(857) Validation of the PROMIS© Medication Adherence Scale for Aspirin Among Pregnant Patients

Low-dose aspirin is an essential tool to prevent pre-eclampsia. A measure validated in the prenatal setting to evaluate adherence, which is essential to efficacy, is lacking.  We investigated the validity and performance of an instrument validated outside a pregnant population to measure medication adherence among pregnant patients of low socioeconomic status who were recommended low-dose aspirin.

Study Design:

This was a cross-sectional analysis using the National Institutes of Health (NIH) Patient-Reported Outcomes Measurement Information Systems (PROMIS©) Medication Adherence Scale (PMAS), which contains two subscales, Medication Beliefs and Knowledge (MBK) and Medication Taking Behaviors (MTB). This was assessed among pregnant people who were recommended low-dose aspirin at an urban tertiary care center clinic between April and July 2023. We assessed internal consistency and reliability, and compared mean PMAS scores using Mann-Whitney U and Kruskal-Wallis tests by self-reported maternal race/ethnicity, history of adverse pregnancy outcome (APO), or history of pre or gestational diabetes.

Results:

Among 40 patients surveyed, the mean PMAS score was overall high at 41.18 (SD 3.90) and the internal consistency reliability was strong at 0.83. Item-to-total PMAS score correlations demonstrated a moderate to good fit (r range 0.39-0.80) for all items except two in the MTB subscale which focused on medication side effects (MedAd7) and cost (MedAd9) (Table 1). Self-reported race, ethnicity, history of APO, or history of diabetes were not associated with a significant difference in PMAS score (p = 0.62, 0.40, 0.36, 0.70 respectively).

Conclusion:

This study revealed that use of PMAS in this population is reliable, but that the MTB subscale may be weakened by inherent characteristics of low-dose aspirin (e.g., few side effects and low cost). It may be preferable to use MedAd5 as a stand-alone item along with the MBK subscale in future evaluations of this population. There was no statistically significant difference in PMAS score by any demographic or clinical characteristic.