Aims The handling and use of concentrated potassium solutions in clinical areas has been acknowledged as a risk to patient safety. In 2002 the National Patient Safety Agency released Patient Safety Alert 1 instructing Trusts and NHS Organisations to take action to reduce the use of concentrated potassium at ward level. Mortality and morbidity as a result of inadvertent intravenous administration of concentrated potassium to patients is categorised as a ‘Never Event’ within the NHS. Management of concentrated potassium is a key performance indicator for quality within the parent organisation.

Routine surveillance of potassium usage identified that usage in paediatric intensive care unit (PICU) increased at the same time that a new balanced crystalloid (Plasmalyte148–5% Glucose) was introduced as the standard maintenance fluid on PICU.

This study aims to:

• Identify and categorise the use of concentrated potassium on PICU.

• Identify possible influences on the addition of potassium to Plasmalyte.

• Identify frequency of concentrated potassium infusions (‘corrections’).

Methods A prospective, observational audit of fluid administration in March and April 2012. Data was collected on fluid currently being administered only. Resuscitation fluids and replacement fluid for drain/stoma losses were excluded. Those patients who had complex fluid requirements (diabetic ketoacidosis, metabolic, congenital hyperinsulinaemia and patients with known pathological electrolyte disturbances) were also excluded as they were managed according to different fluid guidelines.

Results 59 patients received intravenous fluid in the study period. Mean age 3.75 years (range 1 month–17 years) 65% of patients were fluid restricted to 70% of normal maintenance, while the remaining 35% received up to 110% of normal maintenance.

92% of patients were administered Plasmalyte148–5% glucose, or plain Plasmalyte148. Four patients received 0.45% sodium chloride with 5% glucose, and one patient received 0.9% sodium chloride. 17 patients on Plasmalyte148 (30%) had potassium chloride added to their fluids to the maximum 20 mmol/l allowed. The mean trigger potassium level for intervention was 3.3 mmol/l (range 2.5–4.4). Mean post addition potassium values were 3.7 mmol/l (range 2.4–4.4) eight patients were on drug therapy that predisposed them to hypokalaemia.

During the study period, there were 18 parenteral potassium infusions in eight patients. 45% were potassium acid phosphate for hypophosphataemia, and 55% were to correct hypokalaemia. For hypokalaemia the mean trigger level for intervention was 2.8 mmol/l (range 1.9–3.4).

Conclusions Concentrated potassium is added to 31% of administered bags of Plasmalyte. There is evidence of habitual prescribing of fluid with potassium. 5/17 patients with potassium in Plasmalyte had potassium levels >3.5 mmol/l prior to addition. Excluding these five cases, the incidence of addition is 22%. This is in a population pre-disposed to electrolyte derangements.

This presents a challenge in rolling out balanced crystalloids into general paediatric practice as the use of concentrated potassium outside of critical care areas is usually restricted.

The number of parenteral potassium corrections is surprisingly low. Further study should be undertaken to evaluate the impact of balanced solutions on potassium-homeostasis.