Prevention of hospital-acquired hypokalemia in children receiving maintenance fluid therapy ()
Abstract
Objective: It has been suggested that the use of hypotonic intravenous fluid (IVF) puts hospitalized children at a greater risk of developing hyponatremia in children with increased arginine vasopressin (AVP) production. To reduce its risk, the National Patient Safety Agency in UK issued alert 22 in 2007, of which recommendations were to use isotonic solutions for these children at risk of hyponatremia, instead of the previously most commonly used IVF (0.18% saline/ 4% dextrose) for maintenance fluid therapy. Recent observations, however, revealed that hypokalemia are also common in hospitalized patients who do not receive potassium in their IVF. This study was conducted to validate the potassium added IVF for the prevention of hospital-acquired hypokalemia in maintenance fluid therapy. Design: For maintenance fluid therapy, a commercially available IVF solution in Japan named as Solita-T2R (Na 84 mmol/L, K 20 mmol/L, Cl 66 mmol/L, glucose 3.2%) was infused for 41 sick children with a median age of 3.01 years. Its composition is close equivalent to 0.45% saline/5% dextrose (Na 77 mmol/L, K 0 mmol/L, Cl 77 mmol/L, dextrose 5%) except K content. The patients in states of AVP excess were excluded from the analysis. Results: Median serum potassium value did not drop significantly at a median interval of 48 hours (before IVF: 4.30 mmol/L, after IVF: 4.10 mmol/L, p > 0.05), whereas median serum sodium level significantly increased from 136.0 mmol/L to 139.0 mmol/L (p < 0.001). Conclusion: Potassium added (20 mmol/L) IVF solution reduces the risk of developing “hospital-acquired hypokalemia” in children who are not in states of AVP excess in maintenance fluid therapy. It is worthwhile to study prospectively in a larger number of sick children.
1. INTRODUCTION
Intravenous fluid (IVF) prescription practices until recently have been based upon the original description of maintenance fluid requirements by Holliday and Segars in 1957 [1]. Their pioneering work was performed in healthy breastfed children and was based on calorific requirements; they advised that if IVF was necessary, hypotonic fluid should be used at rates based on body weight. In the past decade, however, there have been several articles suggesting that the use of hypotonic IVF puts hospitalized children at a greater risk of developing life-threatening hyponatremia [2-5]. Hyponatremia in otherwise healthy children with acute illnesses, called hospital-acquired hyponatremia, is considered to be mainly caused by increased arginine vasopressin (AVP) production: i.e., dilution of extracellular fluid caused by an impaired free water excretion due to excess AVP seems to contribute to development of hyponatremia [3-5]. To reduce its risk, the National Patient Safety Agency (NPSA) in UK issued alert 22 in March 2007 [6]. The recommendations of the alert were to use isotonic solutions for some children at high risk of hyponatremia, instead of the previously most commonly used IVF (0.18% saline/4% dextrose) for maintenance fluid therapy in children.
On the contrary, hypokalemia defined as a serum potassium concentration <3.5 mmol/L has rarely drawn attention despite our insistence [7,8], while it is a common and potentially serious electrolyte disorder [9,10]. Hypokalemia in hospitalized patients is mostly hospitalacquired and coexistence with hyponatremia [11]. Furthermore, it is of note that one-fourths of hypokalemic children were not receiving potassium in their IVF [12].
Based on these findings, we consider that potassium should be added into IVF and this study was conducted to validate the potassium added IVF for prevention of the hypokalemia in children receiving maintenance fluid therapy.
2. PATIENTS AND METHODS
This retrospective trial was conducted at the Kansai Medical University Hirakata Hospital, a tertiary care pediatric hospital. The medical records of the patients who admitted between November, 2008 and December, 2009 and required maintenance fluid therapy using IVF were investigated. Charts were reviewed for clinical characteristics, such as patient age, gender, diagnosis or details of administration of IVF. Laboratory data were also reviewed for blood urea nitrogen, serum creatinine, and serum levels of sodium (Na) and potassium (K).
Blood sampling were performed at least twice, i.e., just before IVF administration (T0) and at a median of 48 hours after IVF administration (T1). Soon after initial blood sampling, all patients received hypotonic IVF administration because of poor oral water intake: parenteral fluid therapy was continued till they could supply a demand of water requirements orally. Daily water requirements were calculated based on the formula proposed by Holliday and Segar [1]: the amount of daily IVF was decided by deduction of estimated oral intake from the daily water requirements and was infused evenly during parenteral fluid therapy.
Hypotonic IVF named Solita-T2R (Na 84 mmol/L, K 20 mmol/L, Cl 66 mmol/L, glucose 3.2%), which is a commercially available IVF in Japan was used for all patients. Its composition is close equivalent to 0.45% saline/5% dextrose (Na 77 mmol/L, K 0 mmol/L, Cl 77 mmol/L, dextrose 5%) except K content.
The following patients in states of AVP excess were excluded from the analysis: children requiring water restriction because of the high risk of developing syndrome of inappropriate antidiuretic hormone secretion, such as meningitis, encephalitis, acute bronchiolitis or post-operative condition [5]; children with evident dehydration; children demanding only small amount of IVF for maintenance fluid therapy (at a rate of less than 10 ml/hour). Exclusion criteria also included renal disease, cardiac dysfunction, and known adrenal dysfunction.
Hyponatremia was defined as serum Na level < 135 mmol/L and hypokalemia as serum K level < 3.5 mmol/L. The rate of change in serum Na or K was calculated as following: [serum level of Na or K at T1] – [serum level of Na or K at T0] divided by the number of hours between T0 and T1.
Because the serum levels of Na and K, and their rate of change or absolute change were not normally distributed, the difference was tested using the Wilcoxon signed rank test. A level of P < 0.05 by two-tailed analysis was accepted as statistically significant.
3. RESULTS
Table 1 showed both the demographic data of the 41 patients (male 27, female 14) and the summary of the results. The diagnoses of the subjects were Kawasaki disease in 26, acute bronchopneumonia in 10, urinary tract infection in 2, occult bacteremia in 2, epileptic seizure in 1. Median age of patients was 3.01 years (interquartile range [IQR] 1.11, 4.82).