Skip to main content

Table 1 Predictor variables organized in three main themes

From: Predicting hospitalization following psychiatric crisis care using machine learning

Sociodemographics SPI items Psychiatric care
Gender (cat) Suicide risk (cat) Patients’ informal social support system involved (cat)
Age (num) Danger to others (cat) Patient referrer (cat)
Living situation (cat) Severity of psychiatric symptoms (cat) Number of previous face-to-face treatment contacts up to 2 weeks/1 month/3 months/6 months/12 months before the index crisis care contact (num)
Marital status (cat) Problems with self-care (cat) Number of previous psychiatric hospitalizations (last 12 months and last 5 years) (num)
Cultural background (cat) Substance misuse (cat) Number of previous psychiatric day care treatments (last 12 months and last 5 years) (num)
Psychiatric diagnosis (cat) Medical condition(s) (cat) Number of involuntary treatments/hospitalizations (last 12 months and last 5 years) (num)
Global Assessment of Functioning (GAF) score (num) Disturbances in patients’ family connectedness (cat) Days of psychiatric hospitalization (last 12 months) (num)
  Professional functioning (cat) Any earlier psychiatric care referrals (> 1 year and > 5 years before current contact) (num)
  Stability of patients’ living situation (cat)  
  Patient is motivated to receive treatment (cat)  
  Prescription medication compliance (cat)  
  Anosognosia (cat)  
  Patients’ family involvement in informal care (cat)  
  Symptom persistence (cat)  
  1. Data types cat categorical data, num numerical data