Abstract

Early identification of patients with chronic kidney disease (CKD) may allow health-care systems to implement interventions aimed at decreasing disease progression and eventual morbidity and mortality. Primary care in the United Kingdom is computerized suggesting a separate screening program for CKD may not be necessary because identifying data already populates primary care databases. Our study utilized a data set of 163 demographic, laboratory, diagnosis, and prescription variables from 130 226 adults in the regions of Kent, Manchester, and Surrey. The patients were 18 years of age and older in a 5-year study period culminating in November 2003. Estimated glomerular filtration rate was calculated from the four-variable Modification of Diet in Renal Disease equation using calibrated creatinine levels. A valid creatinine value was recorded in almost 30% of this cohort. The age-standardized prevalence of stage 3–5 CKD was 10.6% for females and 5.8% for males. In these patients, the odds ratio for hypertension was 2.1, for diabetes 1.33, and for cardiovascular disease 1.69. Only 20% of the diabetic people with stage 3–5 CKD had a blood pressure less than or equal to 130/80 mm Hg. The proportion of patients with anemia significantly rose as renal function declined. We suggest that stage 3–5 CKD is easily detected in existing computerized records. The associated comorbidity and management is readily available enabling intervention and targeting of specialist resources. Early identification of patients with chronic kidney disease (CKD) may allow health-care systems to implement interventions aimed at decreasing disease progression and eventual morbidity and mortality. Primary care in the United Kingdom is computerized suggesting a separate screening program for CKD may not be necessary because identifying data already populates primary care databases. Our study utilized a data set of 163 demographic, laboratory, diagnosis, and prescription variables from 130 226 adults in the regions of Kent, Manchester, and Surrey. The patients were 18 years of age and older in a 5-year study period culminating in November 2003. Estimated glomerular filtration rate was calculated from the four-variable Modification of Diet in Renal Disease equation using calibrated creatinine levels. A valid creatinine value was recorded in almost 30% of this cohort. The age-standardized prevalence of stage 3–5 CKD was 10.6% for females and 5.8% for males. In these patients, the odds ratio for hypertension was 2.1, for diabetes 1.33, and for cardiovascular disease 1.69. Only 20% of the diabetic people with stage 3–5 CKD had a blood pressure less than or equal to 130/80 mm Hg. The proportion of patients with anemia significantly rose as renal function declined. We suggest that stage 3–5 CKD is easily detected in existing computerized records. The associated comorbidity and management is readily available enabling intervention and targeting of specialist resources. Chronic kidney disease (CKD) is a major public health problem imposing a substantial burden on the patients affected and on the health-care systems caring for them. CKD is now conventionally divided into five stages (Table 1) following the classification proposed by the National Kidney Foundation Kidney Disease Outcome Quality Initiative in 2002.1K/DOQI Clinical Practice Guidelines for Chronic Kidney Disease Evaluation, classification, and stratification.Am J Kidney Dis. 2002; 39: S17-S31Google Scholar Data from the third National Health and Nutrition Examination Survey (NHANES III) demonstrated that the number of people affected in the United States of America is high2Coresh J. Astor B.C. Greene T. et al.Prevalence of chronic kidney disease and decreased kidney function in the adult US population: Third National Health and Nutrition Examination Survey.Am J Kidney Dis. 2003; 41: 1-12Abstract Full Text Full Text PDF PubMed Scopus (2211) Google Scholar and subsequent reports have detailed similar, or higher, estimates of prevalence from various countries.3de Lusignan S. Chan T. Stevens P.E. et al.Identifying patients with chronic kidney disease from general practice computer records.Family Practice. 2005; 22: 234-241Crossref PubMed Scopus (124) Google Scholar, 4Chadban S.J. Briganti E.M. Kerr P.G. et al.Prevalence of Kidney Damage in Australian Adults: The AusDiab Kidney Study.J Am Soc Nephrol. 2003; 14: S131-S138Crossref PubMed Google Scholar, 5Verhave J.C. Hillege H.L. Burgerhof J.G. PREVEND Study Group et al.The association between atherosclerotic risk factors and renal function in the general population.Kidney Int. 2005; 67: 1967-1973Abstract Full Text Full Text PDF PubMed Scopus (85) Google Scholar, 6Drey N. Roderick P. Mullee M. et al.A population-based study of the incidence and outcomes of diagnosed chronic kidney disease.Am J Kidney Dis. 2003; 42: 677-684Abstract Full Text Full Text PDF PubMed Scopus (280) Google Scholar, 7Amato D. Alvarez-Aguilar C. Castaneda-Limones R. et al.Prevalence of chronic kidney disease in an urban Mexican population.Kidney Int Suppl. 2005: S11-S17Abstract Full Text Full Text PDF Scopus (110) Google Scholar The prevalence of CKD increases exponentially with age, and we can expect numbers to rise as the population continues to age and the prevalence of type II diabetes increases. Cohort studies indicate that the risk of mortality in CKD far outweighs the risk of progression to end-stage renal disease. Cardiovascular causes account for nearly 50% of the mortality and CKD is an independent predictor of cardiovascular comorbidity.3de Lusignan S. Chan T. Stevens P.E. et al.Identifying patients with chronic kidney disease from general practice computer records.Family Practice. 2005; 22: 234-241Crossref PubMed Scopus (124) Google Scholar, 4Chadban S.J. Briganti E.M. Kerr P.G. et al.Prevalence of Kidney Damage in Australian Adults: The AusDiab Kidney Study.J Am Soc Nephrol. 2003; 14: S131-S138Crossref PubMed Google Scholar, 8John R.I. Webb M.C. Young A. et al.Unreferred chronic kidney disease: a longitudinal study.Am J Kidney Dis. 2004; 43: 825-835Abstract Full Text Full Text PDF PubMed Scopus (200) Google Scholar, 9Keith D.S. Nichols G.A. Gullion C.M. et al.Longitudinal follow-up and outcomes among a population with chronic kidney disease in a large managed care organization.Arch Intern Med. 2004; 164: 659-663Crossref PubMed Scopus (1283) Google Scholar, 10Go A.S. Chertow G.M. Fan D. et al.Chronic kidney disease and the risks of death, cardiovascular events, and hospitalisation.N Engl J Med. 2004; 351: 1296-1305Crossref PubMed Scopus (8252) Google Scholar, 11Foley R.N. Murray A.M. Li S. et al.chronic kidney disease and the risk for cardiovascular disease, renal replacement, and death in the United States Medicare Population, 1998–1999.J Am Soc Nephrol. 2005; 16: 489-495Crossref PubMed Scopus (746) Google ScholarTable 1National Kidney Foundation KDOQI staging for CKDStage of CKDDescriptionGFR (ml/min/1.73 m2)1Kidney damage with normal or raised GFR>902Kidney damage with mildly reduced GFR60–893Moderately reduced GFR30–594Severe reduction in GFR15–295Kidney failure75 years were overrepresented (Figure 2). The number aged ≥18 years was 130 226, mean age 47.2±18.78 years, female to male ratio 1:1. Only 0.63% (818/130 226) of patients had ethnicity recorded. The mean body mass index in those with both height and weight recorded was 25.8±5.22 kg/m2.Figure 2Study age and sex distribution compared to England and Wales.View Large Image Figure ViewerDownload (PPT) The mean age was 58.1±18.1 years, and female to male ratio was 1.3:1. Height was recorded in 76.1% of the creatinine subset and weight in 80.8%, the mean body mass index was 27.1±5.5 kg/m2, (Figure 1). In the most recent 24 months of the 5-year period, 70% of the study population had an serum creatinine (SCr) recorded. Table 2 shows the age standardized rates for CKD stage 3–5, subdivided by gender. The overall prevalence of CKD stage 3–5 was 8.5% and was higher in females, 10.6 versus 5.8% in males. The effect of creatinine standardization was to increase the proportion of those with stage 3 CKD by a factor of 1.75 and the proportion of those with stage 4 CKD by a factor of 1.6 (data not shown). The proportion of those with stage 5 CKD remained unchanged.Table 2Age-standardized rates for stage 3–5 CKDAge bandsMalesFemalesStudy population*Census populationStudy population*Census populationnProportion with stage 3–5 CKDnExpected stage 3–5 CKDnProportion with stage 3–5 CKDnExpected stage 3–5 CKD18–2482730.01%3 671 800443.829377160.18%3 588 9006511.7425–3412 4240.17%4 215 2007 124.85510 9230.79%4 259 80033 538.735–4413 1150.71%4 381 70031 071.1511 9882.69%4 464 600119 92045–5410 5663.08%3 856 300118 616.199732.79%3 920 300109 27955–6495186.89%3 089 600212 941.5925413.09%3 186 200416 95465–74635617.65%2 307 700407 369.3694327.86%2 639 700735 29975–84388433.16%1 308 300433 854.4575441.68%1 987 300828 21485+99044.75%312 400139 791.1254948.61%817 300397 267Total65 126a23 143 000b1 351 21265 100a24 864 100b2 646 984Age-standardized rate5.8%10.6%CKD, chronic kidney disease.*Based on UK 2001 census data.Expected stage 3–5 CKD by age band=proportion with stage 3–5 CKD *census population.Total for expected CKD (b)=Sum of expected CKD for age bands.Age-standardized rate=b/a *100. Open table in a new tab CKD, chronic kidney disease. *Based on UK 2001 census data. Expected stage 3–5 CKD by age band=proportion with stage 3–5 CKD *census population. Total for expected CKD (b)=Sum of expected CKD for age bands. Age-standardized rate=b/a *100. Table 3 shows the patient demographics and reported comorbidity. There was an increased female preponderance in the three strata of eGFR below 60 ml/min/1.73 m2. In addition, the numbers of those aged ≥70 years increased as eGFR fell; 76.7% of persons with eGFR 60 ml/min/1.73 m2 n=26 531Total n=38 262Demographics n (%) F:M341:1841731:7445710:302113 987:12 54421 769:16 4931.85:12.3:11.89:11.1:11.3:1 Aged >70 years403 (76.7)2009 (81.2)4343 (49.7)4136 (15.6)10 890 (28.5)Characteristics mean±s.d. Age (years)*76.8±14.178.4±10.469.7±13.552.3±16.858.1±18.1 BMI (kg/m2)*27.8±6.127.7±5.427.4±5.426.8±5.627.0±5.5 SBP (mm Hg)*141.3±20.8142.6±20.4138.8±18.9131.8±18.9134.3±19.4 DBP (mmHg)*76.7±11.277.9±10.878.8±1078.5±10.378.5±10.3 SCr (mg/dl)*2.87±1.691.52±0.251.2±0.170.96±0.171.07±0.37 GFR*22.5±6.439.3±4.053.6±4.277.7±18.869.6±21.3 Hb (g/dl)*13.2±1.813.5±1.613.8±1.513.9±1.513.8±1.5Comorbidity n (%) Diabetes121 (23)398 (16.1)1049 (12)2495 (9.4)4063 (10.6) Hypertension461 (87.8)2143 (86.6)6235 (71.4)12 493 (47.1)21 332 (55.8) All CVD266 (50.7)1056 (42.7)2369 (27.1)3929 (14.8)7620 (19.9) Hypercholesterolemia231 (44)1034 (41.8)3751 (43)11 ,014 (41.5)16 030 (41.9) Proteinuria/haematuria144 (27.4)610 (24.6)1086 (12.4)2690 (10.1)4800 (12.5) Renal diagnosis101 (19.2)90 (3.6)51 (0.6)355 (1.3)597 (1.6) BMI372 (70.1)1834 (74.1)6559 (75.1)19 736 (74.4)28 501 (74.5) Blood pressure record503 (95.8)2422 (97.9)8482 (97.1)24 719 (93.2)31 740 (94.4)ANOVA, analysis of variance; BMI, body mass index; CVD, cardiovascular disease; DBP, diastolic blood pressure; GFR, glomerular filtration rate; Hb, hemoglobin; SBP, systolic blood pressure.*P60 ml/min/1.73 m2All‡Hb tested439 (83.6)2057 (83.1)7308 (83.7)22 581 (85.1)32 385 (84.6)‡WHO anemia127 (28.3)379 (18.4)948 (13.0)3024 (13.4)4478 (13.8)‡KDOQI anemia94 (21.4)294 (14.3)590 (8.1)1347 (6.0)2325 (7.2)‡Hb60 ml/min/1.73 m2TotalHypertension defined4612143623512 49321 332Antihypertensive medications None117 (25.4)598 (27.9)2358 (37.8)6530 (52.3)9603 (45) ‡Average number of antihypertensives1.841.761.661.541.6 Use of ACEi/ARB150 (32.5)804 (37.5)2679 (43)3467 (27.8)7100 (33.2)Achieved targets *BP60 ml/ml/1.73 m2 (40.9 vs 44.4%).Table 6Diabetes, treatment, and renal functioneGFR 7.5%642 (40.9)1108 (44.4)P<0.05Treated hypertension1313 (83.7)1590 (63.7)P<0.001BP<130/80 mm Hg in treated hypertensives270 (21)281 (11.2)P<0.001ACEi, angiotensin-converting enzyme inhibitor; ARB, angiotensin 2 receptor blocker; BP, blood pressure; eGFR, estimated glomerular filtration; HbA1c, hemoglobin A1c; NS, not significant.Values are n (%). Open table in a new tab ACEi, angiotensin-converting enzyme inhibitor; ARB, angiotensin 2 receptor blocker; BP, blood pressure; eGFR, estimated glomerular filtration; HbA1c, hemoglobin A1c; NS, not significant. Values are n (%). The proportion of people with hypercholesterolemia was similar in each GFR stratum, (Table 3). After adjustment for age and gender, the OR for hypercholesterolemia in those with eGFR <60 ml/min/1.73 m2 was 1.09 (95% CI 0.99–1.2). The proportion of people in the study cohort with cardiovascular disease (CVD) was 19.9%. CVD was more prevalent in the eGFR <30 ml/min/1.73 m2 stratum (50.7%), (Table 3). The crude OR for CVD in persons with eGFR <60 ml/min/1.73 m2 was 2.92 (95% CI 2.71–3.13) in females and 2.86 (95% CI 2.65–3.1) in males; age/gender adjusted OR 1.69 (95% CI 1.59–1.79). Table 7 details the prescription of ACEI/ARB, antiplatelet agents and lipid-lowering therapy in those with CVD. Those with eGFR <60 ml/min/1.73 m2 were more likely to be prescribed ACEI/ARB and antiplatelet agents. As with the total cohort, the proportion of people with CVD treated with ACEI/ARB, lipid-lowering therapy, and antiplatelet agents decreased in the eGFR <30 ml/min/1.73 m2 stratum (data not shown).Table 7Prescribed treatment in those with cardiovascular diseasePrescribed treatmenteGFR <60 ml/min/1.73 m2 n=3691eGFR ≥60 ml/min/1.73 m2 n=3929P-valueACEi/ARBs1520 (41.1)1352 (34.4)<0.001Antiplatelet agents1899 (51.4)1590 (40.4)<0.001Lipid-lowering agents1910 (50.7)2038 (51.9)NSACEi, angiotensin-converting enzyme inhibitor; ARB, angiotensin 2 receptor blocker; eGFR, estimated glomerular filtration; NS, not significant.Values are n (%). Open table in a new tab ACEi, angiotensin-converting enzyme inhibitor; ARB, angiotensin 2 receptor blocker; eGFR, estimated glomerular filtration; NS, not significant. Values are n (%). Routinely collected primary care computer records can be used to not only detect patients with CKD stages 3–5, but also to describe their comorbidity and current management. We have demonstrated in a large study population, which also included institutionalized people, that the age-standardized adult prevalence of stage 3–5 CKD is 8.5% (10.6% for females and 5.8% for males). This prevalence is based on the assumption that no untested people in the primary care population have CKD stage 3–5 and therefore inevitably underestimates its prevalence. In addition due to the study design this estimate is subject to Neyman bias contributing to the underestimate of prevalence. However, the magnitude of the effect may be less than expected because 70% of patients had a serum creatinine within the last 24 months of the study period. This is discussed in detail later in the discussion. Although because of differences in methodology comparisons with other studies need to be viewed with caution, the gender specific rate for CKD stage 3–5 in this study compares to results from a population study in Iceland17Victorsdottir O. Pallson R. Andresdottir M.B. et al.Prevalence of chronic kidney disease based on estimated blomerular filtration rate and proteinuria in Icelandic adults.Nephrol Dial Transpl. 2005; 20: 1799-1807Crossref PubMed Scopus (89) Google Scholar showing age-standardized rates of 11.55 and 4.71% for females and males, respectively. Other studies have also outlined varying prevalence rates for CKD stage 3–5. A small study from Mexico (n=3564) recorded a prevalence of 8.5% for stage 3–5 CKD,6Drey N. Roderick P. Mullee M. et al.A population-based study of the incidence and outcomes of diagnosed chronic kidney disease.Am J Kidney Dis. 2003; 42: 677-684Abstract Full Text Full Text PDF PubMed Scopus (280) Google Scholar the PREVEND study4Chadban S.J. Briganti E.M. Kerr P.G. et al.Prevalence of Kidney Damage in Australian Adults: The AusDiab Kidney Study.J Am Soc Nephrol. 2003; 14: S131-S138Crossref PubMed Google Scholar from the Netherlands (n=8459) reported a prevalence of 5.84%, and the AusDiab study3de Lusignan S. Chan T. Stevens P.E. et al.Identifying patients with chronic kidney disease from general practice computer records.Family Practice. 2005; 22: 234-241Crossref PubMed Scopus (124) Google Scholar from Australia (n=11 247) reported the highest prevalence of stage 3–5, 11.2%. A much larger study from Northern Ireland (n=337 618), published in abstract only, showed an overall population prevalence of stage 3–5 CKD of 8.0%.18Fogarty D.G. Maxwell A.P. Savage G. et al.There is no population level benefit in using estimated glomerular filtration rate (eGFR) versus serum creatinine for identifying and referring patients with CKD.J Am Soc Nephrol. 2005; 16: 319AGoogle Scholar The North American comparator is the NHANES III study (n=15 625), representative of the noninstitutionalized US population, which reported that 4.7% of the population had stage 3–5 CKD.1K/DOQI Clinical Practice Guidelines for Chronic Kidney Disease Evaluation, classification, and stratification.Am J Kidney Dis. 2002; 39: S17-S31Google Scholar The methodologies of all of these studies clearly differ to that used in our study, but these are the only ones in the literature providing population prevalence estimates in the countries quoted. The change in prevalence engendered by creatinine standardization is significant. In our earlier, smaller study, we predicted a whole population prevalence of stage 3–5 CKD of 4.9% using unstandardized creatinine data.2Coresh J. Astor B.C. Greene T. et al.Prevalence of chronic kidney disease and decreased kidney function in the adult US population: Third National Health and Nutrition Examination Survey.Am J Kidney Dis. 2003; 41: 1-12Abstract Full Text Full Text PDF PubMed Scopus (2211) Google Scholar For the same study period in those aged ≥18 years, the age-standardized prevalence would have been 4.93% (3.16% for males, 6.86% for females). The effect of standardization of creatinine was to lower eGFR and this effect was greatest at lower creatinine levels. The proportion of those with stage 5 CKD remained unchanged, whereas the proportion of those with stage 4 CKD increased by a factor of 1.6, and by a factor of 1.75 in people with stage 3 CKD. There are clear limitations to this cross-sectional study. Although the age and sex profile of the study population were similar to that of England and Wales, ethnicity was unreliably recorded, precluding use of a correction factor for Afro-Caribbean ethnicity in eGFR calculation by the Modification of Diet in Renal Disease (MDRD) equation. Data from the Office of National Statistics suggest that 1.35% of the study population would have been of Afro-Caribbean and 4.11% of Asian ethnicity compared to 2.85 and 4.73%, respectively, in the population of England and Wales as a whole.19National Statistics: 2001. Census. Geographic distribution:by minority ethnic population: Social Focus in Brief: Ethnicity.http://www.statistics.gov.uk/census200Google Scholar This mitigates against a significant overestimate of lower levels of eGFRs despite the lack of recording of ethnicity in the study population. Can we assume that we have captured all those with stage 3–5 CKD? This was a cross-sectional survey of live patients, in other words a patient whose SCr had been checked but who had subsequently died before the data collection would not have appeared in the primary care database at the time of the study (Neyman bias). This will have a significant effect given the fact that large studies have demonstrated that the risk of death in CKD is high.8John R.I. Webb M.C. Young A. et al.Unreferred chronic kidney disease: a longitudinal study.Am J Kidney Dis. 2004; 43: 825-835Abstract Full Text Full Text PDF PubMed Scopus (200) Google Scholar,9Keith D.S. Nichols G.A. Gullion C.M. et al.Longitudinal follow-up and outcomes among a population with chronic kidney disease in a large managed care organization.Arch Intern Med. 2004; 164: 659-663Crossref PubMed Scopus (1283) Google Scholar In addition, when interpreting the comorbidities and management of these patients, the fact that there is an overrepresentation of survivors may impact on their estimates. Also, a SCr value was recorded in 31.5% of the adult population and in predicting the prevalence of stage 3–5 CKD, we have assumed that those who did not have a SCr measured would not have had a GFR of <60 ml/min/1.73 m2. Only a small proportion of patients either had a coded diagnosis of renal disease, or had tests for hematuria, and/or proteinuria recorded. This poor recording was validated by the manual search of 10 979 primary care records.16Anandarajah S. Tai T. de Lusignan S. et al.The validity of searching routinely collected general practice computer data to identify patients with chronic kidney disease (CKD): a manual review of 500 medical records.Nephrol Dial Transplant. 2005; 20: 2089-2096Crossref PubMed Scopus (54) Google Scholar This suggests that a relatively small proportion of people with CKD have been identified and coded as such in primary care. This is an important observation and is an area which requires future study particularly in the United Kingdom where universal eGFR reporting by laboratories was introduced in April 2006 by the Department of Health20Department of Health. Publications Policy and Guidance Article. Estimating glomerular filtration rate (eGFR): Information for General Practitionershttp://www.dh.gov.uk/PublicationsAndStatistics/Publications/PublicationsPolicyAndGuidance/PublicationsPolicyAndGuidanceArticle/fs/en?CONTENT_ID=4133020andchk=HDeM/vGoogle Scholar and the General Practitioner Quality and Outcomes Framework (QOF)21NHS Employers. Primary care contracting. general medical services contract. Revisions to the GMS Contract 2007/07.http://www.nhsemployers.org/primary/primary-902.cfm, NHS-28159-1.Google Scholar included an incentive for setting up registries of patients with CKD stage 3–5 (also in April 2006) which included BP targets for patients with CKD. The impact of these measures on the recording of renal disease will be an important area of study, and so too will be the effect of increased recognition of CKD on the management of both risk factors for progression of CKD and for associated comorbidity. The recording of important data such as BP, diabetes, CVD, Hb, and lipids was sufficiently complete to enable us to describe the associated comorbidity. Prescription data was 100% complete, allowing us to also describe the current management of these patients. What does the data tell us? These data clearly show an exponential increase in prevalence of stage 3–5 CKD with age and suggest that female gender is a predictor of lower level of eGFR. There is a high prevalence of hypertension in patients with eGFR <30 ml/min/1.73 m2 (87.8%) and in patients with higher, but reduced, eGFR. Our data also tell us that very few patients achieve the level of BP control required to prevent progression of renal disease.22Hsu C. McCulloch C.E. Darbinian J. et al.Elevated blood pressure and risk of end-stage renal disease in subjects without baseline kidney disease.Arch Intern Med. 2005; 165: 923-928Crossref PubMed Scopus (307) Google Scholar Furthermore, the management of those with higher eGFR was significantly worse than those with CKD stages 3–5. A major proportion of patients with documented hypertension received no treatment. Of those who were prescribed treatment, patients with stage 3–5 CKD were significantly more likely to be prescribed ACEIs/ARBs (P<0.001) and a greater proportion achieved lower levels of recorded BP (P<0.001 for all levels) compared with those with higher eGFR. However, those with an eGFR <30 ml/min/1.73 m2 were less likely to receive ACEI/ARB than those with an eGFR 30–60 ml/min/1.73 m2. This implies that an opportunity to practice preventive medicine is currently being lost. Similarly, despite the wealth of literature and national guidelines supporting and advocating the importance of control of diabetes and hypertension in patients with diabetes, the levels of diabetic control and achieved BP in patients at all levels of eGFR is suboptimal and dictates the need for improvement. Our data suggest that the potential number of patients with stage 3–5 CKD and anemia defined by Kidney Dialysis Outcomes Quality Initiative in the UK is 399 020 subjects. Although not all of these may be suitable for treatment, we know that treating low Hb improves quality of life, and observational studies suggest that lower levels of Hb are associated with increased mortality and hospitalization. A large proportion of those patients with stage 3 CKD are over the age of 75 in our cohort. We acknowledge that use of the MDRD equation in subjects older than the population from which the equation was derived is also a further potential source of bias and that this area merits further study. Furthermore, there is currently little evidence in the elderly to support the specific use of agents to delay the progression of CKD and although modification of cardiovascular risk may have an impact in the elderly, again concrete evidence for this is lacking in the setting of CKD. The key questions to be answered are what constitutes a normal GFR in the elderly, what level of GFR is associated with adverse outcomes in the elderly, are the other determinants of adverse outcomes in CKD the same for the elderly as for younger age groups, can the risk factors for these adverse outcomes be modified by intervention, and if so are the targets the same in the elderly as for younger age groups? Recent work from our unit demonstrated that patients with significant CKD unknown to renal services have increased levels of mortality with a standardized mortality of 34.5 in those under the age of 60.7Amato D. Alvarez-Aguilar C. Castaneda-Limones R. et al.Prevalence of chronic kidney disease in an urban Mexican population.Kidney Int Suppl. 2005: S11-S17Abstract Full Text Full Text PDF Scopus (110) Google Scholar Cardiovascular mortality was predominant. In this study, the prevalence of all forms of CVD was greater than the general population at all levels of eGFR but was significantly increased in those with stage 3–5 CKD. Prescription data suggested that the management of these patients could be considerably improved upon. Although patients with stage 3–5 CKD and CVD had significantly higher prescribing rates than those with higher levels of eGFR, still only 50% were prescribed antiplatelet agents and lipid-lowering therapy, and even less were prescribed ACEIs/ARBs. We have shown that it is possible to use routinely collected primary care computer data to highlight CKD stage 3–5 and to describe the associated comorbidity and current management of patients with CKD stage 3–5. This approach enables considerable numbers of patients to be highlighted who could have improved primary care management of their risk factors for progression of CKD and of their cardiovascular risk, with appropriate referral to secondary care where indicated. The next phase of this project is to develop an expert system to use this existing data to guide patients with CKD into the appropriate disease management pathway.