Evaluating the effect of digital primary care on antibiotic prescription: Evidence using Swedish register data