A predicting model established for chronic renal failure in children and clinical value research

DU Juan; SHEN Ying

Abstract：Objective To evaluate the prognostic risk factors and establish a predicting model of survival rates of the hospitalized children with chronic- renal failure (CRF). Methods Totally 71 patients with CRF were collected from January 2002 to September 2008 in Beijing Children' s Hospital. All patients were followed up through phone calls, mailing or out-patient reviewing. Kalplan-Meier methods and Cox's proportional hazard regression model were used to analyze prognostic risk factors. Prognostic index (PI) was calculated. Results The average survival time was 47.3 months. The median survival time was 42.0 months. The second and fifth year survival rates of CRK children were 53% and 37% respectively. The risk factors associated with prognosis were treatment modality (symptomatic treatment P= 0.000, RR=53.901, dialysis P =0.047, RR = 4.361), hyperphosphatemia (P =0.000, RR = 10.298), hyperkalemia (P = 0.008, RR = 5.053), water-sodium retention(P = 0.003, RR = 3.714)and hyperproteinuria(P = 0.027, RR = 2.438). Whether receiving renal graft or not was the most important impact factor on prognosis. The predicting formula was PI = -5.556 + 1.620 X1 + 2.332 X2 + 0.891.X3 + 1.312 X4 + 3.987 X5 + 1.473 X6. According to the value of PI, the prognosis of the patients was significantly different among the three subgroups (P = 0.000). Conclusion The prognostic risk factors of CRF in children are treatment modality, hyperphosphatemia, hyperkalemia, water-sodium retention and hyperproteinuria. The established predicting model of survival rates can predict the prognosis of the children with CRF.