Malone 1991.

Methods	Randomised, open controlled study I. Blinding of randomisations ‐ can't tell II. Blinding of intervention ‐ no III. Blinding of outcome measure assessment ‐ no IV. Complete follow‐up‐ yes
Participants	12 infants born at </= 28 weeks GA with hyperkalaemia defined as central non haemolyzed serum K+ of > 7 mmol/L Urine output (ml/kg/hr) was not reported Study period May 1, 1989, and December 31, 1989 Single centre, the US
Interventions	7 infants [mean (SD) GA 24.8 +/‐ 1.5 weeks, BW 780 +/‐ 160 g ] received glucose‐insulin (human insulin with 5% albumin). The insulin was "piggybacked" into the existing i.v. fluids by means of a syringe infusion pump, starting at a rate of 0.1 ml/hr to deliver a dose of 0.05 to 0.1 IU/kg/hr 5 infants [mean (SD) GA 23.8 +/‐ 0.8 weeks, BW 650 +/‐ 60 g] received sodium polystyrene sulphonate (Kayexalate) prepared in a 25% sorbitol solution and given rectally in a dose of 1 gm/kg every 6 hours
Outcomes	Mortality, treatment failure (increase in K+ concentration > 0.5 mmol/L or failure to decrease K+ to < 7mmol/L, cardiac arrhythmia, hypoglycaemia (< 40 mg/dl; 2.2 mmol/L)
Notes	In 4 of the 5 Kayexalate‐treated infants in whom treatment was considered to have failed, a glucose‐insulin therapy was added. All 4 responded with decreased central K+ concentration within 6 hours after glucose‐insulin was added
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Randomised, open controlled study
Allocation concealment (selection bias)	Unclear risk	Blinding of randomisations ‐ can't tell
Blinding (performance bias and detection bias) All outcomes	High risk	Blinding of intervention ‐ no
Blinding of participants and personnel (performance bias) All outcomes	High risk	Blinding of intervention ‐ no
Blinding of outcome assessment (detection bias) All outcomes	High risk	Blinding of outcome measure assessment ‐ no
Incomplete outcome data (attrition bias) All outcomes	Low risk	Complete follow‐up ‐ yes