Using Clinical Classification Trees to Identify Individuals at Risk of STDs During Pregnancy
Article first published online: 6 SEP 2007
Perspectives on Sexual and Reproductive Health
Volume 39, Issue 3, pages 141–148, September 2007
How to Cite
Kershaw, T. S., Lewis, J., Westdahl, C., Wang, Y. F., Rising, S. S., Massey, Z. and Ickovics, J. (2007), Using Clinical Classification Trees to Identify Individuals at Risk of STDs During Pregnancy. Perspectives on Sexual and Reproductive Health, 39: 141–148. doi: 10.1363/3914107
- Issue published online: 6 SEP 2007
- Article first published online: 6 SEP 2007
CONTEXT: Few studies have used classification tree analysis to produce empirically driven decision tools that identify subgroups of women at risk of STDs during pregnancy. Such tools can guide care, treatment and prevention efforts in clinical settings.
METHODS: A sample of 647 women aged 14–25 attending two urban obstetrics and gynecology clinics in 2001–2004 were surveyed in their second and third trimesters. Baseline predictors at the individual, dyad, and family and community levels were used to develop a classification tree that differentiated subgroups of women by STD incidence at 35 weeks’ gestation. Logistic regression analyses were conducted to assess whether the classification tree groups or commonly used risk factors better predicted STD incidence.
RESULTS: Nineteen percent of women had an incident STD during pregnancy. Classification tree analysis identified three subgroups with a high STD incidence (33–61%), one with a moderate incidence (16%) and three with a low incidence (6–11%). Women in subgroups with high STD incidence included those not living with the partner with whom they conceived and those who had a moderate or a high level of depression, a history of STDs and a low level of social support. A logistic regression model using groups defined by the classification tree analysis had better predictive ability than one using common demographic and sexual risk predictors.
CONCLUSION: This classification tree identified risk factors not captured by traditional risk screenings, and could be used to guide STD treatment, care and prevention within the prenatal care setting.