How PRAMS Works: The Pregnancy Risk Assessment Monitoring System

At a glance

The Pregnancy Risk Assessment Monitoring System (PRAMS) is a surveillance project of CDC and health departments. It collects population-based data, specific to each participating jurisdiction, on mothers' attitudes and experiences before, during and shortly after pregnancy.

Who is surveyed

• The sample: women with a recent live birth, drawn from each jurisdiction's birth certificate file.
• Size: each site samples 1,000 to 3,000 women a year — 100 to 250 new mothers a month in a stratified random sample.
• Oversampling: some groups are sampled at higher rates to ensure enough data. Common stratification variables are birth weight, race and ethnicity, Medicaid status and geography, tailored to each site's needs.

How mothers are contacted

PRAMS uses standardized procedures and questionnaires so jurisdictions can be compared and data combined, while letting each site customize some parts. Its mail and telephone methods draw mainly on the research of Don Dillman, emphasizing many varied contacts.

1. Monthly batches: each month a new sample is drawn from the current birth certificates.
2. Mail: a series of mailings starts 2 to 4 months after delivery, giving new mothers time to answer questions about the early postpartum period. Every mailing also invites them to answer online — web surveys began with the 2023 birth cohort.
3. Telephone: mothers who don't respond are called after the last mailing, for 2 to 4 weeks, with up to 15 calls per working number.

More: PRAMS design and methodology overview.

The questionnaire

First developed in 1987, it covers:

• preconception, prenatal and postpartum care;
• obstetric history;
• substance use;
• physical abuse;
• contraception;
• economic status;
• stress;
• early infant development and health.

Weighting

• By jurisdiction: every site samples differently, so data are weighted by jurisdiction; combined PRAMS data are not weighted to give national estimates.
• Linked to birth records: answers are joined to selected birth certificate data (demographic and clinical). Because the sample is drawn from those records, sampling, nonresponse and noncoverage weights let results represent all births in each jurisdiction.
• Analysis: requires statistical software that handles complex sample designs.

Sources

Based on "Data Methodology," Pregnancy Risk Assessment Monitoring System, Centers for Disease Control and Prevention; a work of the United States government in the public domain.