When my patients with acute stroke develop a urinary tract infection, I often prescribe a course of co-trimoxazole (trimethoprim/sulfamethoxazole).  Many of these patients have hypertension and are also taking an angiotensin converting enzyme inhibitor (ACEI) or an angiotensin receptor blocker (ARB). Prior studies show that trimethoprim and renin-angiotensin system blockers  can lead to hyperkalemia, particularly when used in combination, and a study by Michael Fralick and colleagues at the University of Toronto published in The BMJ today finds that in elderly patients taking an ACEI or an ARB, the use of co-trimoxazole is associated with an increased risk of sudden death. Should I avoid the co-trimoxazole and renin-angiotensin blockers in my patients? Should I routinely measure potassium levels in these patients? At what point should I worry?

In an associated editorial, Mahyar Ethminan and James Brophy remind us that we should be careful about making clinical decisions based on data from observational studies, particularly case-control studies) as these can identify associations but not causal links. These studies are prone to several biases owing to unmeasured and residual confounders, imbalances between patients and controls and misclassification of exposures and outcomes. They acknowledge the quality of Fralick’s study and the fact that it brings attention to an important issue but remind us that a plausible course of action for clinicians is to be careful when using the drug combination but await confirmation of the findings before giving up using these essential antibiotics.

Today The BMJ published other papers that, as a vascular neurologist, I found particularly interesting, as they are at the borderlands of internal medicine and neurology. In Endgames, Muhajir Mohamed and Jocelyn Tan describe the evaluation and treatment of a man who presented with a TIA and thrombocytopenia and was diagnosed with TTP. And Abul Siddiky and colleagues published a practical review on the management of arteriovenous fistulas —the lifeline for patients undergoing hemodialysis—to prevent infection, thrombosis, stenosis, and hemodynamic complications, such as steal syndrome and cardiac failure.

José G Merino is US clinical research editor, BMJ.