Predictors of need and use of services
Independent predictors of need for contact with mental health services and actual use were examined using binary logistic regression. Fourteen independent variables were entered into the equation: sex (male=1, female=2), age (18-65), education (1=primary school, 4=university), employment (0=not employed or working from home, 1=employed and goes to work), marital status (0=not married, 1= married), location (0=from Ankara, 1=from outside Ankara), past hospitalization (0=no, 1=yes), accessible doctor/psychologist (0=no, 1=yes), worsening of symptoms (0=no, 1=yes), additional symptoms (0=no, 1=yes), had been PCR positive (0=no, 1=yes), PGI (1-7), COVID-19 Anxiety Score (0-24), and diagnostic group (0=SCZ, 1=BPD, 2=MDD, 3=ANX/OCD); the SCZ group was chosen as the reference group.
Table 2 shows the predictors of need for mental health services (Column 1), and actual use of services (Column 2). The predictors of the need for contact with mental health services were PGI score (self-reported severity of own symptoms), accessibility of a doctor/psychologist, and belonging to BPD, MDD, or ANX/OCD groups (higher need compared to the SCZ group). None of the demographic variables or COVID-19-related variables were predictive of the need for contact.
The same set of predictor variables were regressed onto actual use of mental health services. The predictors of actual use (Column 2) were: accessibility of a doctor/psychologist and belonging to MDD or ANX/OCD groups (lower services use compared to the SCZ group).