KNOW-CKD (KoreaN cohort study for Outcome in patients With Chronic Kidney Disease): design and methods

The KNOW-CKD is a patient-based cohort study. Nephrologists working in nine clinical centers in the major university-affiliated hospitals and epidemiologists, pathologists and biostatisticians in a research modulating center are participating in this study. The basic protocol of the study was approved by the ethical committee of each participating center including Institutional Review Boards of Seoul National University Hospital, Severance Hospital, Kangbuk Samsung Medical Center, Seoul St. Mary’s Hospital, Gil Hospital, Eulji General Hospital, Chonnam National University Hospital and Pusan Paik Hospital. Among the nine clinical centers, six centers are located in the metropolitan city of Seoul, one in a satellite area of Seoul, Gyeonggi-do (a province), and two in the southern area of Korea (i.e., Chonnam-do and Busan). Throughout the national health care system, Korea has 44 tertiary-care general hospitals. The distribution of our participating clinical centers is parallel to the distribution of the major hospitals in Korea. The study is supervised by the CKD Advisory Committee, which comprises members from the KCDC and the Korean Society of Nephrology (KSN) (NCT01630486 at http://​www.​clinicaltrials.​gov).

Each participating center will enroll approximately 250 consecutive individuals over a 5-year period from 2011 until 2015, totaling 2,450 adult patients with CKD who provide written informed consent. The participating individuals will be monitored for approximately 10 years until death or until ESRD occurs.

The KNOW-CKD will enroll ethnic Korean patients with CKD who range in age between 20 years and 75 years. The CKD stages from 1 to 5 (predialysis), based on the eGFR, is calculated using the four-variable Modification of Diet in Renal Disease (MDRD) equation as follows:

eGFR (ml/min per 1.73 m²) = 175 × [serum Cr (mg/dl)] ^-1.154 × [age]^-0.203 × [0.742 if female] × [1.212 if black], using serum creatinine concentrations measured at a central laboratory and an assay traceable to the international reference material [12].

Excluded subjects are those who 1) are unable or unwilling to give written consent, 2) have previously received chronic dialysis or organ transplantation, 3) have heart failure (NYHA class 3 or 4) or liver cirrhosis (Child-Pugh class 2 or 3), 4) have a past or current history of malignancy, 5) are currently pregnant, or 6) have a single kidney due to trauma or kidney donation.

We defined and allocated the specific causes of the CKD into four subgroups: glomerulonephritis (GN), diabetic nephropathy (DN), hypertensive nephropathy (HTN), and polycystic kidney disease (PKD). The definition of the subgroup is defined by the pathologic diagnosis, in the event that the biopsy result is available. Otherwise, the subgroup classification depends on the clinical diagnosis. GN is defined by the presence of glomerular hematuria or albuminuria with or without an underlying systemic disease causing glomerulonephritis. The diagnosis of DN is based on albuminuria in a subject with type 2 diabetes mellitus and the presence of diabetic retinopathy. HTN is defined by the patient’s hypertension history and the absence of a systemic illness associated with renal damage. Unified ultrasound criteria [13] will be used to diagnose PKD. The other causative diseases will be categorized as ‘unclassified’.

The detailed protocol for data and sample collection is described in Table 1. The eligible subjects will be screened and evaluated at baseline for socio-demographic information, smoking, alcohol consumption, medication, and detailed personal and family medical histories. The anthropometric measurements (height, weight, and waist/hip ratios), and measurements of the resting blood pressure (BP) in the clinic and pulse pressure using an electronic sphygmomanometer will be conducted. The subjects will complete questionnaires concerning quality of life, socio-economic status, educational level, physical activity, health behaviors, and health care facility utilization. After enrollment in the study, the laboratory tests, cardiac evaluation and radiologic imaging will be performed according to the specific protocol. DNA, serum and urine samples will be initially collected. Serum and urine samples will be subsequently collected on a regular basis according to the standardized protocol. The details regarding the study visit and test schedule are shown in Table 1.

All of the data, including clinical information, laboratory results and outcome, will be entered into a web-based, electronic, case-reporting form (eCRF). The electronic data management system for the KNOW-CKD was developed by the division of data management in Seoul National University Medical Research Collaborating Center.

A renal event is defined by a >50% decrease in eGFR from the baseline values, doubling of serum creatinine, or ESRD. The estimated GFR for defining renal outcome will be based on the four-variable MDRD formula. However, the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) creatinine equation [14] or the CKD-EPI creatinine-cystatin C equation [15] will also be employed to estimate GFR.

At each visit, the subjects will be queried and evaluated for CV events, including myocardial infarction (fatal and nonfatal), coronary revascularization, stroke and new onset or aggravation of congestive heart failure. The causes of death will be verified and classified as cardiac, non-cardiac and unknown. An independent committee will adjudicate all of the primary outcomes. The subjects who withdraw from the study will be traced for this type of information with the help of the National Health Insurance System and Korea Statistical Information Service. The secondary outcomes are hospitalization, fracture, QOL, and economic status during the follow-up.

Using an EDTA tube, 3 ml of whole blood will be obtained and sent immediately for extracting DNA to the central laboratory. Another 10 ml of whole blood will be obtained using the serum separation tube (SST) and centrifuged within 1 hour for serum separation and sent to the central laboratory for measurements of creatinine, cystatin C, intact parathyroid hormone (iPTH), 25-hydroxyvitamine D3 (25D), 1,25-dihydroxyvitamin D3 (1,25D), and troponin T. Serum creatinine will be measured using an IDMS-traceable method; and cystatin C will be calculated using immunonephelometry calibrated against the reference material [15‐17]. The first-voided urine samples (15 ml) will be collected for assaying albumin, protein, calcium (Ca), phosphorus (P), urea, uric acid, electrolytes and osmolality at the central laboratory. The aliquots of the remaining serum and urine samples will be stored in a deep freezer (−70°C) in two different places for future studies. Stringent rules for the sample harvest and processing will be applied to ensure accurate assays. Blood and urine sampling will be conducted under strict monitoring guidelines throughout the study period.

The participating subjects will visit the center according to the follow-up schedule. At each visit, the subjects will undergo scheduled tests; the investigators will review the subjects’ recent medical histories and occurrence of events. We assumed that 10% of the individuals would withdraw from the study and that ESRD or death would develop subsequently in 10% of the subjects annually [18]. Efforts will be made to prevent drop-outs, i.e., providing free medical examinations, frequent telephone calls and dietary education.

The study protocol was approved by the IRB at each participating clinical center in 2011.

With an overall sample size of 600 subjects, which is the size of the each subgroup (GN, DN, HTN, and PKD), 80% power is achieved to detect hazard ratios of approximately 2.00 and 1.60 for an exposure with prevalence of 0.1 and 0.5 and a dropout rate of 10%.

The baseline characteristics of the study participants for each subgroup were presented with descriptive statistics using a one-way analysis of variance or Kruskal-Wallis test according to the distribution of continuous variables and a chi-square test or Fisher’s exact test for categorical variables.

To analyze the association of outcome variables, the time-to-event analysis will be used to evaluate primary outcomes, such as renal events (initiation of renal replacement therapy and 50% decline of GFR), CV events (myocardial infarction and ischemic stroke) and death. The Kaplan-Meier curves and the proportional hazard models will serve as major statistical methods for the outcomes. The incidence or mortality rates according to each outcome variable will be calculated for each etiologic group; the incidence ratios will be presented with reference to the diabetic nephropathy subgroup. The effect of exposure on the development of outcomes will be analyzed separately in each group and in the overall groups. The heterogeneity of each relative risk within the KNOW-CKD subcohort will be analyzed using the Cochran Q test or using a shared-frailty model, where time-dependent confounders should be adjusted [19]. The CKD progression will be evaluated using decreased GFR calculations. Because it is common for a patient with CKD to have a non-linear decline in GFR or a prolonged period of non-progression [20], we will assess the longitudinal trajectory of GFR for the individual participants of KNOW-CKD using the Bayesian smoothing technique [21]. If the progression of CKD is steady and linear, the generalized estimating equation will be adopted to compare the slope of the GFR with repeated measurements. The progression of CKD will be supplemented with an event-based analysis, whereby the occurrence of ESRD or the substantial decrease of GFR will be treated as the events of interest.

From July 2011 to September 2013, we enrolled 1,470 participants, including 543 GN, 317 DN, 294 HTN, 249 PKD sub-participants and remaining participants with unclassified subtypes. The baseline demographics and the laboratory values of KNOW-CKD are summarized in Tables 2 and 3. The mean age of the total CKD cohort at the time of study enrollment is 53.6 ± 12.3 years, 38.7% are female, and 31.6% diabetic. The proportion of each CKD stage is as follows; 10.9% (stage 1), 17.2% (stage 2), 18.2% (stage 3a), 22.3% (stage 3b), 23.7% (stage 4) and 7.6% (stage 5, nondialyzed). GN, DN, HTN and PKD were the most common causative diseases in descending order. At baseline, 7.3% of the cohort had a history of coronary heart disease, 8.5% had cerebrovascular disease, and 90.6% had hypertension. In terms of the Charlson comorbidity score and various cardiac parameters, the DN subgroup exhibited the worst profiles (Figure 1 and Table 2). In subjects with advanced CKD stages, the elevated levels of pulse pressure, serum phosphorus, and intact PTH were more pronounced than the subjects with early stages. Hemoglobin levels, 25D and 1,25D, were significantly lower and the left ventricular mass index (LVMI), coronary calcium score and valve calcification were higher in subjects with advanced CKD stages.