Research Article: Development of a risk score for earlier diagnosis of chronic kidney disease in children

Date Published: April 19, 2019

Publisher: Public Library of Science

Author(s): Paulo Cesar Koch Nogueira, Tulio Konstantyner, Maria Fernanda Camargo de Carvalho, Cristine Campos de Xavier Pinto, Isabel de Pádua Paz, Vera Maria Santoro Belangero, Marcelo de Sousa Tavares, Clotilde Druck Garcia, Oreste Angelo Ferra Neto, Káthia Liliane da Cunha Ribeiro Zuntini, Marina da Rocha Lordelo, Samira Shizuko Parreao Oi, Renata Trindade Damasceno, Ricardo Sesso, Giuseppe Remuzzi.


To develop a clinical score for the early identification of chronic kidney disease (CKD) in children and adolescents. The early diagnosis of CKD in childhood allows the adoption of measures to slow the progression of the disease, thereby reducing morbidity and mortality. Nevertheless, the diagnosis is often made too late for proper patient management.

We preformed a case-control study of a multicenter Brazilian sample of 752 pediatric patients; the study cases (n = 376) were CKD patients with a median estimated GFR of 37 (IQR = 22 to 57) ml/min/1.73 m2. The control group (n = 376) comprised age-, gender- and center-matched children who were followed for nonrenal diseases. Potential risk factors were investigated through a standard questionnaire that included symptoms, medical history, and a clinical examination. Two multivariable models (A and B) were fitted to assess predictors of the diagnosis of CKD.

In model A, 9 variables were associated with CKD diagnosis: antenatal ultrasound with urinary malformation, recurrent urinary tract infection, polyuria, abnormal urine stream, nocturia, growth curve flattening, history of hypertension, foamy urine and edema (c-statistic = 0.938). Model B had the same variables as model A, except for the addition of the history of admission during the neonatal period and the exclusion of antenatal ultrasound variables (c-statistic = 0.927).

The present scores may serve as a warning sign for CKD diagnosis in children among professionals working in the primary care setting where the symptoms associated with a risk of CKD may be overlooked.

Partial Text

Chronic kidney disease (CKD) in children is a major disorder that leads to health system overload and represents a challenge, mainly in developing countries [1–3]. This disease usually progresses to the loss of kidney functions, leading to the need for renal replacement therapy (RRT), which has a huge impact on patients’ health conditions and family dynamics [4,5]. Particularly in children, CKD is associated with malnutrition, which stunts and delays development, causing a significant impairment in the quality of life and reduction in life expectancy [6,7].

The sample is composed of 752 children, 376 with CKD and 376 controls. The case group consisted of children with median eGFR = 37ml/min/1.73 m2 (IQR = 22 to 57). The causes of CKD were: congenital anomalies of the kidney and urinary tract (CAKUT) in 180 individuals (48%), unknown in 42 (11%), focal segmental glomerulosclerosis in 35 (9%), cystic kidney disease in 35 (9%), acute renal failure sequel in 11 (3%), other glomerulopathies in 10 (3%), and other diseases in 63 (17%) children. With regard to CKD stages, the patients were categorized as follows: 81 (21%) in stage 2, 150 (40%) in stage 3, 101 (27%) in stage 4 and 44 (12%) in stage 5. Median (IQR) time between the diagnosis of CKD and the application of the questionnaire was 4.1 (1.5 to 8.2) years. Table 1 shows the demographic and clinical data of the study groups.

The main contribution of this study was the establishment of clinical signs or symptoms that may be useful to identify children at a greater risk of having a CKD diagnosis. The factors that were selected as predictors of CKD diagnosis in our score could be categorized into 4 groups: a) CAKUT indicators (antenatal ultrasound with urinary tract abnormality, history of recurrent urinary tract infection and abnormal urine stream); b) growth interruption (flat growth curve); c) loss of urinary concentration symptoms (polyuria and nocturia) and d) early neonatal diseases (neonatal admission). All the signs sorted out from the models have a plausible biological role to function as risk markers, and none of them is per se a new discovery. We postulate that the score here described allows one to anticipate the diagnosis and to quantify the risk of CKD in pediatric patients, thereby yielding the support for further detailed laboratory examination.




Leave a Reply

Your email address will not be published.