Utility of Population Attributable Fraction Assessment in Guiding Interventions to Reduce Low Birthweight in the High-Altitude State of Colorado

Abstract

Objectives We evaluated the ability of population attributable fraction (PAF) assessments to alter significant modifiable risks for low birthweight (LBW) and the impact of high altitude as a risk for LBW in Colorado. Methods Logistic regression analysis of birth certificate parameters in 1995–1997 identified risk factors for PAF assessment. PAF for birth at high altitude, multiple births, and LBW in singleton births were determined. Subsequent analysis of singleton LBW risks, using number needed to treat (NNT) analysis, estimated how elimination of major modifiable risk factors could reduce LBW in the study population. Public health interventions were initiated and PAF analysis conducted 12 years afterward to determine the effect of interventions. Results PAF in singleton births revealed low maternal weight gain in pregnancy and maternal smoking as the greatest modifiable attributable risk factors for LBW (12.7/12.5 %, respectively, in 1995–1997 and 12.9/7.1 % in 2007–2009). Significant interaction between these variables resulted in PAF of 34.4 % when the two occurred together in 1995–1997, decreasing to 19.4 % in 2007–2009. NNT analysis of singleton births in 1995–1997 revealed that eliminating low maternal weight gain, smoking, late prenatal care in all women and interpregnancy intervals <1 year in multiparous women reduced LBW by 46.5 %. The respective proportional reductions in PAF of 40.3 and 46.3 % for maternal smoking and weight gain/smoking interaction were associated with a 1.4 % LBW reduction in singleton births between the two study periods. Conclusions for Practice PAF and NNT analyses are valuable tools to predict intervention targets to lower LBW.

from Health via xlomafota13 on Inoreader http://ift.tt/2gtE68C
via IFTTT