Recall accuracy for prescription medications: self-report compared with database information

Area of Science:

• Pharmacovigilance and Pharmacoepidemiology
• Methodological Study Design
• Medication Adherence Research

Background:

• Accurate assessment of medication use is crucial for epidemiological studies and patient safety.
• Self-reported data is frequently used but its reliability for prescription drugs requires evaluation.
• Previous research indicates potential inaccuracies in patient recall of dispensed medications.

Purpose of the Study:

• To evaluate the accuracy of self-reported use of nonsteroidal anti-inflammatory drugs (NSAIDs) and noncontraceptive estrogens.
• To assess the recall of drug name and dosage among patients who were dispensed these medications.
• To identify factors influencing the accuracy of self-reported prescription drug exposure.

Main Methods:

• A methodological study conducted in 1992.
• Sample selection from a pharmacy database: 560 individuals with NSAID/estrogen dispensations and 140 controls.
• Data collection via telephone interviews to ascertain demographic, behavioral, and drug information.

Main Results:

• Significant underascertainment of self-reported prescription drug exposure was observed.
• Recall accuracy for NSAID use was higher with multiple dispensations (85%) versus single (41%).
• Recall of drug name and dose varied: 30% recalled NSAID name, 15% name/dose; 78% recalled estrogen name, 26% name/dose.

Conclusions:

• Self-reported prescription drug exposure is subject to significant underreporting.
• Recall accuracy is influenced by the number of dispensations, age, and time since exposure.
• High specificity suggests overreporting is not a major issue, but underreporting is substantial.