Path analysis on the risk of mortality in very low birth weight infants

The purpose of this work is to clarify by path analysis which of the predictor factors of mortality in a cohort of 423 singletons VLBW infants born at Beilinson Medical Center during the years 1980-1990 (pre-surfactant era) exert direct or indirect effect on mortality. A direct risk factor was defined as a factor having an immediate effect on the outcome without working through an intermediate variable. By contrast, an indirect risk factor was defined as a factor that does not have a direct effect on the risk of mortality but instead works through an intermediate variable. Evaluation of mortality was done at three points over time (pre-delivery, at birth and post-delivery).

In the pre-delivery model, gestational age, appropriateness for gestational age, mother's age and fetal distress work directly on mortality in VLBW infants. At birth, mother's age, birth weight and Apgar score at the fifth minute work directly on the risk of mortality. When the assessment of the risk of death was done post-delivery, mother's age, birth weight, Apgar score at the fifth minute and seizures influence directly on mortality risk, and fetal distress, gestational age and pre-natal induced hypertension (PIH) operate indirectly on the risk of mortality.

The methodology we have used can be adopted in other investigations to distinguish and measure the effect of prognostic factors on the risk of an outcome.