Indian Pediatrics 2000;37: 1348-1353
Electrolyte Abnormalities in Children Admitted to Pediatric Intensive Care Unit
Electrolyte abnormalities are common in children who need intensive care. They occur in a variety of conditions, may remain unrecognized and result in morbidity and mortality irrespective of the primary problem. Timely recognition, a high index of suspicion and a thorough understanding of common electrolyte abnormalities is necessary to ensure their correction.
The present study was undertaken to study the frequency of electrolyte imbalance at admission in children admitted to a Pediatric Intensive Care Unit (PICU). We also examined the relationship between electrolyte abnormalities and the primary illness and their impact on the morbidity and mortality.
All children admitted to the PICU between January 1997 and July 1998 were included in the study. Children with diarrhea and dehydration were excluded, as it is the departmental policy to admit them in a separate ward. At the time of admission, the patient’s age, sex, provisional diagnosis and brief clinical features were recorded.
Two venous blood samples and one urine sample were collected. One blood sample was used for estimation of serum sodium and potassium. The second serum and urine sample were used for the work-up for diagnosis of the syndrome of inappropriate ADH secretion (SIADH). Urine spot sodium and urine and serum osmolality were estimated in patients who were clinically euvolemic with initial serum sodium level less than 125 mEq/L. An ECG was done in patients with hyperkalemia and serum potassium level >6 mEq/L.
Chi-square test was used to test the significance of difference in morbidity and mortality in patients with normal and abnormal electrolytes. The analysis was done using EPI Info Version 6 statistics package(1).
Hyponatremia and hyperatremia were defined as serum sodium concentration below 130 mEq/L and above 150 mEq/L, respect-ively(2-4). Hypokalemia and hyperkalemia were defined as potassium level below 3.5 mEq/L and above 5.5 mEq/L, respect-ively(3,5).
SIADH was diagnosed in the presence of hypotonic hyponatremia, absence of hypo-volemia or dehydration, absence of edema and normal renal excretory function(6,7).
Morbidity was defined as prolonged ICU stay in this study (a duration of stay more than 5 days).
Three hundred and five patients aged between 1 month and 14 years, were admitted to PICU during the period. The mean age in months was 49.1 ± 48.01 months; the ratio of boys to girls was 2 : 1. Ninety nine (32.45%) had electrolyte abnormalities. Of these 24 (7.9%) had mixed electrolyte imbalance, hyperkalemia was seen in 44 (14.4%), hypo-natremia in 29 (9.5%), hypernatremia in 15 (4.9%) and hypokalemia in 11 (3.6%) cases. Metabolic acidosis was present in 23 patients. ECG changes of hyperkalemia were observed in 11 patients with potassium values of >6.5 mEq/L.
Of the 99 patients with electrolyte imbalance, 24 (24.2%) expired. In these 24 patients, 10 (41.6%) had hyperkalemia, 6 (25%) had hyponatremia, 5 (20.8%) had hyper-natremia and 3 (12.5%) had hypo-kalemia.
The electrolyte abnormalities in relation to their primary illness, and morbidity and mortality are shown in Tables I & II.
Of 29 patients with hyponatremia, 19 (65.5%) were clinically euvolemic, 8 (27.6%) were hypovolemic and 2 (6.9%) were hypervolemic. Of 19 patients investigated for SIADH, 10 patients had the abnormality. Of those with SIADH, 6 had infections of the central nervous system (4 with tuberculous meningitis and one each with viral ence-phalitis and cerebral malaria), one patient each had bronchiolitis, septicemia, snake-bite and Steven Johnson syndrome. The remaining 9 cases of euvolemic hyponatremia though clinically suspected to have SIADH could not be confirmed by laboratory values.
Ten patients had a combination of hypo-natremia and hyperkalemia, 5 had hyper-natremia with hypokalemia and 5 had hypernatremia with hyperkalemia. Four patients had a combination of hyponatremia and hypokalemia.
The morbidity in electrolyte disturbances is shown in Table II. When compared to those with normal values of sodium and potassium, the mean duration of hospital stay in children with hyperkalemia and hyponatremia was significantly more, (p <0.001 for sodium and p <0.01 for potassium).
Table II shows mortality in various electrolyte disturbances. Apart from hypo-kalemia all other abnormalities had signifi-cantly higher mortality when compared to those with normal electrolyte values though the underlying disease conditions were similar. Of the 10 patients with hyperkalemia who died, 6 died before their serum potassium could be normalized. Five had serum potassium levels more than 7 mEq/L. Serum potassium values in 4 patients were reduced to below 5.5 mEq/L before they expired.
Electrolyte abnormalities were observed in 32.4% of children getting admitted to PICU. Hyperkalemia was the commonest, found in 14.4% cases. These findings are in contrast to those by Singhi et al.(8) who found hyperkalemia in 5.4% PICU admissions. This difference could be due to the fact that in their study hyperkalemia was defined as potassium level >6 mEq/L. Hyperkalemia was how-ever, defined as serum potassium levels >5.5 mEq/L in this study which probably explains a higher frequency of the abnormality.
Hyponatremia was the second commonest electrolyte abnormality noted in this study. In a study conducted in adults, 6.9% of total admissions had hyponatremia(9). A higher frequency of 29.8% was observed in the prospective study of 727 sick children(10). Twenty per cent children in that study had diarrhea. In another study, 34% of the hospitalized patients were classified as hypo-natremic(11). Since children with diarrhea were not included in our study, this may explain the lower incidence of hyponatremia in the present study. Hypokalemia was the least common electrolyte abnormality. However a significantly higher frequency (13.9-14.8%) was observed in two other studies(8,12). Again this difference is explained by inclusion of patients with diarrheal diseases in these studies.
Spectrum of Illnesses
In a prospective study, 27% of patients with infectious diseases admitted to a PICU, were seen to have hyponatremia(10). Other patients with hyponatremia had either diarrheal disease(20%), pneumonia(19%) or central nervous system disorders (12%)(10). Pizzotti et al.(11) showed that 10.8% of patients with hyponatremia had an under-lying neurologic disorder and hyponatremia persisted in 4.3% of them after treatment of the underlying disorder. This could be due to the difference in the pattern of PICU admissions in these studies where only 10.3% were CNS diseases.
In our study, hypokalemia was almost evenly distributed in diseases of various systems. Similar observations have been recorded in a descriptive retrospective analysis of 290 patient records(12).
Hyperkalemia was found usually asso-ciated with renal diseases and infectious diseases. This is in contrast to observations made by others where respiratory diseases and diarrheal diseases together contributed to 46% of patients with hyperkalemia(8).
The morbidity, as determined by the PICU stay was significantly higher in patients with hyponatremia when compared to those with normonatremia. This observation is similar to the findings reported previously(10,13). The severity of the underlying disease may also contribute to prolonged PICU stay.
The risk of mortality is increased by 3-3.5 times in patients with hyponatremia when compared to those with normal serum sodium(10,13). The predictive factors for mortality in hypernatremic patients are reported to include persistently elevated serum sodium levels in association with protracted hypotension(14). Our study also had similar findings. Mortality rates in patients with mixed electrolyte abnormalities were higher when compared to those with single electrolyte abnormality. Similar obser-vations were made by others(8,10).
In conclusion, electrolyte abnormalities are common even in children without diarrheal diseases, who are admitted to PICU. They contribute significantly to the mortality and morbidity. Since the specific symptoms of electrolyte abnormality often merge with the underlying disease, close monitoring and correction of electrolyte abnormalities is important to reduce morbidity and mortality.
The authors would like to thank Dr. Sylvan John Rego for carefully reviewing the manuscript and giving his valuable suggestions.
Contributors: SDS derived the concept, study design and analysis. BT was involved in case selection, data collection, sampling, analysis, preparation and presentation of manuscript.