Learning Objective #1: Discuss the steps of exploratory factor analysis to a specific client data base

Learning Objective #2: Discuss the use of exploratory factor analysis and Cronbach's coefficient alpha to determine instrument validity and reliability

---

The determination of mothers' prenatal risk underlies successful case management of high risk maternity clients during pregnancy. Instruments must be reliable and valid in order to demonstrate associated perinatal risk reduction. Tulsa Healthy Start Initiative (THS) uses case management to assist high risk, low-income pregnant women during and after pregnancy, where 2/3 represent minority populations. The Client Risk Assessment (CRA), a weighted risk assessment inventory, was developed from the research literature and case managers' experience with clients to assess the extent of mothers' prenatal risk upon admission to the program and at 36 weeks gestation. The CRA has 3 subscales, a social risk subscale (SR) administered prenatally at program admission, at 36 weeks gestation and at specific postpartum intervals, a health risk subscale (PHR) administered prenatally, and an infant risk assessment. Psychometric evaluation of the maternal prenatal subscales (n=107) employed exploratory factor analysis and Cronbach's coefficient alpha. After exploratory analysis, the original 19 prenatal social risk items (PSR) were reduced to 16 items, and 25 prenatal health items (PHR) were reduced to 20 items. Analysis of the PSR, which included a principal components factor analysis with varimax rotation, yielded 5 factors with eigenvalues exceeding 1.0, accounting for 65% of the total variance. Sampling adequacy was good at 0.68. Item communalities ranged from 0.80-.43. Internal consistency reliability coefficient of PSR was relatively low (.69) because of the multidimensional nature of the instrument. Five PHR factors also emerged from the data with eigenvalues >1.0, accounting for 71% of the total variance. Item communalities ranged from 0.96-.54. Sampling adequacy was 0.80. Internal consistency reliability of the multidimensional PHR was .66. Findings provide preliminary evidence for CRA as a reliable and valid measurement of mother's prenatal risk, which may ultimately be used to test effectiveness of risk reduction interventions with at-risk pregnant populations.