INTRODUCTION
Non-adherence to pharmacological treatment is a significant global public health issue, as it is estimated that half of patients do not take their medication as prescribed [1, 2]. Medication adherence is defined as the extent to which medication-intake behavior corresponds with the medication regimen agreed upon with the healthcare professional [1]. Among pregnant women, non-adherence can result in or facilitate the onset of maternal and fetal complications, enhance risks of hospitalization, and subsequently increase healthcare costs [2, 3]. Women tend to overestimate the teratogenic risk of medication [4-6], resulting in reluctance to use medication during pregnancy due to these concerns [6-9]. Incorrect perception of the safety of medication use during pregnancy is widespread [10], and leads to caution among treating physicians in prescribing medication as well [11-13].
Previous studies generally indicate low adherence to prescribed medication during pregnancy, but methods and definitions differed substantially between studies. When comparing dispensing data to self-reported medication use, only 43% of all medications dispensed were reported to be used by Danish pregnant women, with notable differences between medications for chronic conditions (70-100%) and medications used for local or short-term treatment (12-77%) [14]. Similarly, in a Dutch cohort study among pregnant women concordance between self-reported medication use and pharmacy records was moderate to good for medications used for chronic conditions, whereas concordance was low for medications that are used occasionally [15]. Instead of using a self-administered questionnaire to verify actual use of prescribed medication, De Jonge et al. conducted telephone interviews, showing that 84-92% of mothers confirmed taking the medication, but slightly more than half of the medications were used at the prescribed dose and for the duration prescribed [16].
Establishing the extent to which pregnant women are adherent to their medication is challenging. Methods that rely on self-reported medication use, including questionnaires and interviews, are prone to underreporting [17-22], resulting in unfairly low estimates of adherence when being compared to prescription databases. More reliable methods, such as electronic drug monitoring, pill counts, and detection of medication in biological samples, are often not feasible in large-scale studies because of high costs and logistical challenges. Nevertheless, more insight into adherence to pharmacological treatment among pregnant women is needed not only to improve pharmacotherapy outcomes, but also to guide studies on medication safety during pregnancy, which increasingly rely on administrative healthcare databases and registries [23]. Therefore, we aimed to determine to what extent women adhere to treatment recommendations regarding prescribed medication during pregnancy using diaries as the reference standard, which are less prone to underreporting compared to methods that rely on recall over a longer period of time [24, 25].