Table 1

Discharge prediction model variables

Predictor group	Predictor variable
Demographics*	Age; gender; race
Administrative	Origin of admission; prior care locations (emergency department, surgical suite, other inpatient); hospital disposition*; Primary clinical service; observation status; isolation requirement; patient safety attendant; home health services orders
Temporal	Elapsed hospital and unit length-of-stay
Medications	Total counts: continuous infusion (active, new, discontinued), injected, oral, inhaled medications flagged as predictors: antibiotics; anticoagulants; antihypertensives; antiarrhythmics; neuromuscular blockers; diuretics; antihyperglycemics; addiction treatment medications;* analgesics; psychotherapeutics;* steroids; vasopressors
Interventions	Surgical intervention in preceding 24 hours; scheduled surgical intervention; tube thoracostomy; central venous access; mechanical ventilation; renal replacement therapy; plasmapheresis; blood product administration; supplemental oxygen; physical therapy; occupational therapy
Diagnostics	Laboratory study counts: critical care, chemistry, haematology, microbiology Imaging: CT, X-ray, nuclear medicine; interventional radiology; echocardiogram
Monitoring	Telemetry monitoring; vital sign frequency; urinary output measurements; total intake and output measurements
Rehabilitation	Physical therapy mobility assessment; occupational therapy activity assessment; postacute rehab potential
Consults	Addiction counselling;* case management;* cardiology; cardiac surgery; diabetes management; dietary counselling; gastroenterology; hospice; infectious diseases; intensive care; internal medicine; interventional radiology; nephrology; neurology; neurosurgery; oncology; psychiatry;* palliative care; social work;* surgery
Diet	NPO; enteral feeding; dietary nutrition supplements; regular diet
Clinical Flags	Pain management; substance abuse;* sepsis; cardiac arrest; acute kidney injury

*Correlates of social and behavioural determinants of health.