A hierarchical procedure to select intrauterine and extrauterine factors for methodological validation of preterm birth risk estimation

Table 4 Akaike information criterion (AIC) differences among the candidate nested models presented in Fig. 3

Added Risk Factor	Number of Risk Factors	AIC Value	Δ_i
Preeclampsia	8	108.36	4.65
Antihypertensive Medication	9	104.47	0.76
Depression Level	10	104.84	1.13
Fetal Sex	11	103.71	0
Hormones Medication	12	109.62	5.91

We focus on models from size 8 to 12 risk factors. The AIC difference is calculated as Δ_i=AIC_i−AIC_min. Starting from values of Δ_i>4 the level of empirical support of model i, with respect to the best model, is considerably less (see [94] ch.2 for a detailed explanation). We highlight in bold the AIC differences related to the models for which there is a substantial empirical support (i.e. Δ_i<4)

ISSN: 1471-2393