Back to Annual Meeting

Back to Annual Meeting

APHA Scientific Session and Event Listing

Jallah M. Kennedy, RN, BS, Mark C. Fulcomer, PhD, George Rhoads, MD, MPH, and Marcia Sass, ScD. School of Public Health, Epidemiology Program, University of Medicine and Dentistry of New Jersey, 683 Hoes Lane West, Piscataway, NJ 08854, 609-233-1876, jallahmk@hotmail.com

Pregnancy and birth data collection in New Jersey, as in the rest of the US, is now done electronically. Although the Electronic Birth Certificate (EBC) system has improved the completeness and timeliness of data collection, issues of “missingness” and data quality persist. Beginning with revisions of the birth certificates in 1989, the 1,017,301 birth and 7,229 fetal death records of in-state events of NJ residents included in the analyses presented here cover the nine years through 1997. In the design process leading up to the implementation of the EBC, a uniform worksheet was introduced early in 1992, making it possible to compare results across three different “eras” of data collection approaches (i.e., certificates only, worksheets, and EBC). For key variables known to be strong predictors of infant and fetal deaths such as gestational age based on last menstrual period (LMP), prenatal visits, smoking, and weight gain during pregnancy, the EBC leads to much lower levels of missing data when compared to the earlier two data eras. However, compared to survivors, infant and fetal deaths have disturbingly higher levels of missing information in their records regardless of the approaches to data collection. Furthermore, there is also evidence of “digit preference” in LMP and other important predictors. In addition to reviews of more recent findings, a seminal paper by Hotelling and Hotelling (1932) provides some historical insights into statistical issues relating to data quality. Implications of these results for improving future data collection and training of care providers are also discussed.

Learning Objectives:

• Describe the trend in data collection on pregnancy outcomes in New Jersey from 1989 to 1997
• Describe the impact of missing data on the reporting of pregnancy and birth outcomes.
• Discuss the role of the quality and completeness of data collection to determine some predictors of adverse reproductive outcomes.

Keywords: Data Collection, Pregnancy Outcomes

Presenting author's disclosure statement:

Not Answered