The accuracy of tonicity balance formulas in predicting changes in plasma sodium in ventilated infants with respiratory syncitial virus bronchiolitis

Objective

Infants with respiratory syncitial virus (RSV) infection demonstrate avid renal water retention due to raised antididuretic hormone activity and are at risk of hyponatraemia [1]. Fluid restriction (48 ml/kg/day) and diuretics are often required to prevent this occurrence and to maintain fluid and sodium balance. Many methods have been postulated for predicting tonicity balance and the resulting change in plasma sodium (Na) changes. These include calculating the tonicity balance using intake/output of water with net Na flux (tonoNa), Na and K flux (tonoNa+K), and more recently with the Nguyen–Kurtz method [2–4]. The objective of this study was to document the water and electrolyte balance, and to document the accuracy of these different tonicity balance methods to predict Na changes during the first day of intensive care admission in ventilated infants with bronchiolitis.

Method

We prospectively enrolled 30 consecutive infants (median weight 4.3 kg) with bronchiolitis requiring mechanical ventilation to a tertiary PICU. Infants were placed on a fluid-restricted regime of 48 ml/kg/day 0.9% saline intravenously or oral feed when tolerated, and were studied over the first 24 hours following PICU admission. Blood samples were collected on admission and at 24 hours; urine collection over this period was analysed for osmolality and electrolyte content. The volume and content of all fluid inputs, including drugs, were recorded over the same time periods. Data are presented as the median (IQR).

Results

The admission plasma Na was 140 mmol/l (138–141) and after 24 hours was 139 mmol/l (138–143). The net electrolyte and fluid balance is presented in Table 1 with almost neutral balance for Na, K and Cl and fluid, in keeping with the fact that plasma Na did not change over the 24-hour period (P = 0.9). Free water clearance predicted a net gain of water of 10 ml/kg/day (5–17). Electrolyte free water clearance using urine Na and K predicted a net gain of 17 ml/kg/day (7–26) free water. Consequently both the tonicity balance with Na and with Na and K predicted a significant fall in plasma Na at 24 hours: 137 mmol/l (136–139, P = 0.1), and 136 mmol/l (135–140, P = 0.006), respectively. The Nguyen–Kurtz method was most accurate and predicted no change in plasma Na: 142 mmol/l (136–147, P = 0.06). Bias with 95% limits of agreement for predicting plasma Na with the tonoNa, tonoNa+K and Nguyen methods were -2 (-9 to 13), -3 (-10 to 5) and 2 (-9 to 13) mmol/l, respectively.

Table 1 (abstract P182)

Conclusion

Fluid restriction of ventilated infants with bronchiolitis results in a neutral balance of water and elelctrolytes and no resulting change in plasma Na. Tonicity balance methods using the net Na balance or Na and K balance overestimate free water retention and predicted lower plasma Na values. The Nguyen-Kurtz method, although more complex, provides the most accurate method of predicting plasma Na changes in these patients.