Skip to main content

Table 2 Multiple regression model of one-year service engagement vs no service engagement and clinical factors from univariate analyses (i.e., p < 0.2), adjusted for gender and age

From: Don’t blame psychosis, blame the lack of services: a message for early intervention from the Greek standard care model

no SE vs. 1 yr SE

Univariate analyses

Multiple logistic (N = 170)

OR (95% CI)

P

OR (95% CI)

P

Gender (female vs. male)

0.64 (0.36, 1.12)

0.118

0.71 (0.34, 1.48)

0.364

Age on onset (years)

0.99 (0.96, 1.03)

0.748

1.00 (0.96, 1.05)

0.948

Diagnosis

 2 vs. 1

0.87 (0.39, 1.95)

0.740

1.03 (0.37, 2.85)

0.953

 3 vs. 1

3.93 (0.85, 18.1)

0.079

3.81 (0.43, 33.7)

0.230

cannabis use (≥ 1 wk vs. (< 1 wk))

1.52 (0.84, 2.76)

0.166

1.21 (0.59, 2.48)

0.605

Length of admission in days

0.99 (0.97, 1.01)

0.172

0.99 (0.97, 1.01)

0.465

PANSS entrance total

0.99 (0.98, 1.00)

0.150

1.00 (0.98, 1.01)

0.786

  1. Only PANSS entrance total was included in the multivariate analysis, as it was collinear with the PANSS entrance general symptoms. The selection of PANSS total score against general symptoms was made under the assumption that total scores represent a more generic profile of study participants and was found to result to an optimal fit regression model
  2. CI Confidence interval
Back to article page

BMC Psychiatry

ISSN: 1471-244X

Contact us