Response to: ‘Risk of systemic lupus erythematosus in patients with idiopathic thrombocytopenic purpura: a need for a more accurate control group?’ by Mertz and Arnaud