Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Impact of the estimation equation for GFR on population-based prevalence estimates of kidney dysfunction

Impact of the estimation equation for GFR on population-based prevalence estimates of kidney... Background: Estimating equations are recommended by clinical guidelines as the preferred method for assessment of glomerular filtration rate (GFR). The aim of the study was to compare population-based prevalence estimates of decreased kidney function in Germany defined by an estimated GFR (eGFR) <60 ml/min/1.73m using different equations. Methods: The study included 7001 participants of the German Health Interview and Examination Survey for Adults 2008–2011 (DEGS1) for whom GFR was estimated using the Modification of Diet in Renal Disease study equation (MDRD), the revised Lund-Malmö equation (LM), the Full Age Spectrum creatinine equation (FAScre), the Chronic Kidney Disease Epidemiology Collaboration equations with creatinine and cystatin C (CKD-EPIcrecys), with creatinine (CKD-EPIcre) and with cystatin C (CKD-EPIcys). Bland-Altman plots were used to evaluate the agreement between the equations. Results: Prevalence estimates of decreased kidney function were: 2.1% (CKD-EPIcys), 2.3% (CKD-EPIcrecys), 3.8% (CKD-EPIcre), 5.0% (MDRD), 6.0% (LM) and 6.9% (FAScre). The systematic differences between the equations were smaller by comparing either equations that include serum cystatin C or equations that include serum creatinine alone and increased considerably by increasing eGFR. Conclusions: Prevalence estimates of decreased kidney function vary considerably according to the equation used for estimating GFR. Equations that include serum cystatin C provide lower prevalence estimates if compared with equations based on serum creatinine alone. However, the analysis of the agreement between the equations according to eGFR provides evidence that the equations may be used interchangeably among persons with pronounced decreased kidney function. The study illustrates the implications of the choice of the estimating equation in an epidemiological setting. Keywords: Epidemiology, Prevalence, Renal dysfunction, eGFR equation Background excretory kidney function in health and disease [3, 5]. In Chronic Kidney Disease (CKD) is defined by morpho- the epidemiological setting, decreased kidney function logical and functional damage to the kidney [1, 2]. may be defined by a GFR <60 ml/min/1.73m [1]. As Clinical assessment of kidney function is central to the directly measuring GFR is often cumbersome in routine routine clinical practice [3, 4] and glomerular filtration clinical practice, researchers have developed and vali- rate (GFR) is the best overall index-indicator of dated several GFR estimating equations that include demographic and clinical variables as surrogates for muscle mass and unmeasured factors that affect serum creatinine level, such as age, sex and race. Some of these equations are meanwhile recommended by clinical * Correspondence: [email protected] guidelines as the preferred method for assessment of Center of Clinical Epidemiology, c/o Institute of Medical Informatics, GFR in the routine clinical care [1]. Biometry and Epidemiology (IMIBE), University Hospital Essen, Hufelandstr. 55, 45147 Essen, Germany Full list of author information is available at the end of the article © The Author(s). 2017 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Trocchi et al. BMC Nephrology (2017) 18:341 Page 2 of 10 Decreased kidney function has been shown to be an tests. The DEGS1 study has a mixed design, which en- independent marker for major adverse outcomes of ables both cross-sectional and longitudinal analyses. For CKD, including progression to end-stage kidney failure this purpose, a random sample from local population and premature death caused by cardiovascular disease registries was drawn to supplement former participants [6–8]. Patients with decreased kidney function require from the German National Health Interview and Exam- considerable medical attention to prevent deterioration ination Survey 1998 (GNHIES98). To evaluate kidney and the development of complications. If kidney disease function, blood samples were taken from all participants progresses to end-stage kidney disease, renal replace- and serum creatinine concentration (Architect, Abbott ment therapy is an enormously resource consuming Diagnostics, Wiesbaden; IDMS traceable creatinine Assay, condition. Given its high impact on patients’ quality of kinetic Jaffe’s method) and serum cystatin C concentration life and medical resources and rising prevalence esti- (N Latex Cystatin C assay, Prospec, Siemens Healthcare, mates reported from many countries [9], CKD is in- Eschborn) were measured. Participants with diabetes mel- creasingly recognized a major public health problem and litus were identified according to self-reported medical the knowledge of its prevalence is of great importance history and verified current use of anti-diabetic drugs [21]. from both a medical and the economical standpoint. Participants with gestational diabetes were not included Several studies evaluated performance and limitations of among those with diabetes mellitus. Arterial hypertension different estimating equations for GFR against a gold was assumed if the participant reported current treatment standard of kidney function testing in a clinical setting with antihypertensive medications or elevated blood pres- [4, 10–14]. Although it is widely known to nephrologist sure (≥140 mmHg systolic or ≥90 mmHg diastolic) was that the eGFR equations perform differently in relation measured in the survey [22]. to patient characteristics, the behavior of the equations The analyses presented here refer to the sample of in unselected large population-based samples has only 7001 participants of the DEGS1 aged 18–79 years for been investigated in detail in the NHANES study to our whom estimated GFR (eGFR) was calculated using six knowledge. Therefore, a detailed assessment of the be- GFR estimating equations (creatinine was measured in havior of the equations in a European, predominantly mg/dl, cystatin C in mg/l): the isotope dilution mass Caucasian population of 7000 participants is important spectrometry traceable Modification of Diet in Renal for researchers who want to provide population-based Disease study equation (MDRD) [23], the revised Lund- prevalence estimates of kidney dysfunction. To date, lit- Malmö equation (LM) [12], the Full Age Spectrum cre- erature about the prevalence of kidney function in atinine equation (FAScre) [24], the Chronic Kidney Dis- Germany is scarce. Recently, we published population- ease Epidemiology Collaboration creatinine equation based estimates of prevalence of kidney damage in (CKD-EPIcre), the Chronic Kidney Disease Epidemi- Germany based on measures of albuminuria and the use ology Collaboration cystatin C equation (CKD-EPIcys) of an established equation for GFR estimation [15]. Fur- and the Chronic Kidney Disease Epidemiology Collabor- thermore, the Study of Health in Pomerania (SHIP-1) ation creatinine and cystatin C equation (CKD-EPI- and the Cooperative Health Research in the Region of crecys) [4]. Equations are detailed in Fig. 1. Persons with Augsburg (KORA F4) reported results about prevalence missing data on eGFR were excluded (N = 114). of decreased kidney function in Northeast and Southern Germany respectively [16]. Finally, the Berlin Initiative Statistical analysis Study assessed kidney function in Berlin in a cross sec- Participants were classified into four GFR categories tional analysis of people aged 70 years and older [17]. based on the estimated GFR values (expressed in ml/ The present study compares different population-based min/1.73m ) as follows: G1 (≥90), G2 (60 < 90), G3a (45 prevalence estimates of decreased kidney function < 60) and G3b-G5 (<45). We estimated the prevalence among adults in Germany using six different GFR esti- and corresponding 95% confidence intervals (95% CI) of mating equations. a decreased kidney function, as defined by an eGFR <60 ml/min/1.73m , and the prevalence of each GFR Methods category. Furthermore, we calculated the population Study population and design based estimate of the number of persons with decreased The German Health Interview and Examination Survey kidney function in Germany using the population census for Adults (“Studie zur Gesundheit Erwachsener in figures of the Federal Republic of Germany in 2011. Deutschland”, DEGS) is part of the health monitoring Population data were provided by the Federal Bureau of system at the Robert Koch-Institute (RKI). The concept Statistic. and design of DEGS are described in detail elsewhere We evaluated the agreement between the six equations [18–20]. The first wave (DEGS1) was conducted from used for estimating GFR according to the approach pro- 2008 to 2011 and included interviews, examinations and posed by Bland and Altman [25]. Bland-Altman plots Trocchi et al. BMC Nephrology (2017) 18:341 Page 3 of 10 Fig. 1 Equations used to estimate GFR. MDRD: Modification of Diet in Renal Disease study equation; CKD-EPIcre: Chronic Kidney Disease Epidemiology Collaboration creatinine equation; CKD-EPIcys: Chronic Kidney Disease Epidemiology Collaboration cystatin C equation; CKD-EPIcrecys: Chronic Kidney Disease Epidemiology Collaboration creatinine and cystatin C equation; LM: Lund-Malmö equation; FAScre: Full Age Spectrum creatinine equation; Scr: serum creatinine; Scys: serum cystatin C; min: minimum; max: maximum display for each person the difference between two mea- renal disease in our sample (n = 1) was similar to the ex- surements against their mean. To compare the equations pected number of subjects with end stage renal disease with each other, the absolute mean difference (md) be- (n = 7) as derived from a recent report [26]. To take into tween eGFR was used as a measure of the magnitude of account both the weighting and the correlation of the the systematic difference (bias) between the equations. participants within a community, confidence intervals In addition, as a measure of relative change of eGFR, we were determined using the survey procedures in SAS® calculated the mean percent difference between the (SAS Inc., Cary, NC, USA), Version 9.4. measurements. The limits of the agreement between the equations were defined as md ±1.96 standard deviation Results of the differences (SD) and were used as a measure of Table 1 shows details about the main characteristics of the variability of the bias. These values represented the the 7001 participants of the DEGS1 study from whom range within which 95% of the differences were included the GFR was estimated using different eqs. (3364 men (95% CI agreement). Furthermore, for each of the ten and 3637 women). The median age was 46.9 years (men: pairwise comparisons, the analysis of the md was strati- 46.4, women: 47.5). The prevalence estimates of persons fied by the GFR category based on the mean value of with medical history of arterial hypertension and dia- eGFR (G1: ≥90, G2: 60 < 90, G3a: 45 < 60, G3b-G5: <45). betes mellitus were 31.5% and 6.6% respectively. Table 2 Prevalence estimates were weighted by a factor that displays the estimated prevalence of participants with corrects sample deviations from population structure (as decreased kidney function, defined as eGFR <60 ml/ of 31 Dec. 2010) with regard to age, sex, region and min/1.73m , stratified by sex and age group. Overall, the nationality, type of community and education. When prevalence of participants aged 18–79 years with de- calculating the weighting factor for previous participants creased kidney function differed considerably depending of GNHIES98, the probability of repeated participation, on the equation used and was as follows: 2.1% (CKD- based on a multivariable logistic model, was taken into EPIcys), 2.3% (CKD-EPIcrecys), 3.8% (CKD-EPIcre), account. A non-response analysis and a comparison of 5.0% (MDRD), 6.0% (LM) and 6.9% (FAScre). The preva- selected indicators with data from official statistics indi- lence ranged from 1.6% to 5.6% among men and from cate a high level of sample representativeness for the 2.6% to 8.2% among women using the CKD-EPIcys resident population of Germany aged 18–79 years [18]. equation and the FAScre equation respectively. What- In addition, the observed number of subjects end stage ever equation was used, the estimated prevalence was Trocchi et al. BMC Nephrology (2017) 18:341 Page 4 of 10 Table 1 Characteristics of 7001 adults aged 18–79 in Germany 2008–2011 (DEGS1) Characteristic Overall Men Women Sex: N, % 7001 100 3364 49.9 3637 50.1 Age (Years): median (P10, P90) 46.9 (23.6, 70.4) 46.4 (23.3, 69.8) 47.5 (23.9, 70.9) BMI (Kg/m ): median (P10, P90) 26.2 (21.0, 33.4) 26.7 (22.1, 32.9) 25.4 (20.4, 33.8) Serum creatinine (mg/dl): median (P10, P90) 0.82 (0.67, 1.06) 0.92 (0.76, 1.13) 0.75 (0.63, 0.90) Serum cystatin C (mg/l): median (P10, P90) 0.70 (0.57, 0.90) 0.73 (0.61, 0.90) 0.67 (0.55, 0.89) Medical history Hypertension: N, % 2585 31.5 1349 33.4 1236 29.7 Diabetes mellitus: N, % 539 6.6 305 7.0 234 6.2 P10 10th percentile, P90 90th percentile, BMI body mass index higher for women than for men and increased with age persons aged 18–79 years with decreased kidney func- in both sexes: among participants aged <60 years the tion in Germany varied from 1.41 m using the CKD- prevalence varied from 0.2% to 1.2% and increased EPIcys equation to 4.58 m using the FAScre equation. among participants aged 70 < 80 years up to 11.4% using Assuming that the prevalence of persons aged ≥80 with CKD-EPIcys and up to 38.4% using FAScre. The mean decreased kidney function equals that of the study par- eGFR varied from 83.7 using LM to 111.4 using CKD- ticipants aged 75 < 80 years, the total number of persons EPIcys. with decreased kidney function ranged from 2.15 m Based on the age specific prevalence estimates and the using the CKD-EPIcys equation to 6.78 m using the German population in 2011, the estimated number of FAScre equation (Table 3). Table 2 eGFR and estimated prevalence of decreased kidney function (eGFR <60 ml/min/1.73m ) among 7001 adults aged 18–79 in Germany 2008–2011 (DEGS1) according to the equation used eGFR: Mean (SD, CV) Prevalence (95% CI) Overall <60 years 60 < 69 years 70 < 79 years Overall N = 7001 N = 4538 N = 1376 N = 1087 MDRD 88.4 (59.3, 0.67) 5.0 (4.3–5.7) 1.2 (0.8–1.5) 10.9 (8.8–13.1) 20.6 (17.3–23.9) CKD-EPIcre 95.2 (40.9, 0.43) 3.8 (3.3–4.4) 0.4 (0.2–0.6) 7.8 (5.9–9.6) 19.2 (16.1–22.4) CKD-EPIcys 111.4 (34.3, 0.31) 2.1 (1.7–2.5) 0.2 (0.1–0.3) 3.8 (2.3–5.4) 11.4 (8.9–13.9) CKD-EPIcrecys 105.1 (34.5, 0.33) 2.3 (1.9–2.7) 0.2 (0.1–0.3) 3.8 (2.2–5.3) 12.8 (10.4–15.3) LM 83.7 (25.3, 0.43) 6.0 (5.3–6.7) 0.5 (0.3–0.8) 11.7 (9.5–13.8) 31.4 (27.6–35.2) FAScre 91.7 (54.8, 0.60) 6.9 (6.1–7.6) 0.5 (0.3–0.7) 11.9 (9.9–14.0) 38.4 (34.2–42.5) Men N = 3364 N = 2150 N = 667 N = 547 MDRD 91.4 (45.7, 0.50) 4.1 (3.3–4.8) 0.9 (0.5–1.3) 9.4 (6.4–12.4) 18.3 (14.4–22.1) CKD-EPIcre 96.4 (32.2, 0.34) 3.6 (2.9–4.4) 0.5 (.0.2–0.8) 7.7 (4.9–10.4) 19.0 (14.9–23.1) CKD-EPIcys 113.5 (28.1, 0.25) 1.6 (1.1–2.0) 0.2 (0.0–0.4) 2.8 (0.7–4.8) 8.6 (5.8–11.4) CKD-EPIcrecys 106.8 (27.6, 0.26) 1.7 (1.3–2.2) 0.3 (0.1–0.5) 2.3 (0.3–4.3) 10.5 (7.6–13.4) LM 84.0 (28.1, 0.33) 5.6 (4.8–6.5) 0.5 (0.3–0.8) 12.3 (9.1–15.5) 30.5 (25.6–35.5) FAScre 94.2 (43.8, 0.46) 5.6 (4.7–6.4) 0.4 (0.1–0.6) 11.3 (8.2–14.4) 32.1 (27.0–37.1) Women N = 3637 N = 2388 N = 709 N = 540 MDRD 85.5 (43.5, 0.51) 5.9 (4.9–6.9) 1.5 (0.9–2.1) 12.4 (9.3–15.5) 22.6 (17.9–27.3) CKD-EPIcre 94.0 (32.6, 0.35) 4.0 (3.2–4.8) 0.3 (0.1–0.6) 7.8 (5.3–10.4) 19.4 (14.9–24.0) CKD-EPIcys 109.3 (29.6, 0.27) 2.6 (2.0–3.3) 0.1 (0.0–0.2) 4.9 (2.4–7.4) 13.8 (9.7–17.8) CKD-EPIcrecys 103.4 (30.0, 0.29) 2.8 (2.2–3.4) 0.1 (0.0–0.1) 5.2 (2.8–7.6) 14.8 (10.9–18.7) LM 83.4 (27.4, 0.33) 6.4 (5.4–7.3) 0.5 (0.2–0.9) 11.1 (8.2–14.0) 32.2 (27.2–37.1) FAScre 89.3 (40.8, 0.46) 8.2 (7.1–9.2) 0.5 (0.2–0.9) 12.6 (9.7–15.5) 43.6 (38.3–49.0) SD standard deviation, CV coefficient of variation, CI confidence interval, MDRD Modification of Diet in Renal Disease study equation, CKD-EPIcre Chronic Kidney Disease Epidemiology Collaboration creatinine equation, CKD-EPIcys Chronic Kidney Disease Epidemiology Collaboration cystatin C equation, CKD-EPIcrecys Chronic Kidney Disease Epidemiology Collaboration creatinine and cystatin C equation, LM Lund-Malmö equation, FAScre Full Age Spectrum creatinine equation Trocchi et al. BMC Nephrology (2017) 18:341 Page 5 of 10 Table 3 Estimated numbers (millions) of adults with decreased kidney function (eGFR <60 ml/min/1.73m ) in Germany 2011 according to the equation used Age (years) MDRD CKD-EPIcre CKD-EPIcys CKD-EPIcrecys LM FAScre <50 0.179 0.039 0.036 0.022 0.041 0.016 50 < 60 0.373 0.162 0.037 0.058 0.214 0.199 60 < 70 0.944 0.665 0.324 0.320 0.999 1.018 70 < 80 1.772 1.672 1.015 1.119 2.773 3.349 ≥80 1.067 1.070 0.742 0.737 1.930 2.202 Overall 4.335 3.608 2.154 2.256 5.958 6.784 MDRD Modification of Diet in Renal Disease study equation, CKD-EPIcre Chronic Kidney Disease Epidemiology Collaboration creatinine equation, CKD-EPIcys Chronic Kidney Disease Epidemiology Collaboration cystatin C equation, CKD-EPIcrecys Chronic Kidney Disease Epidemiology Collaboration creatinine and cystatin C equation, LM Lund-Malmö equation, FAScre: Full Age Spectrum creatinine equation By use of the estimated prevalence of study participants aged 70 < 80 years Figure 2 presents the prevalence estimates of the GFR Table 4 shows that great absolute differences between categories according to the equation used. Using any eGFR were calculated comparing CKD-EPIcys with LM CKD-EPI equations, the large majority of the partici- (md = 27.6), CKD-EPIcys with MDRD (md = 22.9) and pants were classified as G1. In particular, according to CKD-EPIcrecys with LM (md = 21.4). In contrast, small CKD-EPIcys, almost 9 out of 10 participants were classi- differences were calculated comparing FAScre with fied in this category. Using MDRD or LM the highest MDRD (md = 3.3) and CKD-EPIcre (md = 3.5), MDRD prevalence was estimated for the participants with eGFR with LM (md = 4.7), CKD-EPIcys with CKD-EPIcrecys 60 < 90 ml/min/1.73m (category: G2). The prevalence of (md = 6.2) and CKD-EPIcre with MDRD (md = 6.8). participants with eGFR <45 ml/min/1.73m (category Overall, the absolute md between eGFR were very small G3b-G5) varied from 0.7% using CKD-EPIcys to 1.4% for small values of eGFR and increased considerably using FAScre. with increasing eGFR values. In particular, among Fig. 2 Prevalence estimates of GFR categories among 7001 adults aged 18–79 in Germany 2008–2011 (DEGS1) according to the equation used. MDRD: Modification of Diet in Renal Disease study equation; CKD-EPIcre: Chronic Kidney Disease Epidemiology Collaboration creatinine equation; CKD-EPIcys: Chronic Kidney Disease Epidemiology Collaboration cystatin C equation; CKD-EPIcrecys: Chronic Kidney Disease Epidemiology Collaboration creatinine and cystatin C equation; LM: Lund-Malmö equation; FAScre: Full Age Spectrum creatinine equation Trocchi et al. BMC Nephrology (2017) 18:341 Page 6 of 10 Table 4 Absolute mean differences between GFR estimated by the different equations used among 7001 adults aged 18–79 in Germany 2008–2011 (DEGS1) according to GFR category Overall GFR category G1 G2 G3a G3b-G5 MDRD vs. CKD-EPIcrecys 16.7 (15.9%) 17.5 (15.1%) 16.1 (18.3%) 6.4 (11.1%) 2.2 (6.0%) MDRD vs. CKD-EPIcre 6.8 (7.1%) 7.1 (6.6%) 6.9 (8.4%) 1.7 (3.1%) 0.3 (0.8%) MDRD vs. CKD-EPIcys 22.9 (20.6%) 24.3 (20.4%) 20.0 (22.2%) 4.6 (8.2%) 3.5 (9.1%) MDRD vs. LM 4.7 (5.6%) 9.6 (9.9%) 1.5 (1.9%) 1.4 (2.6%) 2.6 (7.7%) MDRD vs. FAScre 3.3 (3.6%) 5.1 (4.7%) 2.0 (2.5%) 2.9 (5.5%) 0.1 (0.1%) CKD-EPIcrecys vs. CKD-EPIcre 9.9 (10.4%) 10.1 (9.8%) 9.8 (13.2%) 5.4 (10.6%) 1.9 (5.4%) CKD-EPIcrecys vs. CKD-EPIcys 6.2 (5.6%) 6.6 (5.7%) 4.3 (5.3%) 1.1 (2.0%) 0.8 (2.3%) CKD-EPIcrecys vs. LM 21.4 (25.6%) 24.0 (26.1%) 17.6 (24.8%) 8.3 (16.6%) 4.8 (14.9%) CKD-EPIcrecys vs. FAScre 13.4 (14.6%) 12.6 (12.3%) 16.2 (22.8%) 9.4 (19.0%) 2.0 (5.6%) CKD-EPIcre vs. CKD-EPIcys 16.1 (14.5%) 16.8 (14.2%) 14.6 (16.8%) 3.5 (6.2%) 3.4 (9.0%) CKD-EPIcre vs. LM 11.5 (13.7%) 15.0 (15.8%) 8.2 (11.0%) 3.3 (6.3%) 2.9 (8.9%) CKD-EPIcre vs. FAScre 3.5 (3.8%) 2.2 (2.1%) 5.6 (7.4%) 4.6 (8.9%) 0.4 (1.1%) CKD-EPIcys vs. LM 27.6 (33.0%) 30.1 (33.5%) 22.1 (32.2%) 7.0 (13.9%) 6.1 (18.7%) CKD-EPIcys vs. FAScre 19.7 (21.4%) 19.4 (19.4%) 22.3 (32.7%) 8.0 (16.0%) 3.0 (8.7%) LM vs. FAScre 8.0 (8.7%) 13.5 (12.4%) 3.5 (4.4%) 1.2 (2.3%) 2.1 (5.7%) Relative changes (%) of the estimated GFR were calculated as ([first value] – [second value]) / [second value]) MDRD Modification of Diet in Renal Disease study equation, CKD-EPIcre Chronic Kidney Disease Epidemiology Collaboration creatinine equation, CKD-EPIcys Chronic Kidney Disease Epidemiology Collaboration cystatin C equation, CKD-EPIcrecys Chronic Kidney Disease Epidemiology Collaboration creatinine and cystatin C equation, LM Lund-Malmö equation, FAScre Full Age Spectrum creatinine equation a 2 GFR categories are defined according to the mean value of GFR estimated by the equations being compared (expressed in ml/min/1.73m ) as follows: G1: ≥90, G2: 60 < 90, G3a: 45 < 60, G3b-G5: <45 participants with mean eGFR <45 ml/min/1.73m (G3b- of the estimated number of persons with kidney disease G5), the md ranged from 0.1 (MDRD vs. FAScre) to in Germany. 6.1 ml/min/1.73m (CKD-EPIcys vs. LM), while among Prevalence estimates of decreased kidney function vary participants with mean eGFR >90 ml/min/1.73m (G1) substantially both within and between countries and the md ranged from 2.2 (CKD-EPIcre vs. FAScre) to many potential factors leading to these variations have 30.1 ml/min/1.73m (CKD-EPIcys vs. LM). Very good been discussed [27–29]. Our estimated prevalence is levels of agreement between the equations for small lower than those from the SHIP-1 study and the KORA values of eGFR were shown also from the Bland-Altman F4 study, which reported a prevalence of decreased kid- plots that depict the agreement over the whole range of ney function in Northeast and Southern Germany of eGFR values (Figs. 3 and 4, and Additional file 1: Figure 5.9% and 3.1% respectively using CKD-EPIcrecys (vs. S1). The greatest variability of the differences between 2.3% in DEGS1) [16]. If compared with DEGS1, the me- eGFR, estimated as the range between the limits of the dian age of participants as well as the prevalence of agreement, was observed comparing CKD-EPIcys with hypertension and diabetes mellitus in these studies was MDRD (95% CI: -97.6, 143.5). considerably higher, especially for SHIP-1. Therefore, the observed differences are mostly due to differences in age Discussion and in prevalence of risk factors among the study popu- This study shows that prevalence estimates of decreased lations. Prevalence estimates in our study were also kidney function (eGFR <60 ml/min/1.73m ) among lower than those from the US population based on the adults varies considerably depending on the equation National Health and Nutrition Examination Survey used for estimating GFR. Prevalence estimates among of (NHANES), which reported a prevalence estimate of 8% persons aged 18–79 in Germany 2008–2011 (DEGS1) using MDRD [2]. Some reasons for these differences varied from 2.1% using CKD-EPIcys to 6.9% using FAS- have been discussed in our previous publication, includ- cre and the overall number of persons with decreased ing heterogeneity in age distribution and ethnic charac- kidney function ranged accordingly from 2.15 m (CKD- teristics of the study populations [15]. EPIcys) to 6.78 m (FAScre). From a public health stand- As the GFR estimating equations include the same point, the choice of the equation produces a wide range demographic variables, such as age and sex, the Trocchi et al. BMC Nephrology (2017) 18:341 Page 7 of 10 Fig. 3 Bland-Altman plots for comparison between equations that include cystatin C and equations based on creatinine alone among 7001 adults aged 18–79 in Germany 2008–2011 (DEGS1). MDRD: Modification of Diet in Renal Disease study equation; CKD-EPIcre: Chronic Kidney Disease Epidemiology Collaboration creatinine equation; CKD-EPIcys: Chronic Kidney Disease Epidemiology Collaboration cystatin C equation; CKD-EPIcrecys: Chronic Kidney Disease Epidemiology Collaboration creatinine and cystatin C equation; LM: Lund-Malmö equation. Solid, horizontal lines represent the mean difference between the eGFR. Dashed, horizontal lines represent the limit of agreement between the equations. Solid, vertical lines represent the eGFR cut-off value for a decreased kidney function (60 ml/min/1.73m ) observed differences between prevalence estimates could using the equation that includes cystatin C alone as la- be mainly due to the fact that some equations use serum boratory parameter (CKD-EPIcys). These results are in cystatin C (CKD-EPIcys, CKD-EPIcrecys), while other line with published findings [30] and may reflect that equations use serum creatinine alone as biomarker there are less non-renal factors influencing cystatin C (MDRD, CKD-EPIcre, LM and FAScre). In particular, plasma levels than there are for creatinine plasma levels. the prevalence of participants with decreased kidney Higher prevalence estimates using equations with cre- function estimated by those equations that include atinine alone were found also by the Berlin Initiative serum cystatin C was considerably lower than the preva- Study (BIS) for people aged 70 years and older [17]. In lence estimated by those equations based on serum cre- contrast to our study, data based on NHANES showed atinine alone. The lowest prevalence was estimated that equations with creatinine alone yielded lower Trocchi et al. BMC Nephrology (2017) 18:341 Page 8 of 10 Fig. 4 Bland-Altman plots for comparison between equations both based on cystatin C and comparison between equations both based on creatinine alone among 7001 adults aged 18–79 in Germany 2008–2011 (DEGS1). MDRD: Modification of Diet in Renal Disease study equation; CKD- EPIcre: Chronic Kidney Disease Epidemiology Collaboration creatinine equation; CKD-EPIcys: Chronic Kidney Disease Epidemiology Collaboration cystatin C equation; CKD-EPIcrecys: Chronic Kidney Disease Epidemiology Collaboration creatinine and cystatin C equation; LM: Lund-Malmö equation. Solid, horizontal lines represent the mean difference between the eGFR. Dashed, horizontal lines represent the limit of agreement between the equations. Solid, vertical lines represent the eGFR cut-off value of a decreased kidney function (60 ml/min/1.73m ) prevalence estimates if compared with equations that in- added costs to the health care system. Actually, the cluded cystatin C [31]. Interestingly, in agreement with costs of cystatin C tests vary from about 3 to 20 DEGS1, NHANES reported higher mean values of eGFR times those of creatinine tests [32]. It is well known for equations with cystatin C if compared with mean that any estimating equation performs better in those eGFR calculated by equations with creatinine alone. populations which are alike the population in which Our data point out that the seemingly low level of the equation was developed. For example, the CKD- imprecision of the creatinine based calculations may EPIcre equation was validated in a population with a translate into quite relevant differences when using majority of healthy people and therefore this equation the equations for epidemiological questions and sup- provides accurate estimates at higher ranges of eGFR. port the suggestion to use the GFR values estimated In contrast, as the MDRD equation was developed in by equations with cystatin C as confirmatory test for patients with CKD, this equation performs better in people with decreased kidney function as estimated populations with lower eGFR. Finally, the BIS2 equa- by equations with creatinine only [4]. tion was explicitly designed to accurately estimate Giventhe largedifference between theprevalence GFR in persons aged 70 years or older and should be estimates yielded by the different GFR estimating therefore used in older populations. Therefore, in equations, the choice for the equation for assessing order to minimize errors in GFR estimations and to GFR can have a great impact on the assessment of reduce the risk of misclassification the equation public health implications, e. g. projections of disease should be used for which the development population burden or medical resources in relation to CKD. matches best with the population of interest. In our Individually, misclassification of patients as having study, great differences between eGFR were calculated chronic kidney disease can result in unnecessary diag- by comparing equations that include cystatin C nostic and therapeutic interventions with consequent (CKD-EPIcrecys, CKD-EPIcys) with equations based Trocchi et al. BMC Nephrology (2017) 18:341 Page 9 of 10 on creatinine alone (MDRD, CKD-EPIcre, LM and Additional file FAScre). However, the analysis of the agreement be- Additional file 1: Figure S1. Bland-Altman plots for comparison tween the equations stratified by the mean values of between Full Age Spectrum creatinine equation (FAScre) and the other eGFR shows that the absolute and relative change equations used to estimate GFR among 7001 adults aged 18–79 in (percent change) of eGFR was larger among GFR cat- Germany 2008–2011 (DEGS1). MDRD: Modification of Diet in Renal Disease study equation; CKD-EPIcre: Chronic Kidney Disease Epidemiology egories G1 and G2 than G3b-G5. In particular, the Collaboration creatinine equation; CKD-EPIcys: Chronic Kidney Disease systematic differences between eGFR among partici- Epidemiology Collaboration cystatin C equation; CKD-EPIcrecys: Chronic pants classified in the category G3b-G5 can be easily Kidney Disease Epidemiology Collaboration creatinine and cystatin C equation; LM: Lund-Malmö equation; FAScre: Full Age Spectrum considered clinically not relevant. The Bland-Altman creatinine equation. Solid, horizontal lines represent the mean plots showed a similar distribution pattern, with good difference between the eGFR. Dashed, horizontal lines represent the agreement between the estimating equations for low limit of agreement between the equations. Solid, vertical lines represent the eGFR cut-off value of a decreased kidney function values of eGFR and increasing systematic differences (60 ml/min/1.73m ). (PDF 422 kb) with increasing eGFR. These results are consistent with those observed in other studies [33, 34] and Abbreviations suggest that the different estimating equations may be CI: Confidence interval; CKD: Chronic Kidney Disease; CKD-EPIcre: Chronic used interchangeably among persons with moderately Kidney Disease Epidemiology Collaboration creatinine equation; CKD- to severely decreased kidney function (eGFR: <45 ml/ EPIcrecys: Chronic Kidney Disease Epidemiology Collaboration creatinine and cystatin C equation; CKD-EPIcys: Chronic Kidney Disease Epidemiology min/1.73m ). Furthermore, high variability of the Collaboration cystatin C eq.; CV: Coefficient of variation; DEGS: German differences, estimated by the limits of the agreement Health Interview and Examination Survey for Adults; DEGS1: German Health between the equations, was observed by comparing Interview and Examination Survey for Adults, first wave; eGFR: estimated glomerular filtration rate; FAScre: Full Age Spectrum creatinine equation; CKD-EPIcys with equations based on creatinine alone. GFR: Glomerular filtration rate; GNHIES98: German National Health Interview This study has some limitations: First, as we did not and Examination Survey 1998; Hg: Mercury; IDMS: Isotope Dilution Mass measure the GFR by a gold standard, we could not Spectrometry; LM: Lund-Malmö equation; max: Maximum; md: Mean difference; MDRD: Modification of Diet in Renal Disease study equation; determine which equation provides the most valid min: Minimum; RKI: Robert Koch-Institute; Scr: Serum creatinine; Scys: Serum prevalence estimates of decreased kidney function for cystatin C; SD: Standard deviation the German population. Second, information on place of residence of the study participants was not available and Acknowledgements we could not therefore evaluate regional variability in The authors thank Angelika Schaffrath Rosario of the Robert Koch-Institute, Berlin, for helpful comments on the statistical analyses of the data. prevalence estimates. Third, as the first wave of the DEGS study was conducted from 2008 to 2011, serum Availability of data and materials cystatin C concentration was measured using a not stan- Any additional information required by the reader can be obtained from dardized assay which complicates the comparison with authors upon reasonable request. studies that used a standardized assay for cystatin C. A further limitation of our and other cross-sectional stud- Authors’ contributions PT drafted the manuscript and carried out the statistical analysis, MG ies is the lack of a second GFR estimation after 3 months and SM made substantial contributions to the analysis and interpretation which most likely results in a false positive prevalence of of the data and revised the manuscript, CSN collected the data, kidney dysfunction. provided intellectual content of critical importance and revised the manuscript, AS was responsible for the design of the study, data analysis and interpretation of the findings and revised the manuscript. All Conclusions authors read and approved the final manuscript. Our study illustrates the importance of the choice of the GFR estimating equation from an epidemiological Ethics approval and consent to participate All participants provided written informed consent prior to the interview and point of view. Prevalence estimates of decreased kidney examinations. All procedures performed in DEGS1 were in accordance with function in Germany are highly related to the equation the 1964 Helsinki declaration and its later amendments. The study protocol used. In particular, the equations that include serum was approved by the Charité-Universitätsmedizin. Berlin ethics committee in September 2008 (No. EA2/047/08). cystatin C provide lower prevalence estimates if compared with those based on serum creatinine alone. Consent for publication However, the analysis of the systematic differences Not applicable between the eGFR suggests that the equations could be used interchangeably among persons with pronounced Competing interests decreased kidney function. Additional longitudinal The authors declare that they have no competing interests. epidemiological studies are needed to investigate which of the available equations are most useful for Publisher’sNote prediction of CKD and associated complications at Springer Nature remains neutral with regard to jurisdictional claims in the population level. published maps and institutional affiliations. Trocchi et al. BMC Nephrology (2017) 18:341 Page 10 of 10 Author details design, objectives and implementation of the first data collection wave. Center of Clinical Epidemiology, c/o Institute of Medical Informatics, BMC Public Health. 2012;12:730. Biometry and Epidemiology (IMIBE), University Hospital Essen, Hufelandstr. 21. Heidemann C, Du Y, Schubert I, Rathmann W, Scheidt-Nave C. Prävalenz 55, 45147 Essen, Germany. Department of Internal Medicine II, Medical und zeitliche Entwicklung des bekannten Diabetes mellitus. Ergebnisse der Faculty of the Martin-Luther-University Halle-Wittenberg, Ernst-Grube-Str. 40, Studie zur Gesundheit Erwachsener in Deutschland (DEGS1). 06120 Halle (Saale), Germany. Department of Epidemiology and Health Bundesgesundheitsbl. 2013;56:668–77. Monitoring, Robert Koch-Institute, General-Pape-Str. 62-66, 12101 Berlin, 22. Neuhauser H, Thamm M, Eliassen MS. Blutdruck in Deutschland 2008-2011. Germany. Department of Epidemiology, School of Public Health, Boston Ergebnisse der Studie zur Gesundheit Erwachsener in Deutschland (DEGS1). University, 715 Albany St, Boston, MA 02118, USA. Bundesgesundheitsbl. 2013;56:795–801. 23. Levey AS, Coresh J, Greene T, Marsh J, Stevens LA, Kusek JW, et al. Received: 8 June 2017 Accepted: 10 November 2017 Expressing the modification of diet in renal disease study equation for estimating glomerular filtration rate with standardized serum creatinine values. Clin Chem. 2007;53:766–72. 24. Pottel H, Hoste L, Dubourg L, Ebert N, Schaeffner E, Eriksen BO, et al. An References estimated glomerular filtration rate equation for the full age spectrum. 1. Kidney Disease. Improving global outcomes (KDIGO) CKD work group: Nephrol Dial Transplant. 2016;31:798–806. KDIGO 2012 clinical practice guideline for the evaluation and Management 25. Bland JM, Altman DG. Statistical methods for assessing agreement between of Chronic Kidney Disease. Kidney. Int Suppl. 2013;3:1–150. two methods of clinical measurement. Lancet. 1986;1:307–10. 2. Coresh J, Selvin E, Stevens LA, Manzi J, Kusek JW, Eggers P, et al. Prevalence 26. Medical Netcare GmbH. Jahresbericht Datenanalyse Dialyse für den of chronic kidney disease in the United States. JAMA. 2007;298:2038–47. Gemeinsamen Bundesausschuss. In: Berichtsjahr; 2013. http://www. 3. Stevens LA, Coresh J, Greene T, Levey AS. Assessing kidney function - medical-netcare.de/qsd.php. Accessed 6 Nov 2014. measured and estimated glomerular filration rate. N Engl J Med. 2006;354: 27. Glassock RJ, Warnock DG, Delanaye P. The global burden of chronic kidney 2473–83. disease: estimates, variability and pitfalls. Nat Rev Nephrol. 2017;13:104–14. 4. Inker LA, Schmid CA, Tighiouart H, Eckfeldt JH, Feldman HI, Greene T, et al. 28. Brück K, Stel VS, Gambaro G, Hallan S, Völzke H, Ärnlöv J, et al. CKD Estimating glomerular filtration rate from serum creatinine and cystatin C. prevalence varies across the European general population. J Am Soc N Engl J Med. 2012;367:20–9. Nephrol. 2016;27:2135–47. 5. Levey AS, Coresh J. Chronic kidney disease. Lancet. 2012;379:165–80. 29. McCullough K, Sharma P, Ali T, Khan I, Smith WCS, MacLeod A, et al. 6. Coresh J, Turin TC, Matsushita K, Sang Y, Ballew SH, Appel LJ, et al. Decline Measuring the population burden of chronic kidney disease: a systematic in estimated glomerular filtration rate and subsequent risk of end-stage literature review of the estimated prevalence of impaired kidney function. renal disease and mortality. JAMA. 2014;311:2518–31. Nephrol Dial Transplant. 2012;27:1812–21. 7. Go AS, Chertow GM, Fan D, McCulloch CE, Hsu CY. Chronic kidney disease 30. Anderson AH, Yang W, Hsu C, Joffe MM, Leonard MB, Xie D, et al. and the risks of death, cardiovascular events, and hospitalization. Estimating GFR among participants in the chronic renal insufficiency cohort N Engl J Med. 2004;351:1296–305. (CRIC) study. Am J Kidney Dis. 2012;60:250–61. 8. Keith DS, Nichols GA, Gullion CM, Brown JB, Smith DH. Longitudinal follow- 31. Grams ME, Juraschek SP, Selvin E, Foster MC, Inker LA, Eckfeld JH, et al. Trends up and outcomes among a population with chronic kidney disease in a in the prevalence of reduced GFR in the United States: a comparison of large managed care organization. Arch Intern Med. 2004;164:659–63. creatinine- and cystatin C-based estimates. Am J Kidney Dis. 2013;62:253–60. 9. Eckardt KU, Coresh J, Devuyst O, Johnson RJ, Köttgen A, Levey AS, et al. 32. Shlipak MG, Mattes MD, Peralta CA. Update on cystatin C: incorporation into Evolving importance of kidney disease: from subspecialty to global health clinical practice. Am J Kidney Dis. 2013;62:595–603. burden. Lancet. 2013;382:158–69. 33. Delanaye P, Cavalier E, Moranne O, Lutteri L, Krzesinski JM, Bruyere O. 10. Earley A, Miskulin D, Lamb EJ, Levey AS, Uhlig K. Estimating equations for Creatinine- or cystatin C-based equations to estimate glomerular filtration in glomerular fitration rate in the era of creatinin standardization: a systematic the general population: impact on the epidemiology of chronic kidney review. Ann Intern Med. 2012;156:785–95. disease. BMC Nephrol. 2013;14:57. 11. Levey AS, Stevens LA, Schmid CA, Zhang YL, Castro AF, Feldman HI, et al. A 34. Glaser N, Deckert A, Phiri S, Rothenbacher D, Neuhann F. Comparison of new equation to estimate glomerular filtration rate. Ann Intern Med. 2009; various equations for estimating GFR in Malawi: how to determine renal 150:604–12. function in resource limited setting? PLoS One. 2015; doi: 10.1371/journal. 12. Nyman U, Grubb A, Larsson A, Hansson LA, Flodin M, Nordin G, et al. The pone.0130453. revised Lund-Malmö GFR estimating equation outperforms MDRD and CKD- EPI across GFR, age, and BMI intervals in a large Swedish population. Clin Chem Lab Med. 2014;52:815–24. 13. Stevens LA, Coresh J, Feldman HI, Greene T, Lash JP, Nelson RG, et al. Evaluation of the modification of diet in renal disease study equation in a large diverse population. J Am Soc Nephrol. 2007;18:2749–57. 14. Meeusen JW, Rule AD, Voskoboev N, Baumann NA, Lieske JC, Cystatin C. Creatinine-based eGFR equation performance depends on patient characteristics. Clin Chem. 2015;61:1265–72. 15. Girndt M, Trocchi P, Scheidt-Nave C, Markau S, Stang A. The prevalence of renal failure. Dtsch Ärztebl Int. 2016;113:85–91. 16. Aumann N, Baumeister SE, Rettig R, Werner WLA, Döring A, Peters A, et al. Submit your next manuscript to BioMed Central Regional variation of chronic kidney disease in Germany: results from two population-based surveys. Kidney Blood Press Res. 2015;40:231–43. and we will help you at every step: 17. Ebert N, Jakob O, Gaedeke J, van der Giet M, Kuhlmann MK, Martus P, et al. • We accept pre-submission inquiries Prevalence of reduced kidney function and albuminuria in older adults: the berlin initiative study. Nephrol Dial Transplant. 2017;32:997–1005. � Our selector tool helps you to find the most relevant journal 18. Kamtsiuris P, Lange M, Hoffmann R, Schaffrath Rosario A, Dahm S, Kuhnert � We provide round the clock customer support R, et al. First wave of the German health interview and examination survey � Convenient online submission for adults (DEGS1). Sampling design, response, sample weights and representativeness. Bundesgesundheitsbl. 2013;56:620–30. � Thorough peer review 19. Kurth BM, Lange C, Kamtsiuris P, Hölling H. Health monitoring at the � Inclusion in PubMed and all major indexing services Robert-Koch-institute. Status and perspectives. Bundesgesundheitsbl � Maximum visibility for your research Gesundheitsforsch Gesundheitsschutz. 2009;52:557–70. 20. Scheidt-Nave C, Kamtsiuris P, Gößwald A, Hölling H, Lange M, Busch MA, Submit your manuscript at et al. German health interview and examination survey for adults (DEGS) - www.biomedcentral.com/submit http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png BMC Nephrology Springer Journals

Impact of the estimation equation for GFR on population-based prevalence estimates of kidney dysfunction

Download PDF
10 pages

Loading next page...
 
/lp/springer-journals/impact-of-the-estimation-equation-for-gfr-on-population-based-00BUw1ASQE

References (40)

Publisher
Springer Journals
Copyright
Copyright © 2017 by The Author(s).
Subject
Medicine & Public Health; Nephrology; Internal Medicine
eISSN
1471-2369
DOI
10.1186/s12882-017-0749-5
pmid
29183273
Publisher site
See Article on Publisher Site

Abstract

Background: Estimating equations are recommended by clinical guidelines as the preferred method for assessment of glomerular filtration rate (GFR). The aim of the study was to compare population-based prevalence estimates of decreased kidney function in Germany defined by an estimated GFR (eGFR) <60 ml/min/1.73m using different equations. Methods: The study included 7001 participants of the German Health Interview and Examination Survey for Adults 2008–2011 (DEGS1) for whom GFR was estimated using the Modification of Diet in Renal Disease study equation (MDRD), the revised Lund-Malmö equation (LM), the Full Age Spectrum creatinine equation (FAScre), the Chronic Kidney Disease Epidemiology Collaboration equations with creatinine and cystatin C (CKD-EPIcrecys), with creatinine (CKD-EPIcre) and with cystatin C (CKD-EPIcys). Bland-Altman plots were used to evaluate the agreement between the equations. Results: Prevalence estimates of decreased kidney function were: 2.1% (CKD-EPIcys), 2.3% (CKD-EPIcrecys), 3.8% (CKD-EPIcre), 5.0% (MDRD), 6.0% (LM) and 6.9% (FAScre). The systematic differences between the equations were smaller by comparing either equations that include serum cystatin C or equations that include serum creatinine alone and increased considerably by increasing eGFR. Conclusions: Prevalence estimates of decreased kidney function vary considerably according to the equation used for estimating GFR. Equations that include serum cystatin C provide lower prevalence estimates if compared with equations based on serum creatinine alone. However, the analysis of the agreement between the equations according to eGFR provides evidence that the equations may be used interchangeably among persons with pronounced decreased kidney function. The study illustrates the implications of the choice of the estimating equation in an epidemiological setting. Keywords: Epidemiology, Prevalence, Renal dysfunction, eGFR equation Background excretory kidney function in health and disease [3, 5]. In Chronic Kidney Disease (CKD) is defined by morpho- the epidemiological setting, decreased kidney function logical and functional damage to the kidney [1, 2]. may be defined by a GFR <60 ml/min/1.73m [1]. As Clinical assessment of kidney function is central to the directly measuring GFR is often cumbersome in routine routine clinical practice [3, 4] and glomerular filtration clinical practice, researchers have developed and vali- rate (GFR) is the best overall index-indicator of dated several GFR estimating equations that include demographic and clinical variables as surrogates for muscle mass and unmeasured factors that affect serum creatinine level, such as age, sex and race. Some of these equations are meanwhile recommended by clinical * Correspondence: [email protected] guidelines as the preferred method for assessment of Center of Clinical Epidemiology, c/o Institute of Medical Informatics, GFR in the routine clinical care [1]. Biometry and Epidemiology (IMIBE), University Hospital Essen, Hufelandstr. 55, 45147 Essen, Germany Full list of author information is available at the end of the article © The Author(s). 2017 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Trocchi et al. BMC Nephrology (2017) 18:341 Page 2 of 10 Decreased kidney function has been shown to be an tests. The DEGS1 study has a mixed design, which en- independent marker for major adverse outcomes of ables both cross-sectional and longitudinal analyses. For CKD, including progression to end-stage kidney failure this purpose, a random sample from local population and premature death caused by cardiovascular disease registries was drawn to supplement former participants [6–8]. Patients with decreased kidney function require from the German National Health Interview and Exam- considerable medical attention to prevent deterioration ination Survey 1998 (GNHIES98). To evaluate kidney and the development of complications. If kidney disease function, blood samples were taken from all participants progresses to end-stage kidney disease, renal replace- and serum creatinine concentration (Architect, Abbott ment therapy is an enormously resource consuming Diagnostics, Wiesbaden; IDMS traceable creatinine Assay, condition. Given its high impact on patients’ quality of kinetic Jaffe’s method) and serum cystatin C concentration life and medical resources and rising prevalence esti- (N Latex Cystatin C assay, Prospec, Siemens Healthcare, mates reported from many countries [9], CKD is in- Eschborn) were measured. Participants with diabetes mel- creasingly recognized a major public health problem and litus were identified according to self-reported medical the knowledge of its prevalence is of great importance history and verified current use of anti-diabetic drugs [21]. from both a medical and the economical standpoint. Participants with gestational diabetes were not included Several studies evaluated performance and limitations of among those with diabetes mellitus. Arterial hypertension different estimating equations for GFR against a gold was assumed if the participant reported current treatment standard of kidney function testing in a clinical setting with antihypertensive medications or elevated blood pres- [4, 10–14]. Although it is widely known to nephrologist sure (≥140 mmHg systolic or ≥90 mmHg diastolic) was that the eGFR equations perform differently in relation measured in the survey [22]. to patient characteristics, the behavior of the equations The analyses presented here refer to the sample of in unselected large population-based samples has only 7001 participants of the DEGS1 aged 18–79 years for been investigated in detail in the NHANES study to our whom estimated GFR (eGFR) was calculated using six knowledge. Therefore, a detailed assessment of the be- GFR estimating equations (creatinine was measured in havior of the equations in a European, predominantly mg/dl, cystatin C in mg/l): the isotope dilution mass Caucasian population of 7000 participants is important spectrometry traceable Modification of Diet in Renal for researchers who want to provide population-based Disease study equation (MDRD) [23], the revised Lund- prevalence estimates of kidney dysfunction. To date, lit- Malmö equation (LM) [12], the Full Age Spectrum cre- erature about the prevalence of kidney function in atinine equation (FAScre) [24], the Chronic Kidney Dis- Germany is scarce. Recently, we published population- ease Epidemiology Collaboration creatinine equation based estimates of prevalence of kidney damage in (CKD-EPIcre), the Chronic Kidney Disease Epidemi- Germany based on measures of albuminuria and the use ology Collaboration cystatin C equation (CKD-EPIcys) of an established equation for GFR estimation [15]. Fur- and the Chronic Kidney Disease Epidemiology Collabor- thermore, the Study of Health in Pomerania (SHIP-1) ation creatinine and cystatin C equation (CKD-EPI- and the Cooperative Health Research in the Region of crecys) [4]. Equations are detailed in Fig. 1. Persons with Augsburg (KORA F4) reported results about prevalence missing data on eGFR were excluded (N = 114). of decreased kidney function in Northeast and Southern Germany respectively [16]. Finally, the Berlin Initiative Statistical analysis Study assessed kidney function in Berlin in a cross sec- Participants were classified into four GFR categories tional analysis of people aged 70 years and older [17]. based on the estimated GFR values (expressed in ml/ The present study compares different population-based min/1.73m ) as follows: G1 (≥90), G2 (60 < 90), G3a (45 prevalence estimates of decreased kidney function < 60) and G3b-G5 (<45). We estimated the prevalence among adults in Germany using six different GFR esti- and corresponding 95% confidence intervals (95% CI) of mating equations. a decreased kidney function, as defined by an eGFR <60 ml/min/1.73m , and the prevalence of each GFR Methods category. Furthermore, we calculated the population Study population and design based estimate of the number of persons with decreased The German Health Interview and Examination Survey kidney function in Germany using the population census for Adults (“Studie zur Gesundheit Erwachsener in figures of the Federal Republic of Germany in 2011. Deutschland”, DEGS) is part of the health monitoring Population data were provided by the Federal Bureau of system at the Robert Koch-Institute (RKI). The concept Statistic. and design of DEGS are described in detail elsewhere We evaluated the agreement between the six equations [18–20]. The first wave (DEGS1) was conducted from used for estimating GFR according to the approach pro- 2008 to 2011 and included interviews, examinations and posed by Bland and Altman [25]. Bland-Altman plots Trocchi et al. BMC Nephrology (2017) 18:341 Page 3 of 10 Fig. 1 Equations used to estimate GFR. MDRD: Modification of Diet in Renal Disease study equation; CKD-EPIcre: Chronic Kidney Disease Epidemiology Collaboration creatinine equation; CKD-EPIcys: Chronic Kidney Disease Epidemiology Collaboration cystatin C equation; CKD-EPIcrecys: Chronic Kidney Disease Epidemiology Collaboration creatinine and cystatin C equation; LM: Lund-Malmö equation; FAScre: Full Age Spectrum creatinine equation; Scr: serum creatinine; Scys: serum cystatin C; min: minimum; max: maximum display for each person the difference between two mea- renal disease in our sample (n = 1) was similar to the ex- surements against their mean. To compare the equations pected number of subjects with end stage renal disease with each other, the absolute mean difference (md) be- (n = 7) as derived from a recent report [26]. To take into tween eGFR was used as a measure of the magnitude of account both the weighting and the correlation of the the systematic difference (bias) between the equations. participants within a community, confidence intervals In addition, as a measure of relative change of eGFR, we were determined using the survey procedures in SAS® calculated the mean percent difference between the (SAS Inc., Cary, NC, USA), Version 9.4. measurements. The limits of the agreement between the equations were defined as md ±1.96 standard deviation Results of the differences (SD) and were used as a measure of Table 1 shows details about the main characteristics of the variability of the bias. These values represented the the 7001 participants of the DEGS1 study from whom range within which 95% of the differences were included the GFR was estimated using different eqs. (3364 men (95% CI agreement). Furthermore, for each of the ten and 3637 women). The median age was 46.9 years (men: pairwise comparisons, the analysis of the md was strati- 46.4, women: 47.5). The prevalence estimates of persons fied by the GFR category based on the mean value of with medical history of arterial hypertension and dia- eGFR (G1: ≥90, G2: 60 < 90, G3a: 45 < 60, G3b-G5: <45). betes mellitus were 31.5% and 6.6% respectively. Table 2 Prevalence estimates were weighted by a factor that displays the estimated prevalence of participants with corrects sample deviations from population structure (as decreased kidney function, defined as eGFR <60 ml/ of 31 Dec. 2010) with regard to age, sex, region and min/1.73m , stratified by sex and age group. Overall, the nationality, type of community and education. When prevalence of participants aged 18–79 years with de- calculating the weighting factor for previous participants creased kidney function differed considerably depending of GNHIES98, the probability of repeated participation, on the equation used and was as follows: 2.1% (CKD- based on a multivariable logistic model, was taken into EPIcys), 2.3% (CKD-EPIcrecys), 3.8% (CKD-EPIcre), account. A non-response analysis and a comparison of 5.0% (MDRD), 6.0% (LM) and 6.9% (FAScre). The preva- selected indicators with data from official statistics indi- lence ranged from 1.6% to 5.6% among men and from cate a high level of sample representativeness for the 2.6% to 8.2% among women using the CKD-EPIcys resident population of Germany aged 18–79 years [18]. equation and the FAScre equation respectively. What- In addition, the observed number of subjects end stage ever equation was used, the estimated prevalence was Trocchi et al. BMC Nephrology (2017) 18:341 Page 4 of 10 Table 1 Characteristics of 7001 adults aged 18–79 in Germany 2008–2011 (DEGS1) Characteristic Overall Men Women Sex: N, % 7001 100 3364 49.9 3637 50.1 Age (Years): median (P10, P90) 46.9 (23.6, 70.4) 46.4 (23.3, 69.8) 47.5 (23.9, 70.9) BMI (Kg/m ): median (P10, P90) 26.2 (21.0, 33.4) 26.7 (22.1, 32.9) 25.4 (20.4, 33.8) Serum creatinine (mg/dl): median (P10, P90) 0.82 (0.67, 1.06) 0.92 (0.76, 1.13) 0.75 (0.63, 0.90) Serum cystatin C (mg/l): median (P10, P90) 0.70 (0.57, 0.90) 0.73 (0.61, 0.90) 0.67 (0.55, 0.89) Medical history Hypertension: N, % 2585 31.5 1349 33.4 1236 29.7 Diabetes mellitus: N, % 539 6.6 305 7.0 234 6.2 P10 10th percentile, P90 90th percentile, BMI body mass index higher for women than for men and increased with age persons aged 18–79 years with decreased kidney func- in both sexes: among participants aged <60 years the tion in Germany varied from 1.41 m using the CKD- prevalence varied from 0.2% to 1.2% and increased EPIcys equation to 4.58 m using the FAScre equation. among participants aged 70 < 80 years up to 11.4% using Assuming that the prevalence of persons aged ≥80 with CKD-EPIcys and up to 38.4% using FAScre. The mean decreased kidney function equals that of the study par- eGFR varied from 83.7 using LM to 111.4 using CKD- ticipants aged 75 < 80 years, the total number of persons EPIcys. with decreased kidney function ranged from 2.15 m Based on the age specific prevalence estimates and the using the CKD-EPIcys equation to 6.78 m using the German population in 2011, the estimated number of FAScre equation (Table 3). Table 2 eGFR and estimated prevalence of decreased kidney function (eGFR <60 ml/min/1.73m ) among 7001 adults aged 18–79 in Germany 2008–2011 (DEGS1) according to the equation used eGFR: Mean (SD, CV) Prevalence (95% CI) Overall <60 years 60 < 69 years 70 < 79 years Overall N = 7001 N = 4538 N = 1376 N = 1087 MDRD 88.4 (59.3, 0.67) 5.0 (4.3–5.7) 1.2 (0.8–1.5) 10.9 (8.8–13.1) 20.6 (17.3–23.9) CKD-EPIcre 95.2 (40.9, 0.43) 3.8 (3.3–4.4) 0.4 (0.2–0.6) 7.8 (5.9–9.6) 19.2 (16.1–22.4) CKD-EPIcys 111.4 (34.3, 0.31) 2.1 (1.7–2.5) 0.2 (0.1–0.3) 3.8 (2.3–5.4) 11.4 (8.9–13.9) CKD-EPIcrecys 105.1 (34.5, 0.33) 2.3 (1.9–2.7) 0.2 (0.1–0.3) 3.8 (2.2–5.3) 12.8 (10.4–15.3) LM 83.7 (25.3, 0.43) 6.0 (5.3–6.7) 0.5 (0.3–0.8) 11.7 (9.5–13.8) 31.4 (27.6–35.2) FAScre 91.7 (54.8, 0.60) 6.9 (6.1–7.6) 0.5 (0.3–0.7) 11.9 (9.9–14.0) 38.4 (34.2–42.5) Men N = 3364 N = 2150 N = 667 N = 547 MDRD 91.4 (45.7, 0.50) 4.1 (3.3–4.8) 0.9 (0.5–1.3) 9.4 (6.4–12.4) 18.3 (14.4–22.1) CKD-EPIcre 96.4 (32.2, 0.34) 3.6 (2.9–4.4) 0.5 (.0.2–0.8) 7.7 (4.9–10.4) 19.0 (14.9–23.1) CKD-EPIcys 113.5 (28.1, 0.25) 1.6 (1.1–2.0) 0.2 (0.0–0.4) 2.8 (0.7–4.8) 8.6 (5.8–11.4) CKD-EPIcrecys 106.8 (27.6, 0.26) 1.7 (1.3–2.2) 0.3 (0.1–0.5) 2.3 (0.3–4.3) 10.5 (7.6–13.4) LM 84.0 (28.1, 0.33) 5.6 (4.8–6.5) 0.5 (0.3–0.8) 12.3 (9.1–15.5) 30.5 (25.6–35.5) FAScre 94.2 (43.8, 0.46) 5.6 (4.7–6.4) 0.4 (0.1–0.6) 11.3 (8.2–14.4) 32.1 (27.0–37.1) Women N = 3637 N = 2388 N = 709 N = 540 MDRD 85.5 (43.5, 0.51) 5.9 (4.9–6.9) 1.5 (0.9–2.1) 12.4 (9.3–15.5) 22.6 (17.9–27.3) CKD-EPIcre 94.0 (32.6, 0.35) 4.0 (3.2–4.8) 0.3 (0.1–0.6) 7.8 (5.3–10.4) 19.4 (14.9–24.0) CKD-EPIcys 109.3 (29.6, 0.27) 2.6 (2.0–3.3) 0.1 (0.0–0.2) 4.9 (2.4–7.4) 13.8 (9.7–17.8) CKD-EPIcrecys 103.4 (30.0, 0.29) 2.8 (2.2–3.4) 0.1 (0.0–0.1) 5.2 (2.8–7.6) 14.8 (10.9–18.7) LM 83.4 (27.4, 0.33) 6.4 (5.4–7.3) 0.5 (0.2–0.9) 11.1 (8.2–14.0) 32.2 (27.2–37.1) FAScre 89.3 (40.8, 0.46) 8.2 (7.1–9.2) 0.5 (0.2–0.9) 12.6 (9.7–15.5) 43.6 (38.3–49.0) SD standard deviation, CV coefficient of variation, CI confidence interval, MDRD Modification of Diet in Renal Disease study equation, CKD-EPIcre Chronic Kidney Disease Epidemiology Collaboration creatinine equation, CKD-EPIcys Chronic Kidney Disease Epidemiology Collaboration cystatin C equation, CKD-EPIcrecys Chronic Kidney Disease Epidemiology Collaboration creatinine and cystatin C equation, LM Lund-Malmö equation, FAScre Full Age Spectrum creatinine equation Trocchi et al. BMC Nephrology (2017) 18:341 Page 5 of 10 Table 3 Estimated numbers (millions) of adults with decreased kidney function (eGFR <60 ml/min/1.73m ) in Germany 2011 according to the equation used Age (years) MDRD CKD-EPIcre CKD-EPIcys CKD-EPIcrecys LM FAScre <50 0.179 0.039 0.036 0.022 0.041 0.016 50 < 60 0.373 0.162 0.037 0.058 0.214 0.199 60 < 70 0.944 0.665 0.324 0.320 0.999 1.018 70 < 80 1.772 1.672 1.015 1.119 2.773 3.349 ≥80 1.067 1.070 0.742 0.737 1.930 2.202 Overall 4.335 3.608 2.154 2.256 5.958 6.784 MDRD Modification of Diet in Renal Disease study equation, CKD-EPIcre Chronic Kidney Disease Epidemiology Collaboration creatinine equation, CKD-EPIcys Chronic Kidney Disease Epidemiology Collaboration cystatin C equation, CKD-EPIcrecys Chronic Kidney Disease Epidemiology Collaboration creatinine and cystatin C equation, LM Lund-Malmö equation, FAScre: Full Age Spectrum creatinine equation By use of the estimated prevalence of study participants aged 70 < 80 years Figure 2 presents the prevalence estimates of the GFR Table 4 shows that great absolute differences between categories according to the equation used. Using any eGFR were calculated comparing CKD-EPIcys with LM CKD-EPI equations, the large majority of the partici- (md = 27.6), CKD-EPIcys with MDRD (md = 22.9) and pants were classified as G1. In particular, according to CKD-EPIcrecys with LM (md = 21.4). In contrast, small CKD-EPIcys, almost 9 out of 10 participants were classi- differences were calculated comparing FAScre with fied in this category. Using MDRD or LM the highest MDRD (md = 3.3) and CKD-EPIcre (md = 3.5), MDRD prevalence was estimated for the participants with eGFR with LM (md = 4.7), CKD-EPIcys with CKD-EPIcrecys 60 < 90 ml/min/1.73m (category: G2). The prevalence of (md = 6.2) and CKD-EPIcre with MDRD (md = 6.8). participants with eGFR <45 ml/min/1.73m (category Overall, the absolute md between eGFR were very small G3b-G5) varied from 0.7% using CKD-EPIcys to 1.4% for small values of eGFR and increased considerably using FAScre. with increasing eGFR values. In particular, among Fig. 2 Prevalence estimates of GFR categories among 7001 adults aged 18–79 in Germany 2008–2011 (DEGS1) according to the equation used. MDRD: Modification of Diet in Renal Disease study equation; CKD-EPIcre: Chronic Kidney Disease Epidemiology Collaboration creatinine equation; CKD-EPIcys: Chronic Kidney Disease Epidemiology Collaboration cystatin C equation; CKD-EPIcrecys: Chronic Kidney Disease Epidemiology Collaboration creatinine and cystatin C equation; LM: Lund-Malmö equation; FAScre: Full Age Spectrum creatinine equation Trocchi et al. BMC Nephrology (2017) 18:341 Page 6 of 10 Table 4 Absolute mean differences between GFR estimated by the different equations used among 7001 adults aged 18–79 in Germany 2008–2011 (DEGS1) according to GFR category Overall GFR category G1 G2 G3a G3b-G5 MDRD vs. CKD-EPIcrecys 16.7 (15.9%) 17.5 (15.1%) 16.1 (18.3%) 6.4 (11.1%) 2.2 (6.0%) MDRD vs. CKD-EPIcre 6.8 (7.1%) 7.1 (6.6%) 6.9 (8.4%) 1.7 (3.1%) 0.3 (0.8%) MDRD vs. CKD-EPIcys 22.9 (20.6%) 24.3 (20.4%) 20.0 (22.2%) 4.6 (8.2%) 3.5 (9.1%) MDRD vs. LM 4.7 (5.6%) 9.6 (9.9%) 1.5 (1.9%) 1.4 (2.6%) 2.6 (7.7%) MDRD vs. FAScre 3.3 (3.6%) 5.1 (4.7%) 2.0 (2.5%) 2.9 (5.5%) 0.1 (0.1%) CKD-EPIcrecys vs. CKD-EPIcre 9.9 (10.4%) 10.1 (9.8%) 9.8 (13.2%) 5.4 (10.6%) 1.9 (5.4%) CKD-EPIcrecys vs. CKD-EPIcys 6.2 (5.6%) 6.6 (5.7%) 4.3 (5.3%) 1.1 (2.0%) 0.8 (2.3%) CKD-EPIcrecys vs. LM 21.4 (25.6%) 24.0 (26.1%) 17.6 (24.8%) 8.3 (16.6%) 4.8 (14.9%) CKD-EPIcrecys vs. FAScre 13.4 (14.6%) 12.6 (12.3%) 16.2 (22.8%) 9.4 (19.0%) 2.0 (5.6%) CKD-EPIcre vs. CKD-EPIcys 16.1 (14.5%) 16.8 (14.2%) 14.6 (16.8%) 3.5 (6.2%) 3.4 (9.0%) CKD-EPIcre vs. LM 11.5 (13.7%) 15.0 (15.8%) 8.2 (11.0%) 3.3 (6.3%) 2.9 (8.9%) CKD-EPIcre vs. FAScre 3.5 (3.8%) 2.2 (2.1%) 5.6 (7.4%) 4.6 (8.9%) 0.4 (1.1%) CKD-EPIcys vs. LM 27.6 (33.0%) 30.1 (33.5%) 22.1 (32.2%) 7.0 (13.9%) 6.1 (18.7%) CKD-EPIcys vs. FAScre 19.7 (21.4%) 19.4 (19.4%) 22.3 (32.7%) 8.0 (16.0%) 3.0 (8.7%) LM vs. FAScre 8.0 (8.7%) 13.5 (12.4%) 3.5 (4.4%) 1.2 (2.3%) 2.1 (5.7%) Relative changes (%) of the estimated GFR were calculated as ([first value] – [second value]) / [second value]) MDRD Modification of Diet in Renal Disease study equation, CKD-EPIcre Chronic Kidney Disease Epidemiology Collaboration creatinine equation, CKD-EPIcys Chronic Kidney Disease Epidemiology Collaboration cystatin C equation, CKD-EPIcrecys Chronic Kidney Disease Epidemiology Collaboration creatinine and cystatin C equation, LM Lund-Malmö equation, FAScre Full Age Spectrum creatinine equation a 2 GFR categories are defined according to the mean value of GFR estimated by the equations being compared (expressed in ml/min/1.73m ) as follows: G1: ≥90, G2: 60 < 90, G3a: 45 < 60, G3b-G5: <45 participants with mean eGFR <45 ml/min/1.73m (G3b- of the estimated number of persons with kidney disease G5), the md ranged from 0.1 (MDRD vs. FAScre) to in Germany. 6.1 ml/min/1.73m (CKD-EPIcys vs. LM), while among Prevalence estimates of decreased kidney function vary participants with mean eGFR >90 ml/min/1.73m (G1) substantially both within and between countries and the md ranged from 2.2 (CKD-EPIcre vs. FAScre) to many potential factors leading to these variations have 30.1 ml/min/1.73m (CKD-EPIcys vs. LM). Very good been discussed [27–29]. Our estimated prevalence is levels of agreement between the equations for small lower than those from the SHIP-1 study and the KORA values of eGFR were shown also from the Bland-Altman F4 study, which reported a prevalence of decreased kid- plots that depict the agreement over the whole range of ney function in Northeast and Southern Germany of eGFR values (Figs. 3 and 4, and Additional file 1: Figure 5.9% and 3.1% respectively using CKD-EPIcrecys (vs. S1). The greatest variability of the differences between 2.3% in DEGS1) [16]. If compared with DEGS1, the me- eGFR, estimated as the range between the limits of the dian age of participants as well as the prevalence of agreement, was observed comparing CKD-EPIcys with hypertension and diabetes mellitus in these studies was MDRD (95% CI: -97.6, 143.5). considerably higher, especially for SHIP-1. Therefore, the observed differences are mostly due to differences in age Discussion and in prevalence of risk factors among the study popu- This study shows that prevalence estimates of decreased lations. Prevalence estimates in our study were also kidney function (eGFR <60 ml/min/1.73m ) among lower than those from the US population based on the adults varies considerably depending on the equation National Health and Nutrition Examination Survey used for estimating GFR. Prevalence estimates among of (NHANES), which reported a prevalence estimate of 8% persons aged 18–79 in Germany 2008–2011 (DEGS1) using MDRD [2]. Some reasons for these differences varied from 2.1% using CKD-EPIcys to 6.9% using FAS- have been discussed in our previous publication, includ- cre and the overall number of persons with decreased ing heterogeneity in age distribution and ethnic charac- kidney function ranged accordingly from 2.15 m (CKD- teristics of the study populations [15]. EPIcys) to 6.78 m (FAScre). From a public health stand- As the GFR estimating equations include the same point, the choice of the equation produces a wide range demographic variables, such as age and sex, the Trocchi et al. BMC Nephrology (2017) 18:341 Page 7 of 10 Fig. 3 Bland-Altman plots for comparison between equations that include cystatin C and equations based on creatinine alone among 7001 adults aged 18–79 in Germany 2008–2011 (DEGS1). MDRD: Modification of Diet in Renal Disease study equation; CKD-EPIcre: Chronic Kidney Disease Epidemiology Collaboration creatinine equation; CKD-EPIcys: Chronic Kidney Disease Epidemiology Collaboration cystatin C equation; CKD-EPIcrecys: Chronic Kidney Disease Epidemiology Collaboration creatinine and cystatin C equation; LM: Lund-Malmö equation. Solid, horizontal lines represent the mean difference between the eGFR. Dashed, horizontal lines represent the limit of agreement between the equations. Solid, vertical lines represent the eGFR cut-off value for a decreased kidney function (60 ml/min/1.73m ) observed differences between prevalence estimates could using the equation that includes cystatin C alone as la- be mainly due to the fact that some equations use serum boratory parameter (CKD-EPIcys). These results are in cystatin C (CKD-EPIcys, CKD-EPIcrecys), while other line with published findings [30] and may reflect that equations use serum creatinine alone as biomarker there are less non-renal factors influencing cystatin C (MDRD, CKD-EPIcre, LM and FAScre). In particular, plasma levels than there are for creatinine plasma levels. the prevalence of participants with decreased kidney Higher prevalence estimates using equations with cre- function estimated by those equations that include atinine alone were found also by the Berlin Initiative serum cystatin C was considerably lower than the preva- Study (BIS) for people aged 70 years and older [17]. In lence estimated by those equations based on serum cre- contrast to our study, data based on NHANES showed atinine alone. The lowest prevalence was estimated that equations with creatinine alone yielded lower Trocchi et al. BMC Nephrology (2017) 18:341 Page 8 of 10 Fig. 4 Bland-Altman plots for comparison between equations both based on cystatin C and comparison between equations both based on creatinine alone among 7001 adults aged 18–79 in Germany 2008–2011 (DEGS1). MDRD: Modification of Diet in Renal Disease study equation; CKD- EPIcre: Chronic Kidney Disease Epidemiology Collaboration creatinine equation; CKD-EPIcys: Chronic Kidney Disease Epidemiology Collaboration cystatin C equation; CKD-EPIcrecys: Chronic Kidney Disease Epidemiology Collaboration creatinine and cystatin C equation; LM: Lund-Malmö equation. Solid, horizontal lines represent the mean difference between the eGFR. Dashed, horizontal lines represent the limit of agreement between the equations. Solid, vertical lines represent the eGFR cut-off value of a decreased kidney function (60 ml/min/1.73m ) prevalence estimates if compared with equations that in- added costs to the health care system. Actually, the cluded cystatin C [31]. Interestingly, in agreement with costs of cystatin C tests vary from about 3 to 20 DEGS1, NHANES reported higher mean values of eGFR times those of creatinine tests [32]. It is well known for equations with cystatin C if compared with mean that any estimating equation performs better in those eGFR calculated by equations with creatinine alone. populations which are alike the population in which Our data point out that the seemingly low level of the equation was developed. For example, the CKD- imprecision of the creatinine based calculations may EPIcre equation was validated in a population with a translate into quite relevant differences when using majority of healthy people and therefore this equation the equations for epidemiological questions and sup- provides accurate estimates at higher ranges of eGFR. port the suggestion to use the GFR values estimated In contrast, as the MDRD equation was developed in by equations with cystatin C as confirmatory test for patients with CKD, this equation performs better in people with decreased kidney function as estimated populations with lower eGFR. Finally, the BIS2 equa- by equations with creatinine only [4]. tion was explicitly designed to accurately estimate Giventhe largedifference between theprevalence GFR in persons aged 70 years or older and should be estimates yielded by the different GFR estimating therefore used in older populations. Therefore, in equations, the choice for the equation for assessing order to minimize errors in GFR estimations and to GFR can have a great impact on the assessment of reduce the risk of misclassification the equation public health implications, e. g. projections of disease should be used for which the development population burden or medical resources in relation to CKD. matches best with the population of interest. In our Individually, misclassification of patients as having study, great differences between eGFR were calculated chronic kidney disease can result in unnecessary diag- by comparing equations that include cystatin C nostic and therapeutic interventions with consequent (CKD-EPIcrecys, CKD-EPIcys) with equations based Trocchi et al. BMC Nephrology (2017) 18:341 Page 9 of 10 on creatinine alone (MDRD, CKD-EPIcre, LM and Additional file FAScre). However, the analysis of the agreement be- Additional file 1: Figure S1. Bland-Altman plots for comparison tween the equations stratified by the mean values of between Full Age Spectrum creatinine equation (FAScre) and the other eGFR shows that the absolute and relative change equations used to estimate GFR among 7001 adults aged 18–79 in (percent change) of eGFR was larger among GFR cat- Germany 2008–2011 (DEGS1). MDRD: Modification of Diet in Renal Disease study equation; CKD-EPIcre: Chronic Kidney Disease Epidemiology egories G1 and G2 than G3b-G5. In particular, the Collaboration creatinine equation; CKD-EPIcys: Chronic Kidney Disease systematic differences between eGFR among partici- Epidemiology Collaboration cystatin C equation; CKD-EPIcrecys: Chronic pants classified in the category G3b-G5 can be easily Kidney Disease Epidemiology Collaboration creatinine and cystatin C equation; LM: Lund-Malmö equation; FAScre: Full Age Spectrum considered clinically not relevant. The Bland-Altman creatinine equation. Solid, horizontal lines represent the mean plots showed a similar distribution pattern, with good difference between the eGFR. Dashed, horizontal lines represent the agreement between the estimating equations for low limit of agreement between the equations. Solid, vertical lines represent the eGFR cut-off value of a decreased kidney function values of eGFR and increasing systematic differences (60 ml/min/1.73m ). (PDF 422 kb) with increasing eGFR. These results are consistent with those observed in other studies [33, 34] and Abbreviations suggest that the different estimating equations may be CI: Confidence interval; CKD: Chronic Kidney Disease; CKD-EPIcre: Chronic used interchangeably among persons with moderately Kidney Disease Epidemiology Collaboration creatinine equation; CKD- to severely decreased kidney function (eGFR: <45 ml/ EPIcrecys: Chronic Kidney Disease Epidemiology Collaboration creatinine and cystatin C equation; CKD-EPIcys: Chronic Kidney Disease Epidemiology min/1.73m ). Furthermore, high variability of the Collaboration cystatin C eq.; CV: Coefficient of variation; DEGS: German differences, estimated by the limits of the agreement Health Interview and Examination Survey for Adults; DEGS1: German Health between the equations, was observed by comparing Interview and Examination Survey for Adults, first wave; eGFR: estimated glomerular filtration rate; FAScre: Full Age Spectrum creatinine equation; CKD-EPIcys with equations based on creatinine alone. GFR: Glomerular filtration rate; GNHIES98: German National Health Interview This study has some limitations: First, as we did not and Examination Survey 1998; Hg: Mercury; IDMS: Isotope Dilution Mass measure the GFR by a gold standard, we could not Spectrometry; LM: Lund-Malmö equation; max: Maximum; md: Mean difference; MDRD: Modification of Diet in Renal Disease study equation; determine which equation provides the most valid min: Minimum; RKI: Robert Koch-Institute; Scr: Serum creatinine; Scys: Serum prevalence estimates of decreased kidney function for cystatin C; SD: Standard deviation the German population. Second, information on place of residence of the study participants was not available and Acknowledgements we could not therefore evaluate regional variability in The authors thank Angelika Schaffrath Rosario of the Robert Koch-Institute, Berlin, for helpful comments on the statistical analyses of the data. prevalence estimates. Third, as the first wave of the DEGS study was conducted from 2008 to 2011, serum Availability of data and materials cystatin C concentration was measured using a not stan- Any additional information required by the reader can be obtained from dardized assay which complicates the comparison with authors upon reasonable request. studies that used a standardized assay for cystatin C. A further limitation of our and other cross-sectional stud- Authors’ contributions PT drafted the manuscript and carried out the statistical analysis, MG ies is the lack of a second GFR estimation after 3 months and SM made substantial contributions to the analysis and interpretation which most likely results in a false positive prevalence of of the data and revised the manuscript, CSN collected the data, kidney dysfunction. provided intellectual content of critical importance and revised the manuscript, AS was responsible for the design of the study, data analysis and interpretation of the findings and revised the manuscript. All Conclusions authors read and approved the final manuscript. Our study illustrates the importance of the choice of the GFR estimating equation from an epidemiological Ethics approval and consent to participate All participants provided written informed consent prior to the interview and point of view. Prevalence estimates of decreased kidney examinations. All procedures performed in DEGS1 were in accordance with function in Germany are highly related to the equation the 1964 Helsinki declaration and its later amendments. The study protocol used. In particular, the equations that include serum was approved by the Charité-Universitätsmedizin. Berlin ethics committee in September 2008 (No. EA2/047/08). cystatin C provide lower prevalence estimates if compared with those based on serum creatinine alone. Consent for publication However, the analysis of the systematic differences Not applicable between the eGFR suggests that the equations could be used interchangeably among persons with pronounced Competing interests decreased kidney function. Additional longitudinal The authors declare that they have no competing interests. epidemiological studies are needed to investigate which of the available equations are most useful for Publisher’sNote prediction of CKD and associated complications at Springer Nature remains neutral with regard to jurisdictional claims in the population level. published maps and institutional affiliations. Trocchi et al. BMC Nephrology (2017) 18:341 Page 10 of 10 Author details design, objectives and implementation of the first data collection wave. Center of Clinical Epidemiology, c/o Institute of Medical Informatics, BMC Public Health. 2012;12:730. Biometry and Epidemiology (IMIBE), University Hospital Essen, Hufelandstr. 21. Heidemann C, Du Y, Schubert I, Rathmann W, Scheidt-Nave C. Prävalenz 55, 45147 Essen, Germany. Department of Internal Medicine II, Medical und zeitliche Entwicklung des bekannten Diabetes mellitus. Ergebnisse der Faculty of the Martin-Luther-University Halle-Wittenberg, Ernst-Grube-Str. 40, Studie zur Gesundheit Erwachsener in Deutschland (DEGS1). 06120 Halle (Saale), Germany. Department of Epidemiology and Health Bundesgesundheitsbl. 2013;56:668–77. Monitoring, Robert Koch-Institute, General-Pape-Str. 62-66, 12101 Berlin, 22. Neuhauser H, Thamm M, Eliassen MS. Blutdruck in Deutschland 2008-2011. Germany. Department of Epidemiology, School of Public Health, Boston Ergebnisse der Studie zur Gesundheit Erwachsener in Deutschland (DEGS1). University, 715 Albany St, Boston, MA 02118, USA. Bundesgesundheitsbl. 2013;56:795–801. 23. Levey AS, Coresh J, Greene T, Marsh J, Stevens LA, Kusek JW, et al. Received: 8 June 2017 Accepted: 10 November 2017 Expressing the modification of diet in renal disease study equation for estimating glomerular filtration rate with standardized serum creatinine values. Clin Chem. 2007;53:766–72. 24. Pottel H, Hoste L, Dubourg L, Ebert N, Schaeffner E, Eriksen BO, et al. An References estimated glomerular filtration rate equation for the full age spectrum. 1. Kidney Disease. Improving global outcomes (KDIGO) CKD work group: Nephrol Dial Transplant. 2016;31:798–806. KDIGO 2012 clinical practice guideline for the evaluation and Management 25. Bland JM, Altman DG. Statistical methods for assessing agreement between of Chronic Kidney Disease. Kidney. Int Suppl. 2013;3:1–150. two methods of clinical measurement. Lancet. 1986;1:307–10. 2. Coresh J, Selvin E, Stevens LA, Manzi J, Kusek JW, Eggers P, et al. Prevalence 26. Medical Netcare GmbH. Jahresbericht Datenanalyse Dialyse für den of chronic kidney disease in the United States. JAMA. 2007;298:2038–47. Gemeinsamen Bundesausschuss. In: Berichtsjahr; 2013. http://www. 3. Stevens LA, Coresh J, Greene T, Levey AS. Assessing kidney function - medical-netcare.de/qsd.php. Accessed 6 Nov 2014. measured and estimated glomerular filration rate. N Engl J Med. 2006;354: 27. Glassock RJ, Warnock DG, Delanaye P. The global burden of chronic kidney 2473–83. disease: estimates, variability and pitfalls. Nat Rev Nephrol. 2017;13:104–14. 4. Inker LA, Schmid CA, Tighiouart H, Eckfeldt JH, Feldman HI, Greene T, et al. 28. Brück K, Stel VS, Gambaro G, Hallan S, Völzke H, Ärnlöv J, et al. CKD Estimating glomerular filtration rate from serum creatinine and cystatin C. prevalence varies across the European general population. J Am Soc N Engl J Med. 2012;367:20–9. Nephrol. 2016;27:2135–47. 5. Levey AS, Coresh J. Chronic kidney disease. Lancet. 2012;379:165–80. 29. McCullough K, Sharma P, Ali T, Khan I, Smith WCS, MacLeod A, et al. 6. Coresh J, Turin TC, Matsushita K, Sang Y, Ballew SH, Appel LJ, et al. Decline Measuring the population burden of chronic kidney disease: a systematic in estimated glomerular filtration rate and subsequent risk of end-stage literature review of the estimated prevalence of impaired kidney function. renal disease and mortality. JAMA. 2014;311:2518–31. Nephrol Dial Transplant. 2012;27:1812–21. 7. Go AS, Chertow GM, Fan D, McCulloch CE, Hsu CY. Chronic kidney disease 30. Anderson AH, Yang W, Hsu C, Joffe MM, Leonard MB, Xie D, et al. and the risks of death, cardiovascular events, and hospitalization. Estimating GFR among participants in the chronic renal insufficiency cohort N Engl J Med. 2004;351:1296–305. (CRIC) study. Am J Kidney Dis. 2012;60:250–61. 8. Keith DS, Nichols GA, Gullion CM, Brown JB, Smith DH. Longitudinal follow- 31. Grams ME, Juraschek SP, Selvin E, Foster MC, Inker LA, Eckfeld JH, et al. Trends up and outcomes among a population with chronic kidney disease in a in the prevalence of reduced GFR in the United States: a comparison of large managed care organization. Arch Intern Med. 2004;164:659–63. creatinine- and cystatin C-based estimates. Am J Kidney Dis. 2013;62:253–60. 9. Eckardt KU, Coresh J, Devuyst O, Johnson RJ, Köttgen A, Levey AS, et al. 32. Shlipak MG, Mattes MD, Peralta CA. Update on cystatin C: incorporation into Evolving importance of kidney disease: from subspecialty to global health clinical practice. Am J Kidney Dis. 2013;62:595–603. burden. Lancet. 2013;382:158–69. 33. Delanaye P, Cavalier E, Moranne O, Lutteri L, Krzesinski JM, Bruyere O. 10. Earley A, Miskulin D, Lamb EJ, Levey AS, Uhlig K. Estimating equations for Creatinine- or cystatin C-based equations to estimate glomerular filtration in glomerular fitration rate in the era of creatinin standardization: a systematic the general population: impact on the epidemiology of chronic kidney review. Ann Intern Med. 2012;156:785–95. disease. BMC Nephrol. 2013;14:57. 11. Levey AS, Stevens LA, Schmid CA, Zhang YL, Castro AF, Feldman HI, et al. A 34. Glaser N, Deckert A, Phiri S, Rothenbacher D, Neuhann F. Comparison of new equation to estimate glomerular filtration rate. Ann Intern Med. 2009; various equations for estimating GFR in Malawi: how to determine renal 150:604–12. function in resource limited setting? PLoS One. 2015; doi: 10.1371/journal. 12. Nyman U, Grubb A, Larsson A, Hansson LA, Flodin M, Nordin G, et al. The pone.0130453. revised Lund-Malmö GFR estimating equation outperforms MDRD and CKD- EPI across GFR, age, and BMI intervals in a large Swedish population. Clin Chem Lab Med. 2014;52:815–24. 13. Stevens LA, Coresh J, Feldman HI, Greene T, Lash JP, Nelson RG, et al. Evaluation of the modification of diet in renal disease study equation in a large diverse population. J Am Soc Nephrol. 2007;18:2749–57. 14. Meeusen JW, Rule AD, Voskoboev N, Baumann NA, Lieske JC, Cystatin C. Creatinine-based eGFR equation performance depends on patient characteristics. Clin Chem. 2015;61:1265–72. 15. Girndt M, Trocchi P, Scheidt-Nave C, Markau S, Stang A. The prevalence of renal failure. Dtsch Ärztebl Int. 2016;113:85–91. 16. Aumann N, Baumeister SE, Rettig R, Werner WLA, Döring A, Peters A, et al. Submit your next manuscript to BioMed Central Regional variation of chronic kidney disease in Germany: results from two population-based surveys. Kidney Blood Press Res. 2015;40:231–43. and we will help you at every step: 17. Ebert N, Jakob O, Gaedeke J, van der Giet M, Kuhlmann MK, Martus P, et al. • We accept pre-submission inquiries Prevalence of reduced kidney function and albuminuria in older adults: the berlin initiative study. Nephrol Dial Transplant. 2017;32:997–1005. � Our selector tool helps you to find the most relevant journal 18. Kamtsiuris P, Lange M, Hoffmann R, Schaffrath Rosario A, Dahm S, Kuhnert � We provide round the clock customer support R, et al. First wave of the German health interview and examination survey � Convenient online submission for adults (DEGS1). Sampling design, response, sample weights and representativeness. Bundesgesundheitsbl. 2013;56:620–30. � Thorough peer review 19. Kurth BM, Lange C, Kamtsiuris P, Hölling H. Health monitoring at the � Inclusion in PubMed and all major indexing services Robert-Koch-institute. Status and perspectives. Bundesgesundheitsbl � Maximum visibility for your research Gesundheitsforsch Gesundheitsschutz. 2009;52:557–70. 20. Scheidt-Nave C, Kamtsiuris P, Gößwald A, Hölling H, Lange M, Busch MA, Submit your manuscript at et al. German health interview and examination survey for adults (DEGS) - www.biomedcentral.com/submit

Journal

BMC Nephrology – Springer Journals

Published: Nov 28, 2017

There are no references for this article.