Table III. Multinominal regression analysis and two latent class analyses.

Attributes	Multinominal regression		Latent class 1 46%		Latent class 2 54%
	Coeff.	95% CI	Coeff.	95% CI	Coeff	95% CI
Intercept	−5.51		−3.67		−4.64
OHSS per 1% (0–10%)	−0.22	−0.26 to −0.17	−0.075	−1.11 to −0.039	−0.36	−0.41 to −0.21
Cancellations per 1% (0–15%)	−0.086	−0.103 to −0.069	−0.19	−0.13 to −0.25	−0.051	−0.005 to −0.097
30% chance of pregnancy (1–4 treatments)	−0.84	−0.94 to −0.75	−1.41	−1.88 to −0.94	−0.55	−0.02 to −1.08
Number of injections
0	Ref.	−	Ref.	−	Ref.	−
4	−0.52	−0.75 to −0.30	−0.31	−0.61 to 0.01	−0.87	−1.05 to −0.59
20	−0.93	−1.15 to −0.71	−0.55	−0.96 to −0.14	−1.50	−2.23 to −0.87
Costs (per 1000 euro)	−1.16	−1.21 to −1.06	−0.67	−1.09 to −0.26	−1.63	−1.98 to −1.28
2 log likelihood	−645		−601
PseudoR2	0.403		0.418
cAIC*	1549		1397

^*cAIC, consistent Akaike info criterion, Coeff: coefficient,