According to the authors of research published early online in the British medical Journal (BMJ), elderly patients co-prescribed trimethoprim-sulfamethoxazole (co-trimoxazole) and spironolactone are at an increased risk admission to hospital for hyperkalaemia, and this drug combination should be avoided wherever possible. 

The authors note that hyperkalaemia is an important consideration in patients receiving spironolactone in clinical practice.  Research has shown that it can occur in up to a third of patients receiving the drug, emphasising the need for regular electrolyte monitoring and the avoidance of other drugs than can cause hyperkalaemia. 

Trimethoprim has pharmacological similarities to the potassium sparing diuretic amiloride and it can reduce urinary potassium excretion by approximately 40%.  As this antibiotic is frequently prescribed for urinary tract infections (mainly as co-trimoxazole), the likelihood of co-prescription with spironolactone is high; the risk of hyperkalaemia associated with this has however not been systematically studied (evidence for a clinically meaningful interaction is limited to case reports).  The purpose of this population based nested case-control study was to characterise the risk of admission to hospital for hyperkalaemia in elderly patients treated with co-trimoxazole in combination with spironolactone. The study included Ontario residents aged ≥66 years who were treated with spironolactone between April 1992 and March 2010.  A period of continuous use was identified from the available data – beginning with the first prescription following their 66th birthday and continuing until the end of the study period, cessation of treatment, or the occurrence of a study outcome.  The outcome of interest was hospital admission with a diagnosis of hyperkalaemia within 14 days of receiving a prescription for co-trimoxazole, norfloxacin, nitrofurantoin, or amoxicillin (trimethoprim monotherapy was not included as it is invariably used in combination with sulfamethoxazole in Canada). 

The analyses were restricted to antibiotics primarily used for urinary tract infections to avoid potential confounding effects of other systemic infections.  Up to four controls were selected for each case, each of whom had received treatment with one of the study antibiotics.   Odds ratios (ORs) for the association between admission to hospital for hyperkalaemia and receipt of study antibiotic were calculated and adjusted for conditions and other medications known to influence the risk of hyperkalaemia.  Amoxicillin was used as the reference antibiotic as it should not cause hyperkalaemia.  

The main findings reported were as follows:

• A total of 6,903 admissions for hyperkalaemia were identified within the cohort of patients using spironolactone, 306 of which occurred within 14 days of antibiotic use. The majority of these (248; 81%) were matched to at least one control.

• 10.8% of patients taking spironolactone received at least one prescription for co-trimoxazole. Compared with amoxicillin, prescription of this antibiotic combination was associated with a marked increase in the risk of admission to hospital for hyperkalaemia (adjusted OR 12.4, 95% CI 7.1 to 21.6).

• The population attributable fraction was 59.7%, suggesting that approximately 60% of all cases of hyperkalaemia in older patients taking spironolactone and treated with an antibiotic for a urinary tract infection could be avoided if co-trimoxazole was not prescribed.

• Treatment with nitrofurantoin was also associated with an increase in the risk of hyperkalaemia (adjusted OR 2.4, 1.3 to 4.6), but no such risk was found with norfloxacin (adjusted OR 1.6, 0.8 to 3.4)

The authors acknowledge the limitations of their study, including lack of information on serum potassium concentrations, indices of renal function, adherence to drug treatment, or use of non-prescription medicines that may have influenced risk of hyperkalaemia.  They were also unable to identify hyperkalaemia managed in outpatient settings, or that associated with mortality in the pre-hospital setting, and the results may not be generalisable to younger patients with fewer risk factors. 

The authors conclude that increased awareness of this drug interaction is needed to ensure that the potential consequences are minimised.  They suggest that patients with spironolactone should be prescribed alternative antibiotics; if this is not possible they should be closely monitored [this would apply to use of trimethoprim alone where this is common practice, as in the UK].