Evaluation of the Louisiana Fetal and Infant Mortality Review Surveillance System *

BACKGROUND:   Louisiana has the seventh highest infant mortality rate in the US (7.6 deaths per 1,000 infants under 1 year of age in 2010). The Louisiana Fetal and Infant Mortality Review Surveillance System (LaFIMR) is a statewide program that reviews fetal and infant deaths to create recommendations and support actions to decrease infant mortality. Since its inception in 2003, LaFIMR has never been evaluated.

METHODS:   This evaluation follows the 2001 MMWR “Updated Guidelines for Evaluating Public Health Surveillance Systems” and focuses on the attributes of data quality, representativeness, predictive value positive (PVP), and acceptability.  Data quality was evaluated using nine data points from the LaFIMR electronic database for abstracted cases. Representativeness and PVP were assessed using counts of abstracted LaFIMR deaths compared to vital records.  Representativeness was calculated as the proportion of true cases abstracted divided by total true cases, and PVP was calculated as the proportion of true cases abstracted divided by total deaths abstracted. Acceptability was assessed using a questionnaire sent to a convenience sample of stakeholders with an 87.5% response rate. Data sources include 2011 LaFIMR eligible cases from vital records (N=308),  2011 abstracted LaFIMR cases (n=176), and stakeholder questionnaires (n=24).

RESULTS:   Data were missing in the LaFIMR database for more than 10% of deaths for the following indicators: birth weight (13.8%), gestational age (12.5%), BMI (28.4%), mother’s employment (43.2%) and household income (98.9%). Data were missing in the LaFIMR database for less than 1% of deaths for date of birth/ death (0%), mother’s age (0.2%) and cause of death (0.6%). According to the LaFIMR case definition, 29 abstracted deaths in 2011 were not true cases. For 2011, the representativeness was 47.7% (147/308) and the PVP was 83.5% (147/176). Among completed questionnaires, 83.3% of respondents answered that they agreed LaFIMR positively impacts health outcomes in their community (62.5% strongly agreed). Stakeholders identified limited resources for community outreach, lack of timely data, and not having the ability to produce actionable data as LaFIMR limitations.

CONCLUSIONS:   LaFIMR does not provide adequate data for epidemiological analyses (poor data quality) and abstracts a low proportion of cases (poor representativeness) including some ineligible cases (moderate PVP). Cases should be abstracted and reviewed according to a standardized structure to improve data quality, representativeness and PVP. MCH advocates believe LaFIMR activities positively impact health outcomes through providing a valuable forum for collaboration between programs.