Rationale for and design of the CREATIVE-AF trial: randomized, double-blind, placebo-controlled, crossover study of the effect of irbesartan on oxidative stress and adhesion molecules in patients with persistent atrial fibrillation.

Nearly 50 medications have been implicated as inducing hypomagnesaemia, sometimes based on insufficient data regarding clinical significance and frequency of occurrence. In fact, clinical effects attributed to hypomagnaesemia have been reported in only 17 of these drugs. A considerable amount of literature relating to individual drugs has been published, yet a comprehensive overview of this issue is not available and the hypomagnesaemic effect of a drug could be either overemphasised or under-rated. In addition, there are neither guidelines regarding treatment, prevention and monitoring of drug-induced hypomagnesaemia nor agreement as to what serum level of magnesium may actually be defined as 'hypomagnesaemia'. By compiling data from published papers, electronic databases, textbooks and product information leaflets, we attempted to assess the clinical significance of hypomagnesaemia induced by each drug. A practical approach for managing drug-induced hypomagnesaemia, incorporating both published literature and personal experience of the physician, is proposed. When drugs classified as inducing 'significant' hypomagnesaemia (cisplatin, amphotericin B, ciclosporin) are administered, routine magnesium monitoring is warranted, preventive treatment should be considered and treatment of hypomagnesaemia should be initiated with or without overt clinical manifestations. In drugs belonging to the 'potentially significant' category, among which are amikacin, gentamicin, laxatives, pentamidine, tobramycin, tacrolimus and carboplatin, magnesium monitoring is justified when either of the following occurs: clinical manifestations are apparent; persistent hypokalaemia, hypocalcaemia or alkalosis are present; other precipitating factors for hypomagnesaemia coexist; or treatment is with more than one potentially hypomagnesaemic drug. No preventive treatment is required and treatment should be initiated only if hypomagnesaemia is accompanied by symptoms or clinically significant relevant laboratory findings. In those drugs whose hypomagnesaemic effect is labelled as 'questionable', including furosemide and hydrochlorothiazide, routine monitoring and treatment are not required.