Filling in the Gaps with Local Level MCH Data: Developing the Women and Infants Survey for Health (WISH) in Florida

BACKGROUND:   The Florida Department of Health has implemented the Pregnancy Risk Assessment Monitoring System (PRAMS) as its principle MCH surveillance system since 1993. Florida is a large state with regional economic and racial/ethnic diversity. Over 50% of the 2013 FL PRAMS sample resided in the six most populous counties. While FL PRAMS provides state-level data, it cannot provide local level data for the majority of Florida counties, leaving gaps in knowledge about some counties with high infant mortality rates or significant racial health disparities. The FL DOH developed the Women and Infants Survey for Health (WISH) in 2013 with a complex sampling strategy designed to complement the PRAMS sample and fill in the data gap for counties with small to medium-sized populations.

METHODS:   WISH is an annual population-based survey of Florida mothers 15 and older who have given birth 2 to 6 months prior. Approximately 500 new mothers are sampled each month from the birth certificate registry. Mothers receive the questionnaire by mail. It is available in English and Spanish. The sample is stratified first by Healthy Start Coalition (HSC) area, effectively over sampling rural counties and assuring geographic diversity. Each of the 27 HSC samples is further stratified by race and birth weight. The six most populous HSC areas are excluded because they are well represented in the PRAMS sample. The WISH questionnaire contains 150 MCH indicators.

RESULTS:  WISH sampled 5,608 women during 2013 data collection. There were 1,457 completed surveys. Each HSC area comprises between 2-5.5% of the total dataset, attesting to the geographic diversity of the sample. WISH's complex sampling frame over samples women from small and medium-sized counties. As an example, Gadsden County is a rural county with a population of 46,894 and an infant mortality rate of 16 per 1,000. In 2013, PRAMS sampled 10 women from Gadsden County. In contrast, WISH sampled 174 and received 41 completed surveys from Gadsden County in 2013.

CONCLUSIONS:   The WISH sampling strategy allows for data collection in areas of Florida that are often missed by PRAMS due to small populations and low numbers of births. Stratification by HSC area will make it easier for local agencies to use the data to identify areas of need and improve services. Within 3 years of data collection, WISH is expecting to meet the demand for local-level MCH data for a majority of Florida counties.