Prognosis of acute renal failure in children

Transcription

1 Pediatr Nephrol (2006) 21: DOI /s ORIGINAL ARTICLE Prognosis of acute renal failure in children Hasan Otukesh & Rozita Hoseini & Nakissa Hooman & Majid Chalian & Hamid Chalian & Ali Tabarroki Received: 12 May 2005 /Revised: 19 May 2006 /Accepted: 19 May 2006 / Published online: 8 September 2006 # IPNA 2006 Abstract Acute renal failure (ARF) is the acute loss of renal function over a period of hours or days. Given the poor prognosis of ARF among children, there is some urgency to identifying more effective prognostic indicators for detecting disease onset. Such indicators would help provide the means of selecting patients who would benefit the most from early aggressive treatment. In this study we assessed the etiologic and prognostic indicators of ARF, including several risk factors such as sepsis, respiratory distress, age, among others, in 300 children who were admitted to the Ali Asghar Children s Hospital, Tehran, Iran, from 1990 to Statistical analysis was performed using multiple regression and chi-square methods, and a score to determine the prognosis of ARF in children was developed. Result: Based on the results of this study the three common causes of ARF are acute tubular necrosis (ATN, 38%), acute glumerulonephritis (24%) and hemolytic uremic syndrome (24.1%). The overall mortality rate among our patients was 24.7%, with the highest risk group being those patients suffering from ischemic ATN. In addition, the correlation (p<0.0005) between the etiology and mortality rate was particularly high in patients with ischemic ATN. Mortality was also high (68%) in children younger than 2 years. Multiple regression models revealed that among those factors that significantly differed between the survivors and nonsurvivors, only the necessity of dialysis (p<0.0005), the use of mechanical ventilation (p=0.05) and disseminated H. Otukesh (*) : R. Hoseini : N. Hooman : M. Chalian : H. Chalian : A. Tabarroki Ali Asghar Children s Hospital, Tehran, Iran hasanotukesh@yahoo.com intravascular coagulation (p=0.038) can be regarded as independent determinants of ARF prognosis in children. Keywords Acute renal failure. Mortality. Prognosis. Scoring. Risk factors Introduction Acute renal failure (ARF) is defined as the sudden loss of the ability of the kidneys to excrete excess water, regulate electrolytes and acid base status and eliminate waste products. This acute loss of renal function may occur over a period of hours or days [1 3]. Acute renal failure occurs in 2 3% of children admitted to pediatric special care centers and in as many as 8% of infants in neonatal intensive care units [4]. Etiologically, this condition can be placed into three categories prerenal or functional, renal and post-renal with pre-renal ARF being the most common form in children [5 7]. The mortality rates among children with ARF differ among the various studies reported [7 13], depending on the nature of the underlying disease. In contrast to patients with ARF as a result of multi-organ failure, children with ARF caused by conditions solely affecting the kidney, such as postinfectious glumerulonephritis, have a very low mortality rate (90 vs. 1%, respectively), [4]. Many studies have been carried out in adults to identify the prognostic factors of ARF that play an essential role in selecting patients suitable for more aggressive treatments. Risk factors assessed in children and adult ARF consist of oliguria [14 18], age [19], sex, complications such as respiratory distress, pancreatitis, sepsis, hypotension, hypertension, coma, seizure, use of mechanical ventilation, jaundice and hepatic failure, necessity of dialysis, hemor-

2 1874 Pediatr Nephrol (2006) 21: rhage, disseminated intravascular coagulation (DIC) and hemolysis, electrolyte and acid base status disorders, leukopenia, thrombocytopenia, hypoalbuminemia, serum creatinine [20] and Blood Urea Nitrogen (BUN) levels, underlying illness and, finally, the type of treatment and interval between the onset of ARF and the start of therapy [21 24]. However, only limited data are available on the etiology and outcome of ARF in Iranian children. In order to determine etiologic and prognostic indicators of ARF in children, we carried out a retrospective analysis that investigated the causes and outcome of ARF in Iranian children. Methods and subjects This study consisted of a retrospective review of data on children admitted over a 15-year period, from 1989 to 2003, to the department of pediatric nephrology of Ali Asghar Hospital, Tehran, Iran, with the diagnosis of ARF. In total, 300 children, aged from birth to 16 years, were included in the study. ARF was defined as a rapid and progressive decline in renal function manifesting as a creatinine level that increases at a rate of 0.1 mg/dl/day over a 1-week period. The normal creatinine level for each individual can be calculated using the formula below [4]: Pcr ¼ 0:18 þ 0:032 Age ð yrþ Risk factors for mortality were defined on the basis of several evaluations performed in adults and children. Oliguria was defined as a urine volume less than 0.5 ml/kg/h in infants and young children [14] and less than 400 ml/day in older children, while hyperkalemia was defined as serum potassium levels of more than 5.5 meq/l (after correcting for the effect of acidosis and alkalosis or polycytemia) in infants and children. Hyperkalemia was defined, in the neonatal period, as a serum potassium level of more than 7.5 meq/l in the first day of admission and of more than 6.5 meq/l in the remaining days. Hypokalemia was defined as a serum potassium level of less than 3.5 meq/l. Hypotension was determined based on the following formula [25]: Systolic blood pressure<70mmhg þ 2age ð yrþ Shock was defined as severe hypotension and end organ damage due to poor peripheral perfusion. Hypertension was defined as systolic and/or diastolic blood pressure around the 95 percentile for age and sex. Coma was defined as a Glasgow Coma Score (GCS) under 5, and significant metabolic acidosis was defined as a serum bicarbonate level of less than 8 meq/l. The two main treatment modalities were conservative or dialysis [hemodialysis (HD) or peritoneal dialysis (PD)], and the indications for these treatments in our study were severe metabolic acidosis nonresponsive to medical management, persistent hyperkalemia, fluid overload with evidence of hypertension and/or pulmonary edema refractory to diuretic therapy, neurologic symptoms, BUN greater than mg/dl, calcium/phosphate imbalance with hypocalcemic tetany and the inability to provide adequate nutritional intake. A history of chronic renal insufficiency (CRI) was our exclusion criterion. The clinical condition and risk factors were evaluated only during the first visit, and the long-term outcome was not assessed as this did not fall within the framework of our study. Statistical analysis All of the variables in survivors and nonsurvivors were analyzed using multiple logistic regression analysis. In addition, Student s t-test and chi-square test were used for evaluating the quantitative and qualitative variables, respectively. p values less than 0.05 were regarded as being statistical significant. Results ARF was diagnosed in 300 pediatric patients, including 190 (63.4%) boys and 110 (36.6%) girls whose mean age and weight were 4.84 years and 17.8 kg, respectively. The etiologies of ARF among the patient cohort are shown in Table 1. Patients with acute tubular necrosis (ATN) were classified into two groups: ischemic ATN (95 cases, 83%) and nephrotoxic ATN (19 cases, 17%). Of the patients 31.3% (94/300) were under dialysis [PD in 60.6% (57/94) and HD in 39.4% (37/94)]. The need for dialysis among our patients was found to correlate with the mortality rate, and the mean duration of dialysis for these 94 children was 5 days 8.8 days in nonsurvivors and 3.6 days in survivors. In addition, a meaningful correlation was revealed between the duration of dialysis and the mortality rate: mortality increased in patients who had been under dialysis longer (p<0.05). The overall mortality rates of children treated with HD and PD were 97.3% (36/37) and 66.6% (38/57), respectively (p<0.05, HD vs. PD). The mortality rate in patients under 2 years of age was 68% (51/57), with most mortalities occurring in infants under 1 year (except neonates); in other words, the mortality rate was higher in children under the age of 2 years, excluding neonates.

3 Pediatr Nephrol (2006) 21: Table 1 Distribution of etiology of patients by age Age at diagnosis Etiology of acute renal failure a Number of patients (n) ATN HUS AGN Pre-renal Post-renal CN IN TLS Neonate _ 33 1 month 1 year _ year _ 2 _ year year _ 50 Total a ATN, Acute tubular necrosis; HUS, hemolytic uremic syndrome; AGN, acute glomerular nephritis; CN, cortical necrosis; IN, interstitial Of all patients under dialysis, 47.2% were under the age of 2 years. The mean cause of increased mortality in this age group was attributed to the higher incidence of ischemic ATN. However, a delay in putting these patients under dialysis (because of late referral) may also have played a role. Thirty-three patients were in the neonatal age group. The mortality rate in this age group was 48.4% (16/33); overall mortality by diagnosis is presented in Table 2. Our assessments revealed a number of significant relationships between etiologies and outcomes (p<0.0005): 34% of patients with ATN died in contrast to lower mortality rates in patients with hemolytic uremic syndrome (HUS) (25%), glomerulonephritis (7.8%) and acute interstitial nephritis. There was a distinctly significant relationship between ischemic ATN and the eventual outcome (p<0.01). Table 3 presents factors found to be associated to mortality in the present study. The need for dialysis and mechanical ventilation (patients who were not treated appropriately) were found to be important determinants of mortality. Table 2 Etiology and outcome of acute renal failure in children Diagnosis Survivors (n) Non-survivors (n) Mortality (%) Acute tubular necrosis Hemolytic uremic syndrome Acute glomerulonephritis Pre-renal ARF Post-renal ARF Cortical necrosis Acute interestial nephritis Tumor lysis syndrome Neonatal period; ATN Pre-renal ARF Post-renal ARF Based on our results we came to the conclusion that the mean percentage of creatinine increase in ARF patients was remarkably higher in nonsurvivors. Multiple logistic regression models revealed that among the factors that showed a significant difference between survivors and nonsurvivors, only the necessity of dialysis (p<0.0005), use of mechanical ventilation (p<0.05) and DIC could be regarded as independent determinants of the prognosis of ARF in children (Table 4, Fig. 1). In neonates, the most common cause of morality was ischemic ATN secondary to sepsis. We also detected a relationship between oligoanuria and the eventual outcome in this age group (p=0.003). Table 3 Markers of a wrong prognosis in ARF Markers Survivors a Non-survivors a p value Oligoanuria 92 (40.7) 47 (63.5) <0.05* Age (year) <0.05* Male/female 147/79 43/31 >0.05 ns Respiratory distress 30 (13.2) 57 (77) <0.05* Sepsis 22 (9.7) 46 (62) <0.05* Hypotension 30 (13) 34 (45.9) >0.05 ns Hypertension 73 (32.3) 13 (17.5) >0.05 ns Shock 9 (3.9) 13 (17.5) <0.05* Need of dialysis 20 (8.8) 74 (100) <0.05* leukopenia 5 (2.2) 3 (4.05) >0.05 ns Use of vasopressors 7 (3.09) 26 (35) <0.05* Disseminated 26 (35) 8 (3.5) <0.05* intravascular coagulation Hypokalemia 35 (15.4) 5 (6.7) >0.05 ns Hyperkalemia 50 (22) 25 (33.7) >0.05 ns Thrombocytopenia 68 (30) 35 (47.2) >0.2 ns Coma 8 (3.5) 24 (32.4) <0.05* Hemorrhage 18 (24.3) 35 (47.2) >0.05 ns Mechanical ventilation 11 (4.8) 33 (44.5) <0.05* Acidosis 26 (11.5) 18 (24.3) <0.05* *Significant at p<0.05; ns, not significant a Percentages are given in parenthesis

4 1876 Pediatr Nephrol (2006) 21: Discussion Acute renal failure is a life-threatening condition causing significant morbidity and mortality in children, although one study has shown that intensive dialysis is an effective treatment [26]. In our study, the male-to-female ratio was 1.7. The reason for this discrepancy is the high incidence of sepsis in our patients and an underreporting of female patients. The etiology of ARF may be pre-renal, intrinsic or postrenal, and the causes seem to differ among countries. In the present study, ATN was the most common cause of ARF in our patient cohort, which is contrary to the etiologies of ARF in most developed countries where the most common causes are the result of intrinsic renal disease due to nephrotoxins. However, in some other studies, HUS has been reported to be the most common cause of ARF in children living in industrialized countries. In this study, HUS was found in 24% of ARF cases. In Iran, as in any other country with a high prevalence of acute diarrheal syndromes, the most common cause of ARF is ATN or prerenal azotemia; however, in developed countries, the most common cause of ARF is HUS. Ali Asghar Hospital is a referral center, and many patients with severe dehydration are admitted for treatment, which may have caused a bias in our results. Table 4 Multiple regression model a of predictive factors of ARF in children Predictive factors B Standard error Significance EXP (β) Need of dialysis < DIC b < Mechanical ventilation Constant < a Model significance: 0.000; method: backward stepwise ward; R-square: 80% b DIC, Disseminated intravascular coagulation The mortality rate of 24.7% noted in this study is similar to that reported for most other countries [9 11], although mortality rates as high as 40 50% have been reported, depending on the underlying disorder [5, 8, 11]. When kidneys are affected in multiple organ dysfunction, the prognosis is very poor. In contrast, the prognosis improves when primary renal disorders are the cause of the condition. Other studies have found the mortality of childhood ARF to range from 45 to 75% in multiple organ dysfunction and to be 12.5% in primary renal disorders [7, 9, 22, 27]. In the neonate age category, overall mortality has been reported to be between 45 and 60% [17, 28, 29, 30]; our overall mortality rate in this age group was 48.4% and is therefore in agreement with the above studies. After prerenal and post-renal ARF was excluded from the analysis of neonates, the mortality rate was 69.5%, which is in Fig. 1 Receiver operating characteristic (ROC) curve for independent determinants of the prognosis of ARF in children. Diagonal segments are produced by ties 1.00 ROC Curve Source of the Curve Reference Line Sensitivity DIC Dialysis MV 1 - Specificity

5 Pediatr Nephrol (2006) 21: agreement with values reported elsewhere. The most common cause of neonatal mortality in our study was sepsis and shock (65%). The outcome of renal replacement therapy (RRT) in children varies throughout the world. In some studies, mortality was evaluated on the basis of the type of RRT (PD vs. HD). Our evaluations showed that children who received HD had a higher mortality rate than those who received PD. Since only limited data are available at the present time that compares outcomes achieved with different dialysis modalities, the modality suitable for a specific patient is empirical. Some studies advocate continuous techniques of blood purification in patients with ARF [31]; these techniques are not available in our center. We found that dialysis affected our patient outcomes and that patients with ARF were best treated with PD, especially when the condition of the patient was unstable. Oliguria or anuria is a sign of the severity of ARF and a determinant of poor prognosis of this condition in adults [18, 19, 23]. The existence of an association between oligoanuria and the outcome of ARF in children is uncertain and seems to depend on the specific study [8, 14, 32]. Some researchers have reported that oliguria is an indicator of increased mortality in childhood ARF, although the references mentioned are occasionally based on adult investigations [3, 16, 33]. Using multiple regression analysis, we were unable to find an association between oliguria and a poor prognosis of ARF [8, 9]. Loza et al. revealed that in the multiple logistic regression analysis of their data, only oliguria (p=0.07) and age group (p=0.049) were associated with mortality [34], with the latter being significantly higher in patients with hypertension as well in those for whom dialysis was not provided despite there being an indication [35]. The multiple logistic regression models in our study revealed that significant determinants of ARF prognosis in children include being under dialysis, mechanical ventilation and DIC. Among these, DIC was the most significant and the best predictor of mortality. We also concluded that patients with ARF, especially hemodynamically unstable patients with severe metabolic acidosis, are best treated with peritoneal dialysis. References 1. Proesmans W (2002) Acute renal failure in childhood. EDTNA ERCA J [Suppl 2]: Esson ML, Schrier RW (2002) Diagnosis and treatment of acute tubular necrosis. Ann Intern Med 137: Thadhani R, Pascual M, Bonventre JV (1996) Acute renal failure. N Engl J Med 334: Behrman RE, Kliegman RM, Jenson HB (2004) Nelson textbook of pediatrics, 17th edn. Saunders, Philadelphia, pp Edelmann CM Jr (1992) Pediatric kidney disease, 2nd edn. Little Brown Co, Pittsburgh, pp Woroniecki R, Devarajan P (2002) Acute tubular necrosis. emedicine.com, January 22: Kandoth PW, Agarwal GJ, Dharnidharka VR (1994) Acute renal failure in children requiring dialysis therapy. Indian Pediatr 31: Arora P, Kher V, Rai PK, Singhal MK, Gulati S, Gupta A (1997) Prognosis of acute renal failure in children: a multivariate analysis. Pediatr Nephrol 11: Gallego N, Perez-Caballero C, Gallego A, Estepa R, Liano F, Ortuno J (2001) Prognosis of patients with acute renal failure without cardiopathy. Arch Dis Child 84: Moghal NE, Brocklebank JT, Meadow SR (1998) A review of acute renal failure in children: incidence, etiology and outcome. Clin Nephrol 49: Bunchman TE, McBryde KD, Mottes TE, Gardner JJ, Maxvold NJ, Brophy PD (2001) Pediatric acute renal failure: outcome by modality and disease. Pediatr Nephrol 16: Offner G, Brodehl J, Galaske R, Rutt T (1986) Acute renal failure in children: prognostic features after treatment with acute dialysis. Eur J Pediatr 144: Flynn JT, Kershaw DB, Smoyer WE, Brophy PD, McBryde KD, Bunchman TE (2001) Peritoneal dialysis for management of pediatric acute renal failure. Perit Dial Int 21: Barratt TM, Avner ED, Harmon WE (1999) Pediatric nephrology, 4th edn. Lippincott Williams and Wilkins, Philadelphia, pp Liano F, Junco E, Pascual J, Madero R, Verde E (1998) The spectrum of acute renal failure in the intensive care unit compared with that seen in other settings: The Madrid Acute Renal Failure Study Group. Kidney Int Suppl 66:S16 S Honda N, Hishida A (1993) Pathophysiology of experimental nonoliguric acute renal failure. Kidney Int 43: Stapleton FB, Jones DP, Green RS (1987) Acute Renal Failure in neonates: incidence, etiology and outcome. Pediatr Nephrol 1: Cantarovich F, Verho MT (1996) A simple prognostic index for patients with acute renal failure requiring dialysis. French Multicentric Prospective Study on Furosemide in Acute Renal Failure Requiring Dialysis. Ren Fail 18: Schrier RW (2001) Acute renal failure in practice, 2nd edn. Saunders, Philadelphia, pp Wheeler DC, Feehally J, Walls J (1986) High risk acute renal failure. Q J Med 61: Gallego N, Gallego A, Pascual J, Liano F, Estepa R, Ortuno J (1993) Prognosis of children with acute renal failure, a study of 138 cases. Nephron 64: Barretti P, Soares VA (1997) Acute Renal Failure: clinical outcome and cause of death. Ren Fail 19: Schrier RW (1999) Atlas of kidney diseases, 1st edn. Blackwell Science, Philadelphia, pp Schrier RW (2003) Renal and electrolyte disorders, 6th edn. Lippincott Williams and Wilkins, Philadelphia, pp Gunn VL, Nechyba C (2002) The Harriet Lane Handbook, 16th edn. Mosby, St. Louis, pp Filler G (2001) Acute Renal Failure in children: etiology and management. Paediatr Drugs 3: Fargason CA, Langman CB (1993) Limitations of pediatric risk of mortality score in assessing children with acute renal failure. Pediatr Nephrol 7: Chevalier RL, Campbell F, Brenbridge AN (1984) Prognostic factors in neonatal acute renal failure. Pediatrics 74:

6 1878 Pediatr Nephrol (2006) 21: Reimold EW, Don TD, Worthen HG (1977) Renal failure during the first year of life. Pediatrics 59 [Suppl 6] : Norman ME, Asadi FK (1979) A prospective study of acute renal failure in the newborn infant. Pediatrics 63: Barletta GM, Bunchman TE (2004) Acute renal failure in children and infants. Curr Opin Crit Care 10: Bourquia A, Zaid D (1993) Acute renal insufficiency in children, retrospective study of 89 cases. Ann Pediatr (Paris) 40: Sudo M, Honda N, Hishida A, Nagase M (1980) Renal hemodynamics in oliguric and non-oliguric acute renal failure of rabbits. Nephron 25: Loza R, Estremadoyro L, Loza C, Cieza J (2006) Factors associated with mortality in acute renal failure (ARF) in children. Pediatr Nephrol 21: Anochie IC, Eke FU (2005) Acute renal failure in Nigerian children: Port Harcourt experience. Pediatr Nephrol 20: