within 14 days before the start of RRT were not used. Patients with an eGFR >20 ml/min/1.73 m2 were excluded from analysis (n = 553). Data from one centre (Hammersmith and Charing Cross) were excluded from analysis because of errors in the data extraction process of this item (n = 568), leaving 14 462 patients included in the analysis.
|
The log of the eGFR was taken to normalize the data and two-sample t-tests were used to compare the means of the log (eGFR) of those patients with the specific comorbidity against those with none of the comorbidities present. As many tests were being carried out, only a P-value <0.01 was considered statistically significant. This should not imply that these differences imply a clinical significance as they may be only small variations.
The (geometric) mean eGFR prior to starting RRT in patients who are recorded as starting without any comorbidity present is 7.1 ml/min/1.73 m2. Patients starting with different comorbidities were compared against this value.
In each case, eGFR appears to have been slightly higher amongst patients with comorbidity compared with patients without comorbidity, suggesting that patients with more comorbidity tend to be advised to start dialysis earlier than those without comorbidity. If trying to compare patient survival between these groups, then the potential of an ‘earlier start’ may need to be adjusted for in the analyses.
Haemoglobin concentration at the time of starting RRT and comorbidity
The mean haemoglobin prior to starting RRT in patients who are recorded as starting without any comorbidity present is 10.1g/dl, with 52% of patients achieving a haemoglobin >10g/dl. Patients starting with different comorbidities were compared against this value (Table 6.11). Haemoglobin concentrations at the start of RRT were slightly higher amongst patients with ischaemic heart disease than in those without, and lower amongst those with liver disease or malignancy. In addition to the direct influence of comorbidity, erythropoietin (EPO) prescribing patterns and late referral of patients will have an influence on these data.
|
Comorbidity and subsequent kidney transplantation
This analysis was confined to incident patients in each of the years 2000–2005 from centres that had returned
80% complete data for comorbidity in that year (Table 6.2). Table 6.12 shows that patients who underwent transplantation had less comorbidity at the start of RRT than those who died or did not receive a transplant.
|
Figure 6.9 gives the age distribution of those who had received a transplant by the end of 2005 compared with those who remained un-transplanted. Over the age of 65 years, the majority of incident patients are unlikely to undergo kidney transplantation, and this is very rare in patients starting RRT over the age of 75 years.
|
Comorbidity and subsequent survival—Introduction
These analyses were performed on patients starting RRT between 1 January 2000 and 30 September 2005, to allow at least 3 months follow-up from the start of RRT. The 1 year after 90 days analyses only include patients who survived at least 90 days on RRT. The death rate is high in the first 90 days and highly variable between centres, due for instance to variation in policies on inclusion of patients with acute kidney injury requiring dialysis. Use of this ‘90 day rule’ also allows direct comparison of survival statistics with those from other national registries.
The effect within each renal unit of adjusting overall survival for comorbidity can be found in Chapter 12.
Comorbidity and survival within 90 days of commencing RRT
The Registry collects data on all patients with a ‘timeline’ entry that have started RRT for ERF. Patients who present acutely, and who are initially classified as acute renal failure requiring dialysis, but continue to require long-term dialysis can be re-classified as having had ERF from the date of their first RRT. (Most other national registries only start the collection of data at 90 days after the first RRT.) This allows the UK Registry, unlike other registries, to collect data on factors affecting outcomes including survival, in the first 90 days of RRT.
The univariate model (Table 6.13), does not allow adjustment for age, so patients were first stratified by age group (<65 years and 65 years) to make some account for the increasing incidence of comorbidity with age which would otherwise obscure the analysis.
|
On univariate analysis stratified for age, most comorbidities were associated with an increased risk of death both amongst patients aged <65 years and those aged
65 years. However, there was no increased risk of death associated with diabetes mellitus as a comorbidity in the absence of diabetes as a cause of primary renal disease; and smoking was not associated with an increased risk of death (Table 6.13). Some comorbidities may appear not to be associated with an increased risk of death because of low numbers—for instance, liver disease in persons aged 65 years. The observation that the risk of death amongst those 65 years is not greater in the presence of ischaemic heart disease may be due either to competing risks or to negative selection caused by clinicians or patients opting not to start RRT in the presence of severe ischaemic heart disease. Of special interest in this univariate survival analysis was that diabetes was not associated with an increased risk of death amongst patients aged 65 years, possibly due to its close association with other comorbidities in this age group. On multivariate analysis using the stepwise Cox proportional hazards model, age and six of the comorbid conditions were identified as significant independent predictors of the risk of death (Table 6.14). Diabetes did not emerge as an independent predictor, probably due to the close association between diabetes and ischaemic heart disease, cerebrovascular disease and peripheral vascular disease.
|
There were 9047 patients included in the analysis. Variables included in the model included: age per 10 years, angina, MI <3 months ago, MI >3 months ago, CABG or coronary angioplasty, cerebrovascular disease, diabetes (whether as a cause of primary renal disease or as a comorbidity), chronic obstructive pulmonary disease, liver disease, malignancy, claudication, ischaemic/neuropathic ulcers, angioplasty/vascular graft, amputation and smoking.
Comorbidity and survival 1 year after 90 days of commencing RRT
In all analyses, patients starting RRT are only included if they survived at least 90 days on RRT. The death rate is high in the first 90 days, and highly variable between centres, due for instance to variation in policies on inclusion of patients with acute kidney injury requiring dialysis. Use of this ‘90 day rule’ also allows direct comparison of survival statistics with those from other national registries.
On univariate analysis (Table 6.15) stratified for age, most comorbidities were associated with an increased risk of death both in patients starting RRT aged <65 years and in those 65 years. Diabetes as a primary cause of renal failure was not associated with an increased risk of death amongst patients over 65 years, possibly due to its close association with other comorbidities in this age group. COPD was not associated with an increased risk of death in patients aged <65 years.
|
On multivariate analysis using the stepwise Cox proportional hazards model, eight variables were identified as independent predictors of death (Table 6.16). Recent MI was no longer significantly associated with an increased risk of death, possibly because the prognostic importance of this marker is time-dependent, and so would not be any more powerful a predictor than other markers of atherosclerotic vascular disease a year later. Diabetes was a powerful predictor of increased risk of death after the first 90 days.
|
There were 6535 patients included in the analysis. Variables in the model included: age per 10 years, angina, MT <3 months ago, MI >3 months ago, CABG or coronary angioplasty, cerebrovascular disease, diabetes (whether as a cause of primary renal disease or as a comorbidity), chronic obstructive pulmonary disease, liver disease, malignancy, claudication, ischaemic/neuropathic ulcers, angioplasty/vascular graft, amputation and smoking.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
These analyses demonstrate that comorbidity is common amongst UK patients starting RRT, with over 50% of all patients having some recorded comorbidity (using data from centres with >80% returns). Reporting of the presence or absence of these simple markers of comorbidity to the Registry is still poor in many centres, although this situation is gradually improving. Unlike many data items recorded in renal unit IT systems, the recording of the presence or absence of comorbidity is probably not required for the routine day-to-day care of these patients. It is anticipated, however, that the introduction of a system of tariff-based payment by results in England might act to encourage clinicians to improve the systematic recording of comorbidity. The Registry is also exploring the possibility of linking to the Hospital Episode Statistics data set within the Secondary Users Service, which would allow data to be obtained on hospital discharge codes, very much along the lines of the approach used by the United States Renal Data System.
These and other previously published analyses using a variety of comorbidity scores [1–26] also demonstrate that comorbidity is a powerful predictor of survival in patients on RRT. The publication of de-anonymized survival statistics for each renal unit in this year's report should also provide a stimulus to renal unit Directors to ensure that they collect and report complete data on comorbidity.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
- Comorbid conditions and correlations with mortality risk among 3,399 incident hemodialysis patients. Am J Kidney Dis (1992) 20([5 Suppl. 2]):32–38.[Medline]
- Avram MM, Mittman N, Bonomini L, Chattopadhyay J, Fein P. Markers for survival in dialysis: a seven-year prospective study. Am J Kidney Dis (1995) 26:209–219.[ISI][Medline]
- Beddhu S, Zeidel ML, Saul M, et al. The effects of comorbid conditions on the outcomes of patients undergoing peritoneal dialysis. Am J Med (2002) 112:696–701.[CrossRef][ISI][Medline]
- Byrne C, Vernon P, Cohen JJ. Effect of age and diagnosis on survival of older patients beginning chronic dialysis. JAMA (1994) 271:34–36.[Abstract]
- Davies SJ, Phillips L, Naish PF, Russell GI. Quantifying comorbidity in peritoneal dialysis patients and its relationship to other predictors of survival. Nephrol Dial Transplant (2002) 17:1085–1092.
[Abstract/Free Full Text] - Di Iorio B, Cillo N, Cirillo M, De Santo NG. Charlson Comorbidity Index is a predictor of outcomes in incident hemodialysis patients and correlates with phase angle and hospitalization. Int J Artif Organs (2004) 27:330–336.[ISI][Medline]
- Fried L, Bernardini J, Piraino B. Charlson comorbidity index as a predictor of outcomes in incident peritoneal dialysis patients. Am J Kidney Dis (2001) 37:337–342.[ISI][Medline]
- Ganesh SK, Hulbert-Shearon T, Port FK, Eagle K, Stack AG. Mortality differences by dialysis modality among incident ESRD patients with and without coronary artery disease. J Am Soc Nephrol (2003) 14:415–424.
[Abstract/Free Full Text] - Goldwasser P, Mittman N, Antignani A, et al. Predictors of mortality in hemodialysis patients. J Am Soc Nephrol (1993) 3:1613–1622.[Abstract]
- Goodkin DA, Bragg-Gresham JL, Koenig KG, et al. Association of comorbid conditions and mortality in hemodialysis patients in Europe, Japan, and the United States: the Dialysis Outcomes and Practice Patterns Study (DOPPS). J Am Soc Nephrol (2003) 14:3270–3277.
[Abstract/Free Full Text] - Held PJ, Pauly MV, Diamond L. Survival analysis of patients undergoing dialysis. JAMA (1987) 257:645–650.[Abstract]
- Hemmelgarn BR, Manns BJ, Quan H, Ghali WA. Adapting the Charlson Comorbidity Index for use in patients with ESRD. Am J Kidney Dis (2003) 42:125–132.[CrossRef][ISI][Medline]
- Hutchinson TA, Thomas DC, MacGibbon B. Predicting survival in adults with end-stage renal disease: an age equivalence index. Ann Intern Med (1982) 96:417–423.[ISI][Medline]
- Iseki K, Kawazoe N, Osawa A, Fukiyama K. Survival analysis of dialysis patients in Okinawa, Japan (1971–1990). Kidney Int (1993) 43:404–409.[ISI][Medline]
- Johnson JG, Gore SM, Firth J. The effect of age, diabetes, and other comorbidity on the survival of patients on dialysis: a systematic quantitative overview of the literature. Nephrol Dial Transplant (1999) 14:2156–2164.
[Abstract/Free Full Text] - Khan IH. Comorbidity: the major challenge for survival and quality of life in end-stage renal disease. Nephrol Dial Transplant (1998) 13(Suppl. 1):76–79.
[Free Full Text] - Lacson E Jr, Teng M, Lazarus JM, Lew N, Lowrie E, Owen W. Limitations of the facility-specific standardized mortality ratio for profiling health care quality in dialysis. Am J Kidney Dis (2001) 37:267–275.[ISI][Medline]
- Miguel A, Garcia-Ramon R, Perez-Contreras J, et al. Comorbidity and mortality in peritoneal dialysis: a comparative study of type 1 and 2 diabetes versus nondiabetic patients. Peritoneal dialysis and diabetes. Nephron (2002) 90:290–296.[CrossRef][ISI][Medline]
- Miskulin DC, Martin AA, Brown R, et al. Predicting 1 year mortality in an outpatient haemodialysis population: a comparison of comorbidity instruments. Nephrol Dial Transplant (2004) 19:413–420.
[Abstract/Free Full Text] - Miskulin DC, Meyer KB, Athienites NV, et al. Comorbidity and other factors associated with modality selection in incident dialysis patients: the CHOICE Study. Choices for Healthy Outcomes in Caring for End-Stage Renal Disease. Am J Kidney Dis (2002) 39:324–336.[ISI][Medline]
- Miskulin DC, Meyer KB, Martin AA, et al. Comorbidity and its change predict survival in incident dialysis patients. Am J Kidney Dis (2003) 41:149–161.[CrossRef][ISI][Medline]
- Schrander-v d Meer AM, van Saase JL, Roodvoets AP, van Dorp WT. Mortality in patients receiving renal replacement therapy, a single center study. Clin Nephrol (1995) 43:174–179.[ISI][Medline]
- van Manen JG, Korevaar JC, Dekker FW, Boeschoten EW, Bossuyt PM, Krediet RT. How to adjust for comorbidity in survival studies in ESRD patients: a comparison of different indices. Am J Kidney Dis (2002) 40:82–89.[CrossRef][ISI][Medline]
- Van Manen JG, Korevaar JC, Dekker FW, Boeschoten EW, Bossuyt PM, Krediet RT. Adjustment for comorbidity in studies on health status in ESRD patients: which comorbidity index to use? J Am Soc Nephrol (2003) 14:478–485.
[Abstract/Free Full Text] - Vonesh EF, Snyder JJ, Foley RN, Collins AJ. The differential impact of risk factors on mortality in hemodialysis and peritoneal dialysis. Kidney Int (2004) 66:2389–2401.[CrossRef][ISI][Medline]
- Weller JM, Port FK, Swartz RD, Ferguson CW, Williams GW, Jacobs JF Jr. Analysis of survival of end-stage renal disease patients. Kidney Int (1982) 21:78–83.[ISI][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||